--- Log opened Tue Jul 03 00:00:02 2018 --- Day changed Tue Jul 03 2018 00:00 < DLange> Nixola_: usually something grabbed the partition, check dmesg , check dmsetup ls 00:01 < zenix_2k2> guys is there anyhow to reload my default Files system interface ? 00:01 < zenix_2k2> my current GUI looks weird 00:01 < zenix_2k2> and uncomfortable 00:01 < zenix_2k2> anyway i am using Ubuntu 16.04 00:01 < Nixola_> DLange: nothing in dmesg, dmsetup ls says no devices found 00:03 < DLange> Nixola_: if there's no data on the disk / partition, wipe the first few kb (dd if=/dev/zero ..), reboot, try again 00:03 < nineclicks> I have a server with no shell access and no home folder. I am using sftp with it. Is there any way to use keys to login? 00:04 < Nixola_> DLange: there's a few installed software and games, which I'd rather not redownload if possible 00:04 < DLange> you could put the authorized_keys file elsewhere via ssh/sshd_config 00:04 < Sitri> nineclicks: CA keys 00:05 < msiism> is there any way of setting a gradient background in X other than using xsri? 00:05 < Nixola_> lsof also doesn't show anything to do with sda1 00:06 < DLange> Nixola_: well, you need to figure out what grabs the partition, what did you change between it working and it being broken? 00:06 < nineclicks> I don't think I have any kind of access to the remote sever other than putting and pulling files from it. Would CA keys still work? 00:07 < Sitri> Yes 00:07 < DLange> (and you should be able to resync your raid even if you wipe the partition and then re-assemble because otherwise your raid would have been flawed from the very beginning) 00:07 < nineclicks> Would you mind giving me a super overhead rundown of how that would work? 00:07 < Sitri> Basically you make a CA, sign a user key with it, tell the server to trust all keys signed by the CA key 00:08 < Nixola_> DLange: my pc pretty much locked up and I rebooted 00:08 < Sitri> You can use a pre-existing user key, but you have to make the CA key in a special way 00:08 < DLange> Sitri: which means you need to be root on the server ... 00:08 < nineclicks> And I can tell the server to trust it with just sftp? 00:08 < Sitri> DLange: right 00:08 < sbko> what can be 00:08 < Nixola_> not forcefully, but it took a lot of time and /dev/nvme0n1p2 (ext4, / partition) had a lot of errors and I had to fsck 00:09 < Nixola_> (it locked up due to vm and pcie passthrough shenanigans) 00:09 < sbko> what can be the reason that mknodat becomes slower? reboot fixes that 00:09 < DLange> hm, yeah, so something broke, may be your /etc/mdadm/* stuff 00:09 < Nixola_> would that explain why mdadm thinks sda1 is busy though? 00:10 < Nixola_> also, /etc/mdadm.conf looks fine 00:10 < DLange> try to find what broke it or wipe the one RAID member and re-sync (if your data is not important or you have a good backup) 00:10 < nineclicks> So root access on the remote is required to use CA keys for auth? 00:10 < Nixola_> I also tried to stop the array and recreate it, but it still says it's busy 00:12 < DLange> Nixola_: does cat /proc/mdstat tell you something useful? 00:12 < Nixola_> DLange: nope, just that the other array is working without any issue 00:13 < Nixola_> https://hastebin.com/owumugimil.txt specifically, this 00:14 < Sitri> nineclicks: you can do it without root access by writting to the user's authorized_keys file. 00:15 < Sitri> CA auth generally has you add the CA key to the system's authorized_keys 00:16 < DLange> Nixola_: mdadm --stop /dev/md0 ; sleep 1 ; mdadm --assemble --force /dev/md0 /dev/sda1 00:16 < Tazmain> pingfloyd, hmm so my submenu is not being detected 00:16 < Tazmain> or category 00:16 < pingfloyd> Tazmain: check your XDG directories 00:16 < Tazmain> pingfloyd, well my .desktop file was picked up 00:17 < Tazmain> but not the menu item 00:17 < pingfloyd> https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html 00:18 < pingfloyd> if those env vars are empty, then it's only using the default ones. 00:18 < pingfloyd> as you can so, you can define them if you want 00:18 < pingfloyd> if you wanted to use a path outside of the default ones 00:18 < pingfloyd> s/so/see/ 00:19 < Nixola_> DLange: when assembling: mdadm: failed to RUN_ARRAY /dev/md0: Invalid argument 00:19 < Tazmain> yeah which is working , since it picked up the desktop file 00:19 < Tazmain> but its not getting the category 00:19 < Nixola_> but md0 shows up as before (which it didn't after stopping) 00:19 < pingfloyd> probably an issue with your xml file or .directory file, or where you placed those two 00:20 < Tazmain> pingfloyd, won't that be the .menu file ? 00:20 < Tazmain> to add the category 00:20 < Tazmain> so I have the /etc/xdg/menus/application-merged/ folder with my .menu file 00:20 < pingfloyd> .menu is xml file 00:21 < Tazmain> yeah 00:21 < pingfloyd> you have a typo 00:21 < pingfloyd> s/application/applications/ 00:21 < DLange> Nixola_: you only have the raid1 personality, modprobe raid0 and try again 00:21 < pingfloyd> try that 00:21 < Tazmain> pingfloyd, I copied from the freedesktop.org site -_- 00:21 < DLange> wtf of a distro are you using? 00:22 < Nixola_> DLange: it's supposed to be raid1 00:22 < pingfloyd> Tazmain: you're using /etc/xdg/menus/applications-merged or "application-merged"? 00:22 < pingfloyd> the latter isn't going to work 00:22 < Tazmain> oh then the site had a typo 00:23 < DLange> Nixola_: it only has one member ... 00:23 < pingfloyd> looks like it for the code example, but the description shows applications 00:23 < pingfloyd> tias 00:24 < Tazmain> hmmm 00:24 < Tazmain> do I need to logout and in for this ? 00:24 < Tazmain> pingfloyd, no wait it works 00:24 < Tazmain> yeah it was that 00:25 < Tazmain> pingfloyd, whoop !! , thank you ! 00:25 * DLange is afk, good luck Nixola_ 00:25 < pingfloyd> you're welcome 00:25 < Tazmain> now I need to write a script to create all these directories and .desktop files 00:25 < pingfloyd> or make a package, or PKGBUILD 00:26 < pingfloyd> you're using arch iirc 00:26 < Nixola_> DLange: I took one out for reasons 00:26 < Tazmain> pingfloyd, I am 00:27 < Tazmain> pingfloyd, yeah PKGBUILD will make the folders, and base files, and then a script to create .desktop files based on what is installed from a certain repo 00:27 < pingfloyd> if you make a PKGBUILD you'll probably want to distribute to AUR. As far as which hierarchies to use, you should read up their docs on their policy. 00:27 < Tazmain> this will be in the blackarch repository 00:28 < pingfloyd> the PKGBUILD is essentially a bash script with some extra info 00:28 < Tazmain> but I will go read up where to use 00:29 < NightStrike> I'm using an I/O card that can do DMA transfers. There's little documentation on it, but for the "DMA Control" option, I can pick "Constant Read" or "Increment". Does anyone have any idea what that means? 00:34 < Dagmar> How _old_ is this hardware? 00:34 < AppAraat> uh, can a manually edited comment of an ssh pubkey make it so that auth via pubkey is denied? 00:35 < Dagmar> AppAraat: Only if you screw it up very, very badly 00:35 * AppAraat something happened very badly 00:35 < Dagmar> So put it back to what it originally was 00:35 < AppAraat> but then I'll have to manually edit it again lol 00:35 < ayecee> lolol 00:35 < Dagmar> The comment field is actually read by the ssh daemon, but it's not _compared_ against anything. It's sometimes used to limit what the key can be used for 00:36 < Dagmar> Just re-paste the line, man 00:36 < AppAraat> well I removed the old comment 00:37 < AppAraat> so I'm not sure whether it's like a msdos / unix text format thing in vim 00:38 < AppAraat> if it is then whatever I'll edit it to will probably have the same result 00:39 < Dagmar> I'm pretty sure you should be able to paste a line into a text file with vim without managing to somehow insert cr/lf 00:40 < Dagmar> Even so, that shouldn't affect anything. 00:42 < Dagmar> Actually scratch what I said earlier, since it's the first part of the line in authorized_hosts that restricts command usage 00:45 < mattfly> is there any student here that has to read a lot in a day and learn more by listening than reading a boring book? 00:45 < mattfly> what is the speech to text program you use? 00:47 < mattfly> and please no espeak for crazy cli shit 00:47 < mattfly> I need to be able to go back some lines if i need, read some formula or image and so on 00:47 < mattfly> and preferably with ocr support too 00:48 < mattfly> if theres no tool that is usable enough for this on linux lets code it? 00:48 < AppAraat> Dagmar: yeah that's what I know too: Pretty much everything before comment should not be edited. I'm guessing this is a different issue. 00:48 < AppAraat> >2018 00:48 < AppAraat> >still having ssh issues 00:49 < Dagmar> Use ssh -vvv to get an idea if how it's failing 00:49 < AppAraat> I looked at that, but that hinted that I should learn the ssh RFC's by heart :p 00:56 < Dagmar> Allow me to suggest you learn about the ssh-copy-id command then 00:58 < gee111> what is the best utility to scan for bad sectors for linux?? I can't find a bootable CD with one 00:58 < Psi-Jack> badblocks 00:59 < Psi-Jack> Not that great, but a real tool to really get the HDD to handle stuff like that properly: SpinRite. 00:59 < serverwand> Hi, I'm looking to delete my existing ubuntu installation and install a new one over it. I'm trying to figure out what option I should choose for the partition to "use as"? 01:00 < gee111> my computer cannot boot bios 01:00 < AppAraat> Dagmar: done long time 01:00 < koala_man> serverwand: is that the one that lets you choose "/" or "/home" or "swap"? 01:01 < serverwand> swap, ext4 journaling system, ext3 journaling system, ext2 file system 01:01 < serverwand> there's more options too 01:01 < koala_man> serverwand: if you don't know which fs you want, ext4 is a good choice 01:02 < serverwand> okay thanks 01:03 < serverwand> and for the mount point what should it be? 01:03 < koala_man> serverwand: if you're planning on using a single partition for everything, then / 01:04 < Tech_8> hi 01:04 < rcf> gee111: badblocks can run a read-only scan on a running system. 01:05 < Psi-Jack> serverwand: https://linux-help.org/wiki/linux/basics-101/partitions-and-filesystems 01:11 < Tech_8> hi, how are you doing? 01:12 < Psi-Jack> Acceptable levels of functionality. 01:17 < sauvin> "All status indicators within nominal parameters." 01:19 < birdbolt1> hey does anyone here use webpack? 01:19 < Tech_8> no whats it do? 01:19 < birdbolt1> i have a question about passing env vars 01:19 < birdbolt1> Tech_8, asking me? 01:19 < Tech_8> yes 01:20 < Tech_8> birdbolt1: yes 01:20 < birdbolt1> Just started using it a week ago, the internet would answer that question much more clearly than i can, but it "packs" together all your sistes staticfiles, and some dependencies, into a bundle 01:21 < birdbolt1> site's* 01:21 < Tech_8> ok 01:21 < lnnb> like "save page" option in browser? 01:26 < FManTropyx> I thought CentOS is made by some company 01:26 < bazhang> redhat's free version 01:26 < doxinho> lol 01:27 < lnnb> i thought it was made by real life centaurs 01:30 < Psi-Jack> CentOS was originally made strictly as a community, but Red Hat SOftware did acquire the distro as well, and it's been remaining free. 01:43 < qoxncyha> is there a way to force globbing? 01:43 < qoxncyha> i want `ls *.nonexistent` to return 0 lines and exit successfully 01:44 < bls> qoxncyha: look at the nullglob and failglob settings for bash 01:44 < koala_man> qoxncyha: you can shopt -s nullglob , but ls with zero files will not return zero lines 01:44 < ayecee> qoxncyha: it seems like ls wouldn't run at all in that case. 01:45 < qoxncyha> so i'm following https://www.inversoft.com/guides/2016-guide-to-user-data-security#database-server-specific-security 01:46 < qoxncyha> under 4.11.2 Backups, there's `ls *.gz.enc | sort | tail -n +7 | xargs rm` 01:46 < qoxncyha> what do you recommend for that instead? 01:46 < ayecee> so maybe [ -n "*.nonexistent" ] && ls "*.nonexistent" 01:46 < ayecee> qoxncyha: test if any matching files exist before running. 01:50 < Biessie> hey!! how can i use mv command to basically find everything such as *jpg starting from a folder onward into subfolders and move all those files to one specific folder? 01:51 < xamithan> You'd need to use the find command for that, not the mv command 01:51 < Biessie> i have a bunch of mp4/jpg files in a ton of folders i want to basically move all the images to one folder and all the movies into another 01:51 < iflema> pr0n 01:52 < Biessie> or my kids pics broken up into a bunch of folders due to apple 01:52 < qoxncyha> ayecee: are you sure -n is the right one? 01:52 < xamithan> find / -iname "*.jpg" -exec mv {} /to/folder \; ? 01:52 < Biessie> xamithan : Ill give that a shot thanks! 01:52 < AppAraat> xamithan: better quote those {} 01:53 < xamithan> I don't think quotes matter unless you use xargs but it couldn't hurt 01:53 < kerframil> AppAraat: '{}' is exactly the same thing as {} 01:54 < qoxncyha> ayecee: i think -n is right, but it's not an empty string 01:54 < qoxncyha> `[ -n "*.gz.enc" ]; echo $?` is printing 0, specifically 01:54 < qoxncyha> is there a bash builtin to glob a string? 01:54 < kerframil> xamithan: considerably faster would be: -exec mv -t /to/folder {} + 01:55 < kerframil> qoxncyha: [[ $var == *.gz.enc ]]; echo $? 01:55 < qoxncyha> kerframil: i want to determine whether there are any files that glob to *.gz.enc 01:55 < qoxncyha> what is $var? 01:56 < kerframil> qoxncyha: https://mywiki.wooledge.org/BashFAQ/004 01:56 < Biessie> xamithan : Think it worked! 01:57 < Biessie> taking a long time lol 01:57 < xamithan> Well find is pretty slow, theres things you can do to speed it up like what kerframil said 01:58 < xamithan> Also not using it on the root directory helps 01:58 < kerframil> find is fast, it's forking that is slow 01:58 < qoxncyha> kerframil: what is $var? 01:58 < kerframil> it becomes noticeable where processing a significant number of paths 01:58 < Biessie> xamithan i didnt lol 01:58 < xamithan> find is pretty slow compared to locate ;P 01:58 < Biessie> i used specific folder where i mounted the drive to 01:58 < AppAraat> kerframil: it does quoting internally? 01:58 < Biessie> i have locate/updatedb updated 01:59 < Biessie> can i replace find with locate/ 01:59 < Biessie> ? 01:59 < xamithan> Nah 01:59 < kerframil> AppAraat: quoting isn't relevant because find performs the expansion of {}, not the shell. the argument(s) will be conveyed literally to the command that is executed. 01:59 < NightStrike> Dagmar: was that question about old hardware directed to me in response to my DMA question? 02:00 < kerframil> qoxncyha: it's not material to your initial question, which was unclear, but [[ $var == glob ]] allows you to compare the value of a variable to a glob/pattern. 02:00 < NightStrike> if so, it's relatively new, a few years or so. drivers are compatible with RH5, though, which is 2.6.18 02:00 < kerframil> qoxncyha: what you actually want is explained in BashFAQ/004 02:01 < qoxncyha> kerframil: i see the array logic, i don't want to use it because this script will be maintained by those who don't know bash very well 02:01 < qoxncyha> can i use `[[ "" == *.gz.enc ]]`? 02:01 < kerframil> qoxncyha: no 02:01 < qoxncyha> why? 02:01 < kerframil> qoxncyha: that neither makes sense, and will always be false 02:01 < AppAraat> kerframil: I see, I though { and } chars were also interpreted by the shell since you can "echo {0..3}" for example. 02:01 < kerframil> qoxncyha: because it doesn't work 02:02 < qoxncyha> because with no files, *.gz.enc will be ".gz.enc"? 02:02 < kerframil> AppAraat: that's correct. but {} isn't a form of (bash) brace expansion. 02:02 < kerframil> AppAraat: so it's unaffected 02:02 < AppAraat> cool, TIL! Thanks :) 02:02 < kerframil> qoxncyha: [[ whatever == *.gz.enc ]] does *not* do pathname expansion ... at all 02:02 < Biessie> xamithan : Lets say i have duplicate files with the same name but in different folders using that command you provided for me.. will it only move one and leave the duplicate where it's at? 02:03 < kerframil> qoxncyha: therefore, that particular syntax can't be used to solve this problem 02:03 < qoxncyha> i see 02:04 < xamithan> I'm not 100% but I think it'll just leave the duplicates alone since the file already exists 02:04 < Biessie> okay! so it wouldnt rename it as well 02:04 < Biessie> i guess ill find out once its done. i literally have GB's full of pictures and videos over the last 10 years of my kids growing up 02:05 < xamithan> You can do a quick test trying to mv a file that already has the new name 02:05 < Biessie> trying to organize it all and put on my private website for family to have access to 02:06 < xamithan> Hah, well it overwrote it 02:06 < kerframil> qoxncyha: if you really don't want to use an array, a possible alternative (with GNU find) would be: find -name '*.gz.enc' -printf . -quit | grep -q . 02:06 < kerframil> qoxncyha: that's not portable, obviously 02:08 < qoxncyha> i don't even know how to negate the expression in wooledge 02:08 < qoxncyha> `if (shopt -s nullglob dotglob; f=(*); ((! ${#f[@]}))); then` looks like line noise to me 02:08 < kerframil> Biessie: you might want to use mv -i or mv -n because there is otherwise a risk that one photo clobbers another. that's because you're effectively flattening the directory structure that existed by moving them all to the same directory. 02:08 < kerframil> Biessie: obviously, that would only happen where one source file has the same name as another 02:09 < Biessie> kerframil : Too late! lol im already done but ill remember that in the future lol it just finished 02:09 < qoxncyha> any thoughts? 02:09 < Biessie> nothing should be named the same unless its a duplicate anyways. 02:10 < kerframil> qoxncyha: your example is correct. only, you don't need dotglob if you're going to expand *.gz.enc (rather than *). 02:11 < qoxncyha> i want it to match if it's not empty 02:11 < qoxncyha> (negation) 02:12 < kerframil> qoxncyha: you keep changing the question 02:12 < kerframil> qoxncyha: if you're checking for a empty - or non-empty - dir, then there are several examples in the FAQ that will do that. and your use of negation there is correct, btw. 02:12 < qoxncyha> i don't understand. the example says that is empty, not nonempty 02:13 < qoxncyha> okay 02:13 < limarfrek> qoxncyha: there's also a POSIX-compatible example, that works by (ab)using the positional paramaters as a pseudo-array 02:14 < qoxncyha> limarfrek: sure, what is it? 02:14 < qoxncyha> oh, i see 02:14 < limarfrek> qoxncyha: it's beneath the text that says "If your script needs to run with various non-Bash shell implementations [...]" 02:15 < qoxncyha> i'm alright without posix 02:15 < qoxncyha> it's bash-dependent 02:15 < notmike> *gasp* 02:17 < limarfrek> qoxncyha: yet another way, with GNU find: if ! find dirname -mindepth 1 -printf . -quit | grep -q .; then echo "dir is empty"; fi 02:37 < jim> qoxncyha, wooledge is greycat, and he's usually on, in #debian... maybe if he's not busy he's be willing to help 02:40 < jim> qoxncyha, he would definitely require that you'd have read up, both in his guides and probably also the bash manual itself 02:42 < jim> qoxncyha, also you'd want to arrange to speak to him on a channel where the question is on topic, such as #bash 02:50 < markasoftware> How do I find what the ISP DNS servers recommended by the router are? 02:51 < ntd> cat /etc/resolv.conf 02:51 < markasoftware> Nah I have manual ones set there 02:51 < ntd> or look at the dhcp log 02:51 < bls> or /var/db/dhclient..lease 02:51 < ntd> ok, do a dhcp pull to a virtual interface 02:51 < bls> depending on which DHCP daemon you're using 02:52 < markasoftware> No var db folder on here 02:52 < ntd> /var/lib/dhcp 02:52 < jim> markasoftware, pull dhcp from it and then look at your resolv.conf 02:53 < jim> of course, if you're using your own dhcpd, look at what you wrote :) 02:54 < markasoftware> There is nothing of interest in var lib dhcp 02:54 < markasoftware> Just one empty file 02:55 < markasoftware> I use opensuse, networkmanager, nothing interesting 03:03 < markasoftware> OK, let me ask a simpler question...I can't login to the wifi here because the dost of the login page can't be resolved and they block all DNS servers but the ones the router recommends 03:03 < markasoftware> What do 03:04 < oerheks> Nothing, ask them to change their policy 03:06 < bls> or use their DNS to navigate the portal 03:06 < GunqqerFriithian> A while ago I say on reddit a backup tool that would only backup linux system files. I can't find that post or the software itself though my searches, any of you happen to know of anything like this or simmilar 03:06 < bls> "linux system files"? 03:06 < GunqqerFriithian> it backs up most everything except /home and some other things 03:07 < GunqqerFriithian> that's all I remember about it 03:07 < markasoftware> i had to look in the systemd logs (journalctl) 03:11 < bls> GunqqerFriithian: have a look at borg, bup, rsnapshot, git-annex, etc. although most things don't bother to backup files that can be easily downloaded again and focus on /home, /etc, /srv, etc although you can configure any of them how you want 03:13 < GunqqerFriithian> if only the bookmark I made didn't disappear :P 03:15 < masber> good morning, is it possible to tell dd not to use cache for reading? dd if=/share/ScratchGeneral/mansop/test.file of=/dev/zero 03:17 < ayecee> probably no, otherwise it could be readily used for speed testing, and it isn't. 03:17 < Psi-Jack> masber: What reasoning? What are you actually trying to do? 03:17 < bls> masber: look at the direct flag, may or may not do what you want 03:18 < masber> Psi-Jack, /share/ScratchGeneral is a external storage (distributed file system) and I feel it is running slow, I just want to test performance 03:18 < bls> why? 03:19 < masber> bls, so I don't want linux to read from local cache but go to destination to grab the file 03:19 < Psi-Jack> bonnie++ 03:19 < bls> that's the how, not the why 03:20 < Psi-Jack> And/or FIO. 03:25 < masber> bls, because reading from local cache won't give me the real performance of the remote storage 03:26 < masber> bls, not sure if I answered the questions... 03:26 < Psi-Jack> Why do you feel it is running slow? 03:27 < Psi-Jack> That is what bls asked. 03:27 < bls> yes, and what is measuring how fast dd operates going to accomplish? 03:32 < pingfloyd> masber: flush the cache between runs you're comparing 03:32 < pingfloyd> masber: the name of the game is implementing tight controls on your tests 03:33 < boingolov> masber: if you want to benchmark, why not use a benchmark? 03:33 < boingolov> like bonnie++ ? 03:33 < Psi-Jack> and/or fio :) 03:33 < pingfloyd> masber: you have the right idea in filter cache out of results if you're doing benchmarking/comparison 03:33 < boingolov> haven't tried fio 03:33 < pingfloyd> s/filter/filtering/ 03:33 < Psi-Jack> fio's a little more readable than bonnie++, in terms of output resulted. 03:33 < Psi-Jack> results* 03:34 < masber> testing the remote storage was my intention to start checking the system 03:34 < masber> I run a job on the cluster and it is taking 15h when it normally takes 2h 03:35 < pingfloyd> remote is going to be full of ambiguous results 03:35 < boingolov> masber: what is the connectivity between the box in question and the remote storage? 03:36 < pingfloyd> internet connection fluctuates a lot 03:36 < masber> connectivity is 25GBs 03:36 < pingfloyd> conditions 03:36 < pingfloyd> especially if you're on cable 03:36 < pingfloyd> no thanks to shared bandwidth 03:36 < masber> remote means local ethernet connection, it is not going to the internet, sorry I should have mentioned that before 03:36 < pingfloyd> masber: on an isolated network? 03:36 < boingolov> masber: infiniband? 03:36 < masber> yes isolated network 03:37 < pingfloyd> okay, that's good 03:37 < pingfloyd> sounds like you're on the right track 03:37 < boingolov> what's the pint times look like? latency can screw you more than throughput a lot of times 03:37 < boingolov> ping even 03:37 < bls> have you gone on-box to check on the storage system status? 