--- Log opened Thu May 03 00:00:06 2018 --- Day changed Thu May 03 2018 00:00 < dviola> thinking to upgrade my hw to a ryzen 00:00 < dviola> E5500 -> ryzen 00:00 < catphish> bls: ok, i thought they were active-active 00:00 < catphish> though i never considered what the algorithm was, trying to find docs now 00:01 < TheSov> why does vmware not have a native ceph client 00:02 < catphish> i'm also a little confused at why multipathing is used with iscsi when iscsi has its own mechanism for this 00:02 < TheSov> eh? 00:02 < TheSov> what do you mean 00:02 < bls> what's iscsi's mechanism to do that? 00:03 < dfee> is there a pneumonic to remember that `cp -r dir/` copies the contents and `cp -r dir` copies the folder? 00:03 < TheSov> theres only 1 multipathing mechanism for iscsi 00:03 < TheSov> you login multiple times 00:03 < catphish> bls: "Communication between the initiator and target occurs over one or more TCP connections. The group of TCP connections that link an initiator with a target form a session" 00:04 < TheSov> right 00:04 < bls> catphish: yes, but those sessions present indendent paths to the same device, multipathing is what coalesces all those paths into a single block device 00:04 < catphish> "iSCSI targets and initiators MUST support at least one TCP connection and MAY support several connections in a session." 00:04 < Loshki> dfee: the mnemonic is 'man cp' 00:05 < catphish> bls: those sound like the same thing 00:05 < TheSov> they are not 00:05 < bls> if a target presents a LUN over ports A and B, the initiator host will see two different block devices 00:05 < bls> if the host accesses those block devices independently, corruption can occur 00:05 < dfee> Loshki: humerous, perhaps. but probably wasn't worth replying. 00:05 < TheSov> there is also the fact that multiple portals can present the same lun 00:05 < catphish> bls: that seems a bit silly doesn't it? why not just present one block device? 00:05 < bls> so the multipathing layer is what ensure only one session is used for OPs at a time 00:06 < bls> catphish: because you don't need that over in all cases 00:06 < bls> overhead 00:06 < TheSov> catphish, in some cases 1 disk is presented to multiple systems 00:06 < bls> if you've got a single target / init pair with a single session, there's no need for it 00:06 < TheSov> and in other cases the same disk is presented multiple times to spread the bandwidth 00:07 < bls> but if you want to do HA, you want multiple sessions. multipathing is what keeps OP collisions from occurring over those multiple sessions 00:07 < TheSov> in enterprise cases, you have multiple systems hitting the same disk from multiple interfaces 00:07 < catphish> why would you want multiple sessions to do HA though? 00:07 < TheSov> in case 1 session fails 00:07 < bls> so one session can fail 00:08 < TheSov> multipathing isnt just about the disk it can be on completely different networks 00:09 < bls> if you've got session sda and session sdb presented by MP as sdc, you could access the remote resouce via sda or sdc, but if sda fails, you're dead, if you use sdc, MP can reroute you via sdb 00:09 < TheSov> in fact it should be 00:09 < TheSov> bls right on 00:09 < catphish> thinking about it, maybe it's not always practical for all tcp endpoints to be able to share the same session 00:09 < bls> yeah, redundant NICs, redundant switches, separate subnets 00:09 < catphish> though i feel like ideally they should 00:09 < TheSov> proper iSCSI is an art 00:10 < TheSov> our "test" was to turn off entire switch stacks and see if disk connectivity was compromised 00:10 < catphish> the way i always see it used is with multiple sessions, held together by multipath 00:10 < TheSov> catphish, yeah thats it 00:10 < catphish> but reading the spec, that seems silly, when you could just use one session 00:10 < bls> yeah, and you can roll your own SAN using off the shelf software, but you have to have your chops and understand the stack. most people would rather pay a SAN vendor though, despite them only cutting out some of the headaches 00:11 < catphish> with that session spanning multiple tcp connections / networks 00:11 < bls> all that does is elevate the session concept up to the MP layer 00:11 < TheSov> im trying to turn our ceph cluster into a SAN via iSCSI but ubuntu server isnt co-operating 00:11 < TheSov> im thinking of switching distro's for this 00:11 < TheSov> but security will eat my ass for that 00:11 < catphish> mp layer? 00:12 < bls> something has to multiplex multiple TCP connections, the MP layer aka dm-multipath is what does this 00:13 < bls> and iSCSI isn't the only technology to do multipathing, so that multiplexing is abstracted out to a different layer 00:13 < catphish> bls: but iscsi itself supports all this, most importantly including command ordering, so it isn't clear to me why you wouldn't do it at that layer 00:13 < catphish> yes, i agree, multipath means you can do it over any protocol 00:13 < bls> catphish: because you'd have redundant multipathing code in the iSCSI stack, the FCP stack, the FCoE stack, etc, etc 00:14 < catphish> bls: sure, so the iscsi clients just don't bother and delegate it to the next layer, seems sensible enough 00:15 < catphish> which really just leaves my original question of how multipath prevents things getting out of order, couldn't seem to find its manual :( 00:15 < bls> so iSCSI presents its TCP based connections as sessions, FCP presents its FC based connections as sessions, FCoE its ethernet, etc etc and multipathing handles the management and ordering so it doesn't have to know about TCP, ethernet, *and* fibrechannel 00:16 < catphish> it's a shame really, because iscsi seemingly has nice support for multiple tcp connections per session for resiliance, but i can see why it's not used 00:17 < bls> someone could reorg their stack to put it all in an iSCSI driver and it'd probably be more efficient, but most shops I've dealt with used iSCSI and FCP and want a common interface layer to them both 00:19 < bls> similar to the way we don't have different interfaces and tools for different wireless devices 00:19 < bls> (anymore) 00:22 < bls> I used to have docs on the different multipathing algorithms and how they worked, but I forgot to migrate my bookmarks when I left that job 00:24 < kazdax> i hads ome benzos and bacardi and coke 00:25 < kazdax> looks like if i keep going this streek i might not be able to complexte my RHCSA 00:25 < catphish> just reading about how openiscsi uses different tcp connections 00:25 < bls> "The path selector context structure enables a path selector type to track state across multiple ios to the paths of a path group." 00:25 < kazdax> woah mate hold on a sec..this is all giberrish to me 00:26 < catphish> i think it created separare sessions to each destination as suggested 00:26 < kazdax> let me sip my rum and coke 00:26 < kazdax> i took 10 mg of klonopin 00:27 < bls> hmm, section 4.2 is similar to the doc I remember: https://pdfs.semanticscholar.org/5190/d091f525cead0c35c9271dea8a1109be6fbd.pdf 00:27 < bls> it's a bit dated, but includes the fundamentals 00:27 < kazdax> i think vim rocks 00:28 < catphish> i'm embarking on building a san, it's all good fun, but trying to get a handle on what to expect from active-active connections through multiple networks 00:29 < kazdax> is there a social linux channel ? 00:29 < kazdax> where i can dish out my stupidity > 00:29 < catphish> thsnks for the link 00:30 < kazdax> like #linux-social 00:30 < bls> kazdax: /msg alis list social <- lots of folks that'd love to hear about your Hunter S Thompson impressions 00:30 < rypervenche> kazdax: THis is pretty social, but depending on what distro you use there are often distro-chat channels. 00:31 < anickname> am I allowed to ask questions about linux packages here 00:31 < kazdax> i gotta say 00:31 < anickname> like linux software 00:31 < kazdax> ilove my debian 9 00:31 < anickname> if it doesn't really have a respective irc channel 00:31 < bls> anickname: you can try, yes 00:31 < kazdax> so far so no problems 00:32 < anickname> does anyone here use audacity for anything 00:32 < kazdax> except the part where i had to set it up for ATI cards 00:32 < bls> anickname: you might get directed to a better channel if we know the package uses a different channel name or there's a better tangentially related channel 00:32 < royal_screwup21> why is 127.0.0.1 also termed local host? 00:32 < kazdax> then i installed my old nividia card and awallah 00:32 < kazdax> works perfect 00:32 < anickname> thank you bls 00:32 < jml2> kazdax, nividia or nvidia? 00:32 < kazdax> nvidia 00:32 < jml2> kazdax, funvidia! XDXDD 00:33 < kazdax> man so far no problems with my debian on a bare metal install 00:33 < bls> royal_screwup21: because it's a fixed address that always refers to the local machine 00:33 < jml2> kazdax, how many noobs does it take to install a ubuntu? 00:33 < jml2> kazdax, the answer is.. ONE!! 00:33 < kazdax> none because its self instal :D 00:33 < royal_screwup21> ^the real answer 00:34 < kazdax> how can ubuntu not have a root 00:34 < kazdax> how stupid is that shit 00:35 < bls> because it does have one? 00:35 < kazdax> tell me why should i burn my windows installation once and for all ? 00:35 < triceratux> kazdax: it has root, it just doesnt have a password for it. so you have to get to it some way other than logging in 00:36 < kazdax> ubuntu makes me feel like a weenie on a hot dog 00:36 < Loshki> anickname: I used audacity a few times (and cooledit) for cleaning up vinyl recordings before turning them into mp3 00:36 < anickname> Loshki would you have any experience working with audacity modules 00:36 < kazdax> debian makes me feel like pro in work 00:37 < supernovah> hey how can I translate a devpath to a potential symlink in /dev if there is one? eg: DEVPATH="/devices/platform/soc/3f980000.usb/usb1/1-1/1-1.2" => "/dev/ttyUSB0" 00:37 < kazdax> Arch make smkae me seem like an admin of admins 00:37 < kazdax> Gentoo makes me feel like a color changing lizard 00:37 < bls> supernovah: in a generic way or for that specific case? 00:37 < Loshki> anickname: only the noise reduction and compression ones. Which ones are you trying to use? 00:38 < supernovah> bls: generic 00:38 < supernovah> bls: udevadm info -p "$DEVPATH" doesn't return the symlink anywhere 00:38 < kazdax> i am getting my bookk on RHCSA tomorrow and gonna study study study 00:39 < kazdax> someone mentioned getting an oreily subscription 00:39 < kazdax> i might gett hat to 00:39 < catphish> confirmed: openiscsi makes separate sessions per tcp destination and presents one block device for each 00:39 < bls> ah, then I'm not aware of a generic way 00:39 < kazdax> give me all that linux can do 00:40 < kazdax> all that it can do 00:40 < supernovah> bls: in /sys/class/tty I can see for example, all the ttyUSB* are symlinks to ../../devices/platform/soc/3f980000.usb/usb1/1-1/1-1.2///tty/ which I could iterate over and figure out what dev symlink to point to 00:41 < anickname> Loshki I am trying to use the mod-script-pipe module 00:41 < kazdax> i really need a social site 00:41 < catphish> dm-multipath seems undocumented though 00:41 < anickname> however, no matter what location I put it in, it doesn't detect it 00:41 < anickname> it keeps telling me no modules found despite the fact mod-script-pipe.so is in the modules location 00:41 < anickname> where is the location that I need to put it in? I tried ~./audacity-files/modules 00:41 < anickname> and the /usr/share/local/audacity (I think that's what it was) 00:42 < kazdax> cani be as good admins are you guys are ? 00:43 < bls> right, but each subsystem is allowed to organize its /sys/.../devices/... tree as it sees fit. udev is the attempt to wrangle that back to coherent /dev entries 00:44 < triceratux> kazdax: https://www.reddit.com/r/linux/ ? 00:45 < jml2> kazdax, looking for online resources? opensource.com, and lxer.com are some of my bookmarks 00:45 < Loshki> anickname: sorry, no idea. Check your os docs. Gotta be a faq somewhere on adding plugins to audacity, surely? 00:45 < jml2> kazdax, fwiw, it's more effective and cheaper for me to get "ebooks" than print books -- I can easily click on urls from books, and copy/paste commands, package names etc.. 00:47 < anickname> Loshki I think so, I'm on Puppy Linux which is kind of an obscure OS 00:47 < jml2> kazdax, I never found a book that explains anything perfect, there's always some skipping in explanation, so it's good to alwayas google a topic when a particular chapter of a book is rather weak 00:47 < gurki> has anyone in here experience with moving a qcow2 image (not sure if relevant: formatted using llvm) to a physical drive? 00:48 < gurki> i find google hits telling me it can be done in one step using qemu-image and hits telling me it doesnt work 00:48 < jml2> kazdax, some things, /./and- "Everything" --- no book explains everything with the new admin in mind.. 00:48 < gurki> im confused. 00:48 < jml2> a qcow2 image, is just like a file 00:49 < gurki> (i should add that i want to move this virtual machine to a physical server) 00:49 < bls> gurki: there's a qemu channel on here that might know 00:49 < jml2> you can map/mount a qcow image with qemu-nbd 00:49 < jml2> make sure no VM is using it at the same time 00:49 < kazdax> i agree jml 00:49 < kazdax> jml2 00:50 < uplime> how do you format something with llvm? 00:50 < uplime> or do you mean lvm? 00:50 < kazdax> its trail and erorr 00:50 < gurki> oh 00:50 < gurki> yes i do 00:50 < gurki> thanks for pointing that out uplime 00:50 < uplime> np 00:50 < uplime> just wanted to make sure thats what you meant 00:51 < Loshki> anickname: audacity is a standard userland app. Unless your version is very old and/or they've changed the plugin interface (unlikely) chances are your answers are in here somewhere: https://manual.audacityteam.org/ 00:51 < gurki> jml2: im not sure why i would want to mount it if i just kind of want to dd it to some disk? i have a bad feeling about this being bootable if i just copy the files 00:51 < bls> gurki: but unless you've done some serious one-off, manual configuration using proprietary, can't be reobtained software, you're probably better off just installing from scratch and copying over the relevant data 00:51 < kazdax> my being high on bexos and on bacardi rum ..is not going to be a problem to you folks is it ? 00:51 < gurki> bls: ... i did. which is the reason why id want to clone it 00:51 < kazdax> i am preety chill 00:52 < gurki> you implying ill run into trouble? 00:52 < anickname> Loshki thank you 00:52 < mawk> I'm trying the iptables modules to block port scanning 00:52 < mawk> already 8 IPs are glued to my web 00:52 < bls> gurki: if you don't know what you're doing, yeah 00:52 < mawk> and one ipv6 ! the ipv4 could be automated scanning, but the ipv6 means someone was targeting me specifically 00:53 < bls> gurki: getting the partitioning and bootloader right are error prone 00:53 < gurki> the fact that my grasp of this isnt even enough to tell me where i might run into problems probably is a bad sign 00:53 < gurki> where is the difference between cloning hdd a to hdd b to cloning some vm hdd image to a physical hdd? 00:53 < bls> so unless that's one of the things you've heavily customized, it's usually easier to just install the same OS and copy the config files and data over 00:55 < bls> some of the problems you'll hit are the same 00:55 < bls> like different sized disk devices, or disk UUIDs changing, or device order changing 00:56 < jml2> gurki, with linux you can clone any file, just edit the fstab, chroot/rescue it from a live-installer and fix grub... works with any distribution... as long as you can copy the files (hint: use "lsblk -f" to find the uuid) 00:56 < jml2> gurki, ./can clone any distro filesystem./ 00:59 < jml2> gurki, if you use advanced lvm management on the "source" , using it on the target is rather a choice.. 01:00 < gurki> nothing more advanced than what a standard centos install will do 01:00 < jml2> gurki, if you are going from 'flat file'(source)' to a more advanced storage like 'lvm' or 'mdadm' -- that's tricky because you need to make sure that you have the lvm or mdadm support in the system.. 01:00 < de-facto> Hey, which is the best way to track disk writes? i am on openwrt r/w root and want to avoid wearout 01:01 < gurki> i probably shouldve ditched lvm for this setup but i cannot go back to 0 at this point 01:01 < jml2> gurki, if you want to use lvm, you can recreate the lvm things first on the target, map it { then do the file copy } 01:01 < gurki> so dding the whole lvm will break horribly because the uuids changed 01:01 < gurki> did i get this correctly? 01:01 < jml2> gurki, uh -- if you use that qemu-nbd, you will later then be able to use lvm to map the dm devices, then you should be able to do mounting 01:02 < jml2> gurki, from there you can file copy 01:02 < jml2> gurki, to a simple partition (eg, /dev/sda[N]) device containing ext2/3/4 or whatever filesystem that is supported 01:03 < gurki> can you put into simple words why youd prefer this approach to the dd one? 01:03 < mawk> I banned some google bots by mistake I think, it's called g2600-1406-0032-0000-0000-0000-0000-00c2.deploy.static.akamaitechnologies.com. 01:03 < jml2> I have fedora 28 in a vm, and if I wanted to I would be able to do just this.. 01:03 < mawk> the settings look very fine-grained 01:03 < jml2> I've done this over a decade ago XD 01:04 < gurki> this looks a lot more complicated to me than dding :-| 01:04 < jml2> gurki, because using dd is primarily stupid for anything and is largely useless 01:04 < jml2> gurki, you dont want dm to get confused when it sees the same lvm id's all over the place 01:04 < gurki> im fine with stupid if stupid implies "a gurki can get it to work" ... 01:04 < gurki> hm 01:05 < ayecee> that's an unusual word 01:05 < ayecee> oh, that's your nick 01:05 * ayecee gets some coffee 01:05 < kazdax> time for some benzos and some qliour 01:05 < jml2> gurki, someone a week ago described to me what you were doing, but they've been using linux for 20 years 01:05 < jml2> gurki, good luck with that XD 01:06 < ayecee> kazdax: in private, ideally 01:06 < kazdax> eyeeee 01:06 < gurki> now thats encouraging :P 01:06 < kazdax> no way ofr te pirate in here mqtey 01:07 < ayecee> the slurring is not encouraging 01:08 < kazdax> ayeeee noted 01:08 < kazdax> i am going to go play some dead alive 4 01:09 < ayecee> sounds like fun 01:09 < kazdax> yea onlyif my brother give me his system t play with 01:10 < bls> maybe you could trade him some drugs for a gaming sesh 01:11 < kazdax> he only smoooked weed and nothing else 01:12 < kazdax> iwish i was in an ashmram doing my computer studies 01:13 < kazdax> eating dall and rottie ..and writting mean ass code 01:13 < stevendale> Hey, remember that AMD Radeon HD 6450 1 GB I ordered, well it just got here, turns out the seller gave me an AMD Radeon HD 7470 1 GB instead :P Not gonna complain to them haha 01:15 < stevendale> Definitely better than my GeForce 8400 GS now 01:16 < jml2> stevendale, where's that link? hahahaa 01:16 < jml2> stevendale, i'll order one from that idiot XD 01:18 < stevendale> jml2: https://www.ebay.com.au/itm/AMD-ATI-Radeon-HD-6450-1GB-PCIe-DisplayPort-DVI-Video-Card-Low-Profile/182695116679?epid=2261112341&hash=item2a897a3f87:g:V1IAAOSw0axZftb4 Only ships to Australia though 01:19 < jml2> stevendale, ah its used LOL 01:19 < jml2> stevendale, I don't buy second hand things tehehehe 01:19 < jml2> stevendale, I've had bad experience getting second hand hardware.. 01:19 < supernovah> Ok, more simply, is there a way to figure out the /dev file descriptor associated with a device if I know its DENAME eg, /dev/bus/usb/001/002 and I want to find an association with that and /dev/ttyUSB1 01:19 < stevendale> Mine doesn't seem to have any wear at all 01:19 < stevendale> So many I am lucky 01:19 < stevendale> It's actually still shiny haha 01:21 < bls> he accidentally played a game on it and had to get rid of it for fear of bringing shame to his mining operation 01:23 < spear2> is there a way to use fdisk or parted to avoid doing sector arithmetic? like i want to make a 40GB partition starting immediately after an existing partition? 01:25 < catphish> fdisk usually defaults the start address to the next available, i rarely find i need to calculate anything 01:25 < mawk> yes spear2 01:25 < mawk> just read what fdisk is saying 01:25 < mawk> use + to give a GB size 01:26 < spear2> oh thanks! i missed that 01:29 < mawk> why would akamai scan my ip ? 01:29 < mawk> ns3-32.akamaistream.net ns3-32.akamaicache.net. and a ton of other akamai IPs 01:30 < ayecee> hackers 01:30 < mawk> :( 01:30 < mawk> I wonder what they did 01:30 < mawk> maybe the filter is too effective 01:31 < neoncortex> mawk: how are you logging it? 01:31 < mawk> I found an iptables module blocking TCP stealth scan, TCP SYN scan, TCP connect scans, banner grabbing scans 01:31 < bls> why wouldn't they scan your IP? 01:31 < mawk> because my robots.txt say so 01:32 < mawk> also a more configurable module that blocks port scanning, which is effective for ipv6 01:32 < mawk> if 10 probes for privileged ports are found, or 30 probes for unprivileged port, it is blocked 01:32 < mawk> or a combination of both, with respective weights 01:33 < neoncortex> I heard about a guy who owns an isp that akamai are one of the biggest on the internet, like they onws a good part of it (don't ask me why or how, just what the guy said) .. 01:33 < bls> hehe, to be "disruptive", you've got to ignore the laws....like robots.txt 01:33 < ayecee> i bet your firewall is counting FIN or RST from sessions your stateful firewall forgot about as probes 01:34 < ayecee> if you change it to only consider SYNs as probes, you probably won't see that. 01:36 < neoncortex> now .. what's the importance of check for malloc's return? like if it return me NULL, my application are screwed anyway .. I' starting to check it's response but I'm not so sure of it's importance 01:37 < mawk> easier debug 01:37 < mawk> you immediately know where it went wrong 01:37 < neoncortex> mawk: yes, good point 01:37 < ayecee> neoncortex: would be useful to exit with a sane message in that case 01:38 < mawk> ayecee: for ipv6 it's port scanning which is in effect, not the intelligent probes module 01:38 < mawk> so it was a real port scanning 01:38 < bls> neoncortex: not to mention if your application is complex enough, you can potentially free() stuff you've been caching 01:38 < mawk> but yeah I'll try to slow down that filter 01:38 < mawk> you say to only count SYN probes ? 01:38 < mawk> or at least drop "connect" scans 01:40 < Loshki> neoncortex: depends on the app, and whether it's ok for it to just crash, or whether you want to handle this (presumably rare) condition in a mnore graceful way. I'm thinking the software that runs the lasix lasers, for example... 01:40 < mawk> what ? 01:40 < mawk> every sane developer should handler malloc return 01:40 < bls> s/malloc // 01:40 < mawk> not randomly wait for a crash that provides no useful information 01:40 < mawk> yeah 01:41 < mawk> the idioms are pretty simple 01:41 < mawk> if (syscall() < 0) err(EX_OSERR, "foobar failed"); 01:41 < mawk> you can even make a little macro that will include line number etc 01:41 < jim> neoncortex, there also used to be (or maybe still is) something called lclint, which, if you put certain special comments in the code at all the right places, would be able to check for memory allocation problems 01:41 < mawk> but the message is enough 01:41 < Loshki> Lazy programmer, no diet coke! 01:42 < ayecee> :o 01:42 < bls> I get spoiled with languages with exceptions. Now if I could just have strong typing too 01:42 < neoncortex> I'm fearing that if I do not exit as soon as malloc fail, the application can do something 'funny' with data it deals with, but I'm extrapolating here 01:42 < mawk> not a chance I think 01:43 < mawk> unless you're doing some undefined behavior things 01:43 < mawk> a failed malloc will return NULL, a failed mmap will return (void *)0xFFFFFFFFFFFFFFFF 01:43 < bls> neoncortex: if the program isn't structured in a way where you can recover memory yourself, that's often the best approach 01:43 < ayecee> unless you handle the failed malloc, that's a very real expectation 01:43 < stevendale> :P 01:44 < mawk> bls: I made a C++ lib to wrap all these things for me, with exceptions 01:44 < bls> and "failed to allocate memory for foo" is infintely more helpful than a segfault 01:44 < mawk> using std::system_error 01:44 < dannylee> :: 01:44 < mawk> yeah 01:44 < mawk> eg sys::check(open("/dev/net/tun", O_RDONLY), "failed to open the tun device"); 01:44 < bls> nice, I dealt with C++ in the era where exceptions were a no-no 01:45 < mawk> and you have all kind of nice names for the errno values 01:45 < neoncortex> jim: I read about it somewhere, but looks like it's old stuff 01:46 < mawk> try { something that crashes } catch (std::system_error& e) { if (e.value() == std::errc::resource_not_available) do something; else throw e; } 01:46 < Loshki> Segfaults aren't so bad, you get a stack trace, usually with a big arrow saying "HERE". The worst are when your allocation fails but you don't notice and overwrite some random area of memory. You don't notice the corruption for a while, when it finally falls over, your miles away from the cause, mumble, grumble, etc. 01:46 < mawk> that should happen very unfrequently 01:46 < ayecee> it only takes once 01:46 < mawk> often you'll offset the NULL pointer with something like a page or two maximum, it'll still make a segfault 01:47 < mawk> you'd have to make funny things for it to overwrite real data 01:47 < ayecee> "most of the time you'll get away with it for a while" 01:47 < ayecee> is what i'm hearing 01:47 < stevendale> Figured out the GPU thing 01:48 < bls> you get a stack trace if you're writing it/have it compile the correct way. if you've given the program to someone else and the only notes in the bug report are "the program is crashing", you're SoL 01:48 < stevendale> The 7470 is actually a 6450, but with significantly higher clock speeds 01:48 < mawk> I just say that to have enough precautions 01:48 < mawk> in my opinion it will crash 100% of the time if you're not doing undefined behavior stuff with your pointers 01:48 < mawk> but it's not a good method 01:48 < mawk> it could crash 3489843 instructions laters 01:48 < jim> yep, lclint was removed from debian in 2003, as another lint called splint which does the same thing as lclint and had additional features (such as extensible annotations and buffer overflow detection). it's said to be the successor to lclint 01:48 < stevendale> http://www.tomshardware.com/answers/id-1890015/bought-radeon-hd6450-detected-7000-series.html 01:48 < mawk> -s 01:49 < bls> there's also valgrind and I know some people were using boehmgc for it as well 01:49 < neoncortex> I'm using valgrind, it helps a lot 01:50 < jim> splint is still in debian today (in stable, testing and unstable) 01:50 < jim> look into it if you want to, seems like it could also help 01:51 < mawk> in my bash shell the lowest mapped address is 0x55d2b304f000, you'd have to offset your null pointer by that amount to get to any area that won't automatically result in a segfault 01:51 < Loshki> Things that should happen only infrequently are almost as annoying as things that only happen under heavy load in the field 01:51 < neoncortex> yes, I'll look 01:52 < ayecee> i once worked on debugging a crash in a multithreaded program, where the author decided not to lock certain shared structures because "threads would rarely be here at the same time" 01:52 < mawk> lol 01:52 < ayecee> sure enough, it only crashed after several hours of running. 