03:37 < masber> ping to remote storage is ~0.171 ms 03:38 < boingolov> masber: might be useful to let the ping sit there for a while, see if you notice spikes 03:38 < pingfloyd> maybe you should also run a ping at intervals (mark the time you start your dd runs) and redirect results into files for helpful data in your comparison. 03:38 < masber> bls, yes storage system says metadata is nearly full which means, storage used SSDs for metadata and small files and spinning disks for big (> 200KB) files 03:39 < pingfloyd> would at least give you an idea of how much the network may skew your results 03:41 < masber> ping is stable 03:41 < masber> core switch is a mellanox SN2700 100GBs 03:42 < boingolov> masber: what's the i/o wait state look like while running the job? 03:43 < bls> so it could be "thrashing" trying to shuffle data around in a low capacity situtaiont. are other hosts on the netwokrk seeing similar performance? 03:46 < ryouma> when i run dpkg-reconfigure grub-pc in a dumb terminal, it works fine. but when i do the same thing, but in a chroot, it fails to list the drives to install to. clues? 03:46 < boingolov> ryouma: okay, I'll bite: why are you trying to do it in a chroot? 03:47 < boingolov> I'm guessing you didn't copy the dev nodes into your chroot, and also I'm not sure how dev mapper etc would work under chroot 03:47 < ryouma> boingolov: because i want it to affect the target, chrooted root fs 03:47 < ryouma> i did do the dev sys proc 03:47 < ryouma> the crhoot works well 03:47 < boingolov> does, for instance, fdisk -l work from within the chroot? 03:48 < boingolov> using lvm by any chance? 03:48 < ryouma> no 03:48 < ryouma> i will try the fdisk -l 03:48 < boingolov> no to both? 03:49 < boingolov> you might also need to copy in the /etc/mtab into the chroot 03:49 < boingolov> not sure what the grub-pc uses to determine filesystems, but it would seem reasonable that mtab might play a role 03:50 < birdbolt1> can i make a folder /public which doesnt need root access to write to? 03:50 < boingolov> birdbolt1: sure 03:50 < birdbolt1> is that a bad idea? or insecure? 03:50 < Psi-Jack> Of course it is. 03:50 < birdbolt1> lol nested folders it is then 03:50 < CompanionCube> GRUB has a special program to detect a disk's filesystem 03:50 < CompanionCube> man grub-probe 03:53 < boingolov> unfortunately the man pages don't seem to shed light on how it probes / fstests 03:54 < strive> What are some cool things to learn on Linux? (bored) 03:55 < arvut> https://youtu.be/m0Q7eF-buPI?t=2668 sasha and digweed 03:56 < arvut> awesome set, recommend from start 0:00 03:56 < CyberManifest> strive: VIM ? 03:56 < ryouma> boingolov: fdisk -l works fine. i zeroed mtab. you're supposed to for this purpose. 03:57 < ryouma> boingolov: i suspect it is maybe some kind of stdio issue, dunno. 03:57 < arvut> strive: the cli in general, bash, C, python 03:57 < akk> zsh 03:57 < boingolov> ryouma: I'm not sure why zeroing the mtab is recommended? 03:58 < akk> emacs 03:58 < akk> regexps with sed, maybe awk too 03:58 < arvut> strive: the cli is made up of hundreds, sometimes thousands of commands, each of them executing a given task that is tailored for a specific purpose, you can combine them and send data between commands/programs to manipulate workflow in your general direction 03:58 < ryouma> BionicMac: because it will be wrong, so might as well be zero 03:59 < akk> strive: actually I might put regexps first if you don't already know them, they'll give you a feel of godliness. 03:59 < arvut> strive: a router running linux might have a very limited shell, while a fullblown developerbox should have a large toolbox 03:59 < ryouma> if only you didn't have to parse everything 04:00 < boingolov> I know! use regexes to parse HTML! 04:00 < ryouma> me too, and it is gross 04:01 < boingolov> ryouma: the mtab represents the mounted filesystems 04:01 < boingolov> why would it be wrong? 04:01 < ryouma> if emacs is to your likign it will blow all of those other things clean out of the water 04:01 < ryouma> if it isn't, carry on 04:01 < akk> boingolov: haha, then go admit to it on an IRC channel and get flamed for hours. 04:01 < boingolov> the mtab is written as filesystems are mounted 04:01 < boingolov> akk: zalgo... the pony he comes 04:02 < ryouma> boingolov: it's not the only source of that. /proc or smoething has the ifnormation also. the fs discovery commands work fine. 04:02 < CompanionCube> usually mtab's a symlink to /proc/self/mounts 04:02 < ryouma> so it does not seem to be the problem 04:02 < CompanionCube> which is from the kernel and cannot be wrong. 04:02 < CompanionCube> if it's not, then mtab could easily be stale. 04:02 < boingolov> ryouma: sure. my point is, since I can't find documentatino of what that command is actually referencing to discover the filesystems available, having a populated mtab might be a start 04:02 < boingolov> it surely can't hurt 04:03 < boingolov> you can zero it after 04:03 < boingolov> though it will absolutely be rewritten at next boot anyway 04:03 < ryouma> boingolov: ah. well, it used to work is the thing. and i did not change debian much. so i think it is a scripting error on my part or an intermittent issue. 04:04 < ryouma> good idea though and i will try it. takes a while to trythigns. 04:05 < ryouma> ah it is populated. doing the chroot and shelling into it populated it for some reason. so the answer is it is fine. 04:05 < boingolov> ahh 04:05 < boingolov> so you're not doing lvm right? 04:05 < dogbert_2> will need to get 2x4GB of ram for this HP 6000 (DDR3 10600U) 04:05 < ryouma> i /never/ do lvm 04:05 < ryouma> it just adds more stuff to debug 04:05 < boingolov> heh 04:05 < boingolov> I like lvm 04:06 < boingolov> but yes, it is another layer 04:06 < ryouma> i get the snapshot feature, but i'll just pick that up when btrfs or whatever becomes standard on debian stable :) 04:08 < boingolov> ryouma: do you have a separate /boot partition? 04:08 < boingolov> also, are you using uefi? 04:08 < ryouma> yes. no.' 04:09 < boingolov> do you ahve /boot inside of chroot? 04:09 < ryouma> oh, well maybe. but this is an old bios box 04:09 < ryouma> yes 04:09 < boingolov> I'm running out of ideas, heh 04:09 < matsaman> 'bout what? 04:13 < ryouma> this is grasping at straws but i think it might be a redirection buffering issue. basically i want to log my entire script. so i do prepareboot-to-be-logged "$@" 2>&1 | tee ${ftee:?}. then some kidn of buffering issue with stderr not being buffered pops up. dunno why this would be the case. but i just ran without doing that and it worked. maybe it is intermittent though. 04:13 < ryouma> i did try stdbuf to make both stderr and stdout unbuffered and both line-buffered. but that did not fix it. 04:14 < ryouma> but if it /is/ related to that tee, then idk what to do to log it. whatever. i'll do without logging if this is not intermittent. unless anybody knows a fix. 04:22 < ryouma> matsaman: dpkg-reconfigure grub-pc in a chroot did not produce the intro text and list of places to install grub to, but did include the prompt. 04:23 < ryouma> boingolov: the lack of intro also suggests that it is a glitch or redirection or buffering type issue rather than a finding drives issue 04:23 < matsaman> uh... had we been talking to each other? 04:23 < ryouma> 19:09 I'm running out of ideas, heh 04:23 < ryouma> 19:09 'bout what? 04:23 < boingolov> ryouma: you can disable stdout buffering, not sure if that would help 04:23 < matsaman> that's a 'no', then 04:24 < ryouma> boingolov: tried that with stdbuf -o0 -e0 command 04:24 < ryouma> matsaman: great, we are o9n the same page now 04:24 < matsaman> so you say you can't redirect output? 04:24 < ryouma> no 04:25 < ryouma> boingolov: i tried it on dpkg-reconfigure inside teh chroot inside teh redirection. presumably that is the correct thing. i also tried a stdbuf -eL on the whole thing being teed. 04:27 < Amm0n> Hey guys, i got problems running saned. This is what i get: http://termbin.com/fzn6 04:27 < BenderRodriguez> Hello ShotokanZH 04:28 < Amm0n> using init.d instead of xinet.d i get [saned] run_standalone: waiting for control connection 04:29 < ryouma> matsaman: it is either intermittent or not. if it is not, then i can't redirect 2 to 1 and tee to keep a log. 04:31 < matsaman> why not? 04:32 < ryouma> dunno 04:33 < matsaman> what's this got to do with grub, again? 04:45 < pingfloyd> ryouma: what are you expecting as output with your parameter expansion? 04:46 < pingfloyd> I don't see where that var is define 04:46 < pingfloyd> defined 04:51 < Dan39> :o 04:52 < Limp_Trizkit> :O 04:55 < ryouma> pingfloyd: it's a fragment from a shell function that defines it, and is called with a bunch of args 04:56 < ryouma> that it just passes along. it has worked for years. 04:57 < matsaman> worked for years you say 04:57 < matsaman> I hope it doesn't rely on GNU coreutils! =P 05:00 < afidegnum> hello, i m having tought time formating my dual-boot system 05:00 < pingfloyd> ryouma: how about showing the function 05:00 < afidegnum> i have windows with 2 primary partitions, and trying to format linux, 2 root and swap primary partitions, but the system is not allowing me 05:00 < afidegnum> what should i do ? 05:02 < pingfloyd> ryouma: best if you can paste all involved files 05:02 < pingfloyd> ryouma: this will give a better idea of what the problem might be 05:02 < matsaman> afidegnum: GNU/Linux doesn't require primary partitions 05:03 < afidegnum> ok 05:03 < ryouma> 19:57 I hope it doesn't rely on GNU coreutils! =P -- matsaman: what is the issue with that? 05:03 < matsaman> ryouma: oh they've just made >0 ridiculous changes of late 05:05 < ryouma> like what? 05:06 < matsaman> like putting single quotes around files with spaces in the name so people who are two years old won't have their heads explode (or won't they) 05:07 < boingolov> spaces have no places in file names 05:07 < matsaman> spaces are just characters, like everything else 05:07 < matsaman> and look, we use them all the time in our language 05:07 < cmj> what_do_you_mean? 05:08 < matsaman> iswydt 05:09 < ryouma> where do they put quotes? sounds bad. but we live in a heterogenous computing environment so we end up with files with spaces. filesystems should disallow them if they are bad. this is all sh's fault. 05:09 < boingolov> most *nix don't use spaces in any standard directory names / file names, but macOS uses them everywhere 05:10 < boingolov> I *hate* that 05:10 < matsaman> hard to not hate macOS things 05:10 < ryouma> i hated osx too. kept trying to make it look like linux. 05:10 < matsaman> but it's worse they use caps can camelCase 05:10 < matsaman> and cardigans 05:10 < ryouma> work like* 05:10 < boingolov> matsaman: well, I use macOS as a daily driver, I obviously dont' hate all of it 05:11 < boingolov> but I do have hate in my soul for some of it 05:11 < ryouma> coreutils i forgot what programs that is. but they add quyotes in output? that yoh ave to remove? 05:11 < matsaman> like being in a straitjacket 05:11 < matsaman> ryouma: to 'ls' 05:12 < ryouma> that will break a lot of scripts 05:12 < ryouma> ls|whatever 05:12 < ryouma> ls -1 05:13 < boingolov> apple is also anti-developer. they shroud their API's in secrecy, document them poorly / update them rarely even though the underlying sdks have changed drastically, and require you open paid support tickets to get answers. on the other hand, some of it is elegant 05:13 < boingolov> I like Swift 4 05:14 < boingolov> a lot 05:20 < ryouma> the change onlya pplies to interactive, non-usable output right? i don't really care. nice to know where the spaces are. 05:20 < ryouma> but i think the thing that filesystems must not allow is NEWLINES 05:21 < ryouma> filenames should never have spaces on linux, but they will. but newlines? no os should use those. 05:21 < ryouma> perhaps the introsudcution of powershell on windows will discourage newlines 05:29 < sauvin> rename 's/\s+//g' * 05:34 < markasoftware> sauvin: wrong window? 05:35 < peeenis> greetings, earthlings. i am from outer space. my penis is ten feet long. 05:36 < sauvin> markasoftware, no, that's my answer to whitespace in filenames. 05:36 < peeenis> white space is fine in filenames 05:36 < peeenis> as long as you escape it properly 05:36 < afidegnum> hello, is it possible to create an extended partition made of root, swap and home partitions ? 05:37 < ryouma> sauvin: just be sure to attach it as a hook to every downloading software and transfer software or inotify 05:37 < Dan39> afidegnum: why still dos partitions? 05:37 < peeenis> and use octal dump to read it if there is more than one that look the same 05:37 < ryouma> afidegnum: iiuc yes 05:37 < afidegnum> Dan39: dual-booting 05:37 < Dan39> i thought grub files couldnt be in extended 05:38 < ryouma> afidegnum: dunno but if so does root include boot? 05:38 < Dan39> afidegnum: not with windows i guess? i know 10 supports gpt/uefi 05:38 < afidegnum> ryouma: yes 05:38 < peeenis> or you can even use dir | sed s/\ /_/g 05:38 < ryouma> then idk 05:38 < peeenis> to turn all the spaces into underscores 05:38 < ryouma> but i prefer separate boot partition 05:38 < ryouma> i-prefer-this 05:38 < peeenis> and see the whitespace in your file names 05:38 < afidegnum> the primary paritions are limited 05:40 < peeenis> cd ~ ; touch fuck\ this ; ls | sed s/\ /_/g 05:40 < peeenis> you should see the space just fine 05:41 < Dan39> afidegnum: should be fine from the searches i just did 05:41 < peeenis> does the latest build of arch support 3.5" diskettes still? 05:41 < Dan39> afidegnum: LVM is also an option 05:42 < peeenis> LVM is homosexual 05:42 < ryouma> you can jhust stick root on a primary if needed 05:42 < Dan39> peeenis: so...? 05:42 < peeenis> Dan39: just saying 05:42 < peeenis> you want your computer to take it up the ass and die of AIDS 05:42 < peeenis> then have fun 05:43 < ryouma> ok that's enough 05:43 < peeenis> LVM is known to literally set hard drives on fire 05:43 < ryouma> where there are medicines for it aids does not kill you 05:43 < ryouma> i.e. if you have access to medicines 05:44 < peeenis> it re-writes your drive firmware's kernel, and increases the RPM beyond safe limits 05:44 < peeenis> ryouma: life expectancy with AIDS is reduced substantially compared to not having it, even with treatment 05:44 < snkcld> i dont know if it was a pulseaudio person, or a bluez person, but whoever fixed a2dp to allow for my headphones to work, if youre ever in austin, hit me up and i will buy you a beer 05:44 < peeenis> unless you live in a sterile glass bubble 05:44 < ryouma> peeenis: good sources please? 05:44 < snkcld> i will let you stay in my house for free, i will give you a massage 05:45 < peeenis> ryouma: try the motherfucking CDC or mayo clinic, fag 05:45 < Elladan> LVM is the best thing. 05:45 < peeenis> and they say the lord does not strike people dead anymore 05:45 < peeenis> not directly anyway 05:45 < Elladan> Once you try LVM you never go back. 05:45 < afidegnum> what's the differene between GPT and MBR ? 05:45 < peeenis> but the #1 cause of AIDS in the developed world is gay anal sex 05:45 < afidegnum> and UEFU 05:46 < ryouma> so this is about hatred of homosexuzls then. or something. 05:46 < afidegnum> UEFI 05:46 < peeenis> anal sex fags 05:46 < peeenis> ryouma: no 05:46 < peeenis> i will tell you what its about 05:46 < peeenis> its about the lulz 05:46 < peeenis> and goatsex 05:46 < peeenis> we are the Gay Nigger Association of America and we fuck pigs 05:46 < cmj> snkcld: it's all seemingly working well. i use my bluetooth speaker sync without issue now 05:46 < peeenis> !ops all hail the GNAA 05:47 < snkcld> cmj: whoever did this was an angel 05:47 < snkcld> i updated/installed both at the same time, so i dont know which fixed it unfortunately 05:47 < cmj> pulseaudio -l shows many syncs, and just having the mixer open shows how dynamic it is 05:47 < Sitri> afidegnum: GPT is the replacement for the MBR partition-tables. MBR can mean two things, the DOS-style partition tables, and the actual Master Boot Record. UEFI is the replacement for BIOS. 05:47 < snkcld> cmj: i dont understand what you mean about "shows how dynamic it is" 05:48 <@sauvin> !@#%@$# 05:48 < cmj> mpd/ncmpcpp and bluetooth speakers is my life currently 05:48 < Elladan> Thank you. 05:48 <@sauvin> peetaur, that was a mistab, and I apologise. 05:49 < ryouma> nice to see that little o option 05:49 < cmj> they dynamic part is it picking up many applications sending to pulseaudio 05:49 < ryouma> women with hiv are livign to 76 on average in europe 05:49 < afidegnum> here while looking for the partition types, there is a long list, i can't find the primary/logical/ instead GPT, Extended... up to 45 options 05:50 < Elladan> afidegnum, MBR is an old legacy partition format. GPT is a much newer one. 05:50 < snkcld> cmj: pulseaudio -l says that -l is an invalid option 05:50 < Elladan> afidegnum, UEFI ought to be a particular GPT setup, but I'm not sure. 05:51 < afidegnum> so should i chose GPT as the partition type? 05:51 < Elladan> afidegnum, there are no primary/extended/logical etc. with GPT. Instead you just have partitions with ids and names. 05:51 < ryouma> afidegnum: is your machien old or new? 05:51 < afidegnum> recent one, 05:51 < cmj> snkcld: sorry, pulsemixer -l 05:51 < afidegnum> Dell Precision m4500 05:51 < Elladan> You probably want GPT. 05:52 < afidegnum> so root partition is GPT, right ? 05:52 < Elladan> The only issue with GPT that I know of is that you may need to manually reserve some space at the beginning for grub in some versions. 05:52 < afidegnum> ah 05:52 < Elladan> GPT/MBR are partition table types. You format the drive with them and then create partitions. 05:55 < gardotjar> ay hoesers, anyone here try Mint 19 yer? I heard it was supposed to be pretty good 05:55 < gardotjar> yet* 05:55 < Elladan> Nah they haven't released upgrade instructions yet. 05:55 < Elladan> I mean I'm sure it's just "fix your deb resources and do a dist-upgrade" or something. 05:56 < gardotjar> I wouldn't know, normally I do a backup, then use the update as an excuse to reformat... 05:59 < Elladan> I get annoyed digging through dozens of random configuration things. 05:59 < gardotjar> fair enough 06:00 < Elladan> I mean where did I put the hack I had to throw together to work around NetworkManager not reconfiguring dnsmasq correctly when I activate openvpn again? Do I need to rebuild that from scratch? 06:01 < gardotjar> That's normally the problem with customizing everything about your system though 06:01 < bls> if your workflow isn't directly support by NM, the usual approach is to get rid of it and do things yourself 06:01 < Elladan> It's directly supported. 06:01 < Elladan> It's just /buggy/. 06:01 < bls> because it expect to be in control of everything 06:01 < Elladan> Like basically everything NM does. 06:02 < afidegnum> hello, i coul't get the last message, i was disconnected 06:02 < afidegnum> time out 06:02 < bls> yeah, I keep trying to use it, but I know how to make things work without it and the things it does just make me angry 06:02 < afidegnum> the root partition, what do i choose between Linux and GPT ? 06:03 < Elladan> GPT is a partition table format. You format the disk GPT or MBR and then create partitions. 06:03 < jim> afidegnum, how large is the drive? 06:03 < Elladan> Honestly I generally use MBR, actually, because you only really need 2 partitions: /boot and LVM. 06:04 < Elladan> ... and there are less irritating grub things with MBR. 06:04 < Elladan> OTOH you need to set some setting or other in the BIOS to use MBR these days. 06:04 < afidegnum> jim: i have 2 big partitions which 250G each 06:05 < afidegnum> 1 for win and 1 for Linux 06:05 < Elladan> You have to use GPT with modern windows. 06:05 < afidegnum> in which i m splitting root/swap and logical partitions 06:05 < jim> afidegnum, ok, and, how large is the whole drive? 06:05 < afidegnum> 500G 06:05 < Elladan> However, you usually need to install windows first anyway. 06:05 < afidegnum> I did installed already 06:06 < jim> ok, so then there's no longer a choice with partition table type... how much space is left unpartritioned on the drive? 06:07 < afidegnum> 231G 06:07 < afidegnum> i tried setting 10 for root partition 06:07 < afidegnum> and 8 for swap 06:08 < jim> afidegnum, would you like to try lvm? 06:08 < jim> do you understand what it is? 06:08 < afidegnum> no i don't understand much of Lvm, how do i set it ? 06:09 < jim> the basic idea is this... with LVM, instead of putting filesystems in partitions directly, you put 'LVM physical volumes' in partitions, then you make a 'volume group' (which is just a list of physical volumes), and put physical volumes in the volume groups, then you can make 'logical volumes', and these are allocated out of the volume groups 06:10 < jim> are you in the installer now, or what are you running right now? 06:10 < afidegnum> yes, 06:11 < jim> ok, and which piece of the installer are you in? 06:11 < Elladan> Is this a laptop, i.e. do you want encryption? 06:11 < jim> good question 06:11 < afidegnum> well, i trying to install archlinux on the Laptop i don't need encrytion 06:11 < afidegnum> yet 06:11 < jim> ok 06:12 < afidegnum> there is Windows alreday installed 06:12 < afidegnum> i just want to install Linux side by side 06:12 < jim> did you want to do arch or mint? 06:12 < afidegnum> arch 06:13 < jim> ok, I personally don't know much about arch (or what you need to do first) 06:13 < WhiteDevil> i saw a tutorial on arch 06:13 < WhiteDevil> that gives you all the commands to install it with a gnome desktop 06:13 < WhiteDevil> bare minimum install 06:14 < afidegnum> but i read windows do not support LVM, how do i make it compatible ? 06:14 < jim> correct, if you do lvm, windows doesn't know how to do lvm 06:15 < Elladan> Windows doesn't support any linux filesystems, to a first approximation. 06:15 < afidegnum> yes, weird 06:15 < afidegnum> is there a walkaround ? 06:15 < jim> but linux can mount windows partitions 06:16 < Elladan> Windows will just ignore the Linux partitions, because it doesn't understand them. 06:17 < jim> afidegnum, what would you be able to do if they were compatible? 06:17 < jim> hold that thought... be right back 06:18 < cmj> i just ignore windows 06:18 < afidegnum> ok 06:18 < afidegnum> cmj: i wish i could ignore windows but there are other windows utilities i need to use 06:19 < markasoftware> It is possible to setup windows to boot linux from its bootloader 06:19 < markasoftware> I don't recommend it 06:19 < Elladan> Overall I think it's a lot nicer to just run Windows in a VM than to dual boot. 06:20 < afidegnum> i tried it before, it wasn't a good experience 06:20 < afidegnum> not a smooth one 06:21 < jim> ok, back. what do you want to do with a compatibility between them? 06:21 < afidegnum> so i can seemless decide to either use Windows or linux 06:22 < jim> well for one thing lots of games don't run on linux' 06:23 < jim> you mentioned windows utilities you want to run... which ones are those? 06:26 < afidegnum> some office related applications which do not work well with winehq as well as Autocad, People Link Video Confernce utility etc.. 06:26 < markasoftware> Wtf is that last one 06:27 < afidegnum> :D 06:28 < cmj> office related? 06:28 < jim> yeah I've never heard of People Link either 06:28 < cmj> libreoffice should suffice 06:29 < jim> cmj, well we don't actually know what he does, so we don't know if libreoffice would work or not 06:29 < Elladan> If this stuff is for work, the alternate option is to install Linux in a VM. 06:29 < pingfloyd> afidegnum: dual booting isn't going to be seamless 06:30 < cmj> 'office related' is vague, but i would presume ms office 06:30 < afidegnum> for office, i use LibreOffice/OpenOffice 06:30 < Elladan> If you're not very comfortable with partitioning and the installer, a VM is a really good place to start. 06:31 < jim> afidegnum, what attracts you to linux? 