01:52 < mawk> all major vulnerabilities I've seen for linux lately are based on race conditions 01:52 < mawk> like that 01:53 < mawk> linux or programs like sudo 01:53 < markasoftware> sudo had a race condition vuln? 01:53 < mawk> race condition was just some step among the others in the proof of concept 01:53 < mawk> the vulnerability was very dumb 01:54 < mawk> sudo was reading /proc/self/status to get its own command name 01:54 < mawk> /proc/self/stat rather, to read some number in it actually 01:54 < mawk> and that file is space separated, but guess what's in it also: the command name 01:54 < self> proc you 01:55 < mawk> so you could add a command name with many spaces to put the right number into the field sudo wanted to read 01:55 < mawk> that's a variable you control, then using some race condition you were able to execute anything you want 01:56 < neoncortex> mawk: that's amazing, I mean, we do not expect things like that on critical software like sudo 01:59 < mawk> sudo does other dumb things 01:59 < markasoftware> sudo just feels simple. But no 02:00 < mawk> type this: sudo -u nobody python3 -c 'import tty, fcntl, os; fd = os.open("/dev/tty", os.O_RDWR); [fcntl.ioctl(fd, tty.TIOCSTI, bytes((c,)), False) for c in b"id\n"]' 02:00 < mawk> it will type a command as your current user instead of nobody 02:01 < mawk> it's virtually patched, it's just that the default config doesn't mitigate this 02:01 < mawk> maybe because the fix is too heavy, but there are very lightweight fixes as well, and it's not done 02:10 < supernovah> I dream of multithreading... one core arm is sloow 02:11 < spear2> did extraUser options change in 18? before i had my user setting 'home' and 'extraGroups' parameters, but not 'uid', i thought i just uncommented the default generated configuration.nix but maybe not 02:11 < c-c> supernovah: you can get a 4 arm core board for 20-30 - an used android tv box 02:13 < c-c> supernovah: something like https://medium.com/@tomac/how-to-install-libreelec-linux-on-cheap-android-tv-box-51f82cdf10c1 02:14 < c-c> supernovah: if you invest about a hundred, you can get to desktop-ish levels https://cdn-images-1.medium.com/max/1600/1*T4viatIfvFEi0HTjPODUnQ.png 02:14 < dannylee> https://www.youtube.com/watch?v=JSUIQgEVDM4 02:15 < c-c> dannylee: go to ##music 02:15 < dannylee> ok 02:15 < dannylee> man 02:15 < dannylee> ok man your ok 02:16 < c-c> yep, ##music is for sharing your music links and banter 02:17 < c-c> oops. My bad. dannylee sorry, its actually #music 02:21 < ayecee> the official music channel 02:26 < dannylee> does ##emacs still work 02:27 < triceratux> tias 02:28 < ayecee> try it and see 02:28 < dannylee> no 02:29 < ayecee> ok 02:29 < dannylee> maybe i was bande 02:30 < dannylee> baned 02:30 < ayecee> could happened 02:30 < dannylee> does emacs has a channel 02:30 < ayecee> ##emacs appears to be a valid channel 02:31 < dannylee> ok they banded me 6 months ago 02:32 < ayecee> unfortunate 02:32 < dannylee> ya i love emacs..i compile my C programs 02:35 < andy1633> dannylee, why did they ban you? 02:36 < bls> andy1633: don't ask 02:36 < mawk> lol 02:36 < dannylee> i just ask too many question 02:36 < andy1633> :( 02:36 < dannylee> emacs might really be G0D 02:38 < c-c> ok, they must be vim users... 02:38 < Trel> Is there a way to connect to a wifi network with nmcli without it saving a connection profile, so pretty much a one-off? 02:39 < bls> Trel: I usually just use wpa_cli for that 02:39 < snugger> c-c: emacs masterrace 02:39 < bls> snugger: not in here 02:39 < PepeLePew> '/save 02:40 < Trel> bls: cool, I'll give that a look 02:40 < Trel> haven't used it in ages 02:40 < Trel> hope I remember how XD 02:40 < bls> I might have a script for it lying around somewhere 02:42 < dannylee> some time i troll just a little..but i just want too have fun..but not allot..some people are stinkers..life is really short..just have fun/// 02:43 < c-c> dannylee: what if your trolling is not fun to other ppl? 02:44 < ayecee> asking the important questions 02:44 < dannylee> ok linux is not really G0D...the G0D joke have too stop 02:44 < Trel> bls: does using that method require network manager to NOT be running? 02:44 < Starcraftmazter> hi 02:44 < Starcraftmazter> how do i find while module taints kernel? 02:45 < Starcraftmazter> i cant seem to find it 02:45 < ayecee> Starcraftmazter: it would say when the module was first loaded 02:45 < supernovah> Is there a standard place to put IPC communication channels on the filesystem 02:45 < Starcraftmazter> ayecee: ok but what command should I run to find out 02:45 < ayecee> Starcraftmazter: dmesg, maybe 02:46 < ayecee> i don't think it's recorded aside from a log message when it happens 02:46 < Starcraftmazter> ayecee: tried that, theres no info beyond things complained they are tainted 02:46 < c-c> supernovah: I would prefer user home, /tmp/name-of-your-prog is very good in my books, too 02:46 < bls> Trel: ah, not sure about that. probably so. NM doesn't suffer other config methods 02:46 < ayecee> Starcraftmazter: may have to see the messages from shortly after boot 02:46 < ayecee> Starcraftmazter: which may be stored in /var/log/dmesg.log 02:47 < Trel> Darn, I do want to use network manager, just didn't want to save the profile, oh well, guess I can't have everything 02:47 < bls> tried with nmtui? it might have more options exposed for not saving the network selection 02:47 < Starcraftmazter> ayecee: ok ill restart then 02:48 < c-c> supernovah: I use /tmp/progdir when I dev, and once its working ~/.cache/progname... 02:48 < Trel> I'll give it a look (I'm doing this all CLI btw) 02:48 < mawk> where is the routing cache ? 02:48 < ayecee> in memory 02:49 < mawk> yeah but I mean where is the base source for it for userspace ? 02:49 < ayecee> how do you mean? 02:49 < mawk> I'm reading the rtnetlink doc but I don't see anything else than just stuff for getting all routes or specific routes 02:49 < mawk> not routes for a specific address 02:49 < mawk> well I want to do what ip route get 8.8.8.8 is doing, from C 02:49 < mawk> without doing that terrible system("ip route get 8.8.8.8") of course 02:50 < c-c> supernovah: I guess I didn't exactly answer your Q, I don't know of such standard place for IPC fp 02:50 < ayecee> mawk: you'd pretty much need to pull the whole routing table, then replicate how routing is done to figure out which one it uses 02:50 < mawk> but ip route get says "(cached)" when I do ip route get 02:50 < mawk> so I thought there would be a nice cache ready for me to query 02:50 < ayecee> i see 02:51 < ayecee> there could be. i don't know offhand. 02:51 < mawk> instead of building the whole table, figuring out the metrics, going through all the routing policies, executing them with that almost turing complete execution flow, etc 02:51 < mawk> which is pretty complicated 02:51 < ayecee> metrics are easy, they'd only be relevant when you have two routes that match with the same prefix length. 02:52 < mawk> yeah 02:52 < mawk> and well routing policies don't pop every now and then on people's computer but I want a correct program 02:52 < ayecee> cool 02:52 < ayecee> i bet ip uses a library for that 02:53 < mawk> the same library as I probably 02:53 < mawk> or an even lower level one, librtnetlink 02:53 < mawk> so I'm using the same stuff basically 02:53 < mawk> let's strace ip to figure out how it does 02:56 < dannylee> Artificial intelligence might take over the entire computer network..then we will be stuck??? 02:56 < spear2> is "X display server" the same as an "X server" ? 02:56 < mawk> they won't hack through my server dannylee , I have a good firewall 02:57 < bls> spear2: depends on the context, but likely yes 02:57 < dannylee> c0000l 02:57 < c-c> corporations have been artificial intelligence for over 100 years 02:57 < dannylee> no 02:57 < dannylee> 10 02:57 < Starcraftmazter> hey guys, got the dmesg https://paste.fedoraproject.org/paste/pgIPBZKi3Hnd4iLutaTvFw - nothing there about any modules tainting kernel :S weird 02:58 < c-c> also, current cpu's cannot provide you with enough calculating power to come up with meaningfully strong AI - they are good for processing lists of data, only. 02:58 < mawk> apparently just the dumbest operation gives you the hit from the cache, RTM_GETROUTE 02:58 < mawk> alright then 02:58 < spear2> ty 02:59 < c-c> so, you can shove all that AI hype, its same as people fuzzing over having to live on Mars 03:00 < dannylee> i wish all of you success and happyness... 03:00 < c-c> (while living in cities with poor air&water quality, dangerous traffic&crime, etc etc) 03:01 < bls> flying cars + nuclear fusion + the singularity...got it 03:01 < phinxy> How do I close weechat when its opened in console tty (no x)? 03:01 < bls> phinxy: /quit 03:01 < phinxy> What if there was no /quit? 03:01 < PepeLePew> :) 03:02 < andy1633> ctrl-c? 03:02 < bls> you could kill it 03:02 < mawk> switch to another console and kill it 03:02 < bls> /exec pkill weechat 03:03 < jim> phinxy so did /quit /quit? 03:03 < phinxy> y 03:04 < jim> because its good to know! 03:04 < phinxy> how do you mean? 03:05 < jim> didn;t you ask why? 03:10 < stevendale> What the hell 03:10 < stevendale> Lubuntu 18.04 needs a gigabyte RAM 03:10 < stevendale> Apparently 03:11 < stevendale> https://en.wikipedia.org/wiki/Lubuntu 03:11 < stevendale> What happened to being lightweight 03:13 < dogbert2> a gig of ram is standard on even a Raspberry PI these days (or most clones) 03:14 < xamithan> gig of ram is pretty lightweight 03:14 < xamithan> For todays standards anyway 03:14 < triceratux> stevendale: try xubuntu or lubuntu-next lxqt http://cdimage.ubuntu.com/lubuntu-next/daily-live/20180421/ 03:14 < stevendale> What's the difference triceratux o/ 03:14 < xamithan> I thought xubuntu was heavier 03:14 < Adran> is lxqt a thing finally? 03:15 < triceratux> stevendale: they might give you some alternatives before hunting for lightweight distros per se. the solution is generally antix. http://antix.mepis.org/index.php?title=Main_Page 03:17 < triceratux> Adran: theres about half a dozen distros which ship an LXQt spin, & some of them are using v0.12.0 which is exceedingly bugfree & quite usable 03:17 < TheNH813> Allright, I need help. wget keeps failing me and interpreting a URL as several seperate urls. 03:17 < triceratux> http://www.extix.se/?p=393 03:18 < jml2> lol 03:18 < jml2> 1gig of ram and you're complaining? 03:18 < Adran> triceratux: what is one that is prespun? 03:18 < jml2> all distros these days for a desktop need around 1 gig,.. you can get away with it ~ 512mb.. 03:18 < jml2> possibly even less for "light weight" desktops 03:19 < TheNH813> I'm trying to set a GET request to this url with variables like so: "wget https://$userName:$userPassword@dynupdate.no-ip.com/nic/update?hostname=$hostName&myip=$ipAddress" 03:19 < jml2> TheNH813, lol.. save yourself the trouble, just use noip.com's daemon :)) 03:19 < jml2> TheNH813, it makes this a total snap 03:19 < TheNH813> It honestly dosen't work. 03:20 < TheNH813> Which is why I'm writing this script. 03:20 < triceratux> Adran: that one i just posted ^^ ExTiX. ALTlinux sisyphus also has a weekly which is excellent http://mirror.yandex.ru/altlinux-nightly/snapshots/20180502/ & so does fedora but i havent looked at it in awhile 03:20 < jml2> TheNH813, it does work 03:20 < jml2> TheNH813, use the binary 03:20 < jml2> TheNH813, it's distro-agnostic.. 03:20 < TheNH813> Okay. I'l give it a try then. 03:21 < jml2> TheNH813, it should be prompting you the credentials, use the --help iirc to give you insight.. 03:21 < Adran> triceratux: thanks 03:21 < jml2> TheNH813, you need to store and run things at their proper location.. 03:22 < jml2> TheNH813, ./noip2-x86_64 --help (cd /usr/local/bin) 03:22 < triceratux> Adran: np. i use it a lot actually. the pcmanfm-qt is lightyears beyond pcmanfm & supports mtp. & its easy to put in kde apps 03:23 < jml2> TheNH813, /etc/systemd/system/no-ip.service -- make this file https://pastebin.ca/4020364 03:23 < jml2> TheNH813, then systemctl daemon-reload ; systemctl enable no-ip ; -- but do not start it yet 03:24 < jml2> TheNH813, configure it first using ./noip --help things.. 03:24 < TheNH813> jml2: Okay. I'm trying it. 03:24 < jml2> TheNH813, then once you can see it is manually working, then ./noip -- then try to later use systemctl start no-ip 03:24 < jml2> TheNH813, while it is running, you can run ./no-ip -- I use < > because i dont remember the args. 03:25 < jml2> TheNH813, you might have to set something up in your Web portal before this is effective 03:25 < jml2> TheNH813, make sure you setup your domain things first in there 03:25 < triceratux> lubuntu-next is some kind of rogue project. for 2 releases now theyve done nightlies right up to the release & then they didnt release it. so a couple weeks later they get scratched without a trace & the next images are broken during the incipience of the next release cycle 03:25 < jml2> TheNH813, it took me a cycle of times before I just got it.. but once its set it is pretty much unbreakable.. 03:26 < jml2> TheNH813, i have two machines they have it and they never faultered even once 03:26 < jml2> TheNH813, (two years so far) 03:26 < jml2> TheNH813, very robust 03:26 < TheNH813> jml2: Allright. 03:26 < jml2> TheNH813, the first time you setup it up, you will have to wait some time -- after you set it in the Web portal.. 03:27 < jml2> TheNH813, in order for the dns lookups .. maybe 2 minutes -- i think the ttl for the dns record is very low.. 03:28 < dannylee> -- 03:28 < jml2> i can check my old notes on no-ip... 03:28 < stevendale> I wonder if SliTaz was work 03:28 < jml2> a little grep -ri no-ip tehehehe 03:28 < stevendale> Apparently it's rolling release now 03:28 < stevendale> http://www.slitaz.org/en/ 03:28 < jml2> stevendale, you're rolling release on less than 1 gigabyte is spelling for disaster 03:29 < jml2> stevendale, LOL 03:29 < triceratux> stevendale: ive used slitaz a lot on lowend machines. it takes some adjustment but its quite complete & powerful 03:29 < jml2> TheNH813, chown root:nogroup /usr/local/bin/noip2-x86_64 03:29 < jml2> TheNH813, important 03:30 < jml2> TheNH813, (optional - chmod 750 /usr/local/bin/noip2-x86_64 ) 03:30 < TheNH813> chown: invalid group: ‘root:nogroup’ 03:31 < triceratux> stevendale: jml2 is right tho. 6mo from now youll drop another 50 & get an actual 2G machine ;) 03:31 < cluelessperson> TheNH813: "nobody" 03:31 < jml2> TheNH813, you can even set the chown to "nobody" --- depends on what you want to use... 03:31 < TheNH813> That worked. 03:32 < jml2> TheNH813, and do the same thing for reading permission on its configuration file 03:32 < jml2> TheNH813, otherwise 'nobody' will be able to read it 03:33 < stevendale> triceratux: Probably :) 03:33 < TheNH813> Well, the entry on the noip page updated itself now, so that's a good sign. 03:33 < TheNH813> Thanks. 03:34 < jml2> TheNH813, old notes two years ago -- https://pastebin.ca/4020370 03:34 < jml2> TheNH813, ok that's good 03:34 < hatp> If my ssh server is on a VPN, would I use my public IP (before the VPN) still if I want to remote into it? 03:34 < jml2> TheNH813, use ./noip to tell no-ip to shutdown, then try to use systemctl start no-ip 03:34 < TheNH813> I also know why my script is wrong. I didn't escape the ":" character. lol. But I guess using the official client is better anyway. 03:34 < jml2> TheNH813, it's much better because it can manage multiple domains 03:35 < TheSov> wow i havent used slitaz in years 03:35 < TheSov> I did when i was "auditing" wireless networks with aircrack-ng 03:35 < TheNH813> Will say one thing, i got help here faster then anywhere else I ever asked. 03:35 < jml2> TheNH813, lol 03:35 < TheSov> TheNH813, seriously 03:35 < jml2> TheNH813, you're lucky because i'm here :p 03:35 < TheSov> hes right though 03:35 < alexey-nemovff> IRC is so cool 03:36 < TheNH813> Indeed. 03:36 < jml2> TheNH813, 5$ please 03:36 < TheNH813> It's very lucky someone else who uses noip was on. 03:36 < TheSov> i went to the ubuntu channel to ask help with ubuntu.... no one could help. came here helped instantly 03:36 < jml2> TheNH813, I'm a geek. what can i say? XD 03:36 < TheNH813> *Places gold bar into tip jar* is this enough? 03:36 < TheNH813> XD 03:36 < jml2> yw :P 03:36 < TheNH813> Thanks once again. 03:36 < jml2> no-ip is awesome 03:36 < TheNH813> It really is. 03:36 < jml2> TheNH813, and add some portknocking to that too :p 03:37 < jml2> tehehe 03:37 < jml2> TheNH813, I can use my android to portknock+ssh 03:37 < jml2> TheNH813, cuz i'm a wacko 03:37 < triceratux> TheSov: ubuntu channel types think theres nothing wrong with ubuntu. debian channel types too. in here we know better 03:37 < TheNH813> Ssh is the reason I want dns to start with. 03:38 < ayecee> who's we? you got a turd in your pocket? 03:38 < jml2> TheNH813, smart! 03:38 < jml2> TheNH813, that's good good good 03:38 < ArsenArsen> random question, can linux be compiled on windows 03:38 < TheNH813> Hmmmm.... 03:38 < TheNH813> That is a good question. Mingw maybe.... 03:38 < ayecee> ArsenArsen: i don't see why not 03:39 < ArsenArsen> yeah seems realistic 03:39 < ayecee> you'd have to go through all of the hoops involved in cross compiling though 03:40 < TheNH813> Now, my last issue is to deal with the buggy firewall on my gateway. 03:40 < TheNH813> Will likely just throw my PC into the dmz. 03:41 < lemgrago> Just dual booted ubuntu (first OS) and debian (second OS). Didn't install grub during debian installation. Booted back into ubuntu after debian install and ran update-grub. It found the debian partition fine. But, after reboot, while trying to select debian in grub menu, it froze. If I don't do anything, or just hit enter immediately, ubuntu boots fine. How do I troubleshoot? 03:42 < TheNH813> Can you enter in the boot commands to get to debian manually? 03:42 < cluelessperson> TheNH813: yes 03:42 < cluelessperson> oh 03:42 < Dreaman> advanced see in boot kernel 03:42 < lemgrago> TheNH813, is that through grub cli? 03:42 < TheNH813> If you can manage that, run update-grub again in debian. 03:42 < TheNH813> Yes. 03:43 < TheNH813> You'l really just need to specify the kernel, initrd and root partition. 03:43 < triceratux> lemgrago: possibly run the boot-repair reports & see how squirreled it is https://sourceforge.net/p/boot-repair/home/Home/ 03:44 < TheNH813> lemgrago: If updating grub from debian fixes it, then you need to only do one more thing for a permanent fix. Disable update-grub on ubuntu by moving the update-grub file and creating a symlink from /bin/true to /usr/bin/update-grub. I had to do that very thing once. 03:45 < c-c> why not just figure out what the actual problem is - like if grub actually cannot find the debian kernel 03:46 < doubtful> hi 03:46 < TheNH813> Luckily, there should be a link from /vmlinuz to the real location. So, if you know the partition number it should be easy enough to boot. 03:46 < TheNH813> doubtful: hello 03:47 < lemgrago> TheNH813, manual boot worked 03:47 < TheNH813> Okay, so then I would immediately run update-grub in debian. 03:47 < doubtful> using this download link https://paste.ubuntu.com/p/pzZbGS4cPh/ , how can I install these libs in 32-bit ubuntu 03:47 < TheNH813> And see if it's fixed. 03:48 < TheNH813> doubtful: What's the library name? 03:48 < doubtful> TheNH813 https://paste.ubuntu.com/p/pzZbGS4cPh/ 03:49 < lemgrago> TheNH813, fixed 03:49 < doubtful> TheNH813 Sorry it is libgc.so.1 and libgc.so.1.02 03:49 < doubtful> TheNH813 Sorry it is libgc.so.1 and libgc.so.1.0.2 03:49 < TheNH813> Libgc? 03:49 < lemgrago> Thanks for the help, TheNH813! 03:49 < TheNH813> No problem. 03:50 < TheNH813> I would move the update-grub file on ubuntu and replace it witha simlink to /bin/true. 03:50 < TheNH813> To prevent further issues. 03:50 < lemgrago> c-c, out of curiosity, how would I have done what you suggested? I didn't know you could 03:50 < TheNH813> Try this doubtful: sudo apt-get install libgc:i386 03:51 < doubtful> TheNH813 I am already on 32bit 03:51 < lemgrago> TheNH813, will do! :) 03:52 < TheNH813> doubtful: Okay. What version of ubuntu? 03:52 < doubtful> TheNH813 16.04 03:53 < TheNH813> doubtful: If the libgc you want is "conservative garbage collector for C and C++" 03:53 < TheNH813> Try sudo apt-get install libgc1c2 03:54 < c-c> lemgrago: by looking at grub logs? 03:54 < ricochet69> eww. you dig in logs for grubs? 03:55 < ArsenArsen> So I've done something wrong, since I'm now vuln to everything but meltdown, and the inbuilt checks no longer exist 03:55 < ArsenArsen> wrong channel 03:55 < c-c> Well I have the vague recollection that I have fixed booting by looking at the grub2 error - not that I remember where it was written 03:56 < c-c> - perhaps you needed to enable logging, explicitly 03:57 < jml2> ArsenArsen, this is not a honey pot. this is your system. 03:57 < jml2> ArsenArsen, LOL 03:57 * jml2 has all your *bases* 03:57 < jml2> ArsenArsen, wrong channel you say eh? 03:57 < ArsenArsen> jml2, it is a honey pot if I fill it with honey 03:57 < jml2> ArsenArsen, what you smokin' punk? 03:58 < ArsenArsen> bruv i was mid convo with #gentoo which is why this is the wrong place 03:58 < ArsenArsen> I know it's related though don't worryh 03:58 < jml2> ArsenArsen, well at least gentoo is better than ubumtoo so keep up dat hackin' yada yada :p 03:58 < ArsenArsen> will do 04:05 < stevendale> triceratux: Tiny Core Linux looks good \o/ 04:06 < triceratux> stevendale: its not. it gets old. it does run on tiny machines but you cant put in anything but whats in the repo some of which is quite compromised 04:07 < triceratux> its definitely better than it was for awhile 04:08 * stevendale considers Puppy Linux 04:12 < illkitten> Alpine Linux 04:12 < nver_> I think there might be something wrong with my mouse driver 04:13 < nver_> I'm not sure what's happning to my cursor but it keeps "trembling" 04:13 < nver_> Could it be an Ubuntu bug? 04:14 < ||JD||> try it on your UEFI/Live disk/Dual Boot/VM to discard hardware issue 04:14 < nver_> okay 04:15 < jml2> stevendale, a nice distro for 1 gig ram or less is chaletos -- it uses xfce 04:16 < jml2> stevendale, tinycore, the 12megabyte distro ... 04:16 < jml2> nver_, sounds like a video issue 04:17 < jml2> nver_, i would like to see if X picked up the e-did information from the monitor, in the X log 04:17 < jml2> nver_, if it didn't then you need to manually create a sane cvt line, using a cvt command 04:17 < jml2> nver_, (vesa-compatible mode-line) 04:18 < jim> nver_, what kind of monitor do you have? 04:18 < nver_> jim, it's a laptop 04:18 < nver_> Toshiba 04:19 < jim> oh ok 04:19 < nver_> And I'm a noob, so, I don't understand half of what you mean 04:19 < jml2> nver_, I know the solution for you 04:19 < nver_> Yeah? 04:19 < jml2> nver_, buy a laptop with Linux pre-installed on it! :) (Dell has some nice laptop models :P) 04:20 < triceratux> jml2: theres a ton of 16.04 xubuntu out there. lite, voyager, backbox, zorin, runtu, even mintxfce. nothing that special about chalet 04:20 < nver_> I was looking at some of them the other day, but I can't buy now. They looked nice though. 04:20 < jim> doesn't he just need the right driver? 04:20 < jml2> triceratux, most of these "light weight" distros are icing on cake. 04:21 < jml2> triceratux, I'd say that about not only chaletos.. but he is a new user and wants something setup quickly 04:21 < nver_> Yep 04:21 < jim> nver_, ok, let me explain something... 04:21 < nver_> okay 04:21 < triceratux> jml2: yep. if xfce is too heavy antix is the answer 04:22 < iopq> when I made my integrated graphics my primary monitor, Ubuntu went into low graphics mode 04:22 < iopq> is something wrong with my intel drivers? 04:22 < jml2> triceratux, dunno why, but everytime i see that word "antix" it reminds me of the word "vagina". strange eh? 04:22 < jim> your linux installation, when it's running has a number of what are called "virtual consoles" or "virtual terminals" 04:22 < jml2> triceratux, must be the "ix" 04:22 < jml2> triceratux, too much "ix" 04:22 < jml2> LOL 04:22 < c-c> iopq: install, run arandr, play with settings 04:22 < jml2> "It uses Xfce as the default desktop environment, is based directly on Debian Stable," 04:23 < c-c> Too bad LXDE uses xrandr. It sucks compared to arandr. 04:23 < jml2> it uses xfce as well triceratux doofus 04:23 < jml2> :)) 04:23 < jim> nver_, those virtual consoles have numbers, starting from 1 04:23 < triceratux> jml2: nope. thats mx which is from the same team. antix is icewm 04:23 < nver_> jim, how many of them are there? 04:24 * bomb installs Unity on triceratux computer 04:24 < jim> I'm guessing here (because more of them can be created), but if your installation is like most, you have right now a total of seven 04:25 < iopq> c-c, it can't see the integrated monitor 04:25 < nver_> hmm, okay 04:25 < jim> one thing about virtual consoles, they can "support" a graphical screen 04:26 < jim> (so, potentially right now, you -could- have seven graphical screens, but that's not how things get set up) 04:27 < nver_> jim, when I'm trying to open the first two ones it just opens the desktop, so, that means those are the ones with the graphical screens? 04:27 < nver_> tty3 opens up just as a console 04:28 < jim> what you likely have right now, is six textual login screens, and one graphical screen... the textual ones are number 1 - 6 and your graphical one is on 7 04:28 < nver_> oh, okay 04:28 < jim> at least that's how it is on my system, yours could be different 04:28 < nver_> right 04:29 < nver_> ugh, the cursor grabs things up and won't let them go 04:29 < jml2> i think about this point in time with X11/freetype fonts , that I consider it more superior than Mac/Windows own rendering. 04:30 < jim> you can switch which one you're viewing using the alt key and a function key... are you running linux now? 04:30 < jml2> with the right fonts, the clarity to me is very favorable and is something I have seen improved over a long period of time 04:31 < nver_> jim, Yes, I just did it 04:31 < nver_> I do see the other consoles 04:31 < jim> are you running an irc client on the 7th virtual console? 04:31 < nver_> I can use them like I use a Terminal... sort of like a full screen Terminal 04:31 < jim> in the graphical screen that is 04:31 < nver_> Yes, I'm using hexchat 04:32 < jim> ok 04:33 < jim> when you are looking at one of the textual virtual consoles (VCs for short), does the cursor look wavey there? 04:33 < jim> well you won't have an arrow cursor on any of the textual VCs 04:34 < nver_> Yeah 04:34 < nver_> I don't have a cursor there 04:34 < jim> so everything looks solid on the textual VC? 04:34 < nver_> yes, nothing strange happening there 04:35 < jim> but, on the graphical VC, you see weird effects (waveyness or something, you said)? 04:36 < nver_> Yeah, the cursor is doing things that it is not supposed to tho 04:36 < nver_> Well, I'm not touching the mouse at the moment, but it keeps clicking too 04:36 < nver_> scrolling up and down the window in the browser 04:37 < jim> ok, that's pretty weird :) 04:37 < nver_> Won't let me open stuff sometimes, it grabs something and won't let me discard it 04:37 < neoncortex> first linux poltergeist I have ever seen 04:38 < jim> ok, other than the waveyness, that sounds like something's going on with the mouse 04:39 < nver_> I just unpluged the mouse 04:39 < nver_> but it still happens 04:39 < nver_> Maybe it's something with the trackpad? 04:39 < nver_> How can I turn off the trackpad 04:39 < jim> how does your mouse plug into the computer? 04:39 < nver_> USB 04:40 < nver_> I believe it's the track pad 04:40 < jim> ok, right now I don't know how to turn off the trackpad without physically disconnecting it from the usb port 04:41 < nver_> okay 04:42 < jim> which laptop is it? 04:42 < nver_> Toshiba Satellite L755 04:43 < jim> ok, I'll get to that in a sec 04:43 < jim> can you go to one of the other virtual consoles and log in as root? 04:43 < neoncortex> there is used to be a synclient command, looks like it's gone 04:44 < nver_> oh wait, there might be a way to disable it from Ubuntu's Settings 04:44 < nver_> Hang on a sec plz 04:44 < nver_> oh sweet 04:44 < nver_> I turned it off 04:45 < nver_> ah, the problem still exists 04:45 < jim> do you happen to know what kind of touchpad it is? 04:46 < nver_> Nope 04:46 < nver_> I did just turn it off, could it still have effect on the cursor? 04:46 < jim> neoncortex, which dist do you run 04:46 < jim> ?" 04:47 < neoncortex> jim: debian 04:48 < jim> neoncortex, Search for synclient in stretch/amd64: xserver-xorg-input-synaptics: usr/bin/synclient 04:49 < jim> neoncortex, so synclient is in the package xserver-xorg-input-synaptics 04:50 < neoncortex> jim: nice, It's for control touchpads, maybe helps in nver_ case 04:51 < jim> neoncortex, oh ok, I think he's running arch (which I don't know the package manager of) 04:51 < littlepython> when we type ls command, does fork and exec system call get executed? 04:52 < horseface> how do you embed artwork into a flac file? 04:52 < horseface> so that the icon shows the thumbnail? 04:54 < jim> nver_, I think I found something you could look at: https://wiki.archlinux.org/index.php/Touchpad_Synaptics 04:54 < jim> nver_, did you say you're running arch? 04:55 < jim> littlepython, I think so, yes 04:56 < quint> So I've just switched a fairly new laptop over to linux, and the battery seems to be "out of sync" for a lack of better words. It claims to be fully charged, but dies quickly after the power is disconnected. This was not an issue on windows 10. Is there some kind of proprietary driver required for some HP batteries? 04:57 < jim> quint, I don;t know myself, maybe others know here, or on ##hardware 05:00 < kuri0> quint, try calibrating it 05:00 < nver_> sorry I was unable to open hexchat, lol 05:00 < kuri0> or maybe something on linux is eating the battery 05:01 < stevendale> Ah 05:01 < nver_> the cursor is not obeying me! 05:01 < stevendale> Lubuntu 18.04 doesn't require 1 GB RAM 05:01 < nver_> it has openned a lot of other apps because it keeps clicking on random stuff 05:01 < stevendale> That's just to let people know things are changing and the standards are going up 05:03 < jim> nver_, ok, try this: go to VC1, log in as root, come back 05:04 < nver_> jim, I'm running Ubuntu, the latest version 05:05 < jim> oh you are, ok, could you run: apt-get install irssi tmux 05:05 < nver_> I logged in as root 05:05 < jim> in the root login 05:05 < nver_> okay 05:06 < jim> oh, if you're running ubuntu, you should install the package from before, one sec 05:07 < Nver__> I already started this one, is it okay? 05:07 < jim> nver_, let it finish 05:07 < Nver__> It's done 05:07 < jim> ok, now apt-get install xserver-xorg-input-synaptics 05:08 < littlepython> when i do strace ls 05:09 < littlepython> i dont find fork or exec 05:09 < littlepython> jim: ^^ 05:10 < Nver__> Done jim 05:12 < jim> Nver__, ok, go to virtual console 2 (VC2), log in as a plain user, and then, start tmux, and then start irssi when the shell inside tmux starts 05:13 < jim> Nver__, are you connected by your phone? 05:13 < Nver__> Yes 05:13 < jim> ok 05:13 < Nver__> Can I go to VC4 05:13 < jim> sure 05:13 < Nver__> VC2 is the graphical screen 05:13 < Nver__> Okay 05:13 < jim> oh it is? maybe vc4 doesn't exist 05:14 < Nver__> I'm on tty4 now 05:14 < jim> actually, try this: reboot the laptop and come back, let's see if the synaptics driver is any better 05:14 < Nver__> I started irssi inside the shell of tmux 05:14 < Nver__> Okay, one sec 05:15 < stevendale> \o/ 05:15 * stevendale is installing Lubuntu 18.04 32-bit via the alternate installer 05:16 < jim> neoncortex, maybe that was a good call, let's hope 05:16 * triceratux has dnsmasq local caching working on fedora 28 05:17 < stevendale> Swap out Firefox for something like Dillo or Midori... Should be fine... 05:17 < jim> Nver__, how's it looking? 05:18 < Nver__> Oh it's doings great so far 05:18 < jim> any weird mouse clicks? 05:22 < Nver__> Nope jim 05:22 < stevendale> OwO 05:22 < Nver__> It's good 05:23 < Nver__> Thank you! 05:23 < nver_> It's working fine 05:23 < nver_> Finally 05:23 < neoncortex> heh =D 05:24 < nver_> jim, thanks again! 05:24 < neoncortex> stevendale: yes, firefox with less than 59gb of ram is unusable 05:24 < stevendale> I'm eating sandwiches :3 05:25 < ricochet69> good idea: food.. ! (brb`) 05:26 < stevendale> Vegemite :D 05:26 < neoncortex> ok, I feeling I'm beign unfair, it's a lot better lastly 05:29 < cmj> is this #linux? 05:29 < neoncortex> cmj: say the password 05:30 < IPoAC> hunter2? 05:30 < ||JD||> any workaround for wall not broadcasting to gnome-terminal? 05:30 < c-c> 123456 05:31 < cmj> setec atronomy 05:31 < ayecee> it's an old meme, sir, but it checks out 05:31 < cmj> neoncortex: am i right? 05:32 < neoncortex> cmj: sorry, no .. so I can't give you any information ¬¬ 05:41 < promach_> For https://github.com/VectorBlox/linux-xlnx/blob/mxp/block/ioctl.c , could anyone advise on the compilation error https://paste.ubuntu.com/p/4nQJyxQPBc/ ? 05:53 < stevendale> Hey 05:53 < stevendale> 68 MB RAM used 32-bit Lubuntu 18.04 05:54 < baconicsynergy> stevendale, thats pretty impressive 05:55 * ||JD|| installs Netbus trojan in stevendale's XP 05:55 * stevendale stops using XP 05:55 < stevendale> 199 MB of 483 MB used in Lubuntu 18.04 with Firefox open 05:55 < ||JD||> I will DOS jim from there muahaha 05:56 < stevendale> 63 MB swap used 05:57 < aloo_shu> anybody good with xrandr? 05:58 < malina> depends, why don't you just ask? 05:59 < aloo_shu> I'm doing an xrandr --output LVDS-1 --scale 2x2 and get what I expect, but the mouse cursor is confined to the upper left quarter xreen 05:59 < malina> lol.. damn.. of all the things.. I don't know scale 06:00 < aloo_shu> do you know --panning? malina 06:00 < malina> but with that said.. if you are scaling the output to 2x2 it sounds like 1/4 of it (i.e. rthe first quadrant aka the top left square) is where your screen is defined as 06:00 < malina> on top of the 'actual display geometry', which maybe is why the mouse only is working in there? 06:01 < malina> ye, but tbh.. it's been a while and I sorrted things out a bit differently. 06:01 < aloo_shu> I THINK I tried to give it a larger --fb, too 06:01 < malina> are you having issues with scaling the display to the actual monitor size (like with hdmi issues aka over/underscanning)? 06:02 < malina> ya, tbh, every such use-case has many ways which lead to Rome.. : I can't help sorry, as it's just gonna be : have you tried this? have you tried that? keep testing and you will find what you want.. although.. what is your 'initial problem' why are you scaling? 06:02 < malina> xrandr --query -q | awk '/[^dis]connected/ {print $1;}' 06:02 < aloo_shu> no, I am trying to set up a big screen on a small monitor, to zzom in on and to pan around in 06:02 < malina> what does that give you? 06:02 < malina> oh I see 06:03 < malina> one sec. 06:04 < malina> and xorg.conf is not an option? 06:04 < aloo_shu> as soon as I set --panning WxH with the larger screen size, all works as expected 06:04 < malina> so.. what is the problem? 06:05 < aloo_shu> just want to find out why it doesn't work with --scale, I.e. full mouse movility, or if it's a bug 06:06 < malina> no, I don't recall actually how xrandr always works, but as I said above. it seems from your experience, that scale scales the real estate, on top of the output? (there is somethign called viewportout in xorg which might be nvidia specific but might also help).. anyway. 06:08 < aloo_shu> another thing I want to fix, if I do pan around on that large screen, the screen is scrolled whenever I move the mouse pointer, I would like to see movement only when reaching the edges, man xrandr isn't too specific on what the many possible parameters to --panning effectively do 06:08 < aloo_shu> but I'll just re-ask another day, maybe 06:08 < aloo_shu> my life doesn't depend on it 06:09 < c-c> arandr?? 06:12 < aloo_shu> c-c: is arandr an alternative to xrandr ? 06:13 < malina> nah, arandr is just a 'gui' , frontend for xrandr 06:13 < malina> it creates a script though for you, it can help but it helps I think mainly for only size and positiion iirc 06:14 < malina> xorg.conf itself though can help and have some solutions side by side with xrandr commands, yes. 06:14 < c-c> theres also LXrandr (which I earlier confuses with xrandr) 06:16 < aloo_shu> xrandr, I read, is a cli tool to interact with X's RandR 06:17 < aloo_shu> if I find a setting that works, I might put it into xorg.conf instead 06:17 < malina> ya. lxrandr etc all are on top of librandr, it could be lxrandr has some more options than arandr, (and could maybe generate a script wiht more options) but it can't do anything xrandr can't do really.. as I _think_ xrandr kinda exposes the full randr lib 06:18 < baconicsynergy> i love lxrandr 06:18 < malina> ya, I usualyl find a solution between xorg and xrandr.. but ye.. I usually try to shrink screens, not size them so sorry, it's been too long to throw out some solutiuon ye.. man xrandr however, is your friend 06:18 < aloo_shu> https://www.x.org/wiki/Projects/XRandR/ Last edited 2013 :( 06:19 < aBound> Howdy do, doo... :P 06:19 < baconicsynergy> whats going to be the "xrandr"-like tool for wayland? 06:20 < aloo_shu> that would have been my question, too 06:20 < baconicsynergy> wrander? lol 06:20 < baconicsynergy> wrandr* 06:21 < aloo_shu> wrandrslust 06:21 < baconicsynergy> I love xorg but it's like a garbage-can with rockets strapped to the side with duct tape 06:21 < baconicsynergy> aloo_shu, i love it!!! 06:22 < aloo_shu> Ithink that's what everybody should fly 06:22 < baconicsynergy> xorg gives you wingssss 06:23 < aloo_shu> darn sick and tired of getting updated out of my learning progress and configuration successes 06:23 < c-c> malina: arandr > lxrand 06:25 < malina> ok 06:25 < malina> I don't use either so. they are subsets of xrandr. 06:25 < aloo_shu> look, I want to solve a problem, if you have the solution, I will buy your randr frontend, whichever it is xD 06:26 < aloo_shu> but, as I said, my life doesn't depend on it 06:30 < aloo_shu> but wouldn't an --fb to a larger size have fixed it then? 06:30 < aloo_shu> sry 06:30 < aloo_shu> lost in backlog 06:32 < aloo_shu> oh, we're getting nearer: xorg.conf is not by default being used for input devices, it says even in #xorg 's topic 06:32 < aloo_shu> shall read up now 06:33 < c-c> ok so need to set desktop geometry separately for mouse if scaling 06:34 < malina> ya it might be xorg is not the thing you want.. it was just a porposal, (you might even be on those new things, wayland or whatever) 06:42 < aloo_shu> c-c: apparantly so, but wher do I set desktop geometry for mouse? 06:43 < aloo_shu> anyway, I just recall that manpages for xorg were explaining panning parameters in more detail, but I'm on an android device now 06:46 < aloo_shu> c-c: it's funny, the desktop itself, xfce4 here, picks up on the new geometry and re-layouts 07:13 < malina> https://xkcd.com/1988/ 07:13 < malina> highly relevant ,) 07:14 < LordVan> hi 07:15 < LordVan> just wondering: has anyone had a dual boot system before while also having the ability to run it in a virtualbox as well? 07:16 < LordVan> (meaning 1 kernel that has drivers,.. for the machine's hardware and the vm , or 2 kernels maybe? 07:20 < neoncortex> LordVan: speed, virtual machines are slower, they improved a lot recently, but still 07:21 < LordVan> neoncortex, that is not an issue .. i was mainly talking about having it boot ok the same system, but from the hardware itselfa nd from vm - however i need it 07:21 < LordVan> ie during work when i need windows, but want linux too i start up the vm 07:24 < neoncortex> I think I'm too dumb to understand what you said 07:24 < [R]> LordVan: as long as you provide the raw device to the vm, linux will be fine 07:25 < [R]> no sane dist has a special kernel just for yourh ardware 07:25 < Roserin> I have an idea i guess 07:25 < LordVan> [R], i always build the kernel by myself anyway ^^ 07:25 < Roserin> Try using xen? 07:25 < LordVan> meh@xen 07:25 < LordVan> too much work/effort ^^ 07:25 < [R]> LordVan: well then yes, you'll need 2 kenrels 07:25 < Roserin> Xen might be able to do it... 07:25 < LordVan> yeah i looket at xen before but .. well 07:25 < Roserin> Yeah but it's prolly not gonna be easy any way you cut it :P 07:26 < LordVan> the effort outweighs the usefulness 07:26 < Roserin> prolly 07:27 < LordVan> or i just forget about dual boot (90% of the time i am on this machien it will be running windows anyway - need the CAD system for work unfortunately) and just use linux in a vm 07:28 < LordVan> but still with a raw device for the rootfs 07:28 < LordVan> but not bother with direct boot atm 07:28 < LordVan> (ie no grub,..) 07:48 < jim> nver_, curious about some other stuff... have you ever programmed before? 07:50 < jim> Nver__, if you're here, got another question... look above this line (symbol: ^^^) 07:51 < nver_> jim, hey, I've done some javascript before 07:51 < nver_> but that's not even programming. 07:51 < jim> ok... one sec 07:54 < jim> nver_, one thing, is your video still weird? or did that package and reboot take care of the weird mouse and weird video? 07:55 < notmike> Ubuntu poisons drinking water in Africa 07:56 < nver_> jim, it's all fine now and thank you for it! 07:57 < jim> ok... can you type: python --version 07:57 < nver_> Python 2.7.15rc1 07:59 < nver_> Can we upgrade it 07:59 < nver_> to the latest version 07:59 < jim> well one sec 07:59 < jim> try: python3 --version 08:00 < jim> also I got something to help: http://greenteapress.com/wp/think-python-2e/ 08:00 < nver_> Python 3.6.5 08:00 < nver_> oh, I have it apparently 08:00 < jim> perfecct, so you have both 08:00 < nver_> oh, thanks for the book 08:01 < nver_> I will be reading it 08:02 < jim> after think python, the official tutorial is probably a good thing, it's a quick run through... https://docs.python.org/3/tutorial/ 08:03 < jim> probably one of the biggest things linux supports, is coding/programming 08:05 < jim> and if you want an annoying video that covers python2, 08:05 < jim> https://www.youtube.com/watch?v=TV9tSHFAFjg 08:06 < jim> 11 hours of annoying bubblyness 08:06 < ejr> i need to use double quotation marks inside double-nested quotation marks in a command, but this does not work: command "-insidecommand -'insidecommand' -"insideinsidecommand"'" <--- if I leave out the latter command it works (i.e. with two levels, but not with three). Adding \ also doesn't work. Is there any solution to this? 08:06 < jim> but, it's not bad for teaching python2 08:08 < jim> ejr, inside of ".....", you can include a " using a \,,,, like this: "....\"quote this\"...." 08:09 < ejr> jim: thank you, I will try that! 08:09 < jim> welcome 08:24 < KekSi> mode/##linux [-qo $a:yaaic eir] by eir << wtf chanmode is "q"?! and what is the $a:yaaic param to it :o 08:26 < [R]> quiet, and its an account name 08:26 < sauvin> KekSi: +q means "quiet"; it's like a ban, but it doesn't keep the quieted person from joining the channel, just makes him unable to speak. The $a:yaaic part means "this operation is by account on 'yaaic'. 08:29 < KekSi> i see -- i've been on quakenet for more than 20 years and that doesn't have that chanmode (a regular ban would mute a banned person already on the channel but also prevent them from rejoining if they left) 08:32 < nver_> jim, thanks for everything 08:40 < drzacek> Hello there 08:41 < drzacek> So grub has this neat feature, it can start either in text mode or graphic mode (I'm talking about the grub menu, the one where you choose system to boot) 08:41 < drzacek> but it seems it only works with certain graphic cards (or rather, with certain cards the video mode doesn't work) 08:42 < [R]> i've never seen the menu fail... 08:43 < drzacek> Did anyone experienced this problem? That grub menu only start in text mode (black&white only)? 08:44 < [R]> i've also never heard anyone say the menu is "graphic mode" 08:45 < Sitri> Are you talking about where grub can display a background picture on the menu? 08:45 < Sitri> Because it's been able to do that forever, it uses standard VGA stuff. 08:47 < Sitri> Or at least grub1 did. Haven't used grub2 ever. 08:51 < drzacek> [R], how do you call it? 08:51 < [R]> a menu 08:52 < drzacek> Sitri, on almost all machines where I installed linux (and grub), I do get that menu with high resolution (I'm guessing native LCD resolution, but not 100% sure), colors, background image etc 08:52 < drzacek> while on this one pc with intel graphics I end up in small black&white rectangle in lowest possible resolution 08:52 < janco> you can style it like you want 08:53 < janco> what does a grub menu matters anyways; as long as it's functional 08:53 < Sitri> It matters if you're making a distro, or something else that could be a demo 08:54 < drzacek> janco, thats the thing, I can't 08:54 < drzacek> it does matter 08:54 < talx> you can edit it with modules 08:54 < janco> ^ 08:54 < drzacek> what modules 08:54 < janco> grub custommizer 08:54 < janco> afai 08:54 < talx> just read about it 08:54 < janco> grub-customiser or something like that 08:54 < talx> its a really big subject 08:55 < drzacek> yeah I used grub-customizer before 08:55 < janco> but why does the styling matter 08:55 < talx> grub is kind of old 08:55 < talx> you should use grub2 by now 08:56 < drzacek> I thought I was using grub2 08:57 < drzacek> it even says gnu grub version 2.02~beta3-5 09:00 < roxlu_> hi when I use edge-triggerd mode with epoll for writes, and at some point a descriptor becomes available for writes does that mean I can write "as much data" as I want. Or is it possible that e.g. after writing 50% of the data the descirptor becomes unavailable again and I have to (epoll_wait) before it becomes available again? 09:01 < [R]> you can write as much as you want 09:01 < [R]> doesnt mean it wont block 09:01 < talx> drzacek you could change the runlevels on startup with it 09:02 < talx> grub2-mkconfig -o /boot/grub2/grub.cfg 09:02 < talx> I did it on centos tho 09:03 < roxlu_> [R]: I'm using edge-triggered mode. I was in the assumption the descriptors need to be in non-blocking mode. I guess I could receive a EWOULDBLOCK ? 09:03 < [R]> then either ti'll fail or write only a little 09:03 < roxlu_> [R]: ok thanks 09:09 < elichai2> Hey 09:10 < elichai2> Weird thing, when I disconnect my laptop from my C docking and try to reconnect the system hangs unless I reboot the system 09:11 < [R]> that sounds unfortunate 09:11 < [R]> but how can you reboot it if its hung? 09:12 < elichai2> https://www3.lenovo.com/us/en/accessories-and-monitors/docking/universal-cable-docks-usb/Thinkpad-USB-C-Dock/p/40A90090US 09:15 < elichai2> [R]: I force shutdown, or I disconnect the dock and wait for 10-15 secs 09:15 < [R]> well thats not a reboot 09:21 < Sitri> It's a power cycle. 09:21 < Sitri> FYI 09:22 < binTrl> Hi. I want to multiboot with windows, what would be the safest way ? (Do not consider Vbox, that's slow). I'm most afraid of ransomewares encrypting my linux home 09:22 < Sitri> A windows virus wouldn't have access to your linux files 09:22 < [R]> binTrl: huh? 09:23 < sauvin> Modulo shared directories, that is. I run Windows in VBox all the time, seems fast enough. 09:24 < binTrl> Sitri: Provided it has infected my windows install and can get admin priv, can't it easily encrypt all the other partitions ? 09:24 < binTrl> sauvin: You get half the ram, half the cores. 09:25 < sauvin> And absolutely zero access to the outside world, which is just how I like running Windows. 09:25 < Sitri> binTrl: It could in theory, but none of them work like that. They just encrypt all the files in certain directories. 09:25 < binTrl> I was wondering if there would be a way to mark a certain position on the HDD as a new beginning and windows would think that the HDD starts from there but there would be valid data before that. 09:25 < Sitri> Also it's not like there isn't ransomware for Linux 09:26 < Sitri> You could do that with a virtual machine. 09:26 < [R]> you're a virtual machine 09:26 < Sitri> Pick one: speed or security. 09:27 < binTrl> Sitri: This one seems to do it https://www.bleepingcomputer.com/news/security/petya-ransomware-skips-the-files-and-encrypts-your-hard-drive-instead/ 09:27 < [R]> just skip windows 09:28 < [R]> problem solved 09:28 < binTrl> Sitri: Today, I just wanted to play games, after a long long time. :( 09:28 < Sitri> Or just install them on entirely separate disks 09:28 < binTrl> 'cause I plugged in my mouse after a long long time 09:28 < sauvin> Or even an entirely different machine. 09:28 < Sitri> And physically unplug one to boot into the other. 09:29 < binTrl> Sitri: It's a laptop here, so no luck. 09:29 < Sitri> Uhh what? Every laptop I've had has had user-replacable hard-drives. 09:30 < Sitri> I've even had some with a second HDD bay. 09:30 < binTrl> Sitri: Never did that. Well.. I'd have to wait to buy a laptop HDD then. 09:31 < Sitri> You can get one for like 60$ 09:31 < Sitri> Buy used if you need it for less than that 09:31 < [R]> we are buying multiple $5,000+ laptops at work 09:31 < [R]> damn thing cracks me up every time i think about it 09:32 < Sitri> Support contracts! 09:38 < [R]> top of the line i7 09:38 < [R]> 32gb of ram 09:38 < [R]> 1tb ss 09:38 < [R]> d 09:38 < [R]> freakin ridiculous 09:41 < mgolisch> still quite expensive 09:41 < mgolisch> no? 09:42 < [R]> its "rugged" 09:42 < [R]> althugh we did save like $100 by getting it with ubuntu 09:53 * sauvin accidentally dropped a ruggedised iPad at work today and didn't feel very much regret 09:57 < [R]> no ragrets? 09:57 < [R]> and then they sent the quote to me, wanting me to double check it 09:57 < [R]> so i was reading through it... and noticed "no case" 09:58 < [R]> so im like "yeah... we're spending $6k for a laptop, and not getting a case? looks like we missed that one." 09:58 < [R]> haha 09:58 < mhammons> hey all 09:58 < mhammons> i got an xorg.conf problem maybe someone knows how to solve 09:59 < Sitri> Just ask 09:59 < mhammons> so, i got a displaylink usb3 adapter plugged into my computer 10:00 < mhammons> with 2 monitors plugged into it 10:00 < mhammons> it's very fussy, and frequently will pop up that the monitors exist in my display settings, but have no configurable resolution 10:01 < mhammons> if I use these commands: https://pastebin.com/MiY66Cvi I can get them to start displaying content 10:01 < mhammons> and then reconfigure them into a multi-monitor desktop 10:01 < mhammons> now, the question is how to define an xorg.conf file that forces these resolution settings on these monitors 10:01 < mhammons> which are regularly named DVI-I-1-1 and DVI-I-2-2 10:04 < TheWild> hello 10:05 < blaztek> Hello TheWild 10:06 < TheWild> "dd if=/dev/sda4 status=progress | sha512sum" ended at some point with IO error (wtf?). smartctl shows Current_Pending_Sector: 3 (:O :O :O) 10:07 < TheWild> should I backup immediately? 10:07 < mhammons> i'll be back, i'm testing a .xinitrc i just wrote 10:09 * jack_rip_vim o/ 10:12 < mhammons> looks like i can't use an .xinitrc cause the monitors are not always present when i startup 10:12 < jack_rip_vim> mhammons: what happened? .xinitrc isn't working? 10:13 < azizLIGHT> i want to launch 4 terminal windows with a command in each, all launched same time. how can i do it 10:14 < jack_rip_vim> azizLIGHT: tmux command 10:14 < jack_rip_vim> azizLIGHT: you may need to install it 10:15 < mhammons> jack_rip_vim: if I have the commands in .xinitrc and the monitors aren't recognized yet, my desktop won't load 10:15 < jack_rip_vim> mhammons: which command? 10:15 < mhammons> and this set up is fiddly enough that I need to unplug and replug the usb displaylink to get the monitors recognized 10:16 < mhammons> jack_rip_vim: https://pastebin.com/vaajxrKx 10:17 < jack_rip_vim> mhammons: do you try to use systemd to load the command when the system boot up? 10:17 < mhammons> the commands don't work if the monitors aren't recognized, and they typically aren't recognized at boot 10:17 < jack_rip_vim> mhammons: oh, I see 10:18 < mhammons> systemd loads the display link manager at boot, which in theory should make them available 10:18 < mhammons> but in practice it's really buggy and i have to replug the usb3 adapter before it starts working 10:19 < jack_rip_vim> mhammons: you just want to run it when system boot up? 10:20 < mhammons> i would like to, yes, but i can't because i can't be sure i can plug and unplug the usb3 display adapter faster than the command would run at boot 10:20 < chrisphonk> Hey all! 10:20 < chrisphonk> Question - if you were to create a test for aspiring linux admins, would there be something in particular you're looking for? 10:20 < jack_rip_vim> mhammons: try to put it on .xsessionrc 10:20 < spare> you can do until command;do sleep .2;done in .xinitrc but if your connectors loose and re hotplugging itself you should probably fix that 10:21 < mhammons> spare: it's not the connector's loose 10:21 < mhammons> the displaylink software is buggy 10:21 < mhammons> i have to rehotplug it every boot before it works 10:22 < jack_rip_vim> mhammons: try to put it on .xsessionrc, see if it can work 10:22 < spare> if you drop until command;do sleep .2;done itll just halt and re execute til it works but you would need acpid or uevents to rexecute if its re hotplugging itself 10:23 < mhammons> jack_rip_vim: wouldn't .xsessionrc have the same problem if the displays are missing? 10:24 < mhammons> spare: i really don't want to mess with this displaylink driver more than i already have. I just have a script now, when i know the displays are recognized, but not configured, i click it and it works 10:24 < spare> until xrandr --output monitor --mode 1920x1080;do sleep .2;done will just keep trying till monitor exists in .xinitrc 10:25 < jack_rip_vim> mhammons: putting the command in there that will set up when you login to the session 10:25 < mhammons> spare: but this is a laptop, so it may not always exist. for example, if i take it with me to a conference or something 10:26 < spare> probably need acpid to execute script on event 10:26 < spare> have it listen for whatever it spits out and execute the script as and when 10:28 < phre4k> can I address i2c devices in another way than through /dev/i2c-#? Is there something like with disks with a persistent ID I can call? I can write a weird bash script around /sys/bus/i2c/devices/i2c-#/name, but there has to be a more elegant way 10:29 < jack_rip_vim> phre4k: seem like it can't 10:30 < JackMa> when i press ctrl+c, then it don’t stop process, then how can i do? 10:31 < jack_rip_vim> try to press a couple of times 10:32 < jack_rip_vim> hi Armand 10:32 < Armand> Hey, Jack 10:32 < jack_rip_vim> :) 10:33 < phre4k> JackMa: kill -9 `pgrep procname` 10:33 < jack_rip_vim> Armand: haven't seen you for a long time 10:33 < phre4k> JackMa: or even killall -s9 procname 10:34 < JackMa> how can i know process name? 10:34 < JackMa> pwd? 10:34 < phre4k> JackMa: 9 is the signal to be sent and it means "kill the process forcefully without regard to data" (SIGKILL) 10:34 < Sitri> pwd == Print Working Driectory 10:34 < Sitri> The process name is usually the same as the command used to start it 10:34 < phre4k> JackMa: you can also use htop and navigate to the process you want to kill, then press k and 9 10:35 < Armand> jack_rip_vim: I was in the US for 4 weeks on holiday. 