06:31 < pingfloyd> VM is what you want if have vt-x and/or iommu 07:02 < sbef> hello guys 07:03 < sbef> yesterday i-ve installed q4os and i kept getting hash dismatch everytime i was trying to update apt. i tied any solution i found on the net... non was working. So now i was going to reinstall it, but even the installation usb gives me the same error! 07:03 < sbef> why is this happening ??? 07:05 < sh1ro> what the hell is q4os and why are you installing it at the end of q2 07:05 < efloid> does anyone know why 'offline uncorrectable sectors' count would DECREASE? was at 688 for a long time and now went down to 624 07:05 < sbef> sh1ro: its just a debian based distro? 07:05 < sh1ro> uhhh miraculous fix? they should stay dealloced for ever 07:06 * sh1ro would note some advanced malware hides in artificial bad sectors 07:07 < efloid> sh1ro: interesting 07:07 < sh1ro> though as far as i know there isn't a single thing that can dectect such malware 07:07 * sh1ro would live boot and copy those sectors, might be interesting 07:08 < sh1ro> sbef: ask the devs if they have a channel but sounds like repo is kucked 07:08 < zenix_2k2> is there anyhow to restart my nautilus on my Ubuntu ? because my current FIle manager's interface looks so weird 07:08 < sbef> sh1ro: using differen mirrors does not solve the problem 07:09 < sh1ro> zenix_2k2: update and reboot 07:09 < efloid> sh1ro: i'm pretty sure the sectors really *are* bad. i've tried using hdparm to read and write from them and there are errors 07:10 < sh1ro> they probably are, warez that advanced are rare 07:11 < cmj> so use fsck? 07:12 < sbef> ok thanks for not helping 07:17 < afidegnum> hi 07:17 < zenix_2k2> sh1ro: i still don't get it why updating can solve it ? 07:18 < afidegnum> hi, i m back again 07:18 < lnnb> welcome back! 07:18 < afidegnum> i have an initial partition ntfs on which win7 is installed, 07:18 < afidegnum> can i convert that partitionto lvm ? 07:18 < lnnb> ewwwwwwww 07:18 < `7hr34t_hvntr> hello 07:18 < lnnb> hallo! 07:19 < `7hr34t_hvntr> did a remote port fwd a la ssh -R 5555:127.0.0.1:10001 foo@172.16.2.2 07:19 < MLarabel> hey dorks!! 07:19 < `7hr34t_hvntr> killed the terminal session i used to do that, but the port fwd is still up 07:19 < MLarabel> it's me M. Larabel 07:19 < lnnb> haaayyyyyyyyyyy 07:19 < `7hr34t_hvntr> isnt it supposed to die with the terminal session 07:21 < cmj> shell in and kill the process 07:22 < afidegnum> 3 07:22 < `7hr34t_hvntr> so it doesnt systematically die with the closure of the terminal session 07:23 < MLarabel> i love lamp 07:23 < cmj> some will hang, for sure 07:23 < `7hr34t_hvntr> its not just hanging its functional 07:24 < MLarabel> some will hang for their crimes, for sure 07:25 < cmj> good luck 07:34 < phinxy> Whats the one character program or shellscript that adds bookmarks to your terminal? 07:35 < MLarabel> phinxy: first of all, great question 07:36 < phinxy> Might have been "Z", https://github.com/rupa/z 07:40 < micrex22> psi-jack https://upload.wikimedia.org/wikipedia/commons/c/cc/IBM_2210_Router_Interfaces.JPG <-- here you can see the different interfaces for token ring (minus BNC) and AUI near ethernet 07:41 < micrex22> I don't know what those WAN connectors are 07:41 < micrex22> the RJ11s are for *dial-in* VPN connections, lol... 07:41 < phinxy> I found it: Q. https://github.com/cal2195/q 07:42 < taaperotassu> Why when I put program outputs to a file >> it can't be read by less unless I use cat? 07:43 < rasputozen> taaperotassu: you could do ' >> file; cat file | less' 07:44 < kerframil> taaperotassu: what are you actually trying to do (without naming any commands)? 07:48 < MLarabel> phinxy: you tried all 26 and it was the last one? 07:49 < zenix_2k2> is there anyhow to restore the sidebar in my File manager ? i kinda lost it 07:50 < zenix_2k2> cause this is what i only have https://ibb.co/iBFFhJ 07:50 < MLarabel> zenix_2k2: sidebar, your honour? 07:50 < zenix_2k2> what ? 07:51 < MLarabel> zenix_2k2: do you like skateboarding? 07:51 < zenix_2k2> dude i really don't get what are you saying 07:51 < kerframil> zenix_2k2: click the Files menu 07:53 < zenix_2k2> the point is there is no Files menu 07:53 < kerframil> zenix_2k2: ... it's right there in your screenshot 07:54 < zenix_2k2> just look again 07:54 < zenix_2k2> oh wait 07:54 < kerframil> I suggest that you do the same 07:55 < taaperotassu> kerframil: Just tried megatools and thought I could put the output to files. 07:55 < taaperotassu> maybe the output has some strange format 07:55 < zenix_2k2> Hm, why didn't i see that 07:55 < zenix_2k2> anyway thk 07:55 < taaperotassu> wondering if this is usual with other things. 07:55 < zenix_2k2> thank* 07:55 < MLarabel> linux: 2 words: it's the best 07:56 < kerframil> taaperotassu: >> redirects stdout. if the application doesn't convey anything to stdout, nothing will be accumulated in the file to which the direction occurs. 07:56 < kerframil> redirection* 07:58 < MLarabel> Does anyone here have a problem? 08:02 < alexey-nemovff> MLarabel: of any kind? 08:02 < alexey-nemovff> xD 08:05 < MLarabel> alexey-nemovff: yea i am a clinical psychologist 08:06 < alexey-nemovff> I suspected 08:07 < alexey-nemovff> where're you from? 08:07 < MLarabel> i'm a fancy cosmopolite from toronto 08:11 < lnnb> MLarabel: a plane half-loaded 747 leaves chicago at 5:00pm, another plane, fully loaded 747 takes off from arizona at 5:35. there is a two hour layover in atlanta. when the first plane arrives at cruising altitude there are 35,000 gallons on board, the scond plane has 39,000 gallons. when 5/7'th of the remaining fuel has been consumed in the first plane how many gallons remain in the second plane? 08:13 < pingfloyd> what is a cosmopolite? 08:14 < MLarabel> lnnb: that is a dilly of a pickle 08:14 < giby> Hi, I've got a 16GB fasta file, I would like to cut it down for a soft I need has only possibility to support 1GB… what would be the commande tout it, considering the character where it is allow to restart a file is ">" 08:14 < lnnb> you can assume 0.2MPG fuel efficiency 08:14 < MLarabel> pingfloyd: it means 'a cosmopolitan person.' 08:15 < lnnb> if it's too hard to do the other way 08:19 < lilltiger> pingfloyd: it is a word for the unified french culture groups 08:21 < [666]> is there an easy way to get an rt2870 driver for a WUSB600N linksys wireless dongle anymore? 08:26 < Shawn|i3-350M> howdy 08:26 < Shawn|i3-350M> anyone here use peppermintos? 08:30 < dsawr> how do we check which process takes ram 08:30 < dsawr> I can use a top command for this 08:30 < dsawr> but lets see if a process consumes lot of ram, how do we check why that process is taking more ram 08:32 < lilltiger> [666]: mt7601 should be the name of the drivers for it, but not sure there is a working one for kernel 4.x 08:33 < lilltiger> dsawr: Cachegrind 08:33 < lilltiger> but it's not a trivial tool to use 08:35 < afidegnum> is it possible to block all ports by default and allow only specific ones ? 08:36 < afidegnum> just the issue is i dont' know list of know used ports 08:39 < TheWild> hello 08:40 < SysGhost> \o 08:40 < jim> hi 08:43 < masber> good afternoon, is this the right way to format a disk? dd if=/dev/zero of=/dev/nvme7n1 bs=3GB status=progress 08:44 < kremator> guys, with the IP network utility, is different "ip link set dev eth0 down" the same as "ip link set eth0 down" ?? 08:45 < jim> Shawn|i3-350M, the best thing to do is ask a specific question concisely and completely, and include as many informative details as you know 08:46 < jim> masber, you don't need to do that 08:46 < jim> masber, what are you going to use the drive for, and how large is it? 08:47 < TheWild> standard input/output. I want to make two programs communicate with each other and all that s*** hangs, probably because of deadlock. Yes, I'm a newbie. script.sh http://termbin.com/33gt, child.py http://termbin.com/oh59 08:47 < masber> jim, I am planing to redeploy ceph (a distributed storage solution) and need to format the disks, otherwise it will find the old partitions and will fail 08:47 < TheWild> what I should read before? 08:47 < kerframil> masber: then you only need to scrub the metadata that describes any prior partitions. the correct command for that is wipefs. 08:47 < Elladan> masber, don't do bs=3GB, do bs=1M or similar. That will write zeros to the disk. If that's what you want, sure. 08:47 < kerframil> masber: wipefs -a, specifically 08:48 < jim> masber, (it's best to limit writes to an ssd type devive) 08:48 < jim> device 08:48 < lilltiger> masber: just set up the new partition table with your favorite partitioner then use mkfs on the partition you want a fs on 08:48 < Elladan> masber, if you just want to format it, you just need to create partitions / tables and do mkfs. 08:48 < kerframil> masber: alternatively, you can use blkdiscard, if your device supports it. that will instruct the NVMe device to discard all of its blocks, which (usually) effectively wipes everything. 08:48 < HaMsTeRs> Hey my people 08:48 < kerframil> masber: but do not use dd 08:49 < Elladan> masber, also if it's an nvme device, "blkdiscard /dev/nvme7n1" will probably be a thousand times faster. 08:49 < masber> ok 08:49 < jim> hi 08:49 < TheWild> maybe dc is buffering its output, but unbuffer doesn't really help in this case :( 08:53 < CrazyTux> do we have linux alternatives for all or almost all the windows applications? 08:54 < jim> windows? what's that? 08:54 < CrazyTux> why are some linux apps so bland? as compared to those available for windows. 08:54 < CrazyTux> for example, download managers. 08:54 < azarus> there's a thing called wine if you want to execute winodws programs 08:54 < kerframil> afidegnum: "block all ports" is an ambiguous goal 08:54 < jim> I guess windows cooks with garlic? 08:54 < azarus> "download managers" are eh. aria2 ftw 08:54 < azarus> or curl/wget/whatever 08:55 < CrazyTux> I don't use Windows and don't intend to. 08:55 < CrazyTux> why uget or other download managers don't have those features that some windows applications have? 08:55 < jim> CrazyTux, if you knew that in 10 minutes you could never use wnything of windows ever again, would that be ok? 08:55 < Elladan> What's a download manager and why would I want one? 08:56 < azarus> aria2 has all the features a downloading application would ever need 08:56 < kerframil> afidegnum: for instance, if you really did block all "ports" for both ingress and egress traffic, you wouldn't be able to use TCP or UDP at all. 08:56 < CrazyTux> Elladan, to capture downloadable files and download them. 08:56 < azarus> bittorrent support, multiple links, whatever 08:56 < azarus> what kind of file isn't "downloadable" anyway 08:56 < CrazyTux> jim, what? 08:57 < kerframil> afidegnum: I mention this because being clear as to one's requirements is an important place from which to begin when implementing any kind of packet filtering policy 08:57 < Elladan> ... I can just click on links to files in my web browser and somehow it downloads them and puts them in a directory. 08:57 < lilltiger> CrazyTux: as linux dosent crash randomly we dont need anything but wget to download files ;) 08:57 < jim> be right back (hold that thought) 08:57 < CrazyTux> azarus, some download managers have this feature of speeding up the downloads. 08:57 < azarus> CrazyTux: aria2 does too by using multiple links 08:57 < Elladan> I can also copy the uri and give it to wget or curl or whatever. Why would I want some other program to do this? 08:58 < CrazyTux> azarus, ok. I will try that. 08:58 < CrazyTux> Elladan, I will try that too. 08:58 < CrazyTux> lilltiger, I must agree to what you just said. 08:59 < afidegnum> kerframil: thanks for the insight, 08:59 < Elladan> Why would I want to be a bad internet citizen and mis-use TCP congestion throttling by spamming a server with multiple connections for no reason? 08:59 < kerframil> CrazyTux: because the diversity - and in some cases - the quality of amateur desktop applications is higher on windows than on any other platforms. any other questions? 08:59 < afidegnum> recently there was malware attack which sent huge traffic out resulting on my isp blocking my server 08:59 < afidegnum> so to solve the issue is to mitigate/restrict lot of servics 09:00 < afidegnum> as well as fix security patches 09:00 < CrazyTux> kerframil, ok 09:00 < afidegnum> there were ports raning from 40555 to 59855 which i never used before were among those traffics 09:01 < afidegnum> that's why i want to block all ports by defaults and only allow internal and outgoing ones 09:01 < CrazyTux> do kernel versions have anything to do with a linux distro freezing randomly? 09:01 < Elladan> CrazyTux, it's possible. 09:01 < maxxik> Hi There 09:01 < kerframil> CrazyTux: there are many talented amateur developers that simply have no interest in developing on any other platform. until that changes, it's to be expected. 09:01 < sh1ro> maybe? 09:01 < sauvin> Use a VPS. 09:02 < CrazyTux> kerframil, ok 09:02 < CrazyTux> Elladan, ok 09:02 < Elladan> CrazyTux, your machine can freeze for all sorts of reasons. A kernel driver bug is one common reason. 09:02 < kerframil> CrazyTux: and professional developers tend to expect to be able to monetise their offerings 09:02 < Elladan> CrazyTux, bad hardware is another common reason. 09:03 < CrazyTux> ok 09:03 < maxxik> gents, if I have old self signed root certificate CA1 (installed on clients) and would like to replace it by new self-signed CA2 (issuer/subject will be changed) - everything signed by CA1 is invalid and ideally sgould be replaced by ld be replaced by signed by CA2. Question - wil lcross-certification help to aviod this ? 09:04 < HaMsTeRs> hey guys, I just installed ubuntu, and it can definitely replace my desktop 09:04 < CrazyTux> I have two distros intalled on my new laptop. Manjaro which is a rolling release distro and MX 17.1 which is a Debian stable based distro. MX freezes sometimes. I have never had Manjaro freeze, not even once. 09:04 < HaMsTeRs> with Windows 10 09:04 < CrazyTux> both are on the same hardware. 09:04 < HaMsTeRs> I"m currently looking for a clean glassy theme. any recommendation? 09:05 < Elladan> CrazyTux, bad hardware can be exposed by a kernel (or OS!) which does nothing wrong but somehow just randomly hits it often, while it might be stable or rarely crash with another kernel or OS that's no better. 09:05 < xdije> hi 09:05 < kerframil> afidegnum: as in, the destination port or the source port? 09:06 < CrazyTux> ok 09:06 < Elladan> CrazyTux, so it can be quite hard to track down the causes of these things, since negative results don't necessarily imply a causal factor. 09:06 < CrazyTux> ok 09:06 < xdije> is it possible to run a command before a service starts, and depending on the result if ok start the service otherwise stop it or keep it stopped 09:06 < kerframil> afidegnum: filtering on the source port is a non-starter, because applications need to use a standard ephemeral port range from which to originate their (legitimate) connections. filtering outbound traffic on destination port is reasonable, but must be done with care. 09:06 < Dagmar> First question would be, "How frozen is frozen?" Checking that the numlock key still makes the light go on and off is one thing (if it still works it's not 100% crashed). Having a second machine available you can try ssh'ing in with is another way (sometimes people are tricked when the GPU driver falls over and leaves the console apparently dead) 09:07 < CrazyTux> btw, anyone here been using Solus now? 09:07 < kerframil> afidegnum: in short, yes - it's possible 09:07 < Elladan> CrazyTux, however, changing kernel versions and/or disabling some hardware features is a good place to start. 09:07 < sauvin> I've had problems with X freezing but things running in VCs still running normally. 09:07 < CrazyTux> Dagmar, when it freezes nothing on keyboard works 09:07 < Dagmar> xdije: You can do pretty much whatever you want prior to starting a service if you write the service management script yourself 09:07 < CrazyTux> not even mouse. 09:07 < Dagmar> CrazyTux: Not even the numlock light? 09:08 < CrazyTux> Dagmar, not even that. 09:08 < Dagmar> That would be a sign you had a kernel panic then 09:08 < Dagmar> I'd compare the video drivers being used 09:08 < CrazyTux> It happens very infrequently though. 09:08 < Dagmar> If nouveau is involved, it's usually what's gone awry 09:08 < sauvin> Attention: there will be no more unexpected kernel panics without a minimum of five minutes' advance warning. 09:09 < Dagmar> ...particularly if you have one of those wonderful Intel/nVidia hybrid video chipsets 09:09 < Dagmar> They are _touchy_ 09:09 < Elladan> Yeah graphics drivers are a good thing to check too. 09:10 < Dagmar> They're generally an order of magnitude more unstable than the kernel 09:10 < Elladan> Have you looked in your log files to see if there are any interesting kernel oopses etc? 09:10 < CrazyTux> I have this laptop https://www.asus.com/Laptops/X540LA/ 09:10 < CrazyTux> It's a new laptop. 09:10 < Dagmar> ...which is still easily just 100x times a very small fraction 09:10 < TheWild> back to my question, yitz from #bash recommended me to use coproc command. Yay! It worked, an my script has been shortened more than twice. 09:10 < kerframil> afidegnum: consider this: https://bpaste.net/raw/e475dd9b9742. to policy outbound traffic, one would simply filter the OUTPUT chain, in much the same was as the INPUT chain is filtered in that example. 09:10 < kerframil> afidegnum: and, if the host is acting as a router, the FORWARD chain. 09:11 < Dagmar> CrazyTux: The stepson has one of their ROG laptops. That thing is wildly unstable with anything but the exact version of the driver it shipped with 09:11 < kerframil> police* 09:12 < Kremator> bahs is love, bash is life : http://dpaste.com/31P9355 09:12 < azarus> i don't like bash. ksh is my favourite 09:13 < Dagmar> You know you generally don't have to disconnect to query the local SSID list, right: 09:13 < CrazyTux> I will keep updating/upgrading the distro. Probably with newer kernel the problem gets sorted out. 09:13 < SysGhost> *cough* zsh .... 09:13 < azarus> don't even have bash installed on most of my systems 09:13 < Dagmar> You should probably leave. 09:13 < Kremator> Dagmar, I KNOW, BUT I WANTED TO DROP COMPLETELLY THE NIC 09:13 < Dagmar> Filthy preverts 09:13 < CrazyTux> Dagmar, I didn't disconnect. 09:13 < afidegnum> kerframil: from the top question, yes source port were multiplied 09:13 < CrazyTux> probably it was a network issue. 09:13 < SysGhost> what?... no.. I just coughed... did it sound like something? ... =o 09:14 < afidegnum> i m taking into consideration the paste sent 09:14 < Kremator> ooops sorry for the capitl letters 09:15 < CrazyTux> Solus seems to be a good Windows alternative. For Linux newbies. What do you say? 09:15 < azarus> ah, Psi-Jacks favourt 09:15 < azarus> favourite* 09:16 < SysGhost> solus? Yet another distribution... ¯\_(ツ)_/¯ 09:16 < Kremator> CrazyTux, solus? do you mean that cArch Linug distro? 09:16 < CrazyTux> they say it is a curated rolling release distro. 09:16 < Kremator> SysGhost, prone to crash, its based on Arch 09:16 < CrazyTux> https://solus-project.com/ 09:17 < CrazyTux> Kremator, no. It is an independent distro. 09:17 < azarus> "oh no, it's arch based! so unstable" not really 09:17 < Dagmar> you're using a different set of definitions than the rest of the world 09:17 < stevwills> hum arch is not really unstable as people claim it is 09:17 < CrazyTux> https://distrowatch.com/table.php?distribution=solus 09:17 < Dagmar> No. 09:18 < Dagmar> "Stable" and "unstable" have NOTHING to do with whether or not something crashes frequently 09:18 < Dagmar> Zero. Zilch. Nada. 09:18 < azarus> *triggered* 09:18 < SysGhost> Arch is one of the most stable rolling distros I've run. There are however some arch based distros that do things... well.. differently.. and are due to *that* tricky to keep stable. 09:18 < stevwills> anyone use a system like this one??? https://jurassicsystems.com/ 09:18 < CrazyTux> Dagmar, what is the definition of Stable and Unstable? 09:18 < stevwills> arch is actually really stable and bleeding edhe 09:18 < Kremator> CrazyTux, stable = that does not break X11 09:18 < Dagmar> A "stable" platform is one which does not change anything without very solid reasons. Those reasons are generally one or more of the following: 1) The update resolves a bug/exploit. 09:18 < Dagmar> Note that there is no #2. 09:19 < azarus> so basically any major distro with releases. 09:19 < SysGhost> Debian stable then, if that is what one seek. 09:19 < azarus> okay, noted 09:19 < Dagmar> An "unstable" distribution is one which can merrily add new features or just obsess over the newest version numbers of everything 09:19 < CrazyTux> ok 09:19 < Dagmar> These can be a pain in the butt if you have to manage a lot of them over time doing a bunch of different things, because any one of those updates *might* just impact or *break* your line-of-business apps 09:20 < CrazyTux> ok 09:20 < Dagmar> YEah when they call it "Debian stable" they're not making any promises of fitness for any particular purpose. They're telling you they're going out of their way not to break the stuff you're doing on your end. 09:20 < Dagmar> That it happens to be a very stable-doesn't-crash distribution is something they simply assume, like water being wet. 09:21 < SysGhost> If I compare Ubuntu with Arch (Yeah I know. Not a fair comparison), I'd say I had to reinstalls/"start overs" with Ubuntu far more times than any of my arch installations. 09:21 < CrazyTux> for a banking and financial service firm which type of distro is suitable? a fixed release one or a rolling release distro? 09:21 < Dagmar> Anything that makes it crash is something that's addressed rather viciously 09:21 < Dagmar> CrazyTux: A _stable_ release 09:21 < Dagmar> There's literally no way a fincen house is going to use rolling release 09:21 < sauvin> "fincen"? 09:22 < Dagmar> THat would be kind of insane 09:22 < CrazyTux> Dagmar, ok. Suppose the employees there have zero knowledge of linux OS, which distro would you recommend? 09:22 < Dagmar> Financial Center 09:22 < Kremator> CrazyTux, lubuntu 09:22 < CrazyTux> ok 09:22 < Dagmar> CrazyTux: Freakin' RHEL because then you can at least hire some RHCE's and get rid of the customer service reps posing as admins 09:22 < sauvin> And yet, we've seen a persistent trickle of folk stridently clamoring that distros like Arch are perfect for "server use". 09:22 < Dagmar> sauvin: I see people smoking weed all day all the time on YouTube 09:23 < Dagmar> Doesn't mean they're not brain damaged 09:23 < Kremator> Dagmar, sorry to interrupt, but those last ones seems fine to me 09:23 < retran> they don't have Arch on DigitalOcean anymore 09:23 < Kremator> bothing like a good joint while debugging that buggy box 09:23 < azarus> "aww yeah, my webserver config doesn't work anymore since it's a new version that I haven't used at all, it was released this morning" 09:23 < sauvin> That's YouTube. I just ran into a whole batch of people on YouTube obsessed by bodily functions in public, too. 09:23 < SysGhost> CrazyTux: for a mission critical system, one usually run something really stable. like a solid state distribution that only update for security reasons. Often admins tend to stick with debian SUSE or something. And never upgrade the distribution itself and relies only on security updates. 09:23 < CrazyTux> how about Mint 19? 09:24 < Dagmar> Someone who really wants to can put together their platform on Arch, but when it comes to production, they're probably going to be following a strict freeze policy with very strict dev/test/prod separation 09:24 < Dagmar> People who don't wind up having to explain DOWNTIME to management, which quickly convinces management to switch to Windows or hire better people 09:24 < Kremator> CrazyTux, tf? they are already on 19 :( damn man, i remeber when i grabbed up on loonix with Mint 17.1 in 2015 09:24 < retran> and what about security updates, Dagmar 09:24 < azarus> Dagmar: but then you have the problem that if an application happens to have a bug/exploit in it, arch doesn't provide security updates 09:24 < azarus> and partial upgrades are *not* supported 09:24 < Dagmar> retran: Have you ever seen a vuln that wasn't some form of bug? 09:24 < retran> not what I'm getting at 09:25 < Kremator> CrazyTux, Mint is ok, though for some reason i have seen many normal people hating the too green default appearance of mint 09:25 < Dagmar> It sounds to me like you're asking me to explain the rationale of crazy people 09:25 < CrazyTux> Mint now has Timeshift. Its update policy has changed too. 09:25 < HaMsTeRs> I feel cold in here 09:25 < HaMsTeRs> nobody cares about me 09:25 < HaMsTeRs> =< 09:25 < Kremator> HaMsTeRs, your mother does, and probably your father aswell 09:26 * SysGhost sends HaMsTeRs a care package of unknown contents 09:26 < retran> Dagmar, I guess that's the point. in a rolling distro, are the bug fixes ever backported 09:26 < retran> (like Arch) 09:26 < CrazyTux> HaMsTeRs, how are you? 09:26 < Dagmar> I've spent _plenty_ of time manually patching things for security vulnerabilities--usually because I'm a few hours or days ahead of the distro and I know _precisely_ what I'm doing 09:26 < retran> if you're going to "freeze" your Arch distro 09:26 < Kremator> Dagmar, sorry to interrupt but, how many years you have in the linu world? 09:26 < retran> you still have security backports to worry about. do those exist 09:26 < retran> ? 09:27 < Dagmar> retran: Offhand I'd say they mostly roll them out when they arrive, _with whatever codebase they went into_ so it could easily mean a package goes up a minor revision (which might mean someone's favorite feature got refactored, and now there's breakage) 09:27 < heftig> retran: you can't freeze the distro 09:27 < sauvin> What rolling release wouuldn't have "backports"? 09:27 < retran> yeah isn't that neat 09:27 < heftig> retran: there is only one state to update to 09:27 < Dagmar> Debian and RHEL and other "stable" releases *just* apply a patch to the package they've already release, incurring hte absolute minimal necessary change 09:27 < SysGhost> retran: not sure how security backports would work on rolling releases though. Usually it's not recommended to "freeze" such distributions. 09:27 < Dagmar> It's a pain in the ass to backport sometimes, but it beats trying to unravel what just broke in your 50,000 line project 09:27 < sauvin> And thus the more colloquial understanding of "stability", Dagmar. 09:28 < azarus> i do like alpine's release schedule, but sometimes they're a bit tardy ;P 09:28 < Kremator> SysGhost, specially because after 2 months you woudl have a lot fo new updates that all installed at once definitelly could generate problems 09:28 < Dagmar> sauvin: I am seriously considering putting a detailed explanation up on my website, because it seems like we've gotta explain this to new people 2-3 times a week 09:28 < retran> SysGhost, which means you might be faced with either making app changes or doing brain surgery backports 09:29 < sauvin> Feh, just explain there's a difference between the term's technical meaning and its more widely understood meaning. 09:29 < Dagmar> I write in hypertext, man 09:29 < SysGhost> Kremator: indeed. Waiting too long causes package conflicts in the end. a nasty hell to get out of. I've left too many arch installs behind myself =S 09:29 < azarus> 09:29 < heftig> azarus: alpine would be my container distro of choice if it wasn't stuck to musl 09:29 < jim> Dagmar, there are drugs for that... 09:30 < azarus> heftig: i use it exactly because of musl ;) 09:30 < sauvin> Probably some kind of SSRI. :D 09:30 < Dagmar> Like, my interest in hypertext literally had me on Usenet reading posts from Berners-Lee when they were working out how to make it go through gopher, because I was using it to write documentation 09:30 < azarus> gopher:// is experiencing a revival as of late :D 09:30 < sauvin> It is? 09:30 < azarus> kinda, yes 09:31 < jim> gopher... haven't hear that in awhile :) 09:31 < Dagmar> It used to drive me nuts when I'd write up something about how to use say, a green box, and people would read it and not even know what the tones were, and expect me to explain tone generation 09:31 < retran> where is evidence of this gopher revival? 09:31 < azarus> gopher://bitreich.org 09:31 < SysGhost> retran: it begins here. With us. =P 09:32 < retran> I guess gopher will it help me when I'm using a flip-phone on my trip to mars 09:32 < retran> low bandwidth menu driven content 09:32 < SysGhost> retran: no reception. 09:32 < retran> well, assuming I have any reception :p 09:33 < SysGhost> I've heard the radio shadows on mars are.. .extensive 09:33 < retran> oh maybe that's a good point. a mars based gopher server? 09:33 < Dagmar> Not "low bandwidth" so much as plain text and no real way to inline anything else 09:34 < Dagmar> It's amazing how lightweight things become when you throw away all the useless javascript, user trackers, service metrics, and stupid flashing graphics 09:34 < retran> and that's why advertisers love gopher, only plain text :p 09:34 < azarus> you can transmit files via gopher:// 09:34 < SysGhost> There's a few geostationary com satellites over Mars, isn't there? 09:34 < Dagmar> How many times have you followed a news link to find approximately three sentences of useful information 09:34 < azarus> like images and stuff 09:34 < Dagmar> ...and the webpage itself is a 4Mb download 09:34 < Dagmar> Four _million_ bytes just to convey probably 500 bytes uf useful info 09:35 < heftig> Dagmar: with autoplaying video 09:35 < retran> yeah, if the webpage is "optimized" (all javascript and css inline) 09:35 < Dagmar> Yawp 10:01 < taaperotassu> kerframil: Thanks so is there any way to make it show in plain text? 10:02 < kerframil> taaperotassu: I don't know what "it" is. it's still not clear what the actual problem is. 10:03 < peetaur2> [07:42] Why when I put program outputs to a file >> it can't be read by less unless I use cat? 10:03 < peetaur2> that? 10:04 < peetaur2> if your program with >> thinks it's going to a terminal anyway (even though it's not), then it will print out special chars that terminals will interpret...less will break those, but cat won't so the terminal will display them like before. 10:04 < peetaur2> for example rsync output will have lots of outputs for the speed and time remaining changing, but if you cat, it'll animate through the whole thing really fast 10:05 < kerframil> if that's what it is, then there's also less -R 10:05 < kerframil> ideally, the author of the utility being used would adjust it not to emit terminal escape sequences where stdout isn't a tty 10:16 < kuri0> how can i create a cpio archive with files in a directory called foo in it 10:17 < kuri0> i have the files but i want to be able to make them in a folder called foo inside the cpio archive without having to create a directory called foo 10:21 < Triffid_Hunter> kuri0: symlink the folder they're in, name the link 'foo' and cpio it? 10:21 < peetaur2> just give in and make a directory... why go through the trouble of avoiding it? 10:24 < FManTropyx> FileZilla is the worst program ever made 10:24 < Armand> *best 10:25 < Triffid_Hunter> FManTropyx: and ftp is the dumbest protocol, luckily it's easy to stop using both :P 10:25 < retran> you could be using CyberDuck 10:25 < Armand> Eeeeewwww 10:25 < retran> lol 10:26 < retran> i believe FileZilla has supported scp and sftp a long time 10:26 < retran> (if that matters) 10:27 < retran> you shouldn't be distributing artifacts via user-interactive apps like FileZilla on a routine basis anymore 10:27 < retran> (on a serious note) 10:27 < FManTropyx> I am using it for SFTP indeed and I'll look into CyberDuck, thanks 10:27 < Armand> Sometimes, I *have* to use FTP. 10:28 < retran> oh don't use CyberDuck 10:28 < retran> i was kidding bro 10:28 < Armand> Thankfully, plain FTP is becoming rare. 10:28 < Armand> But Filezilla is pretty much one of the best tools for the job. 10:28 < azarus> why not use scp/rsync or sftp? 10:29 < azarus> the command line utilities, i mean 10:29 < retran> you can with FileZilla (minus 'rsync') 10:29 < Armand> Because sometimes those choices aren't available. 10:29 < retran> some people like pretty icons 10:29 < retran> Armand, nonsense 10:29 < Armand> No, sense 10:29 < retran> No way. 10:29 < Armand> Way 10:29 < azarus> if you have openssh, you have scp and sftp 10:29 < Armand> And if we don't have that 10:29 < Armand> ? 10:30 < azarus> well then you don't have filezilla either 10:30 < retran> you can download it 10:30 < Armand> 3rd party = often annoying. 10:30 < retran> it's a smaller dep than "FileZilla" 10:30 < Armand> ¬_¬ 10:30 < retran> and less weird 10:30 < retran> at least you can script with command line utils 10:30 < Armand> What I mean is.. it does occasionally happen that a 3rd party will *only* permit FTP access. 10:30 < Armand> And we have no control over that. 10:31 < azarus> well, curl or wget then 10:31 < Armand> lolz 10:31 < Armand> Ever migrated a whole website and database(s) via wget ? 10:31 < azarus> sure 10:31 < retran> why would you do that 10:31 < rindolf> Hi all! "ls /mnt" as root just hangs there - /mnt/music contains an NFS share. umount /mnt/music also hangs - what can I do? Note that ssh to the nfs server machine works fine. Now I started firefox on the cmd line and cannot C-c, C-z or pkill -9 it. Help! I'm on fedora 28 x64 10:31 < retran> oh, migrate a site? yes yes 10:31 < Armand> I mean, sure.. if the other end has the sense to package everything.. 10:32 < retran> wget is actually great for that 10:32 < retran> it can spider/download all the deps 10:32 < Armand> Hahahaha 10:32 < azarus> exactly 10:32 < retran> if you're talking about FTP 10:32 < azarus> wget is much more advanced than it appears at first sight 10:33 < retran> there's a ton of command line tools for FTP 10:33 < retran> since it's so ancient 10:33 < Armand> I know wget, but it's also subject to rewrites. 10:33 < retran> ... 10:33 < azarus> Armand: you can control that 10:35 < Armand> Well, here comes the real fun. 10:35 < Armand> Certbot. 10:35 < azarus> blergh 10:35 < FManTropyx> hm, bash doesn't work as I expect 10:36 < Armand> azarus: Cent5.11.. haproxy. 10:36 < Armand> Yay! 10:36 < azarus> openbsd's acme-client is like a sane Certbot 10:36 < Armand> The cPanel autoSSL process uses something like that, I believe.. ? 10:37 < sandman13> Hi, should one run both loadbalancer and backend service on the same server? 10:37 < Armand> No.. kinda defeats the point of an LB 10:38 < sandman13> I see 10:38 < sandman13> currently I am trying to run both on same server :D 10:39 < Armand> I snagged myself a hardware LB a few months back. 10:39 < FManTropyx> well, until just recently I didn't know commands like 'scp' and 'sftp' existed :) they were kept from me! 10:39 < sandman13> but I wonder if these packet loops are because of one LB forwarding traffic to another LB 10:39 < Armand> Via Eden 1GHz sitting in that.. so I'm just turning it into a terminal. 10:43 < sandman13> Armand: what are the benefits of hardware LB over something like keepalived/haproxy? 10:46 < afidegnum> i hav inserted the usb modem, using usb_modeswitch -H doesn't connect the modem dongle 10:46 < afidegnum> what's the correct command 10:46 < afidegnum> ? 10:49 < drzacek> Hello there 10:51 < drzacek> I'm still trying to figure out that whole gpg thing. I get the general idea, but struggle with real-life use. I created a key (key pair?), then added two subkeys (one for signing, one for encrypting). Are those subkeys private keys, or are those key PAIRS? Which keys do I use and when? 10:52 < afidegnum> any insight ? 10:52 < drzacek> For example, I want to use enigmail in thunderbird to sign and encrypt messages - from what I see, it imported ALL keys (master and subkeys), how do I tell it to use specific keys? 11:02 < FManTropyx> urf... Linux is so bad for desktop use 11:03 * retran yawns 11:03 * mous dances 11:04 < azarus> FManTropyx: if you linux as you do windows, well it's not really that good 11:04 < peetaur2> what if you Linux in the ways you try but fail to on windows? 11:04 < azarus> but if you adapt to it as you did adapt to windows, you'll have a great time 11:05 < retran> if you want to use iTunes, Photoshop, Word, and play games 11:05 < retran> yeah, you'll need Windows for that 11:05 < azarus> ... or wine 11:06 < retran> yes. wine... and a lot of patience and a love of bloat and wasted CPU cycles 11:07 < searedvandal> or just use rhythmbox, gimp and libreoffice. and play titles available on linux 11:10 < peetaur2> I use linux for everything, except I use a windows virtual machine for games, with kvm and vfio passthrough of a GPU 11:10 < peetaur2> and a 2nd vm for my wife that dual boots linux or windows...so it's a 2 seater 11:11 < peetaur2> not dual dual boot since you don't want windows having access to the disks, but it just attaches the disks you want before starting 11:11 < TheWild> man stdbuf: "Also some filters (like 'dd' and 'cat' etc.) don't use streams for I/O, and are thus unaffected by 'stdbuf' settings." 11:12 < peetaur2> if I worked with silly people that insisted on Word instead of superior things like Libreoffice, then I might even have word in a VM, but I don't, so I don't. 11:12 < TheWild> so if I use read(...) and write(...) (that is on file descriptors), my program might not be affected by stdbuf? 11:12 < retran> peetaur2, thankfully, there's the cloud version of Word nowadays :D 11:13 < retran> (for such situations) 11:13 < peetaur2> if you like giving all your documents to repeatedly proven untrustworthy corporations, yeah, that's a good solution 11:14 < jim> read() and write() aren't the text file operations, fread() and fwrite() (and others like fprintf() fscanf()) are 11:14 < retran> well, sure. I'm not that important or paranoid of a person. so yes. i'm just fine with it 11:14 < TheWild> hmm... I have to be careful then 11:15 < retran> I understand spies and stuff need to be more careful 11:15 < TheWild> In few of my C programs I used read(STDIN_FILENO, ...) and write(STDOUT_FILENO, ...) to process binary data 11:15 < retran> whos to say these untrustworthy corporations don't write a backdoor into the Word binary 11:15 < jim> read() and write() are binary file operations, so those aren't buffered, and they take a fd (file descriptor) that is returned by open() (whereas the other thing, that you need for fread, fwrite, fprintf, etc, is returned by fopen() 11:16 < TheWild> unbuffered! Oh, thanks jim, looks like I'm safe. 11:17 < jim> who's to say there isn't a windows compiler that writes backdoors to every executable it compiles 11:17 < jim> TheWild, make sure you read the man page for the libc func you're going to use 11:17 < peetaur2> jim: :) 11:18 < retran> jim, oh man! 11:19 < jim> moral to that story, is let's save the backdoor discussions to topics about houses 11:20 < jim> retran, just think if gcc had that backdoor thing... then, if you were to compile gcc with gcc, it would have the backdoor, and if you compile it again, then it would have a front door too 11:20 < TheWild> man read, man write (no, not that one.). man fwrite? No man pages? 11:21 < retran> jim, i hate it when that happens 11:21 < TheWild> ugh, it worked out of the box on Ubuntu 16.04.4 11:21 < retran> TheWild, what distro you on then? 11:21 < jim> TheWild, maybe those man pages are in like section 2 or section 3 11:21 < TheWild> Ubuntu 17.something 11:22 < retran> are you missing your bloat 11:22 < jim> try installing libc6-doc 11:22 < jim> not sure that's the package name 11:22 < retran> any reason you still using an odd-numbered ubuntu version? 11:23 < retran> in an even numbered year 11:23 < retran> and the .something part of 17.something makes quite a difference (in non-LTS versions) 11:24 < Nixola_> hi, I'd need some help with mdadm 11:24 < retran> Nixola_, ask your question then 11:25 < Nixola_> I was having issues with mdadm thinking sda1 was busy, so I tried to remake the partition as sda2 instead to see if it'd somehow fix that issue, which it did, but it now complains it's got the wrong UUID 11:25 < Nixola_> can I somehow 1) change the UUID mdadm expects, 2) change the partition UUID to the one it expects (which I forgot to write down), 3) make mdadm ignore the UUID mismatch? 11:25 < Nixola_> just one of the three 11:26 < TheWild> man fwrite would give me first man page it is defined in, but nope. The package name I tried was libc-doc, libc6-doc, man-libc etc. but gave up and googled it. 11:26 < TheWild> it's manpages-dev 11:27 < jim> TheWild, see if you can: sudo apt install glibc-doc 11:27 < Nixola_> retran: https://hastebin.com/amodoqicif.txt (plus what I said above) 11:27 < TheWild> tried it but man still couldn't find fwrite 11:30 < Nixola_> mdadm --detail /dev/md0 after starting it: https://hastebin.com/betayahipe.txt 11:31 < azarus> retran: by the way, wine is much more efficient than a VM 11:32 < Nixola_> (when it works) 11:32 < retran> when you don't mind using old versions of apps because you're afraid to upgrade and break things 11:32 < TheWild> coproc sets up descriptors only for stdin and stdout, but not stderr or any more descriptors :( 11:33 < retran> wine is a pretty awesome project though, lots of work 11:33 < peetaur2> whether WINE or a VM is more efficient depends on which library calls WINE is using, and which sort of VM it is.... eg. most games run waaaaaaaay faster (like no perceptible difference from bare metal) in KVM and vfio passthrough compared to WINE 11:33 < Nixola_> DLange: sorry to ping you again, but I haven't solved the issue yet 11:33 < azarus> DOOM 2016 with Vulkan proved to be much faster for me with an AMD gpu on wine 11:34 < azarus> but then again, other games work better in QEMU/KVM 11:34 * azarus shrugs 11:34 < Triffid_Hunter> peetaur2: that requires a spare graphics card though, no? 11:35 < azarus> not necessarily 11:35 < Nixola_> Triffid_Hunter: that, or shutting X down, unbinding the current GPU, binding it to vfio-pci, starting the vm 11:35 < Nixola_> and then doing the opposite on shutdown 11:35 < azarus> you can always start the VM via ssh or something ;) 11:35 < peetaur2> it's possibly possible you could pass through the primary one and be headless on the linux side...dunno :) 11:35 < azarus> i tried it, that works 11:35 < Nixola_> yup, it can be done, but it's usually a hassle 11:35 < peetaur2> azarus: which gpu, cpu and mobo was it? 11:36 < azarus> peetaur2: Z170i Pro Gaming (or something), i5-6600K, AMD R9 Fury 11:36 < azarus> has a single PCIe slot to an IOMMU group, it's nice 11:37 < azarus> mini-ITX KVM passthrough xP 11:37 < jim> TheWild, try installing manpages-en-dev 11:37 < Triffid_Hunter> hmm, would it be possible to run linux on intel integrated and pass nvidia to the VM, but have both render to the same screen? 11:38 < peetaur2> Triffid_Hunter: get one of those monitors with a switch with multiple inputs 11:38 < TheWild> jim, you mean to not install all the languages? 11:38 < azarus> i need my printf man page in esperanto 11:39 < azarus> /s 11:39 < TheWild> manpages-en-dev not there 11:40 < jim> ok try manpages-dev 11:49 < jcarpenter2> anybody have a command that just outputs all of the posix signals that it receives? 11:50 < jcarpenter2> command or program 11:50 < sauvin> signals such as are listed with a kill -l ? 11:51 < jcarpenter2> yes but i'm looking for a program that outputs signals it receives as it receives them 11:51 < jcarpenter2> eg if it receives SIGUSR1 it outputs 10 11:52 < sauvin> I'm not aware of such a program, but it'd be trivial enough to write. 11:52 < jcarpenter2> ok 11:52 < sauvin> Just write something that traps all the known signals. Caveat: you won't trap a KILL. 11:52 < jcarpenter2> should be fine, i don't think it'll receive one anyway :A) 11:52 < sauvin> Why do you want this? 11:53 < jcarpenter2> a process is exiting when i run it through tmux, and i suspect it might be receiving some sort of signal 11:55 < jim> jcarpenter2, you mean stuff like SIGINT? 11:56 < jcarpenter2> yeah, and SIGHUP etc 11:56 < jcarpenter2> there are multiple signals that could terminate a process 11:56 < jcarpenter2> but, yes 11:56 < jim> yeah, trap the ones you can (some you can't, like SIGKILL 12:03 < jcarpenter2> welpp no signals 12:03 < peetaur2> so write a C program that traps all signals and prints them and otherwise just has a while true { sleep 1; } 12:04 < jcarpenter2> did it 12:05 < jcarpenter2> by "welpp no signals" i mean it didn't output anything when i ran it under tmux, so apparently there is no SIGINT, SIGHUP, or even anything else 12:05 < jcarpenter2> not even SIGKILL, i can tell that because it's still running 12:07 < luke-jr> peetaur2: why sleep(1) instead of pause()? 12:07 < luke-jr> in fact, sleep(1) may be problematic since sleep is sometimes implemented using signals 12:08 < peetaur2> probably the same reason I wrote sleep 1 instead of sleep(1) because it was more bash coming out than C :) 12:11 < hevauq> In man page for "top" command, there is description of various types of memories. It has a table showing a field which is Private and File backed memory, "pgms/shared libs" and CODE field represents "pgms" portion of this table. Does anyone know what does pgms stands for? Or what would be the right place to ask that question. Can't find it through google. 12:11 < pottsy> gandalf999 12:12 < pottsy> eech 12:12 < pottsy> wrong window 12:12 < peetaur2> hevauq: programs? 12:13 < hevauq> @peetaur2: are you sure? or just speculating, my guess was page maps, but that doesn't make sense either. 12:14 < peetaur2> just the first word I thought of that seemed to match 12:14 < hevauq> that won't fit the idea of Type of memory. 12:14 < hevauq> I am not sure if there is an IRC for binutils? 12:14 < hevauq> or core-utils 12:15 < Nixola_> I was having issues with mdadm thinking sda1 was busy, so I tried to remake the partition as sda2 instead to see if it'd somehow fix that issue, which it did, but it now complains it's got the wrong UUID 12:15 < Nixola_> can I somehow 1) change the UUID mdadm expects, 2) change the partition UUID to the one it expects (which I forgot to write down), 3) make mdadm ignore the UUID mismatch? 12:15 < Nixola_> just one of the three 12:15 < Nixola_> https://hastebin.com/amodoqicif.txt 12:15 < Nixola_> mdadm --detail /dev/md0 after starting it: https://hastebin.com/betayahipe.txt 12:23 < Nixola_> well now I feel like an idiot 12:23 < Nixola_> apparently recreating the array fixed it 12:27 < KaPiTaNo> hi 12:45 < msiism> i'm having a little bit of a problem with rsync. it doesn't delete files in the destinantion directory that have been removed from the source directory in the way i thought it would. here's an example: http://paste.debian.net/plain/1031869 12:46 < peetaur2> msiism: that's what you told it to do though... 12:47 < peetaur2> it only deletes things that are not in your set of source files...your set did not include the directory src/, only the files you specified (the shell expanded src/* to just src/file1 and src/file2) 12:47 < peetaur2> so you should do: rsync -a --delete src/ dest/ 12:49 < msiism> peetaur2:ok, why would the set have to include the directory itself? 12:50 < peetaur2> because otherwise it doesn't know what to delete... some things on the dest are not related to your transfer, so they shouldn't be touched 12:51 < peetaur2> if you had a dir1/ dir2/ and you sent dir1 to the dest, and the test had a dir3, should it delete dir3? 12:51 < peetaur2> it knows not to since you didn't tell it that dir3 has anything to do with its job 12:52 < Lindrian> I currently have a raid 5 array. Is it possible to recover the data and skip the array? 12:52 < peetaur2> Lindrian: that has yet to be determined 12:52 < Lindrian> Haha 12:52 < peetaur2> so...what state is it in? 12:52 < Lindrian> Working 12:52 < peetaur2> do you have all the disks? 12:52 < Lindrian> Yes 12:52 < peetaur2> if it works, then why do you call it recovery? 12:53 < Lindrian> Poor choice of words. 