10:35 < jack_rip_vim> Armand: sound great! 10:35 < JackMa> phre4k: ps? 10:35 < Armand> It was :) 10:35 < jack_rip_vim> Armand: have a good rest! 10:35 < phre4k> JackMa: yes, even ps. But htop is way easier because it's a gui task manager 10:36 < Armand> Yeah.. it was ok, jack. 10:36 < JackMa> thanks :) 10:36 < jack_rip_vim> :) 10:39 < jack_rip_vim> Armand: it seems so quiet those days 10:42 < horseface> hi 10:42 < jack_rip_vim> hi horseface 10:42 < horseface> what is a good DE which is a combination of i3 and gnome? with a window switcher. 10:42 < horseface> ? 10:44 < ageis> how can I see literally *all* CONFIG options possible or supported in a kernel version? does `make oldconfig` do that and everything's listed either set or not set? 10:44 < roxlu_> With epoll, when is EPOLLOUT triggered? Is the common approach to toggle the EPOLLOUT every time I want to send? 10:45 < phre4k> horseface: awesomewm + rofi 10:45 < heftig> ageis: not necessarily 10:45 < jim> well i3 has "workspaces", and they made a plugin for xfce... I guess that plugin could be reworked for gnome, if it hasn't already 10:45 < heftig> ageis: options that depend on options that are disabled won't be shown 10:46 < ageis> hrm so how can I see them all regardless of the dependency chain? 10:46 < ageis> grep? :P 10:46 < heftig> if you're looking for something, maybe nconfig's symbol search 10:47 < ageis> gotcha 10:47 < jim> to see all the config options, you could look at the config file too... (yeah, the language might not be so english-like) 10:48 < heftig> ageis: also https://cateee.net/lkddb/web-lkddb/ 10:48 < horseface> i see 10:49 < ageis> cool, this shows which versions an option is found in, which is what i am looking for 10:56 < amelliaa> Anyone knows how to chattr +i a file trough chroot? 10:56 < roxlu_> I guess with edge triggered epoll, one has to toggle the EPOLLOUT flag every time you want to sent something but please correct me if I'm wrong 10:59 < ageis> so what does "changes choice state" mean in regards to a config warning -- how to resolve that ? 11:03 < djph> ageis: err, you'll probably need more context surrounding those three words. 11:03 < ageis> i guess its a warning you'd get from make if you've set a config option that is missing some other option that it depends on 11:04 < ageis> .config:288:warning: override: CC_STACKPROTECTOR_STRONG changes choice state 11:04 < ageis> .config:562:warning: override: LEGACY_VSYSCALL_NONE changes choice state 11:09 < c50a326> can I make a file the function of another file 11:09 < c50a326> I'd like my nvim config to basically be my vim config but with a different first line 11:09 < c50a326> so like a symlink but with a different first line 11:11 < djph> ageis: that's specific to the compiler ... and *probably* means that something forced a different choice on you elsewhere. For example if CC_STACK* requires some other bit, and you didn't have it enabled (or if CC_STACK* and something else conflict, CC_STACK wins...) 11:12 < ageis> ahh, i see now 11:28 < stevendale> \o/ 11:28 < stevendale> Hi 11:34 < stevendale> Hey archpc :D 11:35 < greenit> hi, i have a virtual machine with debian in headless mode installed on it, how can i change the resolution of the terminal? it's a bit tiny... 11:35 < greenit> i'm using virtualbox if it's relevant 11:35 < stevendale> greenit Install VirtualBox guest additions :D 11:36 < greenit> stevendale: they are installed, but it doesn't change anything if i resize the window. i think it's cause i only have a terminal, no X11 11:36 < djph> greenit: guest additions might do it, or try resetting the graphics options (it *should* be a framebuffer tty, so setting to say 1024x768 may make the VM screen itself larger... OR just SSH to it from your host PC. 11:37 < djph> note that the gfx options will likely require a reboot of the VM 11:37 < greenit> djph: how do i set the framebuffer-resolution? 11:38 < djph> greenit: thru vbox. I think without setting any options, vbox defaults a VM to like 640x480, or 800x600 11:38 < djph> can't recall if it's possible to change while the VM is running though 11:41 < greenit> djph: ok, i've shut down the vm... but where should the option to change the resolution be? in "display" i see graphics memory, number of screens and scaling 11:41 < supernovah> Hey guys, what's the accepeted way to report a message to /var/log/messages? 11:44 < amelliaa> Anyone knows how to chattr +i a file trough chroot? 11:48 < supernovah> I'd like to customise what appears there, I see the : , how can I customise the user or hostname when i log? is there a port I can send data to? 11:48 < mawk> just do it amelliaa 11:48 < mawk> it works 11:56 < Sitri> supernovah: syslog 11:57 < supernovah> Sitri: I looked at that but it's not installed on Debian 8 which I run 11:57 < Sitri> You can configure your syslogd to change the message format slightly. 11:57 < Sitri> Then it's some systemd thing. 11:57 < supernovah> I have clevely avoided touching systemd as much as possible and in this instance I might just accept that the processName or username or whatever gets attached to the log, it's okay 11:58 < Sitri> Kill journald and run syslog instead? 11:58 < stevendale> I don't particularly like Windows... But Win XP runs great on my Inspiron 6000 compared to just about any lightweight Linux distro I have tried 11:59 < Sitri> Have you been putting heavy stuff onto your lightwieght distro? 11:59 < supernovah> stevendale: I have spent a lot of time in Window CE and embedded linux on arm processors and by and large, linux is definitely nicer and faster 11:59 < stevendale> I want to play games 12:00 < stevendale> And I can't do that on Linux with 512 MB DDR2 @ 400 MHz 12:00 < bazhang> stevendale, so dosbox then 12:01 < stevendale> Most of my games are Windows 9x unfortunately bazhang 12:01 < supernovah> > come to a linux chatroom > preach about windows gaming 12:01 < bazhang> stevendale, why ask here if you know you dont want linux 12:02 < stevendale> I thought IRC was my personal bloggr page 12:02 < stevendale> \o/ 12:02 < bazhang> stevendale, did you run that by jim or the other ops here first 12:03 < stevendale> o/ 12:03 < mawk> someone played with RTEMS here ? 12:03 < bazhang> /msg alis list social to find a chat channel stevendale 12:03 < mawk> it's a little suspicious that building it takes 0.0000000000000000000001 seconds 12:03 < supernovah> i recently discovered ssh -L :: and I have fallen in love 12:04 < stevendale> I thought I had freedom of speech to share opinions 12:04 < mawk> supernovah: during a ssh session you can do ~C then type -L somethingsomething 12:04 < mawk> on the fly, with the same socket through multiplexing 12:04 < mawk> isn't that even cooler 12:06 < supernovah> mawk: indeed, thx for that 12:06 < mawk> or ~. to terminate a ssh session that's not responding 12:10 < supernovah> mawk: I usually run it with -e none because I use ~. in nested terminals or callups/serial sessions etc 12:11 < mawk> I see 12:13 < djph> greenit: hm, could've sworn it was in the graphics options :/ sorry. 12:21 < kazdax> horaay 12:21 < kazdax> pop 3 ore pills of knolopin 12:21 < kazdax> and ready for the day 12:23 < kazdax> SEO tequiest are absolete right ? 12:24 < kazdax> techqnuies 12:26 < djph> ... I think your chatbot's broken there kazdax 12:26 < kazdax> no its nme eing igh 12:27 < djph> o_O 12:29 < royal_screwup21> I'm reading the dev guide to install android studio on linux here https://developer.android.com/studio/install#linux and I don't understand one thing. They say to unzip the.zip file inside /usr/local, and the very next point says that after you do that, you should go to ~/android-studio/bin. My question is: how does android-studio/bin get created i 12:29 < royal_screwup21> n the home directory, when all the stuff was extracted inside /usr/local? 12:30 < stevendale> \o/ 12:30 < jack_rip_vim> stevendale: o/ 12:30 < stevendale> royal_screwup21 Are you royally screwed up 12:30 < kazdax> mkdir ? 12:31 < djph> royal_screwup21: probably a goof in the instructions 12:31 < prillian5> I have a 16 GB file, which I have to share with another one. I have create a torrent-file wtih qBittorrent. 12:31 < royal_screwup21> djph: huh, seems way more likely that I've misunderstood something 12:31 < paradis> searching for suggestion on a good free vpn for linux 12:31 < prillian5> So whats now... sending the torrent file to the other one 12:32 < prillian5> But how do I see, if he downloads or not? 12:32 < BluesKaj> Hi folks 12:32 < prillian5> must I place the "torrent" file in any way into qBittorent? 12:32 < jack_rip_vim> hi BluesKaj 12:33 < BluesKaj> hi jack_rip_vim 12:33 < djph> royal_screwup21: could be that too - but "tutorials" and "have a mistake somewhere" happen :) 12:33 < kazdax> cintactt eh atuhor 12:34 < oiaohm> royal_screwup21: https://flathub.org/apps/details/com.google.AndroidStudio If this one is current enough it could be the simpler and safer way. 12:34 < royal_screwup21> djph: sure but this is google lol 12:34 < djph> royal_screwup21: so? 12:35 < kazdax> contact uahtor ask hik what he means 12:35 < djph> it's still "people" writing the pages. 12:35 < oiaohm> royal_screwup21: I am getting into the habit of check flathub before going to manual installation. 12:35 < djph> kazdax: fix yo keyboard. 12:36 < royal_screwup21> oiaohm: ah I see, I'll check this out 12:36 < kazdax> got cheery blye mx keyboard 12:36 < kazdax> its fixed 12:36 < kazdax> my high isnt 12:36 < royal_screwup21> djph: yeah, it's just that I'm new to linux, so I just wanted to know if "made sense" for the things to happens the way the instructions laid them out 12:37 < royal_screwup21> oiaohm: what's your experience been like with? Is it reliable? 12:37 < oiaohm> royal_screwup21: the one thing about flatpak/ flathub its it will always cleanly uninstall if it does not work. 12:38 < oiaohm> royal_screwup21: where manual instructions can be a kind of a pain to reverse everything correctly. 12:38 < royal_screwup21> oiaohm: lol. What installations themselves? 12:38 < royal_screwup21> what about the* 12:40 < oiaohm> royal_screwup21: I have had a few issues from nvidia drivers with some programs but most have worked perfectly. 12:41 < oiaohm> royal_screwup21: its like most things third party installs always can bring out a few problems. 12:41 < jack_rip_vim> \o/, got my nick back 12:42 < stevendale> Yay 12:42 * stevendale hugs jack_rabbit 12:42 < royal_screwup21> oiaohm: right on time, my flathub android studio installation just froze up 12:42 < stevendale> jack_rip_vim* 12:43 < jack_rip_vim> stevendale: thanks 12:44 < oiaohm> royal_screwup21: I forgot to mention do it from terminal. If you do it from gui it if asks a source question it can get stuck. 12:45 < stevendale> Gonna order a replacement DVD drive for my Dell Latitude E5400 hopefully o/ 12:45 < stevendale> Took the assembly out finally without breaking anything 12:46 < stevendale> It's a good little laptop, 4 GB DDR2-800 Dual-Channel, Intel Core 2 Duo P8700 @ 2.53 GHz, Intel GMA 4500 MHD, 320 GB 7200 RPM 2.5" SATA-II HDD, 1280x800 screen resolution 12:46 < stevendale> Much, much faster than this inspiron :P 12:47 < oiaohm> stevendale: Dell laptops getting the drives out fairly simple. Getting the new drive in is where you have to be careful not to impact the plug and do the board. 12:50 < royal_screwup21> oiaohm: where are all the downloads located? I just installed android studio and I want to see where it's at 12:53 < noodlepie> I have a box.co.uk custom (no PS) Cube laptop - 8thread i7 - 256GB SSD (/var, /boot, swap), 1TB HDD (/) and Nnvidia/Intel graphics with 8GB RAM. 12:53 < noodlepie> It has 1366x768 screen which is oK 12:53 < noodlepie> I meant "no OS". Just installed Gentoo on it. Works a treat! 12:54 < oiaohm> royal_screwup21: https://github.com/flatpak/flatpak/wiki/Filesystem this gives you the paths where everything is stored. But to run you use the run command "flatpak run com.google.AndroidStudio" or search for it added to your menu. 12:58 < cluelessperson> sigh 12:59 < cluelessperson> I set UFW to allow by default 12:59 < cluelessperson> and blocked two ports 12:59 < cluelessperson> but it's continuing to block other ports 12:59 < cluelessperson> for some reaosn 13:00 < stevendale> oiaohm Decided I'll get a new battery for it as well next week o/ 13:02 < stevendale> New DVD drive, new battery 13:02 < stevendale> \o/ 13:03 < stevendale> $54.06 Australian dollars all up 13:03 < stevendale> 6 well 85 Wh batts aren't cheap :P 13:10 < furrymcgee> royal_screwup21: maybe #replicant has alternatives to android studio 13:11 < royal_screwup21> oiaohm: did you ever get this: "error: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name :1.122 was not provided by any .service files" 13:13 < oiaohm> royal_screwup21: not generally. Seams like something I saw a year ago. 13:14 < royal_screwup21> welp I am truly fucked 13:14 < royal_screwup21> it's sad because there doesn't seem to be many people who use flapack and so there's not much helpful documentation on these errors 13:16 < supernovah> Anyone knw some software that runs on two computers and will let me send/receive data which gets delivered to a serial device associated with a USB device (which is a serial tty) from a server which runs a partner software to deliver those remote serial tty's as file descriptors on the remote system, over the network they're both linked by 13:16 < supernovah> I'm in the middle of writing it but it's going to take a wee while is all 13:17 < supernovah> preference is to be able to uniquely identify the ttyUSBx by its serial number which gets enumerated with its usb subsystem 13:18 < supernovah> generally that gets associated with the tty during that subsystem event though 13:19 < royal_screwup21> I'm trying to install the android sdk with umake and I tried: " umake android android-sdk" -- I got an error: "ERROR: Download page changed its syntax or is not parsable (url missing) 13:19 < royal_screwup21> " I googled but couldn't find anything promising :( How do I go about fixing this? 13:31 < kazdax> Debian 9 rocks 13:31 < kazdax> so far as no problems 13:33 < dgurney> yes, that's something debian is known for 13:33 < kazdax> amnd iam using it as my primary deesktop OS 13:34 < paradis> my name is dupdup 13:35 < kazdax> now gotta work on thatr LFS projwect 13:46 < hackeron> Hi, I have a linux hiring question. I'm looking to hire a person that has experience with building an embedded linux image to be loaded on thousands of devices (small X86 miniPC to be used as an IoT hub) and also experience keeping all those devices updated over the air. I'm thinking something like ResinOS or Yacto for the OS updates and Docker for the application updates. What kind of category would 13:46 < hackeron> such a person fit under? Dev ops? Embedded Linux Engineer? 13:47 < djph> no, just no. 13:48 < SkunkyFone> that's a sys admin job, not a dev job. 13:48 < djph> ^ 13:48 < hackeron> SkunkyFone: djph: Will a sysadmin be able to orchestrate OS updates of thousands of IoT hubs? 13:48 < Armand> Easy 13:48 < SkunkyFone> that's what sys admins DO. 13:48 < djph> SkunkyFone: although, if they're producing the product, it could fall under dev as well, I *guess*? 13:49 < djph> SkunkyFone: in the sense that "we'd better develop something not shit". But then again he was talking about docker, so that boat's already sailed. 13:49 < SkunkyFone> djph: i define it thus: sys admin takes things that other people make, put it together and make it work. dev/engineer makes the stuff the sys admin plays with. 13:50 < djph> SkunkyFone: Agreed. I'm honestly not sure where the line gets drawn though -- I mean 'making a new embedded device, so how do we make a not shit flash image for it" kinda sounds dev-work to me. 13:51 < djph> whereas the sysadmin is the guy in control of "oh hey, $vendor has a new image, let's see what they broke this time" 13:51 < SkunkyFone> djph: i'd still call that a high level system admin job, unless it requires writing, say, a custom bootloader, or something. 13:51 < hackeron> Armand: SkunkyFone: djph: Hmm, I see, thank you :) - it doesn't have to be docker. Any advice where I could find a person like this? - That could build the firmware images and infrastructure to handle safe remote OS updates, etc? 13:51 < djph> SkunkyFone: group effort then - fair enough. 13:52 < compdoc> anyone good with petplan? 13:52 < compdoc> netplan 13:52 < SkunkyFone> djph: there's a LOT of overlap on dev and sys admin on something like this.. 13:52 < hackeron> SkunkyFone: yeh, I doubt they will need to write a custom bootloader, likely just use something like this: https://en.wikipedia.org/wiki/Yocto_Project - but I used to be a sysadmin back in the day and that looks pretty complex to me 13:53 < SkunkyFone> hackeron: you want a "senior" sysadmin, with experience on large distributed systems (ie, your whole bunch of devices all over the place) 13:53 < djph> SkunkyFone: yeah, don't disagree there. I only sysadmin at home ... dayjob is only a handful of specific programs (the systems guys do the server / os / etc. adminning) 13:54 < SkunkyFone> djph: these days, me too. but, long long ago, i was one of a team of sysadmins that covered about 2,000 servers, globally. 13:54 < djph> sweet 13:54 < SkunkyFone> nothing like that 2am page when a server in tokyo goes down. 13:55 < djph> I need to get better at it, really. Wanna get out of this role ... "automation is the future, hurrdurr" 13:55 < kazdax> what OS were you using ? 13:55 < hackeron> SkunkyFone: Got it, and I guess I'll mention Yocto and building bootable linux images with read only root. So far, the sysadmins I asked when you mention read only root, they get confused, lol 13:55 < SkunkyFone> kazdax: solaris, linux, sgi 13:55 < kazdax> cool 13:56 < djph> or rather, the manglement thinks my job can be replaced with a small shell script, which is incorrect, but not a good place to be in, y'know 13:56 < kazdax> i thought solari swas niw open source and without support 13:56 < SkunkyFone> hackeron: root is always read-only on boot... if you mean to keep it that way you'll need some kiund of union filesystem 13:57 < hackeron> SkunkyFone: 2,000 servers is no problem, if it's a pretty standard Linux - the challenge here is the system must remain read only to avoid any possible FS corruption - the end user can unplug the IoT hub at any time, so it's designing something bulletproof and so that if an OS/firmware update failed, it can be rolled back 13:57 < SkunkyFone> djph: support people are always the ones who get the short end of the stick. we don't make money, we just spend it. it takes somebody with a brain to notice that spending money on a support person saves the sales force from wasting time and money on things. 13:58 < kazdax> is solaris still in support ? 13:58 < SkunkyFone> kazdax: not these days, as I said, it was long long ago :) 13:58 < kazdax> ahh 13:59 < djph> SkunkyFone: yup :) ... my direct boss is a decent guy ... but ugh 13:59 < SkunkyFone> kazdax: which is why it was a 2am page, not a 2am text message :) 13:59 < hackeron> SkunkyFone: and sure, for the multiple containers, something like aufs/overlayfs, but that is places on a separate partition so if that gets corrupt, the read only root filesystem can just grab the container again (this is just an idea, this isn't implemented yet) 14:01 < SkunkyFone> hackeron: look at how some of the live distros do things. if you have enough ram to put the rootfs in ram, you don't need an overlay fs 14:04 < BenderRodriguez> I need help 14:05 < BenderRodriguez> My linux server is shutting down without any explainations 14:05 < hackeron> SkunkyFone: ram is pretty limited, so something like Resin.os have a solution for that, they pretty much limit the ramdisk to a set amount, e.g. 100MB and then they configure all services and the Yocto build to stop logging to /var/log but rather log to a process which then keeps the last 10MB of logs either in memory or copies it to a destination partiion. They also have 3 filesystems, so 1 boot 14:05 < hackeron> filesystem, 1 filesystem for OS A and 1 filesystem for OS B - so that when you do an OS upgrde, one of your partitions is always clean and read only. But this is all outside of the scope of every sysadmin I spoke to so far, so I'm not quite sure what to post in the job listing, hmm 14:07 < hackeron> so I tought maybe embedded engineer? - but most of those are way over qualified and they write kernel drivers, etc. Here I need someone that can build linux images, that take the above into account (e.g. multiple root partitions, so 1 is always read only), designing a system to update the operating system safely with rollback, experience with containers, etc 14:08 < solidfox> just get any ol' linux user and train 'em. 14:08 < solidfox> hackeron 14:08 < royal_screwup21> I'm on linux and I'm trying to install the android sdk 14:08 < royal_screwup21> I tried sudo-apt get install android-sdk, it showed a bunch of download and seems to wok 14:08 < royal_screwup21> my question: how do I find the location? 14:08 < solidfox> royal_screwup21, the location of the program file? 14:08 < hackeron> solidfox: haha, if only I knew how to do this myself I would :P 14:08 < solidfox> royal_screwup21, you can use which 14:09 < solidfox> hackeron, just tell 'em to figure it out. 14:09 < hackeron> solidfox: hahaha, who, a sysadmin? a devops guy? 14:09 < rascul> distro developer 14:09 < royal_screwup21> solidfox: basically a folder named Sdk 14:09 < solidfox> royal_screwup21, oh. use locate Sdk 14:10 < solidfox> man locate to read the manual 14:10 < hackeron> rascul: hmm, that's an idea, I wonder where I would find a distro developer 14:10 < solidfox> royal_screwup21, to search some locations root is needed. 14:10 < rascul> they tend to contribute to distros 14:10 < milp_2> hi, if i want to set up bonding over two eth interfaces in ubuntu, do i just need to set it up on the linux machine or also on the switch? 14:11 < rascul> if you're working with yocto, maybe there's a yocto dev who wouldn't mind working on related stuff for some extra pay 14:12 < luckybunny> Hi guys... I've run into a problem which I believe is related to the emoji support in 18.04. Numbers do not display correctly in Chrome. Same is true in firefox 14:13 < luckybunny> They appear as white text (of a different font) with a very light grey outline 14:13 < luckybunny> almost impossible to see against a white backdrop 14:14 < hackeron> rascul: oooh yes, I'll try to reach out to the yocto and resinos teams to see if that's possible 14:14 < solidfox> luckybunny, there is an offical #ubuntu support channel 14:15 < hackeron> rascul: better to use that then to re-invent the wheel - also possibly something like Ubuntu Core, but it doesn't run on all the platforms I need 14:16 < SkunkyFone> luckybunny: 18.04 what? this is the general linux support channel, btw... 14:17 < solidfox> SkunkyFone, 18.04 is the latest release of ubuntu 14:17 < luckybunny> SkunkyFone: sorry about that. I copy/pasted my original question from #ubuntu. It's Ubuntu 18.04 14:17 < solidfox> ah 14:18 < solidfox> luckybunny, I didn't have that issue on chrome. 14:18 < solidfox> luckybunny, what did you do 14:18 < solidfox> to screw it up 14:18 < luckybunny> Well, all was fine before the dist upgrade, and all was not fine after... that's as much as I know 14:19 < rascul> luckybunny font change? 14:19 < bonestormii_> Can anyone tell me what the syntax is called for config files like tmux? Example: https://github.com/gpakosz/.tmux/blob/master/.tmux.conf 14:19 < luckybunny> https://gyazo.com/b9388c81ed102c0017896d8edf5bc564 14:19 < luckybunny> that's what it looks like, the numbers are visible, albeit barely 14:20 < solidfox> luckybunny, try to install fresh. 14:20 < solidfox> luckybunny, you'll need to backup your files first. 14:21 < solidfox> luckybunny, I like your nick btw 14:21 < luckybunny> thanks. Had it since 2008 14:22 < djph> bonestormii_: looks like it's just "freeform". Not sure what you're really asking though 14:22 < solidfox> bonestormii_, do you need documentation on that config file, or are you just curious 14:22 < bonestormii_> djph: Basically it's a bash script, right? 14:23 < djph> no 14:23 < luckybunny> ah... it's not all numbers. looks like it's particularly wikipedia that's the victim of this 14:23 < djph> it's not bash (or any shell that I'm familiar with) at all. 14:23 < luckybunny> might explain why it's taken me a fw days to notice 14:23 < solidfox> luckybunny, if its just wikipedia then hmm 14:24 < helladen> Does anyone want to work on a KDE-connect iOS app? :o 14:24 < djph> luckybunny: perhaps bad font selection (replacements) in wikipedia? 14:24 < bonestormii_> all of those commands though can basically be executed like "tmux bind -r ..." 14:24 < bonestormii_> so I figured it must just be iterating over the file and executing those commands 14:25 < bonestormii_> I'm interested in generating this general type of file, and thought I've seen this kind of syntax before perhaps 14:25 < solidfox> the syntax is simply tmux options then :) 14:25 < bonestormii_> lol cool 14:25 < djph> bonestormii_: that's modifying 'tmux' - not shell 14:25 < bonestormii_> right, so tmux is just parsing it's own config files for these commands and executing them using it's own internal logic, then? 14:25 < djph> correct 14:26 < bonestormii_> its* 14:26 < bonestormii_> thanks guys! 14:26 < solidfox> luckybunny, which wikipedia page 14:27 < luckybunny> https://gyazo.com/9d7d490b080ec294b6d28cae5639fe6c 14:27 < luckybunny> I'm not actually researching hurricane Carmen, but showing that it's a reproducible oddity I'm up against, rather than a one-off 14:27 < solidfox> https://en.wikipedia.org/wiki/Hurricane_Carmen 14:29 < solidfox> font family is just sans serif 14:29 < solidfox> luckybunny, I think you have a fake version of chrome installed 14:30 < solidfox> or some weird ad on 14:30 < luckybunny> also firefox? 14:30 < solidfox> luckybunny, it does on firefox too? 14:30 < solidfox> hmm 14:30 < luckybunny> yup 14:30 < hackeron> solidfox: SkunkyFone: djph: thanks for the advice guys, I came up with this: https://gist.github.com/rgaufman/27771fec514c953288be2ec3d1053779 14:30 < solidfox> strange issue. why would it just be the numbers xD 14:30 < rascul> luckybunny try with a different font 14:32 < helladen> Any Linux programmers want to work with me on an iOS version of KDE-connect. I am going to use Swift and Xcode. I may end up looking for a better IDE. 14:32 < luckybunny> aha 14:32 < helladen> I personally prefer iPhone/Apple Watch over Linux-Android, rather code my own app than to use JVM Android. 14:32 < helladen> :P 14:32 < luckybunny> switching the sans serif font to a different one did help 14:32 < helladen> ' 14:33 < luckybunny> now to find out why my sans serif font is bugged 14:33 < luckybunny> thanks rascul 14:33 < rascul> luckybunny it's probably using whatever you have set for the default font in your desktop environment 14:33 < rascul> for the default sans-serif font i mean 14:34 < luckybunny> that was 'Sans' 14:39 < helladen> Is it wrong to sell software, but provide it open-source for experienced users? 14:39 < dgurney> no, of course not 14:40 < noodlepie> You can sell Free software 14:40 < helladen> I was just wondering. Not sure how Linux community feels about paid software. 14:40 < noodlepie> Many of us just use Free software 100% on our machines. 14:41 < helladen> Well the main reason I ever go open-source is due to contributors. 