12:53 < Lindrian> s/recover/keep/ 12:53 < peetaur2> then what do you mean? random guess: convert it to a non-raid single disk 12:53 < peetaur2> what does "skip" the array mean? 12:53 < Lindrian> I want to remove the array 12:53 < Lindrian> But keep the data 12:53 < msiism> peetaur2: ok, i get the example. if you, e.g., had a backup directory containing backups of several user's home dirs, then you wouldn't want rsync to delete jane's backup when your creating john's, right? 12:54 < peetaur2> will the data fit on just one of the devices? 12:54 < Lindrian> No 12:54 < Lindrian> I have LVM on top of it atm 12:55 < peetaur2> well one strategy is to make 2 things coexist and transition between them...like shrink the array by 1 disk, then make lvm on one, then move the data there, then remove another and extend the lvm, etc. 12:55 < peetaur2> but the simplest is just dump it somewhere else, reconfigure the disks, and copy it back 12:55 < Lindrian> I don't have 4 drives of the same size unfortunately 12:56 < peetaur2> msiism: right 12:57 < peetaur2> and also I wouldn't do multi-drive lvm, except if it's on raid :) 12:57 < peetaur2> lvm supports redundancy, but more people use raid so it's more tested and reliable 12:58 < peetaur2> mdraid I mean 12:58 < Lindrian> I use mdraid yeah 12:58 < jim> Lindrian, I think you just need four drives that have partitions that are the same size 12:59 < Lindrian> But from my understanding, running 4x4TB in raid 5 is almost pointless 12:59 < Lindrian> So I'm losing 4tb of space for false-security for no reason 12:59 < jim> or maybe not four, but depending on the raid level requirements 12:59 < Lindrian> And I don't have critical data on them, so might as well skip raid 12:59 < Lindrian> Or have I misunderstood? 13:00 < peetaur2> raid5 packs more data into the small space than raid10 does 13:00 < peetaur2> not much though... and raid5 kills your iops 13:00 < jim> Lindrian, if you don't want it, you don't have to have it... linux (and other free software) is about choice 13:01 < peetaur2> and raid10 has slightly more redunancy..when you lose one drive, you have only a 1/3 chance that the next drive dying kills the array instead of 100% 13:01 * azarus is a raid6 prophet 13:02 < peetaur2> and if you mean raid6 is better, I agree ... once you lost 1 disk in raid5, you have no redundancy... you have failed to meet your goal to have redundancy; so you should have a raid6 and replace the first failed before the 2nd fails and you lose redundancy 13:02 < peetaur2> also see http://www.zdnet.com/article/why-raid-5-stops-working-in-2009/ 13:03 < peetaur2> (but I disagree with his wording... raid5 stopped working the instant it existed...just larger drives mean far more chance, and where you cut off the safety limit is subjective) 13:03 < epicmetal> Speaking of RAID, has anyone used ADAPT on Lenovo SANs? It's this seemingly proprietary, barely documented "RAID level" that only popped up in the SAN configurator after a firmware update 13:03 < peetaur2> (and then apply that reasoning to raid1/10 and you'll find to have the same standards, you need 3-way raid1's and 10s) 13:04 < msiism> peetaur2: so, if i got that right, src/* dest/ in the above example means "update file1 and file2 in dest/". however, src/ dest/ means "sync the directory structure of src/ with that of /dest" and therefore affects all files in src/. 13:04 < jim> peetaur2, say you have 4 drives, maybe 500g, 750g, 1t and 4t... so you put a 500g raid physical vol on each, and make a raid10 out of them... later, the 500g drive fails, and the only thing available is a 1t drive, so you get it, put a 500g phyical vol on it, then add it to the array, and have it recover, restoring the array... got that picture? 13:05 < peetaur2> msiism: yes...because the shell expands dest/* so rsync does not see the * at all 13:05 < msiism> peetaur2: ok, got it. thanks. 13:06 < peetaur2> msiism: test this..... blah() { for arg in "$@"; do echo "arg is \"$1\""; shift; done; } 13:06 < peetaur2> msiism: eg. https://bpaste.net/show/f471ba0473d0 13:06 < peetaur2> that function doesn't see a * (unless you escape it or stick it in quotes) 13:07 < peetaur2> jim: so you're using 500g out of your 4tb? :D 13:07 < peetaur2> jim: throw out the 2 smaller ones and use the 1TB and 4TB together instead :P 13:07 < jim> yeah, ok, horrible example, but say that's what you have... 13:08 < peetaur2> ok, and the punchline? 13:09 < jim> after the 500g drive fails, the least drive you have is 750g,,, so after you recover with the new drive, can you expand all four partitions to 750g? 13:09 < jim> and then expand the array accordingly? 13:10 < peetaur2> in theory you could even make a raid10 with 500+500+1000+1000 to start 13:10 < peetaur2> and you can use mdadm --grow to reshape it if your disk sizes change, or number of disks 13:11 < peetaur2> and if you are creative, you could do other things like take 500+500 from the first 2 (with 250 left), and then 1000+1000 from the next, then make another with the 250+250 from the 2nd and 4th 13:11 < jim> so, then the answer is yes, you can grow them? 13:11 < peetaur2> but you can also get up to 2.25TB total size also 13:12 < jim> ok... what special thing does raid10 add? 13:12 < peetaur2> jim: yes I believe mdraid is very flexible and basically does anything, as long as you can calculate the right size it'll end up as and then shrink your fs to that before shrink (still --grow), or if you are increasing it 13:12 < peetaur2> raid10 is just a simple way for the tools to manage a raid1 of a bunch of pairs of disks and then raid0 it together for you 13:13 < peetaur2> and for some reason in mdraid, it has an implementation separate from the raid1 and 0 code, so it has its own unique performance characteristics and bugs 13:14 < peetaur2> s/, as long/;as long/ 13:14 < jim> so raid10 is not a striping situation, where if a drive fails, the data is redundant on the other drives, so you can replace a drive and have the array recover? 13:14 < peetaur2> raid10 is a stripe of mirrors....so as long as each mirror is healthy, the raid0 remains healthy too 13:15 < peetaur2> so if 1 disk is alive in each mirror, it doesn't matter how many disks are dead overall, and you can replace them without losing data or availability (assuming no bad sectors) 13:15 < jim> ok, and what's the best raid level to permit the most performant random access (such as what would be required by an sql db)? 13:15 < lupine> but note rebuild time 13:16 < peetaur2> rebuild time is pretty good on raid1-based stuff though 13:16 < peetaur2> jim: raid0 but not redundant, and then raid10 probably 13:16 < lupine> yeah. raid6 is where it gets interesting 13:16 < jim> lupine, yeah, I'm expecting long or at least semi long rebuild times for the examples I'm quoting 13:16 < peetaur2> raid6 has horrendous rand-write sync performance 13:16 < lupine> yup 13:17 < lupine> I'd just spend the money on hardware that's fast enough to not need to bother with raid 13:17 < peetaur2> but you still want to not reduce it by much...so raid1's and 10's 13:17 < jim> so it's generally felt raid is poisonous? 13:17 < lupine> it's just a hassle 13:17 < peetaur2> uh what? 13:18 < lupine> technology has advanced 13:18 < peetaur2> still need redunancy though...so what do you use? 13:18 < azarus> maybe I should try raid10, but performance doesn't even matter to me much 13:18 < lupine> personally, I use backups 13:18 < azarus> so I'll stick with raid6 13:18 < peetaur2> for example ceph doesn't need raid....(I put the os and mons on raid1 though)... the cluster replicates the objects instead of block devices 13:18 < lupine> raid doe have a marginal usecase for improving individual server uptime 13:18 < lupine> does* 13:18 < lupine> but even then, a wide range of disk failures will actually take down the whole machine anyway 13:19 < msiism> peetaur2: ok, i understand. 13:20 < jim> remember those 3ware cards? were they any good for raid? 13:20 < peetaur2> um I don't find that with proper raid a *disk* failure is likely to ever take a machine down...like 0 times ever personally experienced 13:20 < peetaur2> but sure a machine can die for lots of reasons unrelated to disks 13:20 < azarus> like removing RAM while running :P 13:22 < peetaur2> I prefer softraid like zfs and mdraid... slower compared to hwraid+BBU or equivalent, but doesn't blow up and waste your time for no sane reason.... so many super old servers have random io issues when they have hwraid, but never saw such a thing with HBAs and mdraid or zfs yet 13:22 < peetaur2> how can hwraid die more than cpus and RAM ... can't understand that (unless it's intentional ;)) 13:23 < peetaur2> we even have like 40 of a CPU that has known memory controller defects and only something like 2 died in 5 years :) 13:25 < jim> interesting... so over the years has developed an advantage for some forms of software raid? 13:25 < peetaur2> softraid is hard to use, but puts you in control 13:25 < peetaur2> and features of zfs are beyond what any hwraid can achieve (and soon btrfs after it's not so buggy) 13:25 < jim> ok, thanks, appreciate the conversatiuon piece 13:26 < peetaur2> and for whatever reason, I find they are more reliable and less quirky/buggy... software patches superior to firmware upgrades? purposeful malevolence aka "planned obsolescence" aka "sabotage" in the hwraid? 13:27 < peetaur2> so years ago you could debate about the BBU and the RAM cache on there making it superior performance on a hwraid (but of course that cache is non-redundant) ...but now you can stick an NVMe cache on zfs and get some (but not all somehow) benefit, or bcache or something like that 13:27 < peetaur2> long ago it was special vendors that had monopolies on anything fast enough to be used...like netapp had some overpriced PCIe cards only available with their hardware 13:28 < peetaur2> still haven't seen an HBA die... and only saw one onboard SATA controller die before it was even delivered to us 13:32 < peetaur2> we have one machine where a particular HBA didn't like those seagate archive 8TB disks and would randomly drop them until you pull them and put them back in... we put a LSI-9211-8i in there (old 5+ year old controller) now which works fine 13:33 < peetaur2> it was an LSI 9300-4i 13:45 < rafalcpp> me, adminstrating linux VMs host, each time - https://i.imgflip.com/2d98xa.jpg 13:46 < peetaur2> non-sequitur argument 13:49 < afidegnum> does it means if my system support UEFI, the partition type has to be GPT ? 13:51 < peetaur2> afidegnum: if it supports legacy mode, then you can use a MBR/msdos partition table with an old style MBR bootloader 13:52 < peetaur2> but likely GPT and ESP and such are the best option on UEFI hardware...unless your hw sucks (like one machine I have seems to delete the boot menu options in some unknown case and then you have to boot an iso in uefi mode to add it again ...such lovely design) 13:52 < peetaur2> this machine (what the same board) seems to be in legacy mode, and has no such issues 13:53 < afidegnum> well, using dell precision m4500 it has UEFI and Legacy mode, so i want to activate the UEFI since i am going to have multiple ... more than 5 partitions, 13:53 < peetaur2> yeah I remember it was GPT before this board...and then it wouldn't boot GPT+bios_grub on legacy mode, so I made it msdos instead of GPT ...such progress 13:54 < afidegnum> but now i need to find out how to configure the boot option on UEFI 13:55 < peetaur2> to configure UEFI boot and GRUB2, basically you just boot in UEFI mode on some rescue/install media, and then make sure your disk is GPT and has an ESP partition (which is fat32), and then mount it, like at /boot/EFI 13:55 < peetaur2> here's how to do that on arch based https://bpaste.net/show/432ddbb5697d 13:58 < jim> usually installers will make the boot loader work in uefi if it finds that it's been booted in that mode 13:58 < jim> I know debian will do that 13:58 < peetaur2> that's why that snippet is for arch based....where bash is the installer :) 13:59 < afidegnum> ok 13:59 < peetaur2> where else would you need it? ;) j/k, it's highly useful for repairing borked boots (like that machine that forgets the UEFI boot options randomly) 13:59 < afidegnum> arch will need some hand tuning 13:59 < g1itch> I recently reset my user password which required me to reset my gnome keyring. ever since then chrome will not keep me logged in to sites through system restarts regardless of 'remember me' functionalities. Any suggestions? 14:00 < jim> with arch, you'll usually have to learn a lot about each software package you install, in particular how to configure it 14:01 < peetaur2> which is easy since you only install what you need, not the bloat you get most other places ;) 14:02 < jim> yeah, this learning stuff is not seen as bad or wasted effort, you'll usually be glad you did 14:03 < Dagmar> ...and you'll get to spend even more time learning how everything works because quite frequently upgrades will break things you depend on, and you'll have to put on your developer hat and fix them 14:03 < peetaur2> one time my pacman uninstalled itself... that was fun 14:03 < peetaur2> pacman -U /var/cache/pacman/pkg/pacman-... right? command not found 14:04 < epicmetal> peetaur2, so you untarred the tar.xz manually? 14:04 < peetaur2> so instead I just did a tar xf and did some --force commands to make it forget that the package had no db files and add the package anyway, and then it was fine 14:06 < peetaur2> and maybe twice Xorg wouldn't start... one time dbus was broken so I just downgraded it... nothing too hard to fix 14:06 < peetaur2> quite apalling that a thing like Xorg requires a garbage package like dbus at all, let alone fails to even start because of it 14:07 < Dagmar> It doesn't 14:07 < Dagmar> Whatever DE you were using was apparently dependent upon it 14:07 < epicmetal> i really wish things could still have the option of being more static 14:07 < epicmetal> no dynamic starting of things 14:08 < peetaur2> yeah maybe the DE broke Xorg so badly that Xorg was also to blame, but for a separate bug 14:08 < peetaur2> that makes much more sense 14:09 < peetaur2> I don't think static vs dynamic matters much... but we should follow sane design practices, like have loose coupling 14:09 < peetaur2> the adoption of systemd proves people don't give a crap about that though 14:09 < epicmetal> i just like the idea of things not starting without my explicit permission 14:09 < peetaur2> epicmetal: which distro is this where you have actually installed things that you don't want to run? just don't install them... 14:10 < epicmetal> peetaur2, sometimes you want things installed but not starting randomly 14:10 < peetaur2> like I'd love to wipe out dbus and avahi but some things refuse to work without them... like cups refuses to work without avahi even though it is completely 100% worthless unless you also have a 100% worthless printer that requires it 14:10 < epicmetal> peetaur2, it's not a distro-specific grievance 14:10 < epicmetal> there's always gentoo... but they lag on pacakge updates 14:13 < azarus> epicmetal: it's not bad if you're on testing 14:13 < azarus> of gentoo 14:16 < epicmetal> azarus, even testing lags badly 14:18 < Thom1> Hi, I'm just surprised to see there is no patch/ChangeLog for linux-4.4.139. Only the full tarball is available, do you know why ? 14:19 < tvon> I have an odd case here. I'm trying to debug a rather complex set of bash scripts and I noticed that some variables I'm echoing should be paths but have the '/' stripped. E.g., `echo $SHELL` gives me ` bin bash` but if I dump `env` I see slashes in the path. Anyone seen anything like that? 14:20 < tvon> This became apparent because I have a `mktemp` in there that was erroring with "$tmpfile: ambiguous redirect" because the variable seems to have spaces where there should be slashes. 14:21 < kerframil> tvon: that's a symptom of both dicking around with the value of IFS, and not quoting one's variable expansions 14:22 < tvon> hrm 14:22 < epicmetal> Well, not "badly", but it lags 14:23 < tvon> kerframil: yup, I see some IFS foolery elsewhere, thanks! 14:23 < jdmssmkr> I've got a custom embedded board. I have to use xpio to configure WIFI antenna's. When I run this on the commandline: no problem. But when the commands are started in a startup script (/etc/rc5.d/...) they don't seem to have any effect. I don't have a clue at the moment what could be the problem (yes, the path variable is properly defined in the script). Anybody has an idea? 14:24 < Smithe> There's a way so that a machine on one subnet can send packets on a self-assigned IP and that packets go in a tunnel to a 127.0.0.0/8 ip in a second machine in the same subnet? 14:26 < kerframil> tvon: echo "$SHELL" would work as expected, but any broadly-scoped change to IFS is likely a code smell 14:27 < tvon> They seem to set it a lot. 14:32 < jeffree> is it possible to monitor nvme drive temps like I can with other sensors through lm-sensors? 14:33 < jeffree> Ideally, I'd like to have live temperature graphs 14:33 < BluesKaj> Hey folks 14:33 < jeffree> hi 14:34 < jeffree> also, does anyone know a program to do live graphing of lm-sensor data, similar to psensors but better/more capable? 14:47 < g1itch> i'm out of ideas - tried setting keyring password to blank, removed all chrome entires in the login keyring, removed and re-added chrome sync accounts, chrome refuses to keep me logged in to sites through restarts of the application 14:51 < jnor> hi noob question I put an executable in /usr/local/bin, when I start it from term it hijack term, any better way than to create start script and symlink to the executable? 14:52 < peetaur2> g1itch: make a new user, make it work, copy files from the old user ovewriting it until it breaks...then you're still lost, but closer :) 14:52 < leftyfb> jnor: make a systemd unit? 14:52 < jnor> kay 14:52 < jnor> =) 14:53 < jnor> ty 15:16 < triceratux> http://www.freerepublic.com/focus/f-chat/3668007/posts 15:17 < Pentode> i5, ironlake i915 glmark 260fps, 3 sandybridge i915, 1650fps. i wonder how much was due to driver issues and how much is due to improved architecture... 15:18 < Pentode> thats an awful big difference in performance for just a few years 15:18 < Pentode> +i = i3 15:19 < Pentode> not that it matters now but im just curious (i _was_ having driver issues but i don't know how much this affected GL rendering speeds) 15:19 < Pentode> and im in the wrong window 15:19 < Pentode> lol 15:22 < compdoc> I was curious in college 15:22 < Pentode> i bet 15:22 < ayecee> i saw a movie like that 15:27 < TonyWonder> Did somebody say 'Wonder'? 15:41 < ychaouche> Hey ##linux, I have a question regarding this statement : "Finally, administrators should limit directories that are writeable and executable, particularly by service users such as www-apache. If an attacker does not have the ability to execute his exploit, the attack is rendered ineffective". How are "executable directories" and "executable files" related ? 15:41 < ychaouche> if my directory is executable that does not necessarily mean that any file in that dir is executable, no ? 15:41 < twainwek> no 15:42 < deww> ychaouche: directory executable means you can cd into it 15:42 < FManTropyx> I wonder if there are assemblers for Linux 15:45 < ychaouche> so there must be a missing link 15:45 < ychaouche> in the explanation given above "If an attacker does not have the ability to execute his exploit, the attack is rendered ineffective" 15:45 < mnemon> FManTropyx: what kind of assemblers do you mean? for native code can just look at compilers. 15:46 < ychaouche> for the unexperienced reader that I am, I thought that you could make all files inside a directory non-executables by making that directory non-executable, which doesn't seem to be true. Yes ? 15:47 < mnemon> ychaouche: you can mount with noexec and everything under that mount point will be non-executable 15:47 < ychaouche> which the article doesn't mention 15:47 < mnemon> or use acls to mask out exec bit for specific groups/users in directories 15:47 < yn> neat, https://linuxlifecycle.com/ 15:47 < peetaur2> ychaouche: mount -o noexec 15:47 < maximumgeek> Well you can also set the directory to 644 and then you will not be able to execute them 15:48 < mnemon> or use selinux/apparmor to prevent executions ... etc. 15:48 < peetaur2> but if his exploit is in an interpreted language, that won't help...he can just run it like bash /path/to/file 15:48 < triceratux> ychaouche: executable directories are an anomaly. that permission bit determines whether the directory can be made current via cd so it does to some extent control access to the directory itself. but the permissions of its contents remain independent 15:48 < peetaur2> and if you use noexec anywhere, you better use it *everywhere* he can write...such as /dev/shm, /home/whoever, /tmp, etc. 15:48 < maximumgeek> If he is no the machine and can read the file, he will be able to execute it somewhere. 15:49 < maximumgeek> s/no/on/ 15:51 < triceratux> ychaouche: thats not the best description you posted. its replete with some of the most commonplace misrepresentations 15:55 < mnemon> maximumgeek: there's usually pretty strict limitations by default for webservers and the like so the usual exploits often rely on accessible application dirs 15:57 < maximumgeek> @mnemon I missed the part about it being on a web server. Sorry about that. 15:58 < mnemon> just "users such as www-apache" so I'm just assuming it's mainly aimed at webservers and the likes. 16:06 < ayecee> www-data, maybe 16:06 < ayecee> www-apache doesn't sound like a thing 16:11 < mnemon> someone decided to merge www-data and apache for the textbooks maybe ;) 16:16 < tvon> kerframil: Thanks again for that IFS tip, that was indeed the issue (`IFS='/' file_array=(${file_path})` earlier in the code). I don't know how long it would have taken me to find it if you hadn't pointed me in the right direction. 16:21 < kerframil> tvon: I would suggest that be written as IFS=/ read -ra file_array <<<"$file_path". the alteration to IFS then won't persist, and it also prevents some potential side effects from expanding $file_path without quotes. 16:21 < tvon> I changed it to `IFS='/' read -a file_array <<< ${file_path}`, what is the "-r"? 16:22 < kerframil> tvon: stops backslash from being special. you almost always want -r. 16:22 < kerframil> tvon: help read 16:22 < tvon> Aight, thanks 16:23 < kerframil> tvon: use quotes too. by using read, you're no longer relying on word splitting to compose the array. 16:27 < zzz> can anyone explain to me how to make changes highlighted by the watch command permanent? man is not helping 16:28 < peetaur2> zzz: changes highlighted...? 16:30 < ren0v0> Hi, can anyone recommend an application that will allow me to use a proxy for only certain applications? 16:31 < ren0v0> I've tried "proxychains" but getting errors, seems outdated 16:31 < peetaur2> oh I didn't know it does that... and the man page says you can say "permanent" ...did you put that in there? --differences=permanent 16:35 < Dan39> ren0v0: how often do you need it? for the like once every other week that i need to proxy something i just use ssh tunnels 16:37 < ren0v0> Dan39: i have SSH tunnels setup as proxy 16:37 < ren0v0> I have a plugin for chrome that works (pointing to this proxy), now i want select applications (for example IRC, hexchat), to use this proxy 16:38 < ren0v0> i don't want it systemwide, hence the need for something else infront of the application 16:38 < peetaur2> those aliases on rh based are so annoying... cp 10 files and it wants to ask you overwrite 10 times... who can stand those? 16:54 < zzz> peetaur2: well, i found out it works with -d=[whatever] 16:54 < zzz> thanks 16:58 < zzz> what got me was the inconsistent argument syntax. "=" is not needed on other watch options 16:59 < peetaur2> why does yum try bogus urls like http://centos.mirror.net-d-sign.de/7.3.1611/os/x86_64/repodata/repomd.xml: ??? the right one is just 7 instead of 7.3.1611 16:59 < peetaur2> the repo file has stuff like this, no hardcoded 7.3.1611 mirrorlist=http://mirrorlist.centos.org/?release=$releasever&arch=$basearch&repo=os&infra=$infra 17:02 < peetaur2> and going to http://mirrorlist.centos.org/?release=7.3.1611&arch=x86_64&repo=os gives me "Invalid release". But if I give it 7, I get a list of urls with 7.5.1804 in it...so nothing makes sense 17:11 < mawk> wake up 17:15 < ayecee> the matrix has you 17:21 < Gringonar> Hi 17:25 < mnemon> peetaur2: has it been set in yum config? it should otherwise get it automatically from the release package (7Server / 7Client in rhel for an example) 17:26 < Gringonar> I've been trying to get any linux to boot on my new laptop MSI GE73 8RF. I was finally able to (semi)boot ArchLinux after having tried Ubuntu(18.04 LTS and 16.04 LTS) Fedora 28 and others, but still very few things work any assistence would be appreciate. i'm hoping to atleast get the ethernet to work which i found to be eth0: Qualcomm Atheros AR81 17:26 < Gringonar> 6x/AR817x Ethernet . also lspci keeps hanging and timing out with an error i dont know how to terminate since im commandline only and i have no gui :( 17:27 < throwthecheese> I don't know where I can ask general programming questions, but I want to know why won't my patches apply 17:28 < throwthecheese> They fail with a 127 error 17:29 < cannabis_sativa> Gringonar: I haven't heard of such trouble installing linux on (a) new hardware -and- (b) with the latest versions you are using. Ok. one thing is this (1) If you installed (most iso images MUST access the internet during installation).. this means connectivity worked during installation. 17:30 < akk> throwthecheese: Usually either there's an error message, or it partially applies the patches and leaves unpacked parts in the file itself. 17:31 < throwthecheese> It returns 127 17:31 < peetaur2> mnemon: using strace, I found it was going to a certain ip and adding a header Host: mirrorlist.centos.org .... but if you look at the dns for that, it doesn't match ....then found it was in /etc/hosts for unknown reasons... gah! 17:32 < peetaur2> problems that make sense are so much simpler 17:32 < Gringonar> i've booted into a live iso only 17:34 < afidegnum> /3 17:36 < Lope> sometimes when I suspend my laptop, after waking the built in wired ethernet adapter doesn't work. I think it only happens if I suspend after I enable all the power saving stuff in powertop. But if I go back to powertop and disable all the powersaving stuff, the eth adapter still doesn't come up. I have to reboot to get it back. 17:36 < triceratux> Gringonar: thats the way to do it until youve tried out a few distros & found one to your liking. i usually say go with xubuntu lts which these days is both 16.04 & 18.04. if xfce will come up nearly everything will come up. also if your hardware runs arch pretty well 17:36 < Lope> RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller 17:37 < triceratux> Gringonar: ... i tend to recommend this weeks swagarch 18.07 which is linux 4.17, xorg 1.20.0 & firefox 61. works great for me 17:37 < triceratux> https://swagarch.gitlab.io/ 17:37 < throwthecheese> Turns out I didn't have patch /oof 17:37 < Gringonar> the error is get on lspci is long but it says echo 0 > /proc/sys/kernel/hung_task_timeout_secs disables this messagage 17:38 < Gringonar> so its a kernal panic? 17:38 < mawk> wireless USB is a thing 17:39 < triceratux> Gringonar: hrm looks like the arch folks run into this sometimes https://bbs.archlinux.org/viewtopic.php?id=208813 17:41 < triceratux> hrm https://github.com/deadhead420/anarchy-linux/issues/344 17:41 < triceratux> anarchy not good. go with swagarch 17:42 < Gringonar> I honestly prefer debian or ubuntu since its more noob friendly but arch is the only operating that boots that i know off 17:43 < triceratux> "I've installed arch probably on about 100 different computers and my experience is almost every system is different in terms of strange issues that pop up." --the above post 17:44 < triceratux> Gringonar: looks like they see that lspci hang as recently as 2017 & they view it as hardware specific 17:44 < Gringonar> I'm trying to switch away from windows, i'm that new to linux this will be my second linux pc(laptop) 17:44 < WhiteDevil> use debian 17:44 < WhiteDevil> i am new user too and i use it and enjoy it 17:44 < triceratux> mx-17 to be exact 17:44 < Gringonar> tried wont boot 17:45 < WhiteDevil> hasnt failed me yet 17:45 < WhiteDevil> it wont boot from grub ? 17:45 < Gringonar> i cant diagnose anything if it doesnt boot :( 17:45 < Gringonar> i get the boot menu 17:45 < Pentode> Gringonar, you might try disabling secure boot. if that still doesnt work, try using legacy mode (non UEFI) 17:45 < Gringonar> i tried adding some parameters 17:45 < WhiteDevil> ,my computer used to not be able to boot from usb 17:46 < Gringonar> nomodeset and modprobe.blacklist=nouveau but no luck 17:46 < WhiteDevil> and then i went into the bios changed some settings adn it works now 17:46 < Pentode> ^ 17:46 < Gringonar> after grub i get black screen 17:46 < Gringonar> i disabled secure boot right away 17:46 < WhiteDevil> does debian have a media test option for its installer ? 17:46 < WhiteDevil> like in redhat there is an option to check if your media is secure and not messed up 17:47 < Gringonar> yes it gave (is it ok to post an immage link?) 17:47 < Lope> does anyone know how to suspend a kernel module when the computer sleeps in ubuntu 18.04? I want to implement this fix because my wired eth adapter stops working after suspend. http://www.techytalk.info/ubuntu-fix-network-stopped-working-after-resume-from-sleep/ 17:47 < DLange> WhiteDevil: https://www.debian.org/CD/faq/#verify 17:47 < Gringonar> i can give you the message of which i took a picture 17:48 < Pentode> did you try changing the boot device order? 17:48 < WhiteDevil> ya there ya go 17:48 < Gringonar> yes sir 17:48 < WhiteDevil> try checking the intergrety of your install media 17:48 < akk> triceratux: I've only used arch on two systems my my experience was like that, it worked great for a while then strange issues, different on each machine. 17:48 < WhiteDevil> the link Dlang posted 17:48 < Gringonar> i did the md5sum thing 17:48 < Pentode> must be the install media. i assume youve also tried alternate usb ports? 17:49 < Gringonar> after i flashed the iso to usb with rufus is there any way or need to check if its flashed correctly? 17:50 < Gringonar> yes tried multiple usbs 17:50 < triceratux> akk: fact is linux itself is like that, its just that arch makes it seem a little more brittle. thats why the race is on to come up with a robust arch & manjaro is suddenly more popular than mint 17:50 < akk> triceratux: True, that happens with debian and ubuntu too, it just (at least sometimes) gets fixed a little sooner. 17:51 < Gringonar> well i got arch running since nothing else will lol 17:52 < Pentode> well if you are going to get wet you may as well jump in i guess, lol 17:52 < Gringonar> i did some piping of logs like journalctl and dmesg , but i cant make heads nor tails of it :( 17:53 < triceratux> exactly. & its up to us to make ourselves more bulletproof because the exact combination of failures that constitutes a showstopper is the fault of noone. you cant really assign blame when 4 or 5 things have to combine in an unprecedented way to make it seem like linux will never boot on a given computer 17:53 < Gringonar> I know Qualcomm Atheros AR816x/AR817x is my ethernet but if i try to get it to work it says there is no eth0 device :( 17:53 < akk> Computers are super complex. Sometimes I'm amazed any distro works on any random assembly of hardware. 17:54 < triceratux> rofl you clearly know yer linux 17:54 < akk> (I tell myself this when I'm swearing over why-can't-I-run-a-script-on-resume or why-does-my-wifi-reset or something) 17:54 < Pentode> Gringonar, lol. that series of devices is notoriously annoying to get working. ;) 17:54 < Pentode> stupid proprietary firmware 17:55 < cannabis_sativa> Gringonar: what distros have you booted to a live install media? Get the latest version iso of any of the big balled boys. You should have literall zero trouble. 17:55 < akk> On a laptop that can be disassembled, my favorite solution for wifi is to open it up and replace the chip with an intel chip. 17:55 < cannabis_sativa> Gringonar: you wan to sell me that laptop? when you get sick of the errors, let me know. =) 17:55 < akk> Alas my current Asus is very hard to open up and I've been afraid to. 17:56 < akk> (it has a Broadcom, more or less works most of the time) 17:56 < Lope> I've discovered that my question relates to a known bug which was apparently JUST fixed in the new kernel. Will try it. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1752772 17:56 < dgurney> you have to be careful with that though. some laptops have whitelists, so it may not work with a replacement card at all 17:57 < akk> I had an intresting glitch on a trip last weekend: ip addr show (which I use as part of my network setting script) would randomly show a wlan0 17:57 < Gringonar> How can i make lspci stop? 17:57 < akk> which is not the normal name of the wifi interface any more 17:57 < akk> so wlan0 would appear, then I'd try to set it and it would disappear and the next ip addr show would show the normal interface name. 17:58 < akk> (wlp2s0 or something crazy like that) 17:58 < Gringonar> it keeps throwing errors but wont stop 17:58 < akk> Gringonar: ctrl-c? 17:58 < Gringonar> nope 18:01 < akk> ctrl-\ or ctrl-z ? 18:02 < akk> Otherwise, go to another terminal and pkill lspci, or sudo pkill -9 lspci> 18:05 < Gringonar> another terminal? 18:05 < Gringonar> im commandline only no desktop manager runns 18:06 < section1> try holding control-c for a while 18:08 < triceratux> hrm https://superuser.com/questions/1299730/arch-linux-install-lspci-call-hangs 18:10 < Gringonar> this is what i was able to salvage dmesg: https://paste.debian.net/1031927/ and Journalctl: https://paste.debian.net/1031928/ 18:11 < section1> ohh 18:11 < section1> that errors on sdc don't look good.. 18:11 < Gringonar> its the harddrive i wrote the logs to 18:12 < Gringonar> its a usb portable one its not part of the system 18:13 < newbsduser> Hello guys, I want to serve a single installer file only(50 mb), too many people can download it. How should I serve it? I don’t think torrent will be a nice alternative for end users. By using ec2 or google drive? Or without keep-alive vps+ramdisk? Or a cdn? Or amazon s2? What are good alternatives to do that? Do I need load balancer? I know maybe it s not correct channel for this question but If you have knowledge and if yo 18:13 < section1> Gringonar, which errrors throw lscpi ? 18:14 < Gringonar> i only have a picture of it(taken by my fone) since it just freezes after that is it ok to post it here? 18:14 < deww> newbsduser: more like a webhosting question. if it's a lot of downloads, cdn can help. you probably still need an origin for cdn the pull from. 18:15 < section1> Gringonar, upload to some pic paste(like imgur) and paste here the url. 18:15 < Gringonar> https://imgur.com/a/Lt7YHhM 18:15 < section1> :) 18:15 < Gringonar> yeah did that 18:16 < Gringonar> yesterday i asked the arch folks they said update formware call your hw vendor and or buy proper hardware 18:16 < Gringonar> i was like wow thanks 18:17 < section1> Gringonar, that lspci sure its in D state... 18:17 < Gringonar> is that dead state? 18:17 < section1> you cannont kill it...try to unmount sdc1 and unplug it and then try lspci 18:18 < section1> d state is uninterruptible 18:18 < section1> D* 18:19 < Gringonar> the picture is from yesteday. i didnt have sdc connected so it cant be the cause 18:19 < mnemon> Gringonar: well the fact is, if the HW vendor creates new HW that doesn't work with the existing drivers and provides no linux module for it, it will take time until the HW works in linux. 18:19 < Gringonar> but i got to eat dinner rq im starving at this point lol 18:20 < Gringonar> i know 18:20 * triceratux guesses its the vga_switcheroo 18:20 < Gringonar> but what hardware causes it? 18:20 < Gringonar> I read that coffee lake is supported? 18:21 < triceratux> its the nouveau processing in the first log https://devtalk.nvidia.com/default/topic/955951/linux/nvidia-smi-hangs-indefinitely-what-could-be-the-issue-/ 18:22 < triceratux> just a guess. you want to try xubuntu lts with xforcevesa or any sane distro with nomodeset 18:22 < akk> Gringonar: Even in the console, you can get another terminal, with ctrl-alt-f2, ctrl-alt-f3 etc. 18:23 < section1> maybe a bug in the kernbel..its strange the hard lockup 18:32 < Pusteblume> lets say i want to "apt-get download gpm mc htop" then i get these 3 packages. but how do i get the dependencies aswell? 18:35 < Hagrid> hi all 18:36 < cannabis_sativa> why not use "install"? anyhoo, If you must go this route, use 'gdebi' to install them and it will resolve deps. 18:36 < section1> Pusteblume, check apt-rdepends 18:37 < cannabis_sativa> Pusteblume: why not use "install"? anyhoo, If you must go this route, use 'gdebi' to install them and it will resolve deps. or do it manually. 18:38 < Pusteblume> section1, why rdepends and not depends? 18:39 < Pusteblume> cannabis_sativa, i need this way because i try to build a system based on a dvd set 18:39 < section1> Pusteblume, because reverse depends.. 18:39 < Pusteblume> section1, yes? 18:39 < section1> yes 18:40 < Pusteblume> so you dont know but suggest it. not very useful 18:40 < section1> example that packes needs dpkg ..and dpkg needs tar and tar maybe needs somethis...rev deps. 18:40 < section1> easy 18:43 < section1> and most of them needs libc and libc needs libgcc1 :) 18:50 < Pusteblume> so the plan is now something like this: apt-get download $1 && apt-cache depends -i $1 | awk '/Depends:/ {print }' | xargs apt-get download 18:50 < Pusteblume> but it complains "unable to locate package Depends" 19:00 < Gringonar> Did the logs say why my ethernet is not working? 19:00 < Gringonar> i only see 2 mentions of it the second one i dont quite get 19:01 < Gringonar> "enp3s0: renamed from eth0" 19:01 < section1> ah i was wrong with the name... that is recursive depends not rev. depends... i miss in going back in any package its call revers...but not. 19:02 < section1> Gringonar, define "is not working" 19:02 < Gringonar> i found some topics on google that the lan has been supported for a while? 19:02 < section1> do you see eth0 in ifconfig -a or ip ad sh 19:03 < Gringonar> if i do "ping ww.google.com"it says temporary failure in name resolution 19:03 < section1> thats a dns problem Gringonar 19:03 < searedvandal> dns issue then 19:03 < searedvandal> can you ping 8.8.8.8 ? 19:03 < Gringonar> if i type "ping 1.1.1.1"it says adress unreachable 19:03 < section1> try to ping to 8.8.8.8 19:03 < Gringonar> ok 19:03 < section1> ok 1.1.1.1 showd works 19:04 < section1> should 19:04 < searedvandal> what does 'ip link' give you? does enp3s0 have state UP? 19:05 < searedvandal> and check that you have gotten the correct ip from dhcp server 19:05 < section1> and gateway 19:05 < Gringonar> still when i type "ping 8.8.8.8" it also says "network is unreachable" 19:05 < section1> ok check you net config ..ip gateway.. and try first to ping the gateway 19:05 < FreeFull> Gringonar: Does it work on Windows? 19:05 < FreeFull> Is the cable plugged in correctly? 19:11 < Cache_Money> My email delivery service notified me that they’re deprecating SSLv1.0. I’m running a rails app on Ubuntu 14.04. Can I run a command to tell whether I’m on SSLv1.1 or v1.2 currently? 19:12 < FreeFull> SSL 1.0? That's ancient 19:12 < section1> Cache_Money, if its public try some ssl test 19:12 < section1> i think that sslabs have one good 19:12 < FreeFull> Cache_Money: You'll *probably* be fine 19:15 < Cache_Money> section1: I tried a curl request to their test endpoint but I’m getting an ‘Unauthorized’ error message https://www.sparkpost.com/docs/tech-resources/tlsv1-0-test-hostname/ 19:16 * Pusteblume_ locks himself in a finger cage so he wont type for 20 minutes 19:16 < Cache_Money> wasn’t sure if there was a way to tell by running some command on the command line but I’m okay to just wait and see if I have a problem sending emails next week 19:17 < twainwek> Cache_Money: https://stackoverflow.com/a/36925113 19:17 < twainwek> also read the fine print 19:18 < Cache_Money> twainwek: I ran that command but I also read the 3rd comment where he said that it’s not accurate ¯\_(ツ)_/¯ 19:18 < Cache_Money> I guess I’ll just wait and see… 19:19 < Gringonar> sorry it took a while 19:19 < Gringonar> https://paste.debian.net/1031937/ 19:19 < Gringonar> thats what ip link gives me 19:19 < justsomeguy> Is pointing a domain name at my roaming laptop with dynamic dns a terrible idea? 19:19 < justsomeguy> I would like to be able to ssh into it from my home server. 19:20 < Gringonar> yes windows is fine 19:20 < Gringonar> but i hate win 10 lol 19:20 < Pusteblume> that would be the best 19:20 < Gringonar> unfortunatly everything comes with it :( 19:20 < Pusteblume> when i finally make a stay there 19:21 < searedvandal> Gringonar, 'ip ad' do you have the correct ip and everything? 19:23 < Elladan> justsomeguy, your laptop is going to be behind firewalls 99.9% of the time. 19:24 < twainwek> Cache_Money: yes, the comment says that it's the minimum version. so i don't see it being a problem for you 19:24 < sandman13> I am trying clear iptables rule if the service has stopped. But how can I check if service has stopped and start cleanup service using systemd? 19:25 < Elladan> justsomeguy, if you're talking about it being on a work network with "roaming" and you VPN in, then you should give it a static DHCP entry and also a domain name, but not dynamic dns. 19:26 < searedvandal> you want to ssh into your laptop from your home server? where do you have this laptop stashed away? normally people want to do it the other way around :P 19:26 < phinxy> Why is my bash terminal broken, cant do work on it anymore. For example piping something to netcat will echo the pipe input and then just nothing happens. cant put more commands in the terminal anymore. 19:26 < justsomeguy> Elladan: I would be taking it from coffee shop to coffee shop, most likely. Just trying to think of ways to get a ssh connection to my laptop. Maybe dynamic dns is a bad idea after all. 19:27 < rcf> justsomeguy: you should definitely look into hosting a VPN on your home server if you want to do that in a sane manner. 19:27 < Dan39> dont be such a debbie downer searedvandal 19:27 < searedvandal> Dan39, I'm no debbie downer, just curious. 19:28 < justsomeguy> searedvandal: Yes. Sometimes I get a document sent to it from another user, and I want the server to automatically send it to my laptop. Also, I want to be able to get in to it in case of theft. 19:28 < Dan39> lol searedvandal and don't be so serious :P 19:28 < justsomeguy> rcf: That's probably a good idea. 19:28 < searedvandal> justsomeguy, vpn on your home server as rcf suggests is the best solution if you want to ssh in to your homeserver. 19:28 < searedvandal> Dan39, I'm not serious at all. just acting serious. 19:28 < Elladan> justsomeguy, if you want to be able to ssh in while the laptop is on some random network, you need the laptop to connect to your home server or have both connect to some shared VPN or something. 19:28 < Gringonar> ipad gives https://paste.debian.net/1031939/ 19:28 < Dan39> oh, i see 19:28 < bls> phinxy: you've likely dumped characters that've caused is to "stop", try Ctrl+q then run the reset command 19:28 < searedvandal> Dan39, ;) 19:28 < Gringonar> im getting more confused by the minute lol 19:29 < _Super> Hi everyone 19:29 < Dan39> justsomeguy: maybe a SSH reverse tunnel will do what you want? 19:29 < Dan39> that can be used instead of like port forwarding 19:29 < searedvandal> Gringonar, have you set up dhcpcd or any other dhcp client on your machine? 19:29 < Elladan> justsomeguy, an ssh reverse tunnel is probably the simplest approach. 19:29 < bls> VPN would be way more reliable than an ssh tunnel 19:30 < phinxy> bls• the problem is persistent after reboots on all consoles 19:30 < Gringonar> no, i havent installed anything its a livecd flashed unto usb wiithout internet i cant install anything :( 19:30 < Dan39> if you have laptop stashed at location that you don't have access to port forward, but want to be able to connect to it over internet, do a ssh reverse tunnel to a machine you do have access to, it will open a listening port on that renite machine that forward to your ssh port on laptop 19:30 < rcf> bls: yes, but if you're going to be in front of the laptop anyway, it might be worth the simplicity. 19:30 < phinxy> dmesg | tail will act the same way. 19:31 < phinxy> input is echo:ed , only way to get another prompt is CTRL-\ 19:31 < searedvandal> Gringonar, alright. I guess activating dhcpcd or whatever dhcp client that the liveusb have is needed 19:31 < Elladan> justsomeguy, if you just need shared documents, look into syncthing. 19:31 < searedvandal> syncthing is great 19:32 < Dan39> a VPN is a great idea too 19:32 < justsomeguy> Well, I should have also said that part of this is just to experiment with ssh :~} 19:32 < justsomeguy> But a VPN sounds like the sane solution. I'm also looking at a reverse ssh tunnel set up with autossh. 19:32 < triceratux> Gringonar: so youre saying its this weeks arch & it boots but when the installer runs lspci it hangs but once you get a console its clear you dont have a network anyway even tho it looks like its up ? 19:32 < Gringonar> its not by default? 19:33 < Gringonar> yes sir this weeks 19:33 < justsomeguy> Elladan: Syncthing is bae. :) 19:33 < rcf> justsomeguy: also, the 'access it in case of theft' is probably going to require this not be a Linux system. It's a bit optimistic to think that a thief is going to learn your particular flavor and create their own account rather than just wipe the system immediately. 19:33 < Elladan> A reverse tunnel will be fine and reliable, though you need to make sure it has the proper keepalive / timeouts set. 19:33 < triceratux> Gringonar: just making sure. those are the right pastes & stuff ;) 19:33 < Gringonar> oh it says 2018 07 01 19:34 < searedvandal> justsomeguy, if you just want to play around with ssh, look into autossh. that would create a persistent reverse ssh tunnel for you 19:34 < Gringonar> when i downloaded it 19:34 < justsomeguy> searedvandal: Thanks. I'm giving it a look now, actually. 19:34 < Gringonar> how can i confirm the version to be save? 19:34 < Gringonar> hmm let me google that 19:34 < Gringonar> sec 19:35 < triceratux> Gringonar: its at the top of those logs you posted. its also uname -a 19:35 < bls> play around with ssh and keep what you learn in the back of your pocket for quick hacks when you run into issues, then learn the right way to do things before you become reliant on it 19:35 < Elladan> Yeah a thief is just going to do one of two things: not even open the laptop and try to sell it to someone as fast as possible, or they're going to hard-wipe it (maybe look for anything they can blackmail you with first) and install whatever on it so it looks like a legit item for sale. 19:35 < Elladan> If you're worried about thieves make sure you encrypt it. 19:36 < metbsd> so which one is better between ubuntu and centos 19:36 < Psi-Jack> CentOS. 19:36 < bls> metbsd: both, neither 19:36 < searedvandal> you'll get 2279 different answers to that question in here metbsd 19:36 < Gringonar> cat /proc/version says linux 4.17.3.1-Arch 19:36 < Elladan> Ubuntu for most users. 19:37 < Gringonar> oh 19:37 < Elladan> CentOS for people who need Redhat admin stuff on a server. 19:37 < Gringonar> how many ways to do stuff 19:37 < searedvandal> Gringonar, are you plugged in with ethernet or have you set up wifi? 19:37 < justsomeguy> bls: Sage advice, tanks. 19:37 < Gringonar> lol 19:37 < Psi-Jack> Elladan: Incorrect. :p 19:37 < justsomeguy> *thanks 19:37 < metbsd> i find ubuntu and centos quite different. so i want to stick with one of them 19:38 < bls> metbsd: you're the only one that can decide which is better for your needs 19:38 < searedvandal> stick with what fits your needs the best 19:38 < searedvandal> or what you like the best 19:38 < Psi-Jack> metbsd: What's the actual use case scenario? 19:38 < Gringonar> i plugged in an ethernet cable directly into my router 19:38 < searedvandal> or what you're the most proficient in 19:38 < triceratux> Gringonar: thats what im running as well [ 0.000000] Linux version 4.17.3-1-ARCH (builduser@heftig-6515) (gcc version 8.1.1 20180531 (GCC)) #1 SMP PREEMPT Tue Jun 26 04:42:36 UTC 2018 19:38 < metbsd> desktop, linux servers 19:38 < triceratux> (its swagarch with intel graphics of course) 19:39 < searedvandal> swaag 19:39 < Psi-Jack> metbsd: For desktop use, I generally recommend Solus, Fedora, and Debian, in that order. 19:39 < metbsd> will install linux on my old laptop that are not capable of running windows 19:39 < metbsd> whats solus 19:39 < bls> metbsd: either one can do either/both 19:39 < Psi-Jack> metbsd: Solus is a Linux distribution designed specifically for desktop/laptop use in mind. 19:39 < Gringonar> im still unsure if i just hve the configure my ethernet or that it isnt wrking 19:39 < searedvandal> metbsd, I run Arch on my desktops and Ubuntu on my servers. Because that's what fits me and my needs the best. You're the only one who can decide what fits your needs and likes the best. 19:40 < Psi-Jack> metbsd: For servers, I'll generally recommend CentOS and Debian, also in that order. 19:40 < metbsd> can centos do the job? it's almost all i use 19:40 < triceratux> i always say go with the xubuntu lts & dont forget to say xforcevesa on the way up. you never know when theres an nvidia chip in there somewhere 19:40 < Psi-Jack> metbsd: Do what job? 19:40 < metbsd> what's solus based on 19:41 < metbsd> old laptop 19:41 < Psi-Jack> metbsd: Nothing. Solus is a unique distribution. 19:41 < metbsd> i know how good centos is as server os 19:41 < Elladan> If you're used to centos just stick with it. 19:41 < revel> I'm sure it's based on tons of other distros. Just not directly. 19:41 < metbsd> ok 19:41 < searedvandal> Gringonar, check that dhcpcd is running and don't have any errors 19:41 < Psi-Jack> metbsd: CentOS could be OK as a desktop, but you'll likely need a few 3rd party repos to make it useful. nux-dextop for one. 19:42 < Elladan> If you want to learn something new, Debian or a debian-based distro like Ubuntu, Mint, or whatever will give you some widely applicable skills. 19:42 < Gringonar> ethtool eth0 gives "no such device"so i asume its not working? 19:42 < peetaur> Gringonar: ip l to list devices 19:42 < searedvandal> Gringonar, your device isn't eth0, it's enp3s0 19:43 < searedvandal> according to your pastes 19:43 * peetaur didn't look far enough for pastes 19:43 < Gringonar> ah yes 19:43 < Gringonar> weird name 19:44 < Gringonar> now im getting proper output 19:44 < Psi-Jack> Ethernet, Network, Port 3, slot 0. 19:44 < Psi-Jack> Not wierd at all. 19:44 < Gringonar> to a noob like me it is 19:44 < Psi-Jack> Well, to a noob, everything is wierd. :p 19:44 < Gringonar> ty for that makes since now 19:44 < Psi-Jack> "thank you", not "ty" 19:44 < searedvandal> what does eno1 translate to then Psi-Jack ? 19:45 < ][_R_][> Gringonar: sed -i 's,ENV{net.ifnames}=="0",ENV{net.ifnames}!="1",' /usr/lib/udev/rules.d/80-net-name-slot.rules 19:45 < ][_R_][> It'll be fixed on reboot 19:45 < Gringonar> thank you very much sir, i appreciate your explanation 19:45 < Psi-Jack> searedvandal: Ethernet Network Onboard device 1. 19:45 < searedvandal> Psi-Jack, thanks. 19:45 < glix> Hey guys, what would you say if one of my friends want to start using linux and his first distro is Manjaro with i3wm ? 19:46 < searedvandal> nothing 19:46 < Psi-Jack> glix: That's just bad. 19:46 < triceratux> madness ! 19:46 < Psi-Jack> :p 19:46 < Gringonar> let me google how to set op this dhcpd thing 19:46 < searedvandal> Gringonar, https://wiki.archlinux.org/index.php/Dhcpcd 19:46 < Gringonar> thanks 19:46 < jim> searedvandal, sounds like the name of one of your interfaces? 19:47 < searedvandal> oh yes, I'm rocking a eno1 19:47 < Psi-Jack> heh 19:47 < Psi-Jack> o = onboard. :P 19:47 < searedvandal> I should have figured 19:47 < searedvandal> I blame the heat 19:47 < Psi-Jack> How about this heat? 19:47 < searedvandal> I don't function well above 20 celcius 19:48 < Psi-Jack> Celcius? What 3rd world count... Nevermind. :) 19:48 < glix> Psi-Jack: I know right! He was just impressed by the workflow I had going with my laptop using i3 and wants to make the direct leap buy just starting things off with i3 19:48 < Psi-Jack> glix: Generally a bad idea. :p 19:48 < Psi-Jack> Especially starting with Manjaro for one. 19:48 < searedvandal> Psi-Jack, sorry. 68 fahrenheit 19:48 < searedvandal> := 19:49 < Gringonar> yeah 19:49 < glix> Psi-Jack: Why is starting with Manjaro bad though? 19:49 < Gringonar> its 25celcius now here sucks 19:49 < Psi-Jack> glix: There's 255 reasons Manjaro alone is bad. :p 19:49 < searedvandal> glix, hey, if he wants to jump in the deep end without floaties, why not let him 19:49 < phinxy> bls• The problem was a bash directory bookmark shellscript sourced in .bashrc 19:49 < searedvandal> there's at least 1000 lines of reasons why manjaro is bad 19:50 < searedvandal> but hey, if it works it works 19:50 < phinxy> 1. man-jar-o is the 3 syllables? 19:50 < glix> searedvandal: I didn't tell him not to, in fact I helped him with making a bootable DVD. 19:50 < phinxy> oh wow im going to bed, cant type for squats 19:50 < searedvandal> glix, bootable DVD? in 2018? nice 19:52 < Psi-Jack> I still have a nearly full spindle of blank DVD+R 19:52 < glix> serafincpd: I also told him to use a usb but he said he didn't have one lying around xD 19:52 < Psi-Jack> DVD+R's, DVD+RW's, BD-R's, and CD-R's. 19:52 < searedvandal> I only keep CD-R's around 19:52 < searedvandal> they're still pretty useful 19:53 < Psi-Jack> For what? Legacy? :p 19:53 < FreeFull> Psi-Jack: I wonder if they've been exposed to enough UV to be bad by now 19:53 < Psi-Jack> FreeFull: Nope. They're in a dark place. :) 19:53 < searedvandal> Psi-Jack, car stereo, dj cd players, 'see how far you can throw a disc'-contests 19:54 < Psi-Jack> Car Stereo? Bluetooth. DJ Cd players? Bluetooth. 19:54 < Gringonar> it says interface not found or invalid when i do dhcpcd interface 19:57 < searedvandal> Psi-Jack, my car stereo only plays cd and fm radio. my dj players only play cd, usb, sdcard or over ethernet 19:57 < searedvandal> so cd is the ultimate fallback option 19:57 < Psi-Jack> searedvandal: your car stereo takes a line input no? 19:57 < searedvandal> nope 19:57 < Psi-Jack> What 3rd world country do you live in again? ;) 19:57 < searedvandal> Norway 19:57 < searedvandal> ;) 19:57 < Dan39> Gringonar: umm did you replace "interface" with the actual interface name like eth0 or enp0s0? 19:58 < jim> Gringonar, see what interfaces you have: ip l 19:58 < Psi-Jack> Pretty much every car in the US has a line input at least with their factory standard car stereo. Heh 19:58 < searedvandal> or enp3s0 which is the name 19:58 < Gringonar> oh now 19:58 * Gringonar hits his face 19:58 < Psi-Jack> Even a Smart car. Which had the line input in the glove box, but still had it. 19:58 < Psi-Jack> hehe 19:58 < jim> stop hitting my face! 19:59 < Gringonar> mine for being dumb 19:59 < Psi-Jack> Ouch! Mine too! 19:59 < searedvandal> Psi-Jack, my car is 17 years old and the cd player is probably older than that, some aftermarket thing that came with the car. car didn't cost me enough to bother upgrading anything else than the important bits like breaks and stuff 19:59 < bls> no way, you need the sweet CD player mounting arm then you use the casette adapter 19:59 < cloudbud> How can I set up a monitoring server using some script ? 19:59 < meyou> you had to spend like $200 for a box that would impersonate a CD changer to get an aux input on my old prius 19:59 < jim> contrary to popular belief, it's not getting hit in the head lessons in here! 19:59 < Dagmar> cloudbud: by writing the script 19:59 < bls> cloudbud: that's too vague to garner a valid response 20:00 < Psi-Jack> Ahhhh 20:00 < searedvandal> bash monitor.sh 20:00 < Psi-Jack> searedvandal: Well, as for DJ stuff.. Most DJ's here in the US these days don't even use CDs at all. Just laptops, cellphones, tablets. 20:00 < Psi-Jack> Plus their usual audio/video equipment they may need. 20:00 < Gringonar> yay i have internet!! 20:01 < Gringonar> lol 20:01 < Gringonar> i may actually get this working lol 20:01 < jim> Gringonar, what did your interface name turn out to be? 20:01 < searedvandal> Psi-Jack, same here. I never use anything other than my trusty USBs. unless they fail. then I have backups in place, where CDs have saved me at 2 occasions. 20:01 < Gringonar> Thanks a lot guyz!! 20:01 < Gringonar> enp3s0 20:01 < Gringonar> i can ping sites 20:02 < searedvandal> thats good 20:02 < jim> by name? 20:02 < Gringonar> havent actually try using pacman yet 20:02 < triceratux> welcome to linux. wait till it actually *works* 20:02 < glix> I know it's old news, but what do you guys think about irc.com ? 20:02 < Psi-Jack> Ugh, arch. 20:02 < Gringonar> im used to debians package manager 20:02 < jim> don't get eaten by ghosts 20:02 < searedvandal> you'll love pacman 20:02 < Psi-Jack> pacman is literally the worst package manager of them all. :p 20:02 < searedvandal> it installs things. all I need it to do 20:03 < Gringonar> no fighting lol 20:03 < bls> it's a package manager. it either installs packages 20:03 < bls> or it doesn't 20:03 < Gringonar> im just happy it works at this point 20:03 < Psi-Jack> Heh. Well, I've even had to submit fixes for pacman. 20:03 < Psi-Jack> Because the devs refused to accept that it was their problem to deal with properly, and when I showed them, they were like.. Oh.. Okay... 20:04 < searedvandal> it seems pretty functional these days at least. 20:04 < bls> typical of that community 20:04 < Psi-Jack> Yeaaah. 20:04 < Psi-Jack> And it was about how they used gpg. :) 20:05 < Gringonar> so i guess pacman -Sy is like apt-get update? 20:05 < searedvandal> Gringonar, https://wiki.archlinux.org/index.php/Pacman/Rosetta 20:06 < searedvandal> shows you the different options and their equivalent on things like debian 20:07 < Gringonar> oh boy it says it will upgrade the whole system afterwards 20:08 < Gringonar> i barely got it working am i sure i wanne do that? 20:08 < bls> you've got nothing invested yet, and you're expected to do this at least once a week, so yes 20:09 < bls> avoiding the daily/weekly upgrade cycle is one of the best ways to trash an arch system 20:10 < searedvandal> yeah, update at least once a week. if not, troubles 20:11 < Gringonar> i got a bunch of errors :( 20:12 < bls> once you are invested, you're going to want to make sure you employ your data backup scheme and have an upgrade rollback scheme ready 20:13 < Gringonar> oh boy what have i gotten myself into 20:13 < Gringonar> well its a lvcd at this point 20:13 < bls> oh, then you don't want to upgrade on top of that 20:14 < Gringonar> First i like to get a desktop manager running 20:14 < milp> hi, is it possible to run the same x-server on multiple graphics cards, displaying the same desktops? - ubuntu 16.04 x64 20:14 < Dagmar> No 20:14 < milp> aw 20:14 < milp> vnc then i guess 20:14 < Gringonar> i dont mind some cli but exclusivly cli is not for me 20:15 < triceratux> Gringonar: thats why i run swagarch from the iso. its an arch snapshot with xfce already working & i dont have to upgrade or Sync it 20:15 < triceratux> https://swagarch.gitlab.io/ 20:16 < triceratux> your best bet is always to get xubuntu lts going 20:16 < Gringonar> well i need to understand this laptop first since this is the only Livecd that works 20:16 < Gringonar> ive tried debian ubuntu fedora slackex 20:16 < Gringonar> I think i need more understanding first lol 20:17 < Pentode> Gringonar, have you tried disabling UEFI? i know i asked this before but i dont recall if yoy said yes or no. 20:18 < Gringonar> yes 20:18 < Gringonar> but it works better with uefi 20:18 < Pentode> sometimes UEFI will ignore boot orders even if it appears enabled 20:18 < Gringonar> since windows 10 is preinstalled its obviously using uefi 20:19 < Gringonar> running in legacy mode doesnt even give grub 20:19 < Pentode> i bet if you pull the drive with windows out and boot the live media it works. 20:19 < Pentode> if so, get rid of the windows efi partition and try again 20:19 < Gringonar> i disabled the win10 boot drive in bios for now 20:20 < Gringonar> if i cant get this thing working im returning it 20:20 < Pentode> did you ever get the network adapter working? 20:20 < Pentode> if returning is an option i'd get a laptop more conducive to linux 20:20 < Gringonar> yes just did with the help of the guyz here 20:20 < Gringonar> awsome :) 20:20 < Gringonar> yeah i considered system 76 20:21 < Dagmar> You'll have to tell it to boot from the optical media or thumb drive _first_ 20:21 < Gringonar> but im european so importat tax+ shipping adds 30% to the price :( 20:21 < Pentode> the only reason im suggesting this is i've had similar issues dozens of times and it appears having uefi/windows efi partition still active on a disk prohibits some machines from booting from other media for whatever retarded reason 20:21 < Dagmar> Legacy mode or not, "secure boot" disabled or not, it's *going* to default to booting from the internal hard drive first 20:21 < leftyfb> Anyone know if there's a way to run a bash function from my .bash_aliases within a ProxyCommand as part of ssh? 20:22 < Pentode> Dagmar, yeah earlier he tried disabling secure boot _and_ messing with the boot order. 20:22 < Dagmar> Boot device selector is usually just F11 20:22 < bls> you can buy the same hardware system76 sells without their markup 20:22 < Gringonar> thanks Dagmar i did all that 20:22 < Pentode> Gringonar, did you try the quick boot menu or just the bios' internal settings? 20:22 < Pentode> might make a difference 20:22 < Dagmar> Gringonar: You might want to pay a bit closer attention to reality 20:23 < Gringonar> the bios itself i think 20:23 < Dagmar> Gringonar: When you've stuck in a UEFI-capable boot device and it's not what's being booted, _clearly you missed something_ 20:23 < Gringonar> im currently ooted in archlinux livemode and got internet running 20:24 < Gringonar> booted* 20:24 < Pentode> Gringonar, try the f11 menu or whatever it is on that machine. also, did you try a different thumb drive? 20:24 < Dagmar> Who made this laptop? 20:24 < bls> leftyfb: you're better off sourcing your function into a shell script and calling that 20:24 < Gringonar> MSI 20:24 < Gringonar> its an MSI GE73 8RF raider 20:24 < Dagmar> Okay so there's probably going to be an entire menu inside the UEFI setup devoted to which devices you can boot from 20:24 < leftyfb> bls: yeah, I was sure that would be an option but would rather not if possible 20:25 < Gringonar> with the coffeelake/nvidia optimus setup 20:25 < Dagmar> Use it like a bellweather. ANything that's viable for booting via UEFI *should* show up in that menu 20:25 < triceratux> have you tried typing startx ? 20:25 < bls> leftyfb: otherwise, your ProxyCommand is going to end up something complicated like: bash -c 'source ~/.bash_aliases; yourfucntion' 20:26 < leftyfb> bls: I'm ok with the complication 20:27 < leftyfb> bls: HA! That worked! 20:30 < programings> f 20:31 < evanesoteric> Do IRC chats run in RAM? What I am really asking is, if I leave linux chat open for 1 year, and don't clear logs, will my computer crash!? 20:31 < MrElendig> only if the irc client suck 20:31 < dot0x01> o/ 20:31 < revel> If it's a well-written client, then that should't happen. 20:32 < MrElendig> not that you should have a one year uptime 20:32 < scacyn> in linux it will crash your program, not the computer in the worst case 20:32 < MrElendig> that is one year of security updates that you have not applied 20:33 < Dagmar> It represents a year's-worth of people's wasted time 20:35 < Gringonar> i'm reading https://wiki.archlinux.org/index.php/NVIDIA_Optimus#Using_nvidia 20:35 < Dagmar> While the government can privatize an industry if it decides it's necessary to do so, large groups of investors cannot legally force someone to sell controlling interest in their own, exclusively-owned company 20:35 < Loshki> evanesoteric: a year's uptime with current distros would be an achievement 20:35 < Dagmar> @#$#@ wrong channel 20:36 < Gringonar> but i get somewhat confused it sais it can work and cant at the same time? 20:36 < Dagmar> Where does it say this? 20:36 < delsol> Is it possible to set intel_idle.max_cstate=1 post-boot? 20:37 < Gringonar> "Getting Optimus graphics to work on Arch Linux requires a few somewhat complicated steps" 20:37 < delsol> Optimus graphics, barf. 20:37 < Gringonar> "The proprietary NVIDIA driver does not support dynamic switching like the nouveau driver (meaning it can only use the NVIDIA device). " 20:37 < MrElendig> it is a pain no matter which distr 20:37 < Dagmar> delsol: You can set sysctl values at any time you like. Whether or not they'll *do anything* when you set them depends on whatever's mentioned in the documentation 20:37 < MrElendig> is even troublesome on windows 20:38 < delsol> the intel GPU never shuts off... and the memory bandwidth for the framebuffer is ALWAYS burning. 20:39 < Dagmar> Gringonar: Just keep re-reading that sentence until the confusion goes away. 20:39 < Gringonar> So i should go about installing the proprietary driver? 20:39 < Dagmar> Gringonar: It does not say they work and don't work simultaneously 20:39 < Dagmar> Gringonar: By all means if you enjoy sub-standard performance and crashing, continue using the nouveau driver 20:40 < delsol> Gringonar: optimus graphics for linux is about as practical as using a Nimitz Class aircraft carrier to take the family to little league practice. 20:42 < Dagmar> Gringonar: The important takeaway from that bit you pasted is that they're implying that only the nVidia driver is going to actually use the hardware accelleration of the nVidia GPU 20:42 < Dagmar> Undetrstand that the Optimus setup is a very weird configuration 20:42 < Dagmar> You've got two GPUs that have to work together 20:42 < delsol> And inferior to previous and later switchable graphics that used hardware dmux to do the switching. 20:43 < Gringonar> yeah i figured that, but someone else told me the nvidia card passes the rendered image to the iGPU as jpeg so that wont break stuff? 20:43 < Dagmar> One, the Intel one, is just like every other Intel video chipset. It's generally s***ty and only good for 2D stuff like drawing browser windows and xterms. 20:43 < delsol> Gringonar: its not JPEG, its raw pixel buffer. 20:43 < Dagmar> THe other, the nVidia oen, typically sits around and does nothing until you are trying to do some 3D rendering (i.e., play games) 20:44 < delsol> but yes. the nvidia card passes every frame and every pixel to the intel framebuffer by basically exploiting a buffer overflow backwards kindof sortof. 20:44 < delsol> the display is ONLY hooked up to the intel graphics.... and the intel graphics displays whats in the framebuffer, either what IT put there, or what the nvidia put there over the top. 20:45 < Gringonar> yes but the determing "who has to put it there"is part of the optimus thing? 20:46 < delsol> kindof. 20:46 < delsol> in this way, the intel GPU is consuming a bunch of memory bandwidth all the time for the framebuffer, and can't be turned off or powered down... so that hurts CPU turbo performance. 20:46 < delsol> the other issue is WHO put it there, WHO is supposed to be rendering each frame, 20:46 < delsol> and getting the OS and drivers to deal with that. 20:47 < delsol> in SOME cases, supposedly the intel GPU gets to render the background and the OS, and the Nvidia card renders a window mode game on top.... 20:48 < Gringonar> so untill optimus is fully supported(if ever) we get overhead? 20:48 < delsol> even IF it was fully supported, there is tons of overhead 20:48 < delsol> the ONLY advantage optimus has... is that the vendor didn't have to add a couple hardware DMUX's. 20:49 < delsol> it saved the vendor $10 20:49 < Gringonar> ok so you recommand to install nv driver of nouveau anyway? 20:49 < Gringonar> over* 20:50 < Dagmar> Yes 20:50 < delsol> I recommend putting your right foot up on something solid to present your right knee at an acceptable height, firmly grabbing the laptop with both hands, raise it in the air and rapidly move it downward so that your hands go on either side of your knee snapping the laptop into at least two pieces. 20:50 * delsol wouldn't buy anything with optimus just because of how much of a piece of shit it is. 20:51 < Gringonar> isnt all Coffeelake/nvidia optimus? 20:51 < Gringonar> atleast in laptops 20:54 < delsol> Don't think all, but its becoming more common now that the windows drivers are only half fucked. 20:56 < jak2000> hi all 20:56 < jak2000> i am on initramfs how to restart the server? 20:58 < jim> jak2000, what dist? normally it should just come up 20:58 < jak2000> debian 20:58 < jak2000> ich command? 20:58 < lnnb> kill services, unmount everything that isn't /, remount / as read only, sync, then exit 20:59 < jak2000> i need restart the server, how to? 20:59 < jak2000> shutdown command not found 20:59 < lnnb> hit the reset button, or call reboot syscall 20:59 < lnnb> with reboot option* 20:59 < jim> there'd be stuff mounted within the initrd? 20:59 < peetaur> jak2000: ctrl+d probably 21:00 < jak2000> ooo 21:00 < jak2000> peetaur worked 21:00 < peetaur> the initrd probably mounts it all read only, so you don't have to worry about mounts 21:00 < jak2000> thanks 21:00 < jim> sounds to me like the kernel hasn't been read in yet 21:00 < lnnb> initrd can do anything it wants, i make no assumptions 21:01 < peetaur> if you're in initrd, you are running the kernel; it is the kernel that unpacks the initramfs 21:01 < evanesoteric> Hey, thank you for answering the IRC/memory question! Apologies for the delay. 21:02 < TR1950X> is there a format I can use to make notes with linking support? 21:02 < jim> when you have no memory of the question, IRC/memory == infinity 21:02 < TR1950X> linking like html href 21:03 < jim> TR1950X, sounds like you want a wiki? 21:03 < ksk> evanesoteric: yes, an irc client kind of runs in ram. no, if programmed correctly (haha) it should never make your machine crash. 21:03 < TR1950X> jim: yea but I also would like be able have versiong support 21:03 < jak2000> much thanks 21:04 < jak2000> fixed... 21:04 < birdbolt1> hey looking for advice on best practices. I am building my first webapp on docker. Is it better to build my app and then just copy the files into the docker container, or build it in the docker container? 21:04 < ksk> evanesoteric: if you enabled logging to disk you could disable your machine from writing anything though (which is not good). 21:04 < birdbolt1> My problem with the later approach is that (build)x(number of instances) will be costly 21:04 < ksk> evanesoteric: but text-only logs should take a while on a decent drive to fully utilize it ;) 21:04 < dot0x01> opinions on dualbooting vs linux with windows vm? 21:05 < ksk> dot0x01: why do you need windows is the question Id ask 21:05 < birdbolt1> dot0x01, windows sucks in a vm but it depends on what u wanna do with it 21:05 < birdbolt1> ksk, quit hating, people have tehir reasons 21:05 < dot0x01> video editing is the only reason why ive kept windows for such a long time 21:05 < bls> TR1950X: have you checked out org-mode? 21:05 < birdbolt1> ^do u have a gpu? 21:05 < birdbolt1> dual boot 21:05 < ksk> birdbolt1: no hate over here, thats your imagination 21:05 < TR1950X> bls: no idea what that is. will look into it 21:06 < birdbolt1> linux doesnt always play well with passing the gpu to windows through a virtual machine 21:06 < jim> birdbolt1, I'd build a minimal webapp (like a hello world) in the same toolkit you're gonna use, then try to serve it with the final docker container... if that works, you know you can serve the full one, so continue working on it 21:06 < dot0x01> hmm 21:06 < dot0x01> so dual booting would be my best option 21:06 < bls> TR1950X: it's like markdown on steroids. has implementations in emacs, vim, sublime 21:07 < bls> and mobile apps 21:07 < dot0x01> using linux 95% of the time but it'd kinda be nice to edit stuff on windows since im actually used to vegas 21:07 < birdbolt1> jim, I've already got a production setup going; currently, i build on my dev machine, and the files are just copied into the container. Iwondering what people who do this for a living do 21:07 < dot0x01> no clue about any possible alternatives on linux 21:08 < birdbolt1> dot0x01, yes, dual boot for maximum performance as well, since i suppose you may work with large files or complex edits 21:08 < ksk> dot0x01: afaik there are ways to make an dualboot-installed windows available inside virtualbox-vm running on linux- but that might not be too safe, no idea.. 21:08 < bls> birdbolt1: the ecosystem is pretty fast and loose, most people are all doing their own ad-hoc stuff with it 21:08 < birdbolt1> bls, i dont understand 21:09 < ksk> d0tn0t: does your windows software run with WINE maybeß 21:09 < ksk> s/ß/? 