14:41 < milp_2> a lot of companies provide the software open source and make money on support and custom/priority development 14:41 * noodlepie is phill@GNU.ORG , I coded the Konica gPhoto driver and fix gPhoto's runtime dynamic library loader so we didn't need hugh monolithic apps with all drivers linked to the app binary. 14:41 < noodlepie> Also convinced Konica to release their own fully features driver. 14:43 < helladen> That makes sense. 14:47 < helladen> Swift is a little wierd. 14:49 < luckybunny> I dug a little deeper. I did, after all, have a config file hanging about setting Noto Colour Emoji as my default sans serif font. Deleted that file, everything looks fine, even with Chrome returned to its previous settings 14:49 < luckybunny> Thanks guys, that was really beginning to annoy me 14:49 < noodlepie> I like to sell Linux to people as "Fun, Free stuff you can do with a computer and the Internet' It saves getting involved in complicated licensing issue with newbies 14:50 < de-facto> how can i have networking restart on replug ethernet cables? 14:50 < luckybunny> I prefer to sell linux as 'Windows but without the hassle or price tag' 14:51 < noodlepie> Windows is a cancerous virus 14:51 < dgurney> lol 14:51 < noodlepie> All proprietary software is 14:52 < luckybunny> Last time I used Windows for anything important (i.e not gaming), it ran into a networking error. Silently failed to deal with it, so we used a program to sniff the packets being transferred, and managed to glean an error code for the issue, that we could throw into the Microsoft knowledge base 14:53 < luckybunny> Microsoft knowledge base helpfully informed us that the error code translated to 'there was an error' 14:53 < john_rambo> Hi, I did sudo nano /usr/share/applications/firefox.desktop and added firejail to the exec line. When I click on Activities then firefox it opens with firejail but when I click on the FF dash icon it opens without firejail 14:53 < rascul> i've never used windows for anything important, i wonder how it feels 14:53 < john_rambo> Ubuntu 18.04 15:00 < ssarah> Why is it bad to use 777 on dirs? 15:01 < ssarah> it seems like an easy fix 15:01 < djph> I mean if you want them world-writable, that's fine (see /tmp) 15:06 < revel> ssarah: Last time I checked, write access on a dir (without the sticky bit like i.e /tmp and /dev/shm have) means anyone can delete files there. 15:06 < revel> s/Last time I checked/I think/ 15:06 < ssarah> yes, like i got this script to be run by a low privilege user i just created to fetch some metadata and write it somewhere to be fetch by an rsync. I of course used /tmp/user first, but then someone rebooted the machine :( 15:06 < ssarah> and the dir was gone 15:06 < ssarah> now i created /misc/user and gave it 777 15:06 < revel> You could use /var/tmp instad, I think. 15:07 < revel> Or not store anything in a dir marked as temporary :P 15:07 < rascul> or don't 777 ever 15:07 < revel> /var/tmp shouldn't be tmpfs or cleared on reboot. 15:07 < revel> Also, yeah, don't 777. 15:07 < ssarah> i'm just 777 a specific subdir 15:07 < rascul> don't 15:07 < ssarah> why? 15:08 < rascul> you just said you had a script in there 15:08 < rascul> 777 means without you realizing it someone can come along and add bad stuff to the script and you just run it and have no clue 15:08 < ssarah> no. i have no script there, it's where the script writes. the script is on the home of the user i created to run it 15:09 < ssarah> and on his crontab 15:09 < rascul> still, what would make you want 777 on it? 15:09 < ssarah> i could give ownership to the dir to that user, but im really wondering why 15:09 < revel> Why can't the data just be written to somewhere in the user's HOME ? 15:10 < ssarah> Cause it obfuscates where it is, everyone needs to know where the important stuff is. 15:10 < Celmor> I'm having memory issues where even my VM got reaped by OOM, I think at fault is file-roller while an extraction was in progress but I wanna make sure before I replace it https://ptpb.pw/AQKd.html 15:10 < rascul> what is this data used for? 15:11 < revel> ssarah: Why's /var/tmp not an option? 15:11 < ssarah> this is the first time i heard of /var/tmp, seems like a good choice 15:11 < ssarah> rascul, it's some parsed google analytics stuff 15:11 < ssarah> i think 15:11 < revel> It's not cleared on reboot. 15:12 < ssarah> yep yep, i'll consider it from now on. I still didnt get the phobia of 777 15:12 < revel> Well, by default, on distros I've used. 15:12 < ssarah> both /tmp and /var/tmp seem to have 777, right? 15:12 < rascul> 1777 15:12 < revel> They have 1777, which stops arbitrary users from deleting data at their leisure. 15:12 < ssarah> yep, sticky bit, i know i know 15:13 < ssarah> still, it seems more like aesthetics than logic so far to me 15:13 < rascul> how is it logical to let anybody feed you whatever data they want? 15:13 < dgurney> 777 is only okay in situations where there is no other user, ever. and even then, it's quite a security risk 15:15 < rascul> https://en.wikipedia.org/wiki/Principle_of_least_privilege 15:16 < ssarah> Sorry for mentioning it again, but it seems the same risk of having /tmp dirs. I guess the less users can write to a dir the better. There's also the principle of reducing the number of resources you have to manage, so sometimes just a quick dirty solution like this seems easy. 15:16 < mnemon> ssarah: it's mostly bad if for an example some webserver software accesses that place and you can feed scripts for it to execute. So to be on the safe side better not use it unless necessary. 15:16 < ssarah> still, that /var/tmp thing seems peachy 15:18 < ssarah> mnemon, in most debian systems you can read other users' homes right? 15:18 < ssarah> so you can read their scripts 15:22 < rypervenche> ssarah: Why not just give the correct permissions to the thing you need? 15:23 < mnemon> ssarah: With the default rights probably. Usually the biggest security issue with creating new 777 dirs on servers is when you do it inside the directory structure accessible for web servers and such. /tmp tends to be protected somewhat because it used to be such a common attack vector. 15:24 < mnemon> (aka. creating places where anyone can write stuff and make some service execute it) 15:25 < pankaj> I googled the main task of /opt directory and got that it has something to do with proprietary stuff involved with UNIX. But I suspect that it is all what it is for. 15:25 < revel> When /tmp is tmpfs, it generally has some protection in the form of i.e nodev,nosuid,noexec 15:25 < Alexander-47u> hi all 15:25 < rypervenche> pankaj: Yep, pretty much. There are different standards and some don't like to use /opt, others do. 15:27 < pankaj> rypervenche: So, It just does the task that I said and nothing else? 15:28 < pankaj> rypervenche: What are some linux destros that do not use /opt directory or in other words do not want to deal with proprietary softwares? 15:28 < rypervenche> pankaj: It's usually used by people who package proprietary software that is not in the main repositories of a certain distribution. 15:29 < rypervenche> pankaj: That way it won't conflict with the existing files on the machine, possibly with the same name. People do this with newer versions of scripting languages as well. python for example. 15:29 < triceratux> pankaj: a good example of a 3rd party supplier that uses /opt is vivaldi. they take their entire browser & just bolt it onto the system there rather than painstakingly spraying it thru the fhs which these days can differ substantially between distros 15:29 < pankaj> rypervenche: Understood 15:30 < rypervenche> pankaj: https://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard 15:31 < voiter> i am looking for an MTA that is smart enough to deliver a locally sent mail to the local mailbox and nothing more. any hints? 15:32 < voiter> (a bit more context. i develop a web application that sends mail. during development i want those to be delivered to a local mailbox) 15:35 < mnemon> voiter: you can configure all the common ones to do that. 15:36 < voiter> mnemon, hm.. k. will look into it. 15:37 < voiter> postfix has this option: "The only delivered mail is the mail for local users. There is no network." Somehow this implies that any other mail not for local users are discarded. am i right or wrong? 15:37 < Zajt> Hey! How do you remove a line from a file without opening it? So I wanna remove a line from the terminal 15:38 < voiter> Zajt, sounds like a job for sed 15:53 < solsTiCe> hi. I have an issue with usb 1-1.4. It's filling the log with error. How do I know which usb device it is ? from the output of lsusb for example. Is it Bus 001 Device 004 ? 15:55 < terra> solsTiCe: check out by plug/unplug a usb device and watch dmesg 15:59 < solsTiCe> terra: yes but you can't unplug all usb device 15:59 < xdexter> Hello! i need a regex to capture just "content" value in this line: https://pastebin.com/Ai8uWKCL, someone could help me pleasE? 16:02 < solsTiCe> xdexter: sorry but https://regex101.com/ 16:02 < xdexter> solsTiCe, yes, i tested but in Zabbix i think that not work, it's considering " />" too, do you understand? 16:04 < TheWild> hello 16:04 < TheWild> what parameters for dd to use on possibly failing HDD? 16:05 < Psi-Jack> TheWild: Umm.. wut? 16:06 < TheWild> IIRC there were two parameters, one for "don't interrupt on IO error", and another for "despite IO errors, keep sector positions in sync" 16:06 < Psi-Jack> And what.... Are you expecting.... to accomplish with this? 16:07 < TheWild> Psi-Jack: what you're doing when your HDD might fail soon? 16:08 < Psi-Jack> TheWild: Since I have backups of everything always, I'd run SpinRite on it, and see what it says. Often times, if it's not a mechanical failure, it'll fix many issues with an HDD. 16:08 < jhodrien> I think ddrescue is what you're after. 16:10 < TheWild> Psi-Jack: I just hope that out of nothing increasing 'Current pending sector count' is only a connection problem. I tried to disconnect the drive to connect for a while one from another computer. 16:10 < TheWild> it was 0 before I was messing with it. Now it's 9! 16:10 < Psi-Jack> TheWild: What brand HDD? 16:10 < TheWild> Western Digital 16:11 < Psi-Jack> Hmmm... Yeah. That one usually reports pretty decently. Usually (though not always). Blue? Red? Black? 16:11 < TheWild> mmm... give me a while 16:13 < TheWild> WD Scorpio Blue 16:13 < noodlepie> Anyony use CDE or Motif? 16:14 < noodlepie> I like them and they're open now 16:14 < noodlepie> Makes it feel very Solaris-like 16:14 < Psi-Jack> noodlepie: Hello, this is 1980 calling. 16:14 < noodlepie> Or AIX, and a little bit like SGI. 16:15 < noodlepie> eh? Whadaya do you mean? 16:15 < noodlepie> oh 1980's. I read it as 1000 doh 16:15 < Psi-Jack> heh 16:15 < noodlepie> They're still very useful software packages if you want to learn X/ICCM and stuff. 16:16 < noodlepie> And the panel is cute. As if the window manager. 16:16 < jhodrien> Using old DEs *sometimes* reminds you of things the original standards got right. Like not stealing focus. 16:16 < noodlepie> As is, even. 16:16 < turkeyhand> I installed arch, and I can't boot into windows 16:16 < TheWild> another question. RAM in this netbook is damaged above 0x80000000. It has Xubuntu installed, which I configured to just limit RAM to first 2 GB. However, I don't know how to perform this on live distro. 16:16 < revel> turkeyhand: Congratulations! 16:16 < noodlepie> I liked GNOME2. Thankfully its still around in the form of Mate. 16:16 < turkeyhand> it didn't automagically find win10 16:16 < revel> Moreso on the Windows part than the Arch part. 16:16 < turkeyhand> no I have to dual boot 16:16 < noodlepie> The panel and its use of CORBA was cute. 16:17 < TheWild> it's better not to boot up the installed system, and also that I need to copy the partition where Linux is installed. 16:17 < triceratux> turkeyhand: why would you want to boot into windows once you have arch installed ? 16:17 < turkeyhand> becuase occasionally I need to use windows 16:17 < revel> triceratux: Practically everyone I've seen ask about dual-booting has been an Arch user... 16:17 < revel> Why is that? 16:17 < turkeyhand> because grub doesn't find the other OS probably 16:17 < triceratux> non-arch users dont have to ask rofl 16:18 < turkeyhand> because it's done automatically for you 16:18 < turkeyhand> do you know how to fix grub or not? 16:18 < revel> I haven't seen other people even mention dual-booting, it's almost always Arch users. 16:18 < djph> triceratux: and here I figured it was just because the archwiki isn't terrible. 16:18 < revel> It's almost as if only Arch users dual-boot or something. 16:18 < triceratux> turkeyhand: yep. https://sourceforge.net/p/boot-repair/home/Home/ 16:18 < turkeyhand> the archwiki is not good 16:19 < jml2> boot-repair is also supposed to be able to fix Windows boots 16:19 < jml2> turkeyhand, no way! 16:19 < turkeyhand> why are you posting debian software 16:19 < triceratux> if arch would fix their software they wouldnt need so comprehensive a wiki. its a whole new paradigm 16:19 < djph> idk, it's always fixed my problems ... 16:19 < usr123> I just installed ubuntu and cannot find drivers in additional drivers tab 16:19 < djph> triceratux: and here I figured the paradigm was "document it". But I could be wrong. 16:20 < noodlepie> The RIP (Rescue Is Possible) rescue disk ISO is a widely distributed generic Linux live flash which is useful to fix Linux installs with 16:20 < triceratux> more "yer a n00b if you dont understand all the failure modes weve documented". other distros try to minimise the number of failure modes 16:21 < turkeyhand> yes arch people are elitist 16:21 < jml2> noodlepie, RIP is RIP -> https://www.tux.org/pub/people/kent-robotti/looplinux/rip/ 16:21 < jml2> noodlepie, LOL 16:21 < turkeyhand> I just want this fucking OS to work, and the other one too 16:21 < djph> which is frustrating when they just hide it and it's a royal pain to fix. 16:21 < jml2> "I truly loved this project but now has like 2 years orphaned and the website is not current. RIP for RIP." XDXDXD 16:21 < turkeyhand> so far arch IS working, but I need to make sure I can get to windows if I need it 16:21 < turkeyhand> for paid software/work 16:22 < jml2> turkeyhand, a lot of arch users are actually dumb 16:22 < jml2> turkeyhand, truth be told, they thing they can do what no other distro users can do. Use a basic command. 16:22 < turkeyhand> maybe they think to get "cred" they need to use arch 16:23 < turkeyhand> I went through all the distros and hated this one the least, it's still not "best" 16:23 < jken> why is it then dd'ing an image to a USB drive only works every 2nd time on my system 16:24 < jml2> jken, it doesn't 16:24 < jml2> jken, likely you have to do a full power off and plug your usb stick in another usb port 16:24 < TheWild> using internets without mouse (all USB ports are already in use). Oh God, why? tab, tab, tab, tab, tab, tab... 16:24 < jml2> jken, some usb brands are not meant for usb booting 16:24 < revel> jml2: Did you sync? 16:24 < jml2> revel, jken 16:24 < triceratux> the ideal approach is to leave windows & its bootloader alone, ie with an mbr & a bootloader on a separate device like a flashdrive. thats easiest with a live distro which arch is hesitant to create 16:25 < turkeyhand> I think grub messed with the bootloader, I'm not really sure 16:25 < jken> I've even been watching /sys/block/DEVICE/stat to ensure its done syncing buffers before I remove it 16:25 < turkeyhand> I can't boot into windows at all 16:25 < jml2> jken, and true to what revel said, you should type "sync" before pulling out the usb -- or use hte eject command before pulling out the usb 16:25 < turkeyhand> I wouldn't know how to 16:25 < turkeyhand> I need to edit grub to include windows 16:26 < turkeyhand> any good ideas? 16:26 < jml2> turkeyhand, some windows installs but hte windowsbootmgr in the mbr, you'll need to recreate it --- or sometimes grub finds windows part#1 instead of using windows part#2 16:26 < jken> jml2, I do that. Every time it actually works (every 2nd time) gnomes file manager actually throws a DBUS error when I click "eject" 16:26 < jken> I've come to know that error as "its going to work when I boot this usb drive" 16:26 < jml2> turkeyhand, that's as far as i go to fixing winbloze XD 16:26 < turkeyhand> you just pretend other people aren't required to use windows against their will 16:27 < triceratux> turkeyhand: if the windows bootloader is intact sometimes grub2 can perform an ordinary chainload http://www.gnu.org/software/grub/manual/grub/html_node/Chain_002dloading.html#Chain_002dloading 16:27 < turkeyhand> you know like work and shit 16:27 < turkeyhand> that thing that means rent and expenses can be paid? 16:28 < turkeyhand> windows is on the first portion of sdb 16:28 < revel> turkeyhand: It depends on what sort of job you have. And with some, you may be working with data that you can't really work with at home anyhow. 16:29 < turkeyhand> paid software doesn't work in linux 16:29 < turkeyhand> a lot of graphics stuff doesn't either 16:29 < jml2> turkeyhand, you'll need to find out if you need to create "bootmgr" on C:\ .. for that you use your windows installer --- grub needs to chainload the C:\ parition 16:30 < revel> Define "paid software" 16:30 < jml2> revel, he's pimping 16:30 < revel> Pimpin'? 16:30 < turkeyhand> software that costs money and doens't work through wine 16:30 < jml2> punkin pimpin 16:31 < revel> I don't know, VMWare is paid software and works on Linux. 16:31 < MrElendig> for some defenition of "works" 16:31 < MrElendig> (often it doesn't) 16:31 < revel> Just an example. 16:32 < turkeyhand> stop giving examples and also opinions 16:32 < jml2> turkeyhand, if you pay me I can help fix it for you :)) 16:32 < MrElendig> vmware is a good example of how much of a pain closed source software can be 16:32 < turkeyhand> no 16:33 < jml2> turkeyhand, good luck to getting windows help on a linux channel XD 16:33 < MrElendig> only supports a handfull of outdated er.. I mean "stable" distro version etc 16:34 < revel> MrElendig: You deal with Arch users more, right? Why is it that so many Arch users specifically talk about dual-booting, as compared to users of other distros? 16:34 < MrElendig> revel: that is not my experience 16:34 < jml2> revel, it must be because all the ubuntu users now are in school XD 16:34 < solidpizza> revel, I've noticed that too 16:35 < solidpizza> revel, I've noticed many arch users that also like to dual-boot specifically with windows 16:35 < revel> MrElendig: I haven't noticed users of other distros ask about help with dual-booting and the people who talk about dual-booting somehow always seem to be Arch users. 16:35 < revel> solidpizza: I meant with Windows, yes, not another distro or some BSD. 16:35 < solidpizza> traitors. 16:36 < jack_rip_vim> windows is gone, microsoft joins linux 16:36 < triceratux> for a long time the archwiki had easily the most accurate & comprehensive discussion of multiboot installations 16:40 < jack_rip_vim> triceratux: linuxfromscratch has the best manual to start up a linux system 16:40 < jml2> arch wiki is pretty good, not sure why this particular noob says its bad... and i'm not even an arch user lol 16:42 < Nussi> well this channel gather a lot of trolls these days 16:42 < jack_rip_vim> jml2: you should try it 16:42 < jml2> jack_rip_vim, why? 16:42 < jml2> jack_rip_vim, I already did 5 years ago :) 16:42 < jml2> jack_rip_vim, I find nothing satisfying about it.. 16:42 < jack_rip_vim> jml2: archlinux changed 16:43 < jml2> jack_rip_vim, oh it did? well I'm here to help linux users (any distro), not to fix "Windows boots". 16:43 < jml2> jack_rip_vim, try reading some punkin pimpin poopie 16:44 < jml2> yeah that's what I thought.. 16:44 < jack_rip_vim> jml2: lol, actually, i already try most of the distro to get a better server 16:44 < triceratux> even better, arch derivatives have improved vastly. swagarch ftw 16:45 < jml2> triceratux, I tried swagarch yesterday evening because you kept insisting it.. this distro doesn't even install a cron package.. 16:45 < jml2> triceratux, it's xfce with the droid font... 16:46 < jml2> triceratux, I don't know what's so special about it than installing one of those material themes for it.. 16:46 < jack_rip_vim> jml2, I believed you have experience with plan9 16:46 < jml2> triceratux, i'm pretty the defaults are slim for light weight machines... 16:47 < jml2> triceratux, ./quite sure of./ hte "idea" that it is meant for older things... which looks like the intent of it 16:47 < jml2> it has a quick and easy installer iirc... which distinguishes itself from the stock arch 16:48 < triceratux> jml2: whatz special about it is that it boots at all. & into a vanilla xfce4. & yeah you have to replace the theming & the fonts right away like on most distros. i noticed theres no cron rofl. doesnt keep it from asynchronously scheduling the mandb rebuild. i have to track that down someday. must be a systemd thingy 16:48 < jml2> triceratux, that's the first thing I did :p.. "apropos ." --- no result, I do "mandb" and that updated the man cache... 16:48 < jml2> triceratux, I do find /etc/ 16:49 < jml2> triceratux, |grep cron -- and nothing, so I search with pacman -Ss cron , and see there's 2 or 3 flavours of cron to use 16:49 < triceratux> jml2: a little easier than lfs huh ? 16:49 < jml2> triceratux, the default theme is good 16:49 < jml2> triceratux, this is probably good for users who want to dive into arch.. 16:50 < jml2> triceratux, because the installer is rather simplistic -- but it still requires "a bit" to know what is /dev/sda, filesystems... 16:50 < jml2> triceratux, it's not noobie-noobie.. 16:50 < triceratux> jml2: the only real issue is pacman comes up borked & you have to learn how to reinit the keyring. thats just because the projects servers are so iffy 16:51 < iodev> don't get me started on Arch ... bluez is all that comes to mind 16:51 < iodev> bluez segfault edition that is 16:52 < rigid> Is it possible to password-protect editing of boot-parameters with grub2? 16:52 < jml2> rigid, yeah you use the lockdown option in the /etc/default/grub ... then use your update grub script 16:52 < iodev> rigid: just use LUKS 16:53 < rigid> I just see password-protecting specific menu entries or the complete menu. I want a default entry to boot without password and every manual edit should require a password (including pressing "e" on the default entry) 16:53 < iodev> who cares if anyone edits if the drive is encrypted :D 16:53 < Toaster_Strudel> What is a good chan for syslog support? 16:53 < iodev> Toaster_Strudel: use dpaste.com, ask here 16:53 < Toaster_Strudel> I want to make a proof of concept test syslog server to audit a linux DFS system 16:54 < Toaster_Strudel> I want to audit file access 16:54 < Toaster_Strudel> Do I just setup a samba server, or do I need to implement another system for syslog? 16:54 < jml2> rigid, google grub-mkpasswd-pbkdf2 16:55 < rigid> jml2: that's what I did. I added the parameters to /etc/grub.d/10_linux. It locks the whole boot process 16:55 < rigid> jml2: manually editing grub.cfg let's me lock specific menu entries but I can still "e" the default entry 16:57 < rigid> jml2: ah, i guess I need to do "set superusers=..." 16:57 < rigid> thank you 16:58 < Toaster_Strudel> It's looking like samba only audit logs ntlm/kerb, but not individual file access (by default) 17:01 < Toaster_Strudel> I see auth_audit, but not access_audit 17:09 < turkeyhand> if I copy /home to a new install will all the same shit be there 17:09 < turkeyhand> wallpapers and all that stuff 17:10 < turkeyhand> software 17:16 < djph> turkeyhand: well, yeah, because you copied it ... 17:16 < emberquill> Your documents, user-specific config settings, some installed things like steam games or other things that install to ~/.local will still be there. Anything that installs to /usr will be gone though. 17:17 < jml2> turkeyhand, you can also copy your documents from your ntfs partition into Linux and then remove your Windows partition 17:17 < jml2> you can do that too 17:17 < jml2> you know.. 17:17 < jml2> software XDXDXD 17:17 < jml2> "dem dam dee software software" 17:18 < anddam> hello 17:22 < jml2> anddam, hoodiee 17:23 < jml2> anddam, you coming to tell us some trump news again? XDXD 17:23 < rumpel> yay, grandpa news \o/ 17:24 < anddam> jml2: drunk already? 17:24 < jack_rip_vim> grandpa news? 17:25 < MrElendig> jml2: breaking news: http://i0.kym-cdn.com/photos/images/original/001/010/280/5dd.jpg 17:28 < CtrlC> does anyone know a good socks server software to run a proxy server? 17:29 < jml2> CtrlC, squid is pretty popular 17:29 < CtrlC> jml2, something simpler maybe? and I thought it is http. 17:29 < jack_rip_vim> like some kind of VPN stuff 17:30 < jml2> sock servers don't suck *wink wink* 17:30 < jack_rip_vim> :) 17:32 < jml2> MrElendig, the next pm of canada :) -> https://upload.wikimedia.org/wikipedia/commons/thumb/e/ea/MPP_Jagmeet_Singh_at_his_annual_community_BBQ_in_2014_%28cropped%29.jpg/1200px-MPP_Jagmeet_Singh_at_his_annual_community_BBQ_in_2014_%28cropped%29.jpg 17:32 < jml2> MrElendig, a leader of a major political party here in canada.. 17:32 < jml2> MrElendig, it'd be interesting to see if it actually happens.. 17:32 < jml2> MrElendig, lol 17:33 < jml2> CtrlC, there's many of them, you want to do a reverse proxy to a backend-server? 17:33 < CtrlC> jml2, I just want to change my IP. :D 17:34 < jml2> CtrlC, you can only change it with a external provider --- nordvpn has a very simple linux support -- i'd recommend them... 17:34 < jml2> CtrlC, prices are decent -- and speed is "great" 17:34 < CtrlC> jml2, I don't want to use a service. I don't trust people. 17:34 < jml2> CtrlC, you can always setup your own proxy --- but i've tried this and I am capped at "3 megs /second".. 17:34 < jack_rip_vim> CtrlC: You can buy a server, then chat with us on that server 17:35 < CtrlC> jack_rip_vim, it's not for here and I already have servers. 17:35 < CtrlC> shadowsocks does the job I guess. 17:35 < jml2> CtrlC, i understand, but it's one of the most respected vpn's out there.. 17:36 < jml2> CtrlC, ahem even respected by torrent sites XD 17:36 < jml2> CtrlC, even though here I don't even use torrents.. 17:36 < CtrlC> jml2, I see. but still. it's really complicated. 17:36 < CtrlC> Kim doesn't like those services they're blocked in NK. 17:36 < jml2> CtrlC, no it isn't, you can run the openvpn in terminal, it does its magic, and when you are done, just do a ctl-c in terminal and your routing table is back to normal.. 17:37 < jml2> CtrlC, they have an opepvn file generator ... 17:37 < CtrlC> jml2, it doesn't work when they block the protocol. 17:37 < jml2> CtrlC, iirc you basically download and feed it to openvpn and it works.. 17:37 < CtrlC> VPN won't work. but thanks. 17:37 < jml2> openvpn is open 17:37 < CtrlC> Believe me I know what I need. 17:38 < CtrlC> I just needed software. 17:38 < jml2> people use openvpn to setup their own vpn service 17:39 < CtrlC> jml2, I know what openvpn is. still ISP can block it. 17:39 < jml2> you can host it on your own site... not sure why you're stumbled that "openvpn" doesn't work.. 17:39 < CtrlC> the only thing really they can't block is tor. 17:39 < CtrlC> but socks can also be fine. 17:39 < jml2> CtrlC, and what isp is that? 17:39 < jml2> CtrlC, they're not allowed to block it 17:39 < CtrlC> jml2, North Korea national ISP. it doesn't matter. 17:39 < jml2> ok 17:39 < jml2> so you don't know how to setup openvpn 17:39 < jml2> LOL 17:40 < jml2> and you say the ISP blocks it which is 100% nonsense. 17:40 < jml2> jc. 17:40 * jml2 XD 17:40 < djph> jml2: DPRK (and China) do fun and interesting things with packets. 17:41 < jml2> his english is too good to be coming from china XD 17:41 < CtrlC> yeah my engrish is great 17:41 < djph> or perhaps he got sent there on biz? that happens, you know. 17:41 < CtrlC> sorry for racist joke. :( 17:41 * jack_rip_vim is learning on how to manage relationship with 6500+ on LinkedIn 17:41 < jml2> yeah I was right.. he's not from there.. 17:42 < jml2> he's some yankee doodle who doesn't know how to setup openvpn and just blamed his ISP for his lack of competency.. 17:42 < jml2> ahem 17:42 < jml2> probably even a trump supporter 17:42 * jml2 *hides* :)) 17:43 < jack_rip_vim> recently, trump is on news often 17:43 < CtrlC> I have a wall size poster of him hanging next to my bed. 17:45 < jack_rip_vim> anyone who has 10K relationship on LinkedIn? Can anyone teach me how to keep relationship with them? 17:45 < CtrlC> jack_rip_vim, I never really used linkedin too much. in your experience does it really help with career? 17:46 < CtrlC> I don't have much faith in it. :D 17:46 < jml2> linkedin does very heavy data mining.. 17:46 < CtrlC> yeah who doesn't! 17:46 < jack_rip_vim> CtrlC: I never try to apply any jobs in there. But I like to chat with people in there 17:46 < jml2> people who use linkedin sharing my internet connection -- I get messages all the time from linked-in asking if "I know this person" 17:47 < jml2> they also make it known that they collect data from "ip addresses" -- but they do so intrusively.. 17:47 < jml2> it's almost like spam... 17:47 < CtrlC> Tor baby. Tor. 17:48 < jml2> they won't stop sending reminders until you actually connect them in your linked-in account 17:49 < jack_rip_vim> those message are often, Today, it is first year who in new company, today, someone have news career 17:49 < jml2> I think anyone who uses linkedin in here knows this all too well 17:49 < jack_rip_vim> .... 17:49 < jack_rip_vim> s/news/new 17:50 < jml2> jack_rip_vim, if you want to get a lot of spam from linked-in, simply go to as many hotspots as you can and connect to linked-in :) 17:50 < CtrlC> I like hot spots. 17:50 < jack_rip_vim> But, no matter what, chatting with people in there is fine 17:51 < jack_rip_vim> jml2: my mailbox already has a lot of it 17:54 < jack_rip_vim> Aha! By the way, my nick in LinkedIn is Jack Vim, if you see snoopy with his buddy woodstock, that is me. 17:56 < AE-35> jack_rip_vim: are you using emacs now? 17:57 < jack_rip_vim> AE-35: I use both emacs and vi/vim 17:57 < absurdistani> no one should be using emacs. 17:57 < absurdistani> vim is love. vim is life. 17:57 < phre4k> do I discuss business webapps here? Wondering if Nextcloud can support 7000+ contacts and not go tits up 17:58 < phre4k> ah wait, ##networking should be better for this, sorry 17:58 < jack_rip_vim> phre4k: the problem is how to keep relationship with all the people 17:59 < jack_rip_vim> my relationship is growing, everyday, I have new relationship. 18:00 < CtrlC> and I'm single af 18:01 < jml2> jack_rip_vim, that does not make you look more outstanding by autobotting your connections in linked-in 18:01 < jml2> lol 18:01 < jml2> millenials! 18:02 < jack_rip_vim> jml2: I don't know I just use it to chat with people. I am not thinking to keep myself outstanding. 18:03 < jml2> linked-in is supposed to be use for professional purposes, not for some lame quest like facebook 18:03 < davr0s> is there a way to expose a directory for read access with no authentication requred (what I have in mind is rcp/scp -ing from it from elsewhere) 18:03 < jml2> lol 18:04 < ayecee> davr0s: kind of like what a web server does, eh 18:04 < davr0s> i guess so, are you saying 'setup a web server on it' 18:05 < ayecee> sure, that would be an option 18:05 < jack_rip_vim> jml2, few minutes ago, someone in LinkedIn just ask me if it is a good to learn C language if he want to become a system administrator 18:05 < ayecee> couldn't hurt 18:05 < jack_rip_vim> s/a good/a good idea 18:06 < jack_rip_vim> the guy seem more interested in system security 18:06 < jack_rip_vim> he has kali intalled 18:08 < jml2> Mr Robot fan boys.. 18:08 < jack_rip_vim> Mr Robot 18:08 < jml2> wow linked-in is becomig the plague for social media frenzy poopie 18:09 < tangarora> How do you mean? 18:09 < seven-eleven> hi 18:09 < jack_rip_vim> hi seven-eleven 18:09 < seven-eleven> how can I create a virtual interface with networkmanager? i noticed you can add multiple IPs to one network interface in networkmanager, but i don't see where you can add eth0:1 eth0:2 etc 18:09 < seven-eleven> hi jack_rip_vim 18:10 < jack_rip_vim> seven-eleven: do you have dual netwok card? 18:10 < seven-eleven> jack_rip_vim, nope, just one 18:11 < jack_rip_vim> seven-eleven: it can't work 18:12 < seven-eleven> jack_rip_vim, yeah looks like NM can't do this then 18:12 < seven-eleven> jack_rip_vim, so I'd just use ifconfig/ip to create virtual interfaces eth0:1 etc 18:12 < davr0s> i gather sshkeys are a solution to my scenario, but I can' get it working http://www.linuxproblem.org/art_9.html 18:13 < jack_rip_vim> seven-eleven: yeah 18:14 < davr0s> any attempt to login gives a warnign "the ECDSA host key for ... differs from the key for the IP address ..." 18:15 < jack_rip_vim> davr0s: seem like the key is changed 18:15 < davr0s> i see something saying 'remove the cached..' 18:15 < davr0s> 'ssh-keygen -R ...' 18:16 < jack_rip_vim> davr0s: you can create a new key to replace the old one 18:18 < terra> Guys, when I terminate X session, gvfs-monitor services doesn't get killed. Will they terminated after a timeout reached? 18:19 < jack_rip_vim> terra kill command isn't work? 18:19 < turkeyhand> how do I patch up grub so it includes windows 18:19 < davr0s> differing IP addresses would cause this ? it's caching them per IP adress? i regularly add/remove things from the network 18:19 < turkeyhand> I just reinstalled arch and it didn't find windows10 18:19 < terra> jack_rip_vim But it is not convenient way to terminate them. 18:19 < dgurney> turkeyhand, install os-prober 18:19 < turkeyhand> os prober? 18:19 < turkeyhand> from where 18:20 < dgurney> that's the package name, install it with pacman 18:20 < turkeyhand> yup 18:20 < turkeyhand> then what 18:20 < dgurney> then afterwards you just re-run grub-mkconfig 18:20 < turkeyhand> that's safe? 18:20 < jack_rip_vim> terra: have no idea with this problem 18:20 < dgurney> safe how? 18:21 < turkeyhand> it won't explode my shit, I dunno 18:21 < dgurney> it won't break anything 18:21 < turkeyhand> I just type grub-mkconfig 18:21 < dgurney> yes, you type the same command you used while installing 18:22 < terra> For every X session logoff then login, these gvfs-monitor services are gathering in process list. 18:22 < jack_rip_vim> terra: why you have to kill it? 18:22 < LtL> turkeyhand: os-prober will tell you if it detects another OS, iirc it will also make necessary changes 18:22 < davr0s> everything i try fails 18:23 < turkeyhand> grub-mkconfig seemed to find windows 10 18:23 < dgurney> good :) 18:23 < turkeyhand> I'm not sure if it's automatically made the menu? 18:23 < dgurney> if it mentioned it, it did 18:23 < turkeyhand> oh cool. thanks! 18:23 < LtL> turkeyhand: win, it should have 18:23 < terra> jack_rip_vim: I don't mean kill them.. but they are launch new instances for every X session.. this is the problem 18:23 < davr0s> its not got the cached one anymore but it still makes me give passwords et 18:23 < jack_rip_vim> davr0s: try to restart the dhcpd server 18:23 < turkeyhand> great, I'll try that when I can reset 18:24 < za1b1tsu> Hello, does bleeding edge and rolling release mean the same thing? 18:24 < Frith> No. 18:24 < jack_rip_vim> terra: try to see if systemd can disable it 18:24 < davr0s> jack_rip_vim (at which end, mine or the other machine?) 18:25 < dgurney> bleeding edge would be something like fedora rawhide, with absolute latest, even development versions of everything 18:25 < terra> jack_rip_vim: I actually use them and I don't use systemd either. Thanks btw. 18:25 < jack_rip_vim> davr0s: the machine you want to perform it 18:25 < Frith> za1b1tsu: You can have rolling releases of very stable software, that have gone through extensive testing and soaking prior to rollout. 18:26 < jack_rip_vim> terra: you can try service command, if you are in ubuntu 18:26 < dgurney> opensuse tumbleweed and solus are good examples of more stable rolling release distributions 18:26 < jack_rip_vim> or debian 18:26 < Frith> Bleeding edges tend to be covered by a lot of bandaid solutions. 18:26 < turkeyhand> new arch install is good 18:26 < dgurney> arch is also not as bad as some people would have you think reliability-wise 18:26 < turkeyhand> somehow things are better 18:27 < turkeyhand> how do I upgrade things I install from aur ? 18:27 < turkeyhand> dgurney, I've been using it 4 years (poorly) but had maybe 2 times I had to repair something manually 18:27 < za1b1tsu> @Frith, I see like OpenSuse tumbleweed... so I've read 18:27 < turkeyhand> or downgrade a package 18:29 < DigNicker> turkeyhand: depends on the size of the package, I personally don't like to downgrade a 2GB package. 18:29 < dgurney> my experience with it was similar when I've used it in the past, very little breakage, which was usually caused by me messing around 18:29 < Frith> If starting from the point of "what should I use?", you likely need to have the answer of, "What do I want to do with this system?" and "What do I not want to do with this system?" answered first. If you know what apps you want to run, you can find the distribution that best handles those. 18:29 < turkeyhand> I don't think I have a 2gb package 18:29 < turkeyhand> I mean system breaking ones, like networkmanager being broken 18:30 < turkeyhand> no internet = no point in booting this usually 18:30 < DigNicker> so your package is small then? 18:30 < DigNicker> can't you solve it in another eway? 18:30 < turkeyhand> my package is adequate 18:30 < turkeyhand> I've had only compliments 18:31 < davr0s> ok seems what is going on is i made another sshkey and it's always trying to use that, and that one had a passphrase required 18:31 < DigNicker> the packages tend to grow when you upgrade them 18:31 < davr0s> when i actualy use the passphrase it allows me to login without using the *password* for the target 18:31 < davr0s> so thats a step froward but what I really want to do is to be able to copy some stuff from a bunch of similar machines in a script 18:31 < Tahlwyn> Anyone here have any experience using Matomo (formerly piwik)? 18:32 < jack_rip_vim> davr0s: scp? 18:32 < davr0s> right, but without having to authenticate e.g. for all 8 of them 18:33 < davr0s> if i have different keys in ~/.ssh , can i manually specify one to use in an ssh or scp command? 18:33 < jack_rip_vim> davr0s: setup some server like svn, download the files directly 18:34 < jack_rip_vim> davr0s: -F or -I 18:34 < jack_rip_vim> -i* 18:34 < davr0s> i guess your earlier suggestion of 'a web server' would be appealing in being able to browse it too 18:35 < jack_rip_vim> davr0s: yeah, because you didn't want to auth it 18:35 < jack_rip_vim> davr0s: scp with -F or -i is what you need now 18:36 < davr0s> ok let me try that,thans 18:36 < jack_rip_vim> :) 18:39 < davr0s> jack_rip_vim Thanks! that actually worked. 18:40 < davr0s> thats great i can finally copy from a script now, I have what I needed. 18:43 < jack_rip_vim> davr0s: you are welcome! 18:49 < iopq> How do I get my intel integrated graphics to work while I have a GPU in? 18:49 < dgurney> iopq, see if your bios has an option to enable it 18:50 < iopq> dgurney, it already works - but when I set it to the default screen in the bios, Ubuntu shows a low graphics mode pop-up 18:51 < funksh0n> Hi all. 18:51 < iopq> in other words, there's an issue with drivers 18:52 < funksh0n> I'm wanting to compare the read/write speeds of a USB stick and M.2 to sata to USB. Is dd a suitable tool? 18:52 < dgurney> I think so 18:54 < jack_rip_vim> Gonna go, see your guys later 18:54 < jack_rip_vim> ! 19:00 < tim_the_tool> Hi channel, looking for help re restoring my system after some unmounting of my /home drive I think cause everything has disappeared bar a screensaver. I have access to terminal and programs under sudo 19:02 < koala_man> tim_the_tool: disappeared from where? obviously if you unmount /home, nothing on it will be accessible until it's remounted 19:03 < TheWild> going to use ddrescue someday, maybe tomorrow. I've read 'man ddrescue' quickly (guess, tl;dr). Is that 'ddrescue infile outfile' enough, or I'm encouraged to use some options? 19:04 < TheWild> I don't see anything that could make the recovery better 19:05 < tim_the_tool> @koala_man I don't know how it got unmounted, plus it was encrypted. When I log back in as main user, I ls in the and the old /home data seemed to be stored in a .Private file located in /home/.ecryptfs/user/.Private 19:08 < obiwahn> xpad: Unknown symbol input_ff_create_memless 19:08 < obiwahn> hi i have tried to install the xpad driver and got that message in dmesg 19:08 < obiwahn> https://github.com/paroj/xpad 19:08 < funksh0n> What 'type' does on use for a partition if I want the filesystem to be exfat? Is it 11 (Microsoft basic data)? 19:08 < obiwahn> what can i do to fix it? 19:08 < koala_man> tim_the_tool: ecryptfs is file level encryption so yes. is 'main user' the user whose data you're trying to access? 19:10 < bls> obiwahn: do you have the v4l kernel module loaded yet? 19:10 < turkeyhand> I rebuilt grub, it mentioned windows 10, but it didn't write it 19:10 < bls> obiwahn: scratch that, wrong thing 19:10 < turkeyhand> also I'm using grub 2.02 squiggle something 19:11 < nemesys> o/ 19:11 < bls> obiwahn: looks like that driver might be out of date 19:12 < triceratux> TheWild: dont forget theres both ddrescue & dd_rescue https://askubuntu.com/questions/211578/whats-the-difference-between-ddrescue-gddrescue-and-dd-rescue & documentation is good but always follow up with some googling. no manpage is an island 19:12 < turkeyhand> I installed arch, there's only arch and advanced options in the grub menu 19:12 < tim_the_tool> @koala_man yes main_user wants access back to original setup as before with all the data that is in the decrypted tmp /home file but also the same custom desktop settings/config before this accident happened. 19:12 < turkeyhand> I used os-prober 19:12 < turkeyhand> and grub-mkconfig 19:12 < turkeyhand> and it said it found windows 10 19:13 < koala_man> tim_the_tool: desktop settings/config should be stored in /home with everything else 19:13 < turkeyhand> no entry in the menu though. 19:13 < noodlepie> I;ve just moved to EFI boot rather than BIOS Legacy boot and it works. Had to remove VGA support due to screen freeze but EFI Framebuffer works. I'm dead happy with this. 19:14 < koala_man> tim_the_tool: why is encrypted home not being automatically mounted when you log in as that user? 19:18 < TheWild> ok, thanks triceratux 19:20 < tim_the_tool> @koala_man well yes, you would think so but no. I get following errors on reboot / login as main user. "could not update ICEauthority file /home/user/.ICEauthority", well that's becuase that has disappeared. Also another popup message states that "the path for the directory containing caja settings needs read and write permissions /home/user/.config/caja" 19:20 < mrig> hello 19:20 < TheWild> if it's really failing HDD problem and not the connection then this is the first occurence of failing HDD I own 19:20 < konsolebox> rarr 19:21 < sudo_halt> SMART is your friend. 19:21 < TheWild> yup, I check SMART periodically 19:21 < sudo_halt> Good then. 19:21 < mrig> Having a little trouble booting a debian box, just gives a recovering journal message then 'clean' but seems to freeze after that. 19:22 < saberu> hey guys had a huge attack on my vps yesterday. Looking at auth logs I can see they were trying to break in, can someone tell me how I can find out if they got in? 19:22 < sudo_halt> Only HDDs that i ever failed where ones i threw out of the window. 19:22 < koala_man> tim_the_tool: normally the distro will make sure it's mounted on that user's login. did you change their password, modify the partitions or fstab, dist-upgrade or something? 19:22 < sudo_halt> IF they got in, there should be a backdoor or Unix malware or something. 19:22 < bls> saberu: did you set up anything to detect if they got in? 19:23 < sudo_halt> No hacker comes in witouth leaving a window open. 19:23 < saberu> fail2ban 19:23 < koala_man> sudo_halt: you can check for successful auth attempts, and run rkhunter 19:23 < bls> saberu: that's not going to detect if they succeeded 19:24 < bls> best way is to take the thing offline and analyze the disk via a different system 19:24 < saberu> This looks dodgy as hell 19:24 < konsolebox> mrig: try to force-check the filesystem offline. e.g. fsck.ext4 -f /dev/xyz if your filesystem is ext4 19:24 < Psi-Jack> Strike fail2ban, upgrade to OSSEC. :) 19:24 < saberu> 2018-04-30 18:14:30,449 fail2ban.actions.action: ERROR iptables -D fail2ban-ssh -s 221.194.47.243 -j REJECT --reject-with icmp-port-unreachable returned 100 19:24 < mrig> I have tried running gparted from a live cd to check the disk but that is giving an error :/ 19:24 < TheWild> my dad once had a netbook and its drive failed before even it had a full month... so this "doesn't count" ;) 19:24 < saberu> is that somoene executing an iptables command in my system? 19:24 < Psi-Jack> saberu: No, that's fail2ban failing to run iptables. 19:24 < mrig> konsolebox, I will try running that from a live usb boot 19:25 < koala_man> tim_the_tool: the files should all be fine if you mount ecryptfs manually, but obviously that's not something you want to do each time 19:25 < konsolebox> mrig: so it really has ext4? 19:25 < saberu> what about this? 19:25 < saberu> Apr 30 18:05:37 scrape2 sshd[17657]: reverse mapping checking getaddrinfo for clt-176-120-25-126.z1.netpoint-dc.com [176.120.25.126] failed - POSSIBLE BREAK-IN ATTEMPT! 19:25 < tim_the_tool> @koala_man no nothing bar I was trying to mount a new version 4 of Porteus.iso on a usb and something went all wrong when I umounted the usb drive /media/usb folder. 19:25 < bls> saberu: it is what it says it is 19:25 < saberu> I have a whole bunch of logs with login attempts every 1 second for user root and various random usernames 19:26 < koala_man> that's the internet for you 19:26 < bls> saberu: that's pretty typical and not really going to tell you anything 19:26 < Psi-Jack> saberu: That's a fail2ban log entry that is, by itself, useless. 19:26 < saberu> the hack attempt actually crashed my ssh server 19:26 < Psi-Jack> I highly doubt that.. 19:26 < bls> they probably ddos'd fail2bain 19:26 < tim_the_tool> @koala_man ok just me a moment to see if I can locate the desktop settings as you suggested in a previous post. Thanks 19:26 < saberu> i rebooted and removed public IP so im trying to determine if im safe to reattach networking 19:27 < bls> saberu: then take the system offline and run rkhunter via a different system 19:27 < bls> that's no guarantee, but if this really mattered you'd have an IDS running 19:28 < saberu> it's not like i have business-crtical systems running on it, it's a web server. i could probably risk putting it back on 19:28 < mrig> konsolebox, there is a partition on the disk already, not sure that it is, it is supposed to be but gparted says that there is a an issue with the byte syse 19:28 < mrig> size 19:29 < mrig> sorry block size ... 19:29 < bls> the only risk there is you serving malware to others or becomming a botnet zombie yourself 19:33 < keeposin> i am running an nfs share on arch linux, which is advertised through an avahi service file 19:33 < mrig> konsolebox, I have a 250G SSD that is partitioned, the system is I think on a 30G ext4 partition. 19:33 < keeposin> problem is that vlc on android sees the share, but fails to connect to it 19:34 < bls> didn't they remove the nfs client from android? 19:34 < keeposin> it tries to connect as nfs://address:2049, which fails 19:34 < mrig> but fdisk is also listing sdb1 as a 7.5G fat partition :| 19:34 < keeposin> but manually connecting as just nfs://address, without the port succeeds 19:35 < konsolebox> mrig: that might be a problem on the partition table itself. like the recorded disk size is different from the actual size of the disk. i'm not sure what to comment on that. it's pretty dangerous. 19:35 < keeposin> i am at my wit's end. but caanot find a solution 19:35 < tim_the_tool> @koala_man so was looking through the decrypted older /home contents and directories. I take it that all my old config setup is located in the .config hidden folder? 19:35 < bls> keeposin: you probably need to talk to the vlc channel then 19:35 < keeposin> I talked to them, and to arch liunx 19:35 < keeposin> no result 19:35 < kope> i make my website and i upload my sources in my cpanel , when i open my website it show index of/ and my file sources 19:35 < mrig> konsolebox, sounds like it could be the SSD loosing space? 19:36 < konsolebox> mrig: not sure, but i encountered a problem like that with flash drives 19:36 < keeposin> the last resort is posting on the VLC forum. but I did not want to create a new account 19:37 < toothe> I am grepping the command apparmor_status 19:37 < mrig> konsolebox oh, could deal with that as a flash drive, not so nice on the system disk :| 19:37 < toothe> And making sure that no profiles/commands are in complain mode. 19:37 < milp_2> is it possible to use network manager to dynamically create bonded interfaces in ubuntu? 19:37 < toothe> Is there a way to grep multiple lines to make sure they are 0 ? 19:37 < koala_man> tim_the_tool: depends. a bunch of programs have started storing config there, but many still use other hidden dirs in ~ 19:37 < konsolebox> mrig: can you backup? 19:37 < tim_the_tool> @koala_man so all I have to do is mount the /tmp/ecryptfs.kasdkja back to current /home and reboot? Or is this tool simple? 19:37 < bls> toothe: grep only operates on individual lines 19:38 < mrig> konsolebox, what is the name of the check disk command again, I have forgotten it? 19:38 < toothe> bls: yes....I think I'm going to search for not 0 19:38 < mrig> konsolebox, I have 3 thankfully :) 19:38 < konsolebox> mrig: well if you worry about your files in the system partition, you can temporarily copy the whole partition to another drive. you can use dd. 19:39 < saberu> can i delete the user sync? looks dodgy 19:39 < mrig> oh its fsck 19:39 < tim_the_tool> @koala_man as you stated earlier, will this remount persist? 19:39 < koala_man> tim_the_tool: mounts don't survive reboots 19:39 < mrig> konsolebox ah cool yes that would be an exact copy, would I need to reinstall grub? 19:40 < tim_the_tool> @koala_man so do I just copy over everyting? 19:40 < konsolebox> mrig: probably 19:40 < mrig> I'm going to try fsck direct from the terminal 19:41 < mrig> What is the difference between running fsck on /dev/sda and on /dev/sda1 ? 19:42 < noodlepie> fsck won't run on /dev/sda - only partitions therein. 19:42 < noodlepie> so like /dev/sda1 19:42 < koala_man> tim_the_tool: you should decide on whether you want your new home dir to be encrypted, then you should get the distro to go along with it. if you copy your files to the plain home dir while the distro thinks you want an encrypted home or vice versa, things won't be awesome 19:42 < mrig> ah ok, thank you 19:43 < jim> mrig, /dev/sda is a whole drive, /dev/sda1 is a partition on that drive 19:43 < konsolebox> mrig: also make sure /dev/sda1 is the system partition; not the boot drive if you have any 19:43 < konsolebox> *boot partition 19:44 < mrig> curious, it is giving the message: e2fsck: Get a newer version of e2fsck 19:44 < saberu> omg someone really has hacked into my server look at this - postfix/smtpd[2226]: connect from calfreight.plus.com[81.174.163.117] 19:44 < saberu> they accessed root via postfix 19:44 < mrig> /dev/sda1 has unsupported features 19:45 < milp_2> sorry, ill just ask again: is it possible to use network manager to dynamically create bonded interfaces in ubuntu? 19:45 < saberu> and this - CRON[2323]: (root) CMD ( [ -x /usr/lib/php5/maxlifetime ] && [ -x /usr/lib/php5/sessionclean ] && [ -d /var/lib/php5 ] && /usr/lib/php5/sessionclean /var/lib/php5 $(/usr/lib/php5/maxlifetime)) 19:45 < tim_the_tool> @koala_man ok look I have backup anyway of older /home data, I will try and copy over the decrypted info and see what happens. Thanks for help and time ;) 19:45 < koala_man> saberu: just sounds like someone trying to use you as a spam relay. 19:46 < konsolebox> mrig: maybe your version of fsck[.ext4] is too old 19:46 < koala_man> tim_the_tool: home it works out 19:46 < milp_2> saberu: did you check crontab if that cronjob exists somehow? 19:46 < saberu> milp_2, yes it doesn't exist on the root user, the crontab was empty maybe they put it on the postfix user 19:46 < tim_the_tool> @koala_man #metoo :) 19:47 < milp_2> saberu: but doesnt it says root in that log message? 19:47 < jim> mrig, some people have chosen to put their filesystems on the whole drive (so they did mkfs /dev/sda, so they would run fsck on that filesystem by running fsck /dev/sda), and others have chosen to put their filesystem on a partition of the drive, like this: mkfs /dev/sda1, so in that case they would fsck just the partition, by running fsck /dev/sda1 19:47 < saberu> they managed to gain root access through the postfix user somehow 19:48 < jim> mrig, notice the pattern: it's where the filesystem has been written 19:48 < konsolebox> saberu: does postfix run as root? 19:48 < koala_man> saberu: how did you determine this? nothing you've posted so far is suspicious 19:48 < milp_2> saberu what does that php job do though? im not familiar with php but is it possible that this is some kind of routine maintenance job? 19:49 < mrig> jim, hi, ah ok, yes this is on a partition. One curious thing is that my partitioned disk is wrapped entirely by another partition, I suppose that is the live os on the usb key that is doing that in order to read the system disk? 19:49 < saberu> they already locked me out of ssh, ill have to VNC in and see whats happening 19:49 < bls> you really need to know what should be installed on the server, what shouldn't be, and what's normal vs abnormal logging. just picking random things you don't understand and offering them as proof you've been hacked isn't going to get you any close to being back online 19:49 < mrig> running apt upgrade on the usb OS :/ 19:51 < konsolebox> mrig: you only have to upgrade the package that contains fsck.ext4. perhaps it's something like e2fsprogs. 19:51 < saberu> well that PHP command is dodgy because I'm running PHP7 19:51 < saberu> php5 should not be there 19:51 < turkeyhand> dgurney, this grub-mkconfig doesn't seem to write it 19:52 < dgurney> well it should 19:52 < mrig> konsolebox, that would be a much more reasonable approach, now I better stop this thing before it fills the entire USB drive! 19:52 < saberu> anyway it's obvious they are getting in through the mail proc so ill just remove it 19:52 < dgurney> and I don't know why it isn't doing it for you 19:52 < bls> no, it's not obvious 19:53 < turkeyhand> maybe I should manually put it in there? 19:53 < turkeyhand> can I paste the output 19:53 < turkeyhand> somewhere 19:54 < turkeyhand> is there a preferred paste service for this channel 19:54 < Dominian> check the /topic 19:54 < jim> mrig, oh? is that disk sda? if so could you do: (lsblk /dev/sda ; echo =========== ; fdisk -l /dev/sda) | nc termbin.com 9999 19:55 < mrig> jim yes it is sda6 19:56 < mrig> konsolebox oh dear, it says that I have the latest version ... 19:56 < turkeyhand> https://paste.linux.community/view/b993fe96 19:56 < turkeyhand> what's this 19:56 < turkeyhand> it was the last thing it did 19:56 < ayecee> a url 19:57 < ayecee> pointing to a script 19:57 < konsolebox> mrig: try jim's suggestion 19:57 < mrig> What does that do jim? 19:58 < jim> pastebins a report consisting of running two commands 19:58 < mrig> ok cool, I will run that and post a link 19:58 < jim> so, it will return a url, give us that :) 19:59 < mrig> gret thanks 19:59 < saberu> why is it when i try to remove php5-fpm then type apt-get -f install it prompts me to remove php7 and INSTALL php5? 19:59 < jim> welcome 20:00 < jim> saberu, is the php7 version not for your version of the os? 20:00 < bls> likely because you've installed something else that requires it 20:04 < triceratux> https://www.phoronix.com/scan.php?page=news_item&px=LLVM-Rafael-Espindola 20:05 < cthulchu_> hi folks, I know it's probably not good to ask mac-related questions here, but I'm having troubles doing trivial stuff in its console. Using local console, when I login with my user and do echo $HOME, I get /var/empty 20:05 < jim> mrig, is it in the middle of the fsck right now? 20:06 < cthulchu_> therefore, I don't think I can set this user .bash_profile 20:06 < iopq> any idea why putting in a dedicated card makes my integrated intel graphics not be detected by Ubuntu 16.04? 20:06 < bls> cthulchu_: this isn't a mac support channel, try #macosx 20:06 < saberu> oh wow they enabled permit root login without password in sshd_config 20:08 < mrig> jim, just decomposing the command, was not sure how to use the brackets there ... 20:08 < mrig> http://termbin.com/n80y 20:10 < jim> mrig, oh... (...) groups commands by running a "subshell" and running the commands in it, in that subshell... then, the output of that subshell gets fed to nc 20:10 < roxlu_> When I use epollo with EPOLLET and non-blocking sockets how I am supposed to write data on a socket? 20:10 < saberu> they also added a few SSH keys to the sshd_config which meant i couldnt login to ssh without their keys 20:11 < konsolebox> mrig: your drive is 240G? 20:11 < mrig> jim oh nice, must look into that. 20:11 < mrig> konsolebox well it is 250G so it would actually be a little less than that I guess. 20:13 < jim> mrig, ok, notice (if you look at the output yourself) that sda2 is an "extended" partition, which is windows-speak for a partition that contains the storage for other partitions... sda5 is one of the partitions whose storage is in sda2... sda1 is its own partition 20:13 < plexigras> should i just put setxkbmap in my .xinitrc file if i want a specific keyboard layout for a specific user? 20:13 < konsolebox> mrig: can you also run this? dmesg | grep ' sd ' | nc termbin.com 9999 20:14 < jim> that, you might have to run as roto 20:14 < jim> root 20:14 < bls> plexigras: that's what I've always done 20:15 < konsolebox> or: sudo dmesg | grep ' sd ' | nc termbin.com 9999 20:15 < konsolebox> but i doubt it's needed since it's a live system 20:15 < saberu> guys when i do 'su -' to try to ssh into root i get auth error, what could be wrong? 20:16 < jim> saberu, wrong password probably... maybe caps lock is on? 20:16 < saberu> no i already reset the password with sudo passwd root 20:17 < bls> su - has nothing to do with sshing into root 20:17 < saberu> yeh it uses sudo to access root user shell 20:17 < saberu> so my users password should have worked 20:17 < saberu> but it didnt cos the hacker has done something to root 20:17 < mrig> konsolebox: http://termbin.com/70ck 20:17 < plexigras> ok thanks bls i will just do that then 20:17 < bls> su doesn't use your user's password 20:18 < saberu> what password would it use? 20:18 < bls> root's 20:18 < jim> saberu, su would use root's, sudo would use yours 20:18 < saberu> ah ok works now i just reset root pass again 20:18 < konsolebox> mrig: all reports say 240G 20:18 < konsolebox> anyway i asked it just to confirm that your drive is not shrinking in size 20:19 < konsolebox> so the problem might as well just be in the filesystem itself 20:19 < mrig> Ah ok, will need to look at the dik its self to see what it was sold as ... 20:19 < mrig> sorry my bad, that is 240G 20:19 < konsolebox> now as for the error, are you sure your filesystem is ext4? 20:20 < mrig> konsolebox, could be that I have forgotten write the file system type when formatting initially. I did have a problem with this on a drive before 20:21 < shinka> I'm trying to print a line with variable without resolving the variables, I have "echo "export PATH=$HOME/.cargo/bin:$PATH", I want to keep $HOME and $PATH as is, I don't want these variables resolved. How can I do it? I try googling it but I probably don't use the right terms because I cannot find anything. 20:21 < bls> shinka: escape the '$s' 20:21 < uplime> or put it in single quotes 20:22 < bls> shinka: or use 's instead of "s 20:22 < shinka> oh dang.. Single quote works... Thanks! 20:22 < bls> echo "export PATH=\$HOME/.cargo/bin:\$PATH" or echo 'export PATH=$HOME/.cargo/bin:$PATH' 20:22 < mrig> konsolebox disks reports the 30G system partition as Ext4 20:22 < bls> shinka: https://mywiki.wooledge.org/Quotes 20:23 < mrig> But is shows an unknown partition for sda2 20:24 < mrig> curious that it has different sda values ? 20:27 < konsolebox> mrig: have you tried mounting the system partition again? 20:27 < konsolebox> mrig: sda2 is your extended partition, which contains the logical partitions sda5 and sda6 20:27 < mrig> konsolebox I could try dd'ing the system partition to the storage disk reformatting and them dd'ing it bask perhaps? 20:27 < mrig> ah ok, yes I see 20:28 < konsolebox> mrig: that wouldn't fix the data in the system partition itself 20:28 < mrig> Yes seems to have mounted ok 20:29 < konsolebox> your partition table looks fine as well, so i don't think that would help. but it's not bad to do a backup 20:29 < mrig> ok, so same problem would be copied back and forth, that would be pretty annoying 20:29 < konsolebox> mrig: and have you tried checking if you can access the files normally? 20:30 < mrig> konsolebox, just in the file system now, yes seems to be fine using the live OS 20:31 < mrig> oh so something has screwed up the file system records? 20:32 < konsolebox> mrig: probably. maybe you failed to unmount it properly, or it's a subtle disc failure 20:32 < konsolebox> mrig: do you see /etc/fstab in that partition? 20:33 < mrig> yes that is here 20:33 < konsolebox> ok i guess it is the system partition indeed 20:34 < konsolebox> now all that's left to wonder is that why fsck reported unsupported features from your partition 20:34 < saberu> is it normal for a cron job to backup passwd and shadow file to a directory called /var/backups? to .bak files? 20:34 < lesshaste> when I run ifconfig I see docker0 and enp2s0 where I was expecting to just see eth0. What are those? 20:35 < mrig> Yes that is curious, it has worked well on other backup data disks in the past, strange that it does not work here 20:36 < mrig> It says: errors=remount-ro 0 1 next to sda1 in fstab, is that normal? 20:37 < konsolebox> yes 20:37 < konsolebox> it means to remount the filesystem read-only if it fails 20:37 < konsolebox> or an error is detected 20:38 < mrig> ah ok 20:39 < konsolebox> mrig: the latest version of e2fsprogs is 1.44.1. is that your current version? 20:40 < mrig> no it is not, this is an ubuntu live cd, I guess that it does not have the testing repo or something like that 20:40 < mrig> 1.42.13 here 20:41 < mrig> says the metadata_csum is not correct, how do I get the newer version? 20:42 < obiwahn> Is there a channel for people that write / debug modules? 20:42 < konsolebox> mrig: maybe you can ask #ubuntu. it's been a while since i used that distro 20:43 < uplime> obiwahn: kernel modules? 20:43 < obiwahn> yes 20:43 < mrig> yes I am actually using debian now, but had the ubuntu live usb to hand as mine is the net install of debian. Ok thanks I will ask over there. 20:46 < mrig> ok, I have the latest version, I will install and try running that :) 20:50 < konsolebox> mrig: make sure you unmount the filesystem first 20:50 < mrig> ok, will do 20:54 < mrig> Ok, that ran fine, says clean, then gives the file count and the blocks ... 20:55 < mrig> Could perhaps be a hardware failure elsewhere on my system? 20:55 < mrig> old motherboard and it does play up sometimes 20:57 < mrig> ah yes a different message on trying to boot this time: ata2: SError: { DevExch } 21:00 < eqw> i mounted vfat with default iocharset(iso-8859-15) and now when i mount with iocharset=utf-8 filenames are like глÑпоÑÑÐ. How to rename them? 21:04 < bls> eqw: mv 21:05 < SmashingX> how can I make sure a file is writeable by php if php is not a user or group 21:05 < bls> it's not possible to execute without a user or group 21:05 < Psi-Jack> eqw: Perhaps mount the filesystem with the correct charset, and not the wrong one? 21:06 < eqw> Psi-Jack: utf-8 is the correct charset. The only portable one. 21:07 < mrig> Thanks for your help konsolebox, jim, trying to boot into recovery mode, will try over at #debian see if they recognise any of the symptoms. 21:07 < Psi-Jack> So, why did you mount it with iso-8859-15? 21:08 < eqw> i just didn't know that i should specify iocharset in mount options 21:08 < bls> is utf-8 vfat portable? I thought windows needed cp437 or utf-16 21:08 < dgurney> I doubt it has such limitations today 21:09 < Psi-Jack> NOTE: "iocharset=utf8" is not recommended. If unsure, you should consider the following option instead. 21:09 < eqw> bls: no idea. I don't have windows, I have ubuntu and android. 21:09 < Psi-Jack> https://www.kernel.org/doc/Documentation/filesystems/vfat.txt 21:10 < eqw> Psi-Jack this doesn't solve the problem of existing files. 21:11 < konsolebox> mrig: besides disc problem, it could also be memory error 21:12 < nfshr_> Hi all, I'm looking for a way to remove every zip-file that's older than 5 days in the subdir 'log' of any subdir of 'one'. Now find does not allow wildcards as in /one/*/log/. Can anyone think of a quick way to go about this maybe? 21:13 < bls> nfshr_: find does allow wildcards like that 21:14 < nfshr_> bls: does it? When I issue it w/ the wildcard, I get 'No such file or directory'... 21:14 < mrig> konsolebox, ah ok, will try swapping those about or removing some 21:14 < bls> nfshr_: "issue it"? 21:14 < noodlepie> nfshr man find specifically -atime option 21:14 < nfshr_> bls: fire the command, yeah 21:15 < bls> nfshr_: :| 21:15 < nfshr_> noodlepie: thx, I know about the time options 21:15 < Psi-Jack> eqw: Likely, easier to mount it in the former iocharset, move the file(s) off in their correct names, and remount with the appropriate actual correct iocharset, (which iocharset=utf8 is not recommended as the vfat documentation states), and move them back. 21:15 < nfshr_> bls: what? 21:16 < eqw> Psi-Jack: i don't have such a huge hdd at my disposal now... 21:16 < konsolebox> mrig: you can try memory tests like memtest86 or memtest86+. i think some live cd's/usb's have that. you can select them from boot through grub. 21:16 < Psi-Jack> That is... unfortunate. 21:16 < bls> nfshr_: it's a lot easier to get help if you can provide the actual commands you're using and your expected and actual outputs 21:17 < noodlepie> OK 21:20 < mrig> konsolebox, I will look at that to see if there are any bad areas, will be an interesting exercise. The machine has just booted, some fellow over in #debian suggested that I clumsily thump at all the keys to generate entropy whilst booting, seems to have worked :D 21:21 < mrig> Like starting my old motorbike :D 21:21 < bls> nfshr_: or check out https://mywiki.wooledge.org/UsingFind and read up on matching entire paths 21:21 < ArsenArsen> time to do a random LFS install 21:22 < Psi-Jack> That's... random. 21:22 < Psi-Jack> Going to go through various versions of the book and randomly pick parts to follow randomly? 21:22 < konsolebox> mrig: i see. that's nice. 21:22 < ArsenArsen> Psi-Jack, no, just ditch the book altogether 21:22 < ayecee> better yet, tear out and discard random pages before starting 21:23 < ArsenArsen> I'll discard all of it 21:23 < Psi-Jack> ArsenArsen: That's not very random... 21:23 < nfshr_> bls: ok, thanks for your help. It had to do with me issuing it with sudo instead of root. Sudo seemingly did not read the content of the folder-tree correctly 21:23 < macwinner> I have a cluster of linux servers that each have their own server keys for SSH. I want to run SFTP on them, but I want the sftp servers to have the same server key so the SFTP client doesn't warn if a new server comes into the cluster. is it possible to have the SFTP service have different keys than the SSH service? 21:23 < ayecee> macwinner: nope 21:23 < nfshr_> bls: using the wilcard that is 21:23 < ArsenArsen> Psi-Jack, yeah, but just sitting down one day and going 'Oh I know, let's install LFS without the book' is 21:24 < bls> nfshr_: likely to do with escaping the quoting or the current directory more than sudo 21:24 < mrig> konsolebox, rather curious, I know that the kernel has an entropy pool and requirement and I do tend to just stream video on this machine and not use the keyboard much, could be that it was causing this ... Scratching my head on this one though. 21:25 < nfshr_> bls: the commands I issued were the same in both cases. The paths were absolute 21:25 < bls> so you weren't using wildcards or quotes anywhere? 21:25 < nfshr_> I was 21:25 < nfshr_> in both cases 21:26 < nfshr_> sudo find xyz vs. sudo su -; find xyz 21:26 < turkeyhand> if I drop a copy of the home folder into a new arch install, will all my old software and shit be there? 21:26 < turkeyhand> and the gnome setup and everything else? 21:26 < bls> turkeyhand: no 21:26 < turkeyhand> ok 21:26 < turkeyhand> I really just need my cv which was written in libre writer? 21:26 < bls> turkeyhand: your home will just have your files and settings, not the actual programs 21:27 < turkeyhand> that's ok I moved the files, how can I recover what I was writing in word 21:27 < turkeyhand> the libre writer recovery stuff 21:27 < bls> install writer again and open the file 21:27 < turkeyhand> what file 21:27 < ayecee> to paraphrase a recent ad, what's a file? 21:28 < bls> you entered your entire cv into libre writer without ever saving it then switched distros and expect it to still be there? 21:28 < curiousx> Yo! 21:28 < curiousx> Could you please help me installin' Arch on a la UEFI ? 21:28 < turkeyhand> more like a colossal fuckup forced me to reinstall 21:28 < turkeyhand> something I didn't do 21:28 < ayecee> mind the language please 21:28 < turkeyhand> ok 21:29 < bls> if you saved the file, it'll be in the same location in your $HOME as you left it. if not, it's likely gone 21:29 < turkeyhand> where does it save stuff to recover 21:30 < konsolebox> mrig: i find that too unlikely, but then again i'm not really knowledgeable when it comes to entropy stuff and how userspace applications would need such 21:31 < bls> turkeyhand: I'd recommend asking in the libre office channel 21:32 < Psi-Jack> curiousx: That's documented in the Arch Wiki for installing. 21:33 < mrig> konsolebox, yes me too, just a thought, I still have so much to learn about linux and programming in general; Getting there slowly. 21:34 < Psi-Jack> curiousx: Also, cross-posting the same question in multiple channels, is frowned upon. (I see you in #archlinux) 21:34 < curiousx> Indeed!!! 21:34 < curiousx> Psi-Jack: already watching a youtube vid :p 21:35 < Psi-Jack> Arch Wiki doesn't refer to youtube. 21:36 < curiousx> I have installed Arch a la mbr, a la gpt (non efi), now ima install it a la EFI on a UEFI system :D 21:36 < Psi-Jack> You don't seem to understand what "a la *" is... 21:36 < curiousx> Trust me, i do 21:37 < phinxy> since movie players like mpv can draw to framebuffer, how come other programs like FireFox requires X? 21:37 < Psi-Jack> phinxy: Because. 21:37 < bls> phinxy: because firefox uses gtk to do it drawing, and gtk targets X instead of the FB 21:37 < phinxy> There isnt anything wrong with X is there? 21:37 < curiousx> there are lot of ncurses browsers tho 21:37 < dgurney> no 21:38 < bls> phinxy: depends on who you ask 21:38 < dgurney> but they aren't that useful for regular use 21:38 < curiousx> there are lot of ncurses web* browsers tho 21:38 < Psi-Jack> curiousx: "though" not "tho" 21:38 < curiousx> Ok teacher 21:38 < qman__> X certainly has shortcomings, but it's been around for decades and we still use it 21:39 < Psi-Jack> curiousx: Not a teacher. Just correcting you. SMS-speak is frowned upon here, as mentioned in the channel rules. 21:39 < dgurney> I'll continue using X for a long time, I think 21:40 < qman__> If and when something drastically better than X comes along, people will use that instead 21:40 < curiousx> are you a mod Psi-Jack ? 21:40 < Psi-Jack> Ops are mentioned on the channel website as well, see /topic 21:41 < curiousx> Ok. then please let the ops do their job, and if you won't help, please stfu 21:41 < ayecee> curiousx: don't be rude, please 21:41 < curiousx> yeah! i know, sry 21:41 < Psi-Jack> "sorry" not "sry" 21:41 < phinxy> "no offence" 21:42 < curiousx> i said i'm sorry :'( 21:43 < phinxy> freaking birds chirping from 4 in the morning until well past my bedtime 21:43 < ayecee> http://i.imgur.com/WTDnnwE.gif 21:45 < phinxy> If gcc, or the linker?, complains about undefined reference XGetWindowAttribute and the .h and .so which has that function is included what do I do? 21:46 < ayecee> what is the error message 21:47 < crng_init> curiousx: 'I' not 'i' 21:48 < phinxy> termbin.com/xsag is the program or the relevant makefile 21:48 < kraftb> hi 21:48 < ayecee> ... 21:48 < ayecee> phinxy: what is the error message 21:50 < phinxy> ayecee? tried to pipe the 'make' but got this instead of the errors.. do I need to redirect stderr? http://termbin.com/l4pn 21:50 < ayecee> probably yes 21:52 < phinxy> http://termbin.com/bcn9 21:53 < ayecee> which library is supposed to have these functions? 21:53 < bls> which library supposedly provides XGetWindowAttributes, XGetTransientForHint, XMoveResizeWindow, etc? 21:53 < bls> and am I seeing correctly that you're trying to link with: -l:libX11.so ? 21:54 < ayecee> yeah, that doesn't look right 21:54 < ayecee> probably doesn't break anything because -lX11 is right before it 21:55 < phinxy> My guess right now is that the architecture (arm64) isnt supported 21:56 < bls> that really shouldn't have anything to do with this, unless you're trying to use libraries from another architecture 21:56 < phinxy> Wouldnt the XGetWindowAttributed etc be in Xlib.h? 21:56 < bls> don't know, don't program in that layer 21:57 < ayecee> phinxy: the header might be, but the function is in a .so 21:57 < ayecee> and it's the function that the linker can't find 22:06 < phinxy> xorg-dev sounds like a good candidate for XOpenDisplay() 22:07 < toothe> What is the difference between stat(1) and ls -l ? 22:07 < bls> toothe: they're two different tools that do different things, with some overlap 22:07 * Dan39 rages about the trend of light-grey text on white backgrounds. wtf happen to at least a minimum standard contrast!? 22:07 < phinxy> is dir a third option? 22:08 < toothe> dir > * 22:08 < bls> dir is a DOS command isn't it? 22:08 < toothe> no, its a universal thing. DOS > Linux!!! 22:08 < Dan39> some distros have a dir command iirc 22:09 < bls> ah, looks like GNU added it 22:09 < Dan39> toothe: i find stat useful sometimes in scripts 22:10 < toothe> Dan39: How so? Because I'm writing a script. 22:10 < bls> yeah, ls is really for human-only interaction. stat provides way more control and formatting for proper scripting 22:10 < Dan39> like to get a file's size 22:10 < toothe> Dan39: I find grep/awk more involved than with ls -l 22:10 < bls> you should not be parsing ls -l output 22:11 < toothe> Dan39: This STIG is saying check to make sure that /etc/issue is Uid/Gid of 0 and 644 22:11 < toothe> oh? 22:11 < bls> and you don't need to use grep or awk with stat 22:11 < toothe> Do tell? I'm brand new to this command. 22:12 < Dan39> o_O 22:13 < toothe> ahh, the --format flag? let me give this a shot. 22:14 < uplime> toothe: /msg greybot !ls 22:14 < Dan39> maybe like... myfilesize=$(stat -c %s "$myfile") 22:14 < bls> https://mywiki.wooledge.org/ParsingLs and $ stat -f '%u' ini stat -f '%g' ini 22:14 < za1b1tsu> I want to automate the setup that I do after a fresh install (package installs, databases, git cloning, configs) etc. Is it possible to do this with bash scripting? Is it something that power users usually do? 22:14 < bls> za1b1tsu: it's pretty ugly to do something like that in bash. most people use a config management tool for that 22:14 < djph> sounds like something you'd use ... oh what's it called ... ansible(?) 22:15 < uplime> ansible, chef, puppet, etc 22:15 < toothe> cool, that works! 22:15 < cxc99> anyone know why hostnamectl wouldn't be running on boot? i keep getting Failed to issue method call: Did not receive a reply. 22:15 < bls> za1b1tsu: check out ansible, chef, puppet, salt, etc 22:15 < uplime> or roll an image with all of that stuff baked in 22:15 < uplime> but config mangement is sexier 22:15 < bls> "golden images" are a massive pain in the ass 22:15 < za1b1tsu> oh nice, thank you 22:15 < cxc99> i noticed a bug filed and the reply was not to put it in /etc/rc.local, so i put it in a systemd unit 22:15 < djph> cxc99: because it probably has a dependency on something else; and whatever your boot script is runs too early. 22:16 < Dan39> and just in case, add the -- like bashguide says... myfilesize=$(stat -c %s -- "$myfile") 22:16 < Dan39> i always forget about that :| 22:17 < cxc99> funny thing is it will set with nmcli 22:18 < cxc99> i put After network-online.target in the systemd file 22:19 < za1b1tsu> I see that configuration management tools seem like tools used for serious server setups. Do people use these on their home computers? 22:19 < ayecee> rarely 22:20 < GNU\colossus> there's always going to be outliers ;) 22:20 < bls> za1b1tsu: yes 22:21 < bls> za1b1tsu: I have ansible roles for my personal laptop, firewall, rpi, build box 22:22 < toothe> I'm not that familiar with gdm. Anyone know what /etc/dconf/profile/gdm is? 22:22 < toothe> I am STIG'ing a machine and saw that I need to create that file and put in some basic settings. 22:22 < za1b1tsu> bls: The Self-Support offering starts at $5,000 per year, and the Premium version goes for $14,000 per year for 100 nodes each. Regarding ansible... 22:22 < toothe> But I didn't follow what it was. 22:22 < toothe> and there's no man page for 'gdm'. 22:22 < bls> za1b1tsu: if you want support. if not, you can use the free version. think redhat vs centos 22:23 < bls> toothe: gdm is Gnome's login/desktop manager 22:23 < toothe> ahh 22:23 < toothe> does it have a man page? 22:23 < za1b1tsu> bls: how would you describe the learning curve? 22:23 < cxc99> za1b1tsu, is that for tower? or just ansible? 22:23 < za1b1tsu> ansible 22:23 < cxc99> ansible will take u like a few hours to learn 22:24 < bls> za1b1tsu: pretty low 22:24 < cxc99> versus puppet.....stick with ansible ;) 22:24 < uplime> who manages https://freenode.linux.community/? 22:24 < Johnjay> is a text interface like midnight commander or emacs with -nw technically a "GUI"? 22:25 < uplime> Johnjay: its usually referred to as a terminal user interface 22:25 < RayTracer> Johnjay: no, its TUI 22:25 < Johnjay> ok. i was saying i don't care about GUIs but then I thought well what if you had a gui from the command line like that 22:25 < Johnjay> I use those and appreciate them 22:26 < bls> za1b1tsu: this is a portion of my setup https://paste.linux.community/view/474bbc79 22:26 < neoncortex> I find funny that we still did not manage to create a more interactive and unified tool as command line 22:26 < uplime> neoncortex: hrm? 22:26 < Johnjay> like midnight commander? 22:26 < Johnjay> you can install that from the repos i think in ubuntu 22:27 < neoncortex> uplime: yes, we have beautiful interfaces, but generally they lack power, one application does not interact with the others like in the shell 22:27 < RayTracer> neoncortex: the command line usually provides you with a 104 key interface on the pc side and 10 fingers at your side. beat that with a mouse. 22:27 < cxc99> with ansible you can get started very quickly 22:27 < uplime> neoncortex: wat 22:28 < za1b1tsu> cx99, bls thank your for the information 22:28 < neoncortex> uplime: at this point I'll just stop =D 22:28 < bls> neoncortex: have you ever tried to use a tool to automate inter-gui process communication? they end up very limited in ability 22:28 < uplime> non command line applications can easily interact with each other 22:28 < uplime> sockets, pipes, etc 22:29 < neoncortex> bls: to be honest I never did, because I see no point in it since I have the shell 22:29 < triceratux> toothe: https://linux.die.net/man/1/gdm ? 22:29 < iopq> so I have a problem, I want to use two monitors, but my GPU only has one DVI output, what can I do? 22:30 < neoncortex> uplime: yes, not not simple and elegant as text input/output and a pipe 22:30 < uplime> neoncortex: its pretty simple 22:30 < hippy> iopq> get a new video card 22:31 < bls> you can make GUI applications communicate with one another, but they have to be designed specifically to use your chosen IPC method and agree on a data interchange format. no one has been able to standardize that in a useful way 22:32 < iopq> hippy video cards got more expensive, so I'm probably going to get one, but maybe I'll wait until the prices come down (will they?) - but let's say we don't want to spend money just yet 22:32 < xamithan> You are going to wait until cryptocurrency completely dies out ? 22:32 < RayTracer> bls: I think dbus would be the current attempt 22:32 < xamithan> Might be waiting a while mate 22:33 < iopq> xamithan, I hope not, I own some 22:33 < bls> RayTracer: for *nix, yeah, it's the closest thing we've got to say powershell on Windows or Automater on mac 22:33 < Psi-Jack> Personally... This "cryptocurrency" stuff, needs to hurry up and fall off the planet. 22:34 < xamithan> If they really wanted to they could lock down the GPUs. But they don't have any incentive to as profits keep going up 22:34 < meyou_> it's not a requirement of cryptocurrencies that mining require GPUs and a giant wattage vacuum 22:34 < hipp> iopq> video cards come in a wide range of prices and capabilities ... do some research and find one that works for you 22:34 < d1z> lol 22:34 < d1z> zip can't compress files from stdin 22:35 < Psi-Jack> Why even zip? :p 22:35 < koala_man> gzip is the same algorithm but in a unix friendly package 22:35 < iopq> hipp, yes, at a certain point I will buy a GPU. But I tried using onboard + GPU and onboard doesn't get detected by Ubuntu while it works just fine on the load-out screen (when I hit reboot) 22:35 < Psi-Jack> No, gzip is not the same algorithm. 22:36 < koala_man> what's the difference? 22:36 < iopq> the g 22:36 < xamithan> Hah 22:37 < xamithan> zip does both archive and compression, gzip only does compression and it does it a whole lot better 22:37 < koala_man> pkzip and zlib/gzip are both listed as encoder implementations of DEFLATE on wikipedia 22:37 < meyou_> xamithan, google tells me they both use DEFLATE, so idk how it could be a whole lot better 22:37 < xamithan> That is the simple answer. Also zip is more geared towards windows users 22:38 < koala_man> xamithan: better how? 22:38 < xamithan> zip compresses the files one at a time 22:38 < RayTracer> zip is an archiver/compressor while gzip is a compressor and needs eg. tar as archiver 22:38 < Psi-Jack> They both share the LZ77 algorithm, that I will agree. But yeah, zip is both a container and compression, while gzip is just compression. 22:39 < xamithan> gzip does it as one large file so it works better for the redundancy to reduce the size 22:39 < xamithan> Kind of like how dedup works 22:39 < meyou_> O_o 22:40 < meyou_> wouldn't that be sort of the opposite? gzip compressed individual files, zip puts them in a container and compresses the whole thing? 22:40 < koala_man> xamithan: I don't think gzip even compresses a single file as a single file. it can't do the huge cross file leaps that modern algorithsm do 22:40 < iopq> all right I'm going to install some weird experimental shit and see if my onboard works (spoilers: probably won't) 22:41 < xamithan> No, zip does them individually. Thats why you can extract individual files easily as compared to gzip which can't 22:41 < Psi-Jack> True. 22:41 < Psi-Jack> As with zip you can see the compression ratio of /each file/. 22:41 < seven-eleven> can i use consumers routers that ISPs hand out to customers also just as a switch? i tried it and it seems to work as a switch, are there any caveats though? 22:41 < koala_man> this only applies for tiny files that fit in the same windows 22:42 < mattfly> hi 22:43 < mattfly> is there any port rediector that works easy and intuitive on CLI like a ncurses based thing r something based on socat? 22:43 < mattfly> im not lucky with iptables 22:43 < Psi-Jack> iptables. 22:43 < Dagmar> There's no such thing as "luck" with iptables 22:43 < mattfly> everything i try with it fails... 22:43 < Dagmar> Either you read the documentation and learn the simple commands, or your stuff doesn't work. 22:44 < xamithan> Well what are you trying, did you read the guides on netfilter.org ? 22:44 < mattfly> I want to redirect port 2222 to port 22 on my nainer 22:44 < Psi-Jack> iptables. 22:44 < mattfly> my lxc container* 22:44 < Dagmar> A google search would reveal how to do that 22:44 < mattfly> i have 22:44 < mattfly> -A PREROUTING -i eth0:0 -p tcp -m tcp --dport 2222 -j DNAT --to-destination 172.16.0.110:22 22:44 < mattfly> doest work 22:44 < Psi-Jack> ethX:Y is the wrong way to do IP aliasing these days with iproute2. 22:44 < Dagmar> Also probably the wrong interface 22:45 < mattfly> can i place my ip there or what? 22:45 < neoncortex> After all that years I did not memorize the iptables syntax 22:45 < Psi-Jack> neoncortex: Who does? LOL 22:45 < xamithan> https://www.netfilter.org/documentation/HOWTO//NAT-HOWTO.txt is the best go-to guide I've found 22:45 < Dagmar> You put the rule to match the interface the packet will come in on 22:45 < mattfly> its impossible to memorize this 22:45 < mattfly> eth0:0 is the interface 22:45 < neoncortex> Psi-Jack: Well, I think we supposed to =D 22:45 < Psi-Jack> neoncortex: Nah. I use wrappers for it primarily. :) 22:45 < mattfly> i also have -A PREROUTING -i eth0:0 -p tcp -m tcp --dport 2222 -j DNAT --to-destination 172.16.0.110:22 22:45 < Dagmar> Impossible? Pfft. 22:45 < Psi-Jack> shorewall, firehol, proxmox-firewall, etc. :) 22:46 < xamithan> firewalld? ^_^ 22:46 < Psi-Jack> No. 22:46 < mattfly> -A PREROUTING -i eth0 -p tcp -m tcp --dport 2222 -j DNAT --to-destination 172.16.0.110:22 *** 22:46 < Psi-Jack> Configuration files with XML files is stupid. 22:46 < Psi-Jack> ^ firewalld uses such xml files. 22:46 < xamithan> Least it doesn't use YAML like networkmanager 22:46 < Psi-Jack> XML for computers. YAML for humans. 22:47 < toothe> gah! sed is so difficult to use. 22:47 < Dagmar> XML isn't that hard to write by hand 22:47 < toothe> rather, sed with regex 22:47 < toothe> I'll come back to this damn sig. 22:47 < toothe> stig* 22:47 < Dagmar> If it was too hard for humans to write, _we'd have never had the world wide web_ 22:47 < Psi-Jack> XML isn't hard to write. just hard as hell to read. :p 22:48 < xamithan> xml reminds me too much of excel 22:48 < Psi-Jack> How?!? 22:48 < mattfly> socat is much eaiser than iptables 22:49 < mattfly> i think i will add a crontab to do the socat work on every boot 22:49 < Dagmar> So is tic-tac-toe 22:49 < xamithan> Because all those data crunchers pull xml into it 22:49 < Psi-Jack> mattfly: cron isn't a service manager. 22:49 < xamithan> Then ask you how to use it, and you have no idea 22:49 < dgurney> I think xlsx (and other modern MS office formats) does use XML internally 22:49 < mattfly> what would be better? 22:49 < Psi-Jack> mattfly: Curious, what distro, and version? 22:49 < mattfly> ubuntu 22:49 < Dagmar> xamithan: That's where DTDs and so forth come in 22:49 < mattfly> 14.04 22:49 < Psi-Jack> And... iptables, would be better. 22:50 < xamithan> Would be better to use iptables directly 22:50 < mattfly> iptables doesnt work i dont know why 22:50 < xamithan> Then ask for help, or visit #netfilter ;P 22:50 < Dagmar> xamithan: like, I could hand you a page full of numbers which is actually a struct, but until you have something dictating how that struct is put together it would be kind of useless 22:50 < mattfly> i have 2 ips on this vps and then there are 2 interfaces eth0 and eth0:0 22:50 < Psi-Jack> xamithan: pastebin (not pastebin.com, see /topic for paste site you can use): ip l 22:50 < Dagmar> ...and eth0 is what 22:51 < xamithan> You need established,related rules. And a DNAT. Then either masquerade or SNAT 22:51 < Psi-Jack> "VPS" is a bling word that by itself has no meaning. 22:51 < Dagmar> Psi-Jack: Not true. It tells me that I was right to question the -i argument 22:51 < Psi-Jack> Which still, by itself, has no meaning. :) 22:52 < Psi-Jack> I question the eth0:0 myself. Because. It's wrong. :) 22:52 < Dagmar> I'm sayin' it's the wrong bloody interface 22:53 < Dagmar> ...but for some people, packets arrive magically on the backs of faeries 22:53 < toothe> apparently kali and ubuntu treat eth0:0 differently. 22:53 < catphish> eth0:0 is a horrible legacy way to configure interfaces 22:53 < Psi-Jack> toothe: ethX:Y is the old ifconfig-style way of doing IP aliases. iproute2 just applies aliases to the actual interface. 22:54 < toothe> When I was a kid we had broadband internet. I remember I used to run multiple PUBLIC ips using eth0:1 et al. 22:54 < toothe> Psi-Jack: Right! 22:54 < catphish> not even aliases, just multiple IPs on an interface 22:54 < Psi-Jack> Aka: Aliases. :p 22:54 < toothe> Psi-Jack: Having read the FreeBSD ifconfig source and looked into iproute2's justification, I think they the latter has a superior argument. 22:54 < catphish> one interface can now have as many ipv4 and ipv6 addresses as it likes, no mess, no aliases 22:54 < mattfly> how to tell iptables to use eth0:0 as a interface for this? 22:54 < toothe> well, can you have multiple mac addresses with eth0:0 ? 22:55 < Psi-Jack> catphish: Still called aliases. :p 22:55 < Dagmar> There would be no point 22:55 < catphish> eth0:0 isn't a real interface, iptables probably just sees it as et0 22:55 < catphish> *eth0 22:55 < Dagmar> You can publish whatever mac addresses you like with the arp tool 22:55 < toothe> Dagmar: I work in Pentesting. Those types are things are often "useful" from our perspective. 22:55 < Psi-Jack> toothe: How many pens have faulty ink today? :} 22:56 < toothe> 7 22:56 < Dagmar> toothe: Then get comfy with the arp tool 22:56 < neoncortex> haha 22:56 < toothe> I use pencils. 22:56 < Psi-Jack> heh 22:56 < Dagmar> I'm just over here building firewalls and routers without IP addresses bound to any interface and stuff because I know how arp works 22:56 < toothe> Some of the PenTesters I work with don't know how to run Linux. 22:56 < toothe> like, basic things. 22:56 < Psi-Jack> That's.... just... sad. 22:56 < toothe> very. 22:57 < neoncortex> toothe: And how they managed to get there? 22:57 < catphish> plenty of pen testers just know how to run acunetix :) 22:57 < toothe> They think I'm a genious. 22:57 < Dagmar> It's because the requirements have been reduced to "download kali, click things, send reports to management" 22:57 < toothe> neoncortex: I honestly do not know. I think they were short-staffed. 22:57 < tds> Dagmar: is there any advantage of messing with arp/ndp manually rather than just binding IPs to interfaces? 22:57 < tds> unless you want to do weird proxy arp style stuff I guess 22:57 < Psi-Jack> Mmm, proxy-arping can be fun. 22:58 < Dagmar> tds: Yes. Specifically I used to be on a frac-T with 10 static IP addresses bound to it. The gateway box had no IP addresses bound to it's external interface at all. 22:58 < catphish> i've hired a couple of pen testers, one got a shell on one of my app servers, the other basiclly just send me a report about some cookie properties 22:58 < Dagmar> tds: ...and if you're wondering about "10" the answer is that yes, they were not a "clean" netblock, it was just ten IP addresses in series 22:59 < catphish> tds: there really shouldn't be many reasons to manually manipulate arp tables 22:59 < alexandre9099> hi, can pulseaudio/alsa "normalize" the amplitude of the audio? for some reason my laptop fails when the sounds gets too loud, but there are some audio sources where the range of amplitudes is to big 22:59 < Dagmar> tds: The only thing required to get the packets into the gateway box was for it to repond to WHO-HAS queries for those IP addresses. Once it had them, it knew to send them along on the internal interface 22:59 < Dagmar> Actually very simple 22:59 < tds> Dagmar: hmm, could you still not just add your own internal routes then enable the magic kernel parameter for proxy arp? 22:59 < Dagmar> because what you're describing is complex 23:03 < Dagmar> Mind you, you *have* to emit responses to who-has queries if you want traffic to come in _at all_ 23:03 < Dagmar> Unless you're connected to a godforsaken hub of some kind 23:04 < catphish> Dagmar: that doesn't make sense 23:04 < catphish> Dagmar: you have to respond to those so that l3 devices will transmit to you 23:04 < Dagmar> catphish: Are you familar with the OSI seven-layer cake? 23:05 < tds> even if it's a hub, stuff wouldn't be directed to your mac address and would get dropped by the NIC surely? 23:05 < catphish> at l2, it doesn't matter, the switch will just flood it of doesn't know where you are 23:05 < Dagmar> If it were a hub, the traffic would just be spammed to the segment you're on no matter who it's for 23:05 < catphish> Dagmar: same with a switch 23:05 < catphish> unless it knows the destination 23:05 < Dagmar> No, switches only send packets down a wire if they think the host is actually there 23:06 < catphish> i don't think that's true 23:06 < Dagmar> routers will just send back host unreachables if no one's answering the who-has 23:06 < catphish> that part *is* correct :) 23:06 < catphish> my understanding was that most switches would flood frames for unknown MACs 23:06 < Dagmar> catphish: Not pointlessly flooding wires is the main reason switches exist, man, so that one host on the switch doesn't even have to listen to the crap being generated by say, the other 23 23:07 < Dagmar> catphish: Some will, generally _bad_ ones 23:07 < Dagmar> There's lots of behaviours that vary between switches, some of which can merrily be abused 23:08 < catphish> i'm pretty sure normal behabiour is to flood 23:08 < Dagmar> Some will automatically decide you're some kind of uplink if you spew a few dozen strange MAC addresses at them 23:08 < Dagmar> Some will decide you're broken and _turn off your segment_ 23:09 < catphish> more advanced switches will have clever secutity policies to disable bad ports, certainly, and may limit flooding 23:09 < catphish> but i maintain that the default position is to flood frames to unknown destination 23:09 < Dagmar> Again, it all depends on the switch 23:10 < Dagmar> Decent ones will only broadcast the arp queries and make a note of who responds from where 23:10 < Dagmar> THey won't generally relay a packet to all ports just because the MAC is unknown 23:10 < catphish> i'm quite sure you're wrong 23:10 < Dagmar> That's what broadcast addresses literally are for 23:11 < Dagmar> Spend more time trying to break themn 23:11 < catphish> i cant find any citation for them doing anything other than flooding them 23:11 < Dagmar> Lots of people basically rely on switches meaning you can't sniff traffic 23:11 < catphish> no they don't 23:12 < Dagmar> ...and vendors have a lot of leeway into what they let the engineers that designed their firmware do 23:12 < catphish> they provide almost no protection against that 23:14 < catphish> cisco's documentation notes "Most switches implement no special command to detect flooding", though it is a feature of more advanced switches to rate limit it 23:15 < catphish> don't feel bad, i visit from ##networking 23:16 < iopq> any way to use integrated and discrete graphics at the same time on Ubuntu 16.04? 23:17 < catphish> iopq: in what way? 2 monitors? or are you talking about prime? 23:17 < iopq> catphish, yes, two monitors 23:18 < catphish> i'd think that would work with no special configuration, but i've not tried so can't comment with certainty 23:18 < iopq> catphish, I tried it and my integrated is not working in Ubuntu, even though it otherwise works 23:19 < iopq> catphish, when I set the integrated as the main monitor it tells me to fix my stuff and won't start x 23:19 < iopq> catphish, when I set the AMD gpu as main, the intergrated one shows a black screen and won't be detected by xrandr 23:19 < catphish> :( i'm afraid i don't know exactly how it should be configured, good luck though 23:22 < NonSecwitter> can I determine the max queue length for my disks from cli? 23:22 < catphish> you can see what it's configured as 23:23 < catphish> /sys/block/*/queue/nr_requests 23:23 < Dagmar> I'm running out of time here, but no... One can't normally sniff arbitrary traffic on a switch becuase the switch only passes traffic down a wire that's _meant_ for the things on that particular wire 23:23 < Dagmar> That is literally the core difference between a switch and a hub 23:23 < twainwek> iopq: have you looked into your xorg log 23:24 < tds> Dagmar: my understanding was that a switch started out by flooding and then learnt a mac-port association, while a hub just flat out always broadcasts traffic to every port 23:24 < Dagmar> Ethernet broadcast traffic is the one exception to this, for obvious reasons (because it's the _broadcast address_ and meant for everyone) 23:24 < catphish> Dagmar: that's true, once a switch knows where a mac is, it will send traffic only to that port, but it's not used as a security mechanism, too trivial to redirect with arp spoofing 23:25 < Dagmar> catphish: I can tell you've not spent the requisite sleepless nights after having to audit people's crap 23:25 < iopq> twainwek, yeah, when I boot with integrated set as main (with the AMD in) it has errors 23:25 < catphish> tds: that's correct, a switch floods anything where it doesn't yet know the destination's port 23:25 < catphish> Dagmar: i don't understand why you'd think that 23:25 < catphish> i'm just providing facts 23:25 < Dagmar> ...but they won't flood the traffic belonging to things they already know what port they're on 23:25 < catphish> Dagmar: correct 23:26 < Dagmar> catphish: because I've _seen_ people relying on a switch to prevent eavesdropping literally _dozens_ of times 23:26 < twainwek> iopq: and the errors are? 23:26 < Dagmar> That's why 23:26 < Dagmar> It's not entirely safe, but people do it anyway 23:26 < tds> Dagmar: then you're doing it wrong ;) 23:26 < mawk> hi 23:26 < tds> unless you're relying on VLANs on the switch to isolate ports 23:26 < Dagmar> tds: what part of the word "audit" are you unfamiliar with? 23:26 < mawk> now that I'm netlink-fluent I want to do cool things with it 23:26 < catphish> Dagmar: oh, i meant nobody *successfully* relies on this for security, i didn't mean people aren't misinformed :) 23:26 < mawk> what about a web based top for processes and networking stuff 23:26 < NonSecwitter> catphish: is that the same queue as reflected in avg queue depth in atop ? 23:27 < meyou_> a web based ajaxy top would be sweet 23:27 < mawk> polling the netlink socket for new processes, route/addresses/neighbor changes, and more stuff 23:27 < catphish> NonSecwitter: i believe so 23:27 < meyou_> i bet somebody's done it 23:27 < mawk> yeah but with ugly code I'm sure 23:27 < mawk> polling is the way to do it 23:27 < iopq> twainwek: https://pastebin.com/GBgqpYdc 23:28 < NonSecwitter> catphish: I'm curious because my avg queue depth is 1.5 to 3, and the file you suggested showed a max queue length of 128. My drive is sitting at 99% utilization 23:28 < iopq> twainwek, this one too https://pastebin.com/9Uu7Tszw 23:29 < catphish> NonSecwitter: that would happen if you just had one process that was waiting for each response before making the next one 23:29 < bls> what if your drive receives a near constant stream of OPs, but can respond to them before the queue backs up? 23:29 < twainwek> iopq: do you have an xorg.conf file youve created? 23:29 < NonSecwitter> catphish: ah. that's what it is 23:29 < NonSecwitter> what's the goto solution for software striping on linux? 23:29 < NonSecwitter> I want to add some more disks to this logical volume 23:30 < bls> "software striping"? 23:30 < cousteau`> how feasible is it for a (GNU/)Linux distribution to NOT have a POSIX shell? 23:30 < catphish> the queue would only fill up id you had lots of things waiting for responses, or maybe lots of writes waiting to be flushed to disk 23:30 < catphish> *if 23:30 < bls> cousteau`: very 23:30 < mawk> cousteau`: chsh /usr/bin/fish 23:30 < cousteau`> is sh required in Linux, or not really? 23:30 < mawk> here it is 23:30 < NonSecwitter> catphish: it's a sql database that is being hammered by back to back, poorly written queries. 23:30 < catphish> cousteau`: pretty sure it is needed because posix 23:30 < koala_man> cousteau`: Linux doesn't care, but it's a POSIX requirement and a lot of tools depend on it 23:30 < bls> cousteau`: it's not a formal requirement, but it is a POSIX and de-facto one 23:32 < catphish> NonSecwitter: hard to guess, but most likely the database is just making one IO request at a time, it may be the sql server can be configured to push the disks harder that that, mysql has a configurable io queue depth, but it may not be able to max it out with one query at a time 23:32 < bls> pretty sure there was a linux/plan9 hybrid that used rc for the system shell 23:32 < catphish> posix is a written standard, posix de-facto seems like a contrdiction 23:32 < catphish> but i don't know the details 23:33 < koala_man> catphish: GNU/Linux is not officially POSIX compliant 23:33 < bls> I didn't say POSIX de-facto, I said it's a POSIX requirement and a de-facto Linux requirment 23:33 < alexandre9099> tell me a light distro (to be install on a vm to test sofware) 23:33 < lupine> alpine 23:34 < cousteau`> tinycore? 23:34 < bls> alexandre9099: distrowatch.com <- pick one yourself 23:34 < djph> LFS? 23:34 < cousteau`> I'm quite out of the state of the art 23:34 < lupine> debian mins down to a couple of hundred MiB 23:34 < alexandre9099> bls, was asking for sugestions :) 23:34 < alexandre9099> lupine, i'll check 23:34 < alexandre9099> djph, well, i want a ready to use distro :D 23:34 < bls> alexandre9099: probably don't want alpine or tinycore 23:34 < cousteau`> is slitaz still a thing? 23:35 < bls> alexandre9099: if you're testing things, those two are going to be very different from what would be considered a normal Linux distro 23:36 < iopq> twainwek, there are some files in that folder, yes 23:36 < alexandre9099> bls, hmm, i'll see distrowatch, i may stay with ubuntu or linux mint 23:36 < twainwek> iopq: backup, remove, reboot 23:37 < cousteau`> alexandre9099, if you want a light ubuntu you could try Lubuntu 23:37 < bls> alexandre9099: your best bet for being lightweight is to take something you know and use the minimalist/network installer and get the bare minimum 23:37 < Dagmar> If you don't mandate sh, people get weird ideas like the system should be driven by tcl or something 23:37 < iopq> twainwek: I'll try that, brb 23:37 < triceratux> alexandre9099: mx-17 23:37 < cousteau`> (or even Xubuntu; I use Xubuntu as my main distro) 23:37 < bls> Dagmar: hah, is that guy still around here? 23:37 < twainwek> only good thing about ubuntu is its larger official repo 23:38 < alexandre9099> triceratux, well, never heard about that one, surely i'll give it a look :) 23:38 < cousteau`> what did you mean with "light" anyway? small, or low resource usage? 23:38 < Dagmar> bls: I dont' think he ever demanded the system be driven by tcl, but he'd managed to use tcl for just about everything else. Considering that he was actually successful at it, I can't really say it was wrong, just umm... perverse and sick. ;) 23:38 < alexandre9099> cousteau`, low resource usage (and small image, capped internet :/) 23:39 < cousteau`> if low resource usage… ah, then light it is 23:39 < bls> always got a good laugh out of the TCL and Forth zealots 23:39 < cousteau`> then maybe ubuntu minimal (for example) 23:39 < Sitri> DSL was very lua heavy (which nearly killed it because of silliness) 23:39 < cousteau`> alexandre9099, do you want a desktop, or just command line? 23:39 < catphish> koala_man: oh i thought it was :) 23:40 < alexandre9099> cousteau`, desktop (i could have said from the begining :D i want to demo on a presentation teamviewer, anydesk and those kinds of remote control software) 23:40 < alexandre9099> (the software needs to be cross-platform) 23:40 < cousteau`> Dagmar, as much as I like Tcl… use it as a replacement for sh? really? 23:40 < catphish> although i guess switches like POSIX_ME_HARDER imply it's not by defult 23:41 < cousteau`> s/like/tolerate/ 23:41 < sake> hello 23:41 < Dagmar> cousteau`: In theory you can replace init with perl if you wanted to, which would probably work fine until you tried to call a perl function that wrappers sh. ;) 23:41 < bls> alexandre9099: then definitely stay away from most of the advertised "lightweight" distros, as there's no way they'll be able to run proprietary software 23:41 < phogg> Dagmar: driving a system with tcl isn't that crazy, not like driving it with guile 23:41 < iopq> twainwek, back, didn't help. I can probably put that file back in and edit it to have both onboard and discrete at the same time? 23:41 < Dagmar> phogg: lisp 23:42 < alexandre9099> bls, oh, so i might go with a full ubuntu install :D 23:42 < alexandre9099> not really lightweight but it *should* work 23:42 < cousteau`> alexandre9099, well then I suggest Ubuntu or another Debian thing (not sure if the TeamViewer installer is available for anything other than .deb and maybe .rpm) 23:42 < twainwek> iopq: errors? 23:42 < Dagmar> On the good side, if your BSD init is driven by lisp, you can then measure system performance in parenthesis-per-second 23:42 < cousteau`> so I'd say Lubuntu, or maybe Xubuntu 23:43 < iopq> twainwek just the integrated GPU not showing in xrandr, I'm sure if I boot with it as primary I'll get the same errors 23:43 < bls> alexandre9099: yeah, you're going to be on a debian/*buntu/mint type distro or a fedora/centos/redhat distro if you want to run proprietary stuff 23:43 < twainwek> iopq: what's the xorg log look like 23:44 < alexandre9099> cousteau` bls: thanks guys :) 23:45 < cousteau`> alexandre9099, check out those two sub-distros' "minimal installs" 23:46 < bls> you can get that stuff working on one of the more obscure distros, but it'd take some battling which you don't want to need to do mid-presentation 23:46 < twainwek> alexandre9099: try opensuse and during installation uncheck all the "recommended" packages 23:47 < angelo_ts> hello experts 23:47 < iopq> twainwek, nothing exciting 23:48 < twainwek> no errors? 23:48 < iopq> no, just some warnings 23:49 < angelo_ts> question: my company wants me to use yocto to build some embedded stuff. Yocto informs me that he doesen't like my debian unstable version, and wants an ubuntu ... otherwise, it fails his job. Is there some other way to make it running in an ubuntu OS except virtualbox ? 23:49 < iopq> this is the only relevant one, but it's because I enabled tear free [ 3.457] (WW) RADEON(0): KMS Pageflipping: disabled because of ShadowPrimary/TearFree 23:49 < bls> angelo_ts: there are some yocto users here that might have an answer. there's also a yocto channel on here 23:50 < cousteau`> alexandre9099: it would seem that the ubuntu "mini.iso" is only 64MB; then once installed you run `sudo apt-get install lubuntu-core` or `xubuntu-core` to get the desktop 23:50 < twainwek> iopq: dunno. have you google'd your problem 23:50 < angelo_ts> bls, thx. Do you think is possible to run another distro in a container ? 23:50 < cousteau`> er, xubuntu-core^ with a ^ 23:50 < alexandre9099> cousteau`, well and what is the size of "lubuntu-core"? 23:50 < iopq> twainwek, I've been googling it for three days and counting 23:50 < twainwek> :( 23:50 < cousteau`> heck if I know 23:50 < alexandre9099> :D 23:51 < bls> angelo_ts: don't know, not my area of experience 23:51 < angelo_ts> ok will ask more info on #yocto :) thanks 23:51 < alexandre9099> maybe it will not be too distinct from the image, i'll just download the whole image, that way i'll put on the pendrive, might come in handy 23:51 < bls> angelo_ts: but other than virtualbox, there's KVM 23:51 < angelo_ts> bls, ok thanks 23:52 < cousteau`> probably not much though... I'd ask in #lubuntu for details 23:52 < cousteau`> alexandre9099, no no, the "core" is a slimmed down version 23:52 < cousteau`> not the full ISO 23:52 < twainwek> iopq: vga? and how do you know it's functional 23:53 < cousteau`> you don't need Firefox or Gimp or LibreOffice (and if you do then you install them) 23:53 < iopq> twainwek, if I take the GPU out the onboard works just fine 23:54 < cousteau`> -core gives you a version with none of that; only the basics 23:54 < alexandre9099> cousteau`, oh, 2 minutes left to download the image, thanks anyway :) i might also put the mini iso in my pendrive :) 23:54 < cousteau`> ah... well then :) 23:55 < cousteau`> which one? lubuntu? 23:55 < alexandre9099> yep it was the smallest (1GB vs 1.3GB :D) 23:55 < TheNH813> Allright, anyone know why SSH refuses connections until I log in first? 23:56 < TheNH813> I had to call someone and have them log into my PC as the guest account before ssh worked remotely. 23:56 < twainwek> iopq: have you verified with another os that the two work together by any chance 23:56 < bls> TheNH813: did you encrypt your home dir? 23:56 < TheNH813> bls: No. 23:56 < meyou_> TheNH813, maybe it was just asleep? 23:56 < cousteau`> one damn GB?! damn, back in MY day a full Ubuntu would be 690-ish MB 23:56 < iopq> twainwek, I don't have another OS, but it used to work with a DIFFERENT AMD card 23:56 < TheNH813> meyou_: That's... actually a possibility I didn't think of. 23:57 < alexandre9099> cousteau`, :D 23:57 < alexandre9099> cousteau`, i remember that ubuntu was almost 2GB if not more 23:57 < cousteau`> (and the mini is 64 MB, plus whichever desktop you get) 23:58 < alexandre9099> the bad of the mini iso is that it needs an internet connection 23:58 < alexandre9099> the full isos can be installed without internet :) 23:58 < twainwek> iopq: maybe not related to linux 23:58 < TheNH813> Hmmm... no I don't think suspend is enabled on my account. Maybe it still does if nobody is logged in. 23:58 < alexandre9099> cousteau`, (ubuntu is 1.8GB) 23:59 < twainwek> iopq: also check your bios to see if there's a setting for it --- Log closed Fri May 04 00:00:07 2018