21:10 < birdbolt1> idk why people always recommending wine like it works with any major software 21:10 < birdbolt1> chances are, if theres something you need from windows which u cant quite get from linux, wine wont be able to run it. 21:10 < dot0x01> birdbolt1: edits aren't really that complex but files are quite large so it would be a win-win to dualboot windows w/ arch 21:10 < bls> I was saying that there's no established best practice or even common practice. people are all making it up as they go along 21:11 < birdbolt1> Games, office suite, adobe suite, autodesk suite, etc 21:11 < bls> linux runs any of these? https://appdb.winehq.org 21:11 < dot0x01> dont really play games anymore, mainly just photoshop/premiere/vegas 21:11 < fugee> can i read data from a file encrypted.block with encrypted.key present 21:12 < fugee> i deleted a bunch of chromeos partitions i thought i wouldn't need lol 21:12 < birdbolt1> bls, ahh i see. I think it might be wiser for me to build and copy files in for now 21:14 < bls> I see everything people automating their entire build, treating the containers as black boxes, and respinning on every change, and I see people using containers to run an init system and ssh then managing the container internals by hand, and everything in between 21:16 < ldlework> birdbolt1: have you seen multi-stage builds? 21:17 < shambat> I have the following intry in my route -n table: 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 p1p1 does this look like it does anything? 21:17 < dot0x01> brb 21:18 < ldlework> bls, while people in general are trying lots of different things with regards to containers, I would argue that Docker and Google and some others have figured out quite a bit with regards to the advantages of using them in certain ways 21:19 < ldlework> they are very flexible technology so there is no intrinsically correct way to use them, but there are only certain ways to use them that best utilize the properties of modern scheduling and orchestration techniques and tools 21:20 < bls> ldlework: that's true and I see that building around things like kubernetes. but there's a lot of sloppieness and ignorance in the ecosystem that isn't being chased out 21:21 < ldlework> yeah, like if you run multi-process containers with init or whatever, you can't scale the boxed components separately or easily conceptualize their independent resource demands 21:22 < bls> things that a sysadmin would never allow, but because it's in a container, it's somehow becomes OK 21:22 < ldlework> docker works best as an abstraction around "running a unix process". you can now treat all unix processes as essentially having a single api, the docker api 21:22 < ldlework> bls: heh 21:22 < ldlework> bls: I built the first automated build service for DockerHub so I have seen some things :) 21:30 < metbsd> can docker run cluster mangodb? 21:31 < metbsd> i mean mongodb 21:31 < bls> haha https://www.hulu.com/watch/10316 21:32 < jim> didn't saturday night live use mangodb? 21:33 < metbsd> i used to setup cluster mongodb on about 10 vms 21:33 < jim> metbsd, probably it can... I've heard stories about serious breakages with mongodb, and I normally recommend that people look at postgres 21:35 < metbsd> postgresql seems awesome 21:35 < metbsd> but i doubt they will migrate 21:35 < metbsd> they are using mysql with middle ware 21:36 < bls> my current setup is mysql based, and when we needed a document store, we went with postgres and are now looking to migrate the relational stuff 21:36 < jim> remember that oracle now owns mysql :) I don't know where its future lies :) (and -yes-, that's -FUD-) 21:37 < bls> not unfounded though given past history 21:38 < jim> bls, in which direction? 21:38 < metbsd> i know nothing about oracle. another dba is on it 21:38 < jim> founded FUD :) 21:38 < MrElendig> oracle tried to keep ssl support a pay to play extension to mysql so.... 21:39 < jim> metbsd, oracle is horribly expensive... to give you an idea, major banks use it 21:39 < bls> jim: move our realational stuff from mysql to postgres so it can live alongside the document store 21:40 < jim> bls, do you folks expect that to be pretty easy (albeit time consuming)? 21:41 < Salsha> Hello guys! 21:41 < jim> hi 21:42 < Salsha> i'm in trouble. i tried to install q4os and debian 9.4 both 32 bit and i keep getting hashes mismatch when trying to usa apt. i tried any solution, i reinstalled debian several times... but still getting this error 21:42 < bls> yeah, we only use a couple mysql specific features, and we've slowing been converting our codebase to use sqlalchemy 21:42 < Salsha> wghat is it caused by? it never happenedto me, with any pc 21:42 < jim> Salsha, is it a 64 bit machine? 21:42 < Salsha> Nope 21:42 < Salsha> it's a prescott 21:43 < Dagmar> Betcha the issue is installing two apt-based distros in the same filesystem without actually wiping first 21:43 < Dagmar> i.e., a _thoroughly_ corrupted package database 21:44 < Salsha> well i always wipe the hdd before reinstalling. And on this hard disk there was xp before 21:44 < Salsha> So not an apt based distro! 21:44 < Dagmar> So why is q4os even relevant to this? 21:45 < Salsha> Dagmar: i'm just saying it's just not an issue met with the latest debian release, but also on other deeply debian based distros 21:45 < Salsha> and i do not get why it happens 21:45 < Dagmar> Strangely mine all work fine 21:46 < Salsha> Dagmar: guess? if it wasn't strange i would not being asking here? 21:46 < jim> Salsha, ok, how exactly are you installing? start with: what are you using to install? 21:46 < Dagmar> I'm saying outright, it's something you're doing wrong that you've not mentioned. 21:46 < evanesoteric> Anyone here familiar with IBM Mainframes or the "IBM Mainframe Environment"? 21:46 < Salsha> jim: i'm using a bootable usb, created with rufus, and using classical install. 21:47 < Salsha> evanesoteric: i'm not i'm sorry 21:47 < Dagmar> evanesoteric: I will *cut you* 21:47 < triceratux> evanesoteric: wut like submitting jcl to compile cobol for cics & stuff ? 21:47 < evanesoteric> Is that an inside joke Dagmar? I'm not in the know yet. 21:47 < Dagmar> Well, for one thing, "mainframe" isn't something that runs Linux 21:48 < Dagmar> For the other, "mainframes" are obsolete 21:48 < Dagmar> ...and _nasty_ 21:48 < scacyn> so z-series is not a mainframe? 21:48 < triceratux> mainframes run plenty of linux. & linux compiles plenty of cobol 21:48 < scacyn> those run linux if i'm not mistaken 21:48 < Salsha> Dagmar: you must be some kinda of genius or some asshole in a really bad mood tonight 21:49 < xamithan> What cipher do I need for tls 1.3 ? 21:49 < Salsha> ok let's stop trying debian based distros since nobody knows how to solve this damn issue 21:50 < phogg> Salsha: What issue? 21:50 < Salsha> phogg: i keep getting hash missmatch error when trying to update apt 21:50 * phogg scrolls up 21:50 < jim> Salsha, when you say "classic", what does that mean exactly? 21:50 < phogg> Salsha: How sure are you that your hardware is good? 21:50 < Salsha> jim: no gui 21:50 < evanesoteric> I ask because, I have an interview Monday for a NYS job. It's considered a bonus to know "IBM Mainframe Environment, MVS, JCL 370/390, JES2, Z/OS Operating System" - I don't know those things. The things I do know I left out of that quote. 21:51 < Salsha> phogg: windows 7 runs smoothly, it's usually the first one failing when there is any hardware problem eheheh 21:51 < jim> Salsha, what's the exact name of the installation file you downloaded to install debian? 21:51 < phogg> evanesoteric: there's no way to learn enough about those things between now and Monday to make a significant difference. 21:51 < evanesoteric> Was wondering if it IBM mainframes were BSD or Unix. If there are many varieties, and what not. 21:51 < phogg> evanesoteric: neither 21:51 < triceratux> Salsha: i can vouch for both debian 9.4 & q4os 2.5r1 64bit as being robust viable distros. but i dont run their installers or their 32bit versions. something wildly strange is going on 21:51 < bls> well you're not going to snag one and get some good learning time in before then, so be honest in the interview 21:51 < phogg> evanesoteric: Z/OS is a completely different beast that Unix 21:52 < bls> VMS is going to be the closest thing 21:52 < phogg> s/that/than/ 21:52 < Salsha> this: https://q4os.org/dnt2.html and this: https://cdimage.debian.org/debian-cd/current/i386/iso-cd/debian-9.4.0-i386-netinst.iso 21:52 < evanesoteric> Interesting. What's MVS, JCL, and JES2? 21:52 < jim> that sounds right 21:52 < triceratux> evanesoteric: at least go thru the wikipedia on mvs/jcl & jes2 21:52 < evanesoteric> ahh gotcha 21:53 < triceratux> if they offer you a job take it 21:53 < phogg> Salsha: it's not impossible that there's something fundamentally broken wit the 32bit discs, but it's hugely unliekly for it to have gone unreported. I'd believe any other explanation first. 21:53 < Salsha> triceratux: i see. I have installed q4os over any pc i had by hand, and this is pretty much surprising 21:53 < jim> Salsha, you did get the correct file... it's both the latest stable debian, and correct for your arch 21:53 < phogg> evanesoteric: JCL is IBM's job control language: https://en.wikipedia.org/wiki/Job_Control_Language 21:53 < Salsha> jim: eheheh if i didn't i'd be pretty much dumb :P 21:53 < twainwek> what's nys? new york state? 21:53 < evanesoteric> & yeah 21:54 < evanesoteric> ^ 21:54 < evanesoteric> Sooo like... people use IBM mainframes? Or are those entirely obsolete? 21:54 < jim> Salsha, do you think the machine has hardware issues? 21:54 < Salsha> phogg: i'm gonna try again with an older release and let's see! 21:54 < phogg> Salsha: I can say for sure that the Debian jessie netinst 32bit disc works OK. That's the last one I used myself. 21:55 < Salsha> jim: it shouldn't... i actually switched the processor two days ago but stress tests gave no problem 21:55 < Salsha> maybe i'll try changing the installation device .... let's see if that's the issue! it's the only thing that comes to mind 21:55 < Salsha> and btw, thank you a lot for your support guys 21:56 < phogg> Salsha: if your works-OK windows install is on a different physical disk that could explain it. A bad disk or bad cable can result in write errors. 21:56 < jim> Salsha, (the thing is, that particular file has worked over and over for folks who wanted to install debian... I'm surprised it didn't work for you) 21:56 < triceratux> evanesoteric: https://www.google.com/search?q=ibm+mainframe+fortune+500+usage 21:57 < Angelica> Hi, I just installed gentoo and nouveau was causing bad video output so I replaced it with nvidia. The problem I'm having now is that X uses Wayland when nvidia is configured. I know that one possible solution is to install gdm and add WaylandEnable=false, but the 212 dependencies are going to take hours to compile. Is there any other way to tell X not to use Wayland? It's really slow due to my graphics card not being Intel. 21:57 < phogg> I've used some IBM big iron before, but never as an administrator. As far as I can tell the entire design of it is completely alien if you're used to a "normal" Unix-like system such as e.g. Windows. 21:58 < jim> Salsha, that file is also a "hybrid", meaning that it works when written to a usb memory stick, or you can burn it to a cd blank 21:58 < phogg> Angelica: X should not "use wayland"; Wayland should be used *instead of* X. 21:59 < phogg> Angelica: and then Wayland can run the xwayland program to provide an X server for X clients that will display within wayland. 21:59 < Salsha> jim: phogg i had to change the hdd cable too! Right! didn't think about it! 21:59 < Salsha> good hint 21:59 < bls> aka "what'd you do to your USE flags" 21:59 < Salsha> let's see 22:00 < Angelica> phogg: Thanks, however this is the process, it says X is what's running, but I had this same problem on other distros. Disabling Wayland in gdm solves it. root 4805 24.2 0.5 1925472 368632 tty1 Sl 19:22 8:49 /usr/bin/X -nolisten tcp :0 -auth /home/luz/.serverauth.4788 vt1 22:00 < phogg> Angelica: I don't know what wayland has to do with it, then. 22:00 < leftyfb> bls: The bash -c as part of ProxyCommand partially worked. I get as far as "-bash: line 1: SSH-2.0-OpenSSH_7.2p2: command not found" It looks like it's trying to run the output of the ssh command that the function does as opposed to just ssh'ing 22:01 < Angelica> phogg: OK *-* 22:01 < bls> gdm just decides what environment to launch into 22:01 < Evidlo> is "auto" a valid fs type in fstab? 22:01 < twainwek> i've been hearing about wayland for over 7 years and have yet to see it first hand 22:02 < bls> leftyfb: there may be some constraints/expectations on what and how the command consumes and produces 22:02 < phogg> Angelica: from the Arch wiki: "GDM uses the Wayland backend by default which conflicts with NVIDIA driver. Turning off the Wayland backend could enable proprietary NVIDIA driver. " Sound like your scenario? 22:03 < bls> so you don't even have gdm/wayland running? 22:03 < phogg> twainwek: you can get it on almost any distro and run it within a window in X. Not the full experience but you can get the idea. My problem is that it's "Just" a low level display mechanism. 22:03 < phogg> Evidlo: yes 22:03 < jim> twainwek, I tried it once, looks just like X 22:04 < leftyfb> bls: The same thing happens if I stick the function in a script and call that instead. Looks like the bash -c wasn't causing the issue 22:05 < phogg> Evidlo: when will attempt to detect the correct filesystem type. This could fail, though. "if the auto type is specified, mount will try to guess the desired type. Mount uses the blkid library for guessing the filesystem type; if that does not turn up anything that looks familiar, mount will try to read the file /etc/filesystems, or, if that does not exist, /proc/filesystems." 22:05 < Angelica> phogg: Yes, exactly :) 22:05 < Angelica> phogg: Oh, actually, I misread. No, it's not that. 22:06 < Angelica> I'm not currently using gdm. I need to wait for it to emerge, which will take ages, so I'm asking if there's a better solution than installing all this junk. 22:06 < phogg> Angelica: then I don't understand what your problem actually is 22:06 < phogg> what does not work? 22:07 < jim> Salsha, do you remember which debian mirror you picked? also, did you try just installing the base from the image (by doing "leave network unconfigured")? 22:07 < bls> also, did you verify the checksum of the ISO 22:07 < Angelica> phogg: Performance is slow. The fix I described solves it. I've just heard on #gentoo that using sddm could be another solution, I'll try that now. 22:08 < jim> yeah that's a good question too 22:08 < Salsha> bls: mmmm i have no idea of how to verify it 22:09 < bls> how do you get from slow performance to wayland might be running somewhere and the only way to turn it off is via a DM? 22:09 < Salsha> jim: i tried all the ones of my state 22:09 < Salsha> Angelica: ariosto? 22:10 < jim> Salsha, hmm... nothing really jumps out at me... it sounds like you did everything right... 22:11 < jim> Salsha, do you have a different machine you could try it on? 22:13 < Salsha> jim: i think i'm gonna try in a virtual machine 22:13 < jim> Salsha, that sounds like a great idea 22:13 < jim> I know I've tried it and it worked in a qemu/kvm setup 22:15 < Salsha> nothing i still get the same problem. md5sum and sha are different 22:15 < Salsha> "weak" 22:16 < Salsha> let's try on a virtual machine 22:16 < evanesoteric> Fascinating (the IBM stuff). Thanks for the info everyone! 22:17 < evanesoteric> What are you trying to achieve Salsha? Verify a sha checksum, and create a bootable ISO? 22:17 < Salsha> evanesoteric: i'm trying to understand why i keep getting hashes missmatch wehn trying to use apt update 22:18 < Salsha> i noticed i get it even using a live usb. it must be definitely some kind aof hardware problem.... but what could it be? 22:18 < bls> did you run a RAM test? 22:18 < Dagmar> Bad RAM or broken CPU 22:18 < bls> in addition to your CPU test 22:19 < Salsha> bls: i did a long time ago and it was fine. before changing the cpu it used debian though and never had these prbolems. But cpu did not give any error 22:20 < bls> ah, so a socketed CPU. tried to reseat it? checked the thermals for a mislaid headsink? 22:20 < jim> evanesoteric, the debian netinstall images are already bootable, he just has to write it to a usb, or for the virtual machine, there was a way to make a "pretend" usb or "pretend" cdrom, and use the image on that 22:21 < koala_man> if it's a cpu issue then surely something simple like "seq 1 100 | md5sum" would tend to trip it 22:21 < koala_man> which is d632eba71107bf7bc3ec423eab256d78 on my machine 22:22 < Salsha> so probably a cpu problem 22:22 * bls sees the lights flicker as everyone wants to make sure their cpu is ok too 22:23 < twainwek> uh oh 22:23 < twainwek> oh no it's fine 22:25 < Salsha> i cannot run i686 images on virtual box in a 64 bit system for real? 22:25 < Dagmar> Youc an 22:26 < Salsha> Dagmar: it says the kernel does not support my cpu 22:26 < Dagmar> Which CPU are you talking about 22:26 < jim> Salsha, you're better off trying an amd64 image 22:27 < Salsha> jim: but the amd64 has evidently no problems, i used it two days ago 22:27 < Salsha> oh ops sorry i forgot to enable pae 22:27 < jim> Salsha, what kind of thing is your machine? laptop? desktop? something else? 22:27 < syb0rg> Hi guys. I have a machine behind a firewall that only allows network traffic over a vpn. However, I would like to send traffic on ports 80 and 443 out over the actual network interface rather than the vpn. What is the easiest approach to make this happen? 22:27 < Dagmar> Adding a static route for the host in question 22:28 < Dagmar> A host route is always going to wind up above the default route in priority 22:28 < Salsha> jim: desktop. prescott based (yeah i like grilling some good steaks while programming) 22:28 < Dagmar> Otherwise, you're looking at a very complex netfilter rule because you're doing something very ill-advised 22:28 < bls> syb0rg: read up on how to configure split routing 22:28 < evanesoteric> Sounds like you want to use IP tables to forward traffic syb0rg 22:29 < triceratux> Salsha: the general trend of the entire linux ecosystem has been towards 64bit systems. your experience may be demonstrating that theres already little left 22:29 < Salsha> 3 gb ddr400, radeon 9600 256 mb, 160 maxtor ata hdd and 160 wd blue sata hdd 22:29 < syb0rg> ok, thanks all. Dagmar, why is this ill-advised? 22:29 < Salsha> triceratux: i know 22:29 < Dagmar> For one thing, you're likely to be making connection decisions based on the view of DNS available from the other end of the VPN 22:29 < Smithe> How can I get a list of all the devices connected to my rasp access point? 22:30 < bls> syb0rg: bypassing a VPN is considered a bad practice as there's a good chance of a mistake leaking sensitive data 22:30 < Dagmar> For the other, you're pretty much planning on _causing_ what's generally referred to as VPN leaks 22:30 < Smithe> Obviously from the rasp and not in active ways (ie ping nmap) 22:30 < Salsha> triceratux: but i just need a second computer for te second desktop in my office. it's the last 32 bit system left in there... and btw back inb 2004 it was a gaming pc i don't wanna throw it in trash :P 22:30 < jim> Salsha, before, you said you were using i386 images for the prescott... what's the actual bit width of the machine? 22:30 < MrElendig> depends on the software stack you are running on ti 22:30 < syb0rg> Dagmar, I will doublecheck my config but I am pretty sure that I am already doing DNS over the physical interface 22:30 < Dagmar> Your VPN isn't set up right then 22:31 < syb0rg> I don't really care about VPN leaks so long as the specific traffic I care about is routed over the vpn 22:31 < Dagmar> Smithe: Just ask hostapd to tell you 22:31 < Salsha> in virtualbox runs perfectly 22:31 < Salsha> damn 22:31 < Salsha> it's the hardware 22:31 < Salsha> jim: 32 22:31 < bls> time to take it out behind the shed 22:31 < Smithe> Dagmar, how can I ask to hostapd? 22:32 < jim> yeah, an amd64 image isn't gonna work 22:32 < Dagmar> Smothe: hostapd_cli all_sta IIRC 22:32 < Smithe> Failed to connect to hostapd - wpa_ctrl_open: Permission denied 22:32 < Smithe> uhm wat 22:33 < bls> magic 22:33 < Dagmar> This isn't the sort of thing you're likely to be allowed to run as a joe user 22:33 < Smithe> Oh so only sudo user? 22:33 < Dagmar> I can assure you that hostapd knows what's connected to it if you're running your RPi as an AP 22:33 < registeredmail> syb0rg, in that case, look into policy routing 22:33 < Dagmar> If it *didn't* it couldn't work 22:33 < syb0rg> ok registeredmail 22:34 < Smithe> Nice to know 22:34 < Lope> where can I lookup linux kernel compile config parameter definitions? 22:34 < bls> Lope: most distros save them alongside the kernel in /boot 22:34 < Dagmar> Eh, like /proc/config.gz? 22:34 < bls> Lope: that or there's a file in /proc if the kernel was compiled with it 22:35 < registeredmail> Salsha, it might be the that the usb stick you're using is bad, I've had that happen before 22:35 < Lope> bls, thanks, but this is on a samsung galaxy android phone 22:35 < MrElendig> (often built as a module) 22:35 < bls> Lope: ah, then you want to be asking #android 22:36 < Dagmar> No, #android-root 22:36 < Lope> https://github.com/LineageOS/android_kernel_samsung_smdk4412/blob/b7ffe7f2aea2391737cdeac2a33217ee0ea4f2ba/arch/arm/configs/lineageos_i9305_defconfig#L1403 Does this look like USB ethernet adapters will not be supported? 22:36 < phogg> Dagmar: is that for people who have rooted their phones? 22:36 < Lope> On my Samsung Galaxy S5, USB ethernet devices work. On my S3, they don't. 22:36 < Dagmar> ...and the simple answer to "Can I get the kernel config used in a stock Samsung firmware?" the answer is no 22:36 < Lope> The above config file might be for my s3, not 100% sure. 22:37 < Lope> Dagmar, I don't want the kernel config for stock samsung firmware. 22:37 < Lope> Dagmar, I'm running cyanogenmod and will upgrade to lineageOS 22:37 < Dagmar> That only applies to a very specific model of S3 22:37 < phogg> Lope: what part of CONFIG_USB_USBNET=y makes you think they're not supported? 22:38 < Lope> Dagmar, My S3 is the i9305 (the specific model that has LTE) 22:38 < Lope> phogg, what about "# CONFIG_USB_IPHETH is not set" 22:38 < Lope> phogg, maybe some of those unset params represent popular USB ethernet adapters? 22:39 < phogg> Lope: https://cateee.net/lkddb/web-lkddb/USB_IPHETH.html 22:40 < Dagmar> I can pretty muich guarantee you that one isn't ht eone 22:40 < Dagmar> ...unless you're somehow running Android on an iPhone 22:40 < phogg> indeed 22:42 < Dagmar> I think you need Android 6 as a baseline requirement or you wind up having to adjust the routing yourself 22:43 < phogg> I don't know from Android, all I know is kernel config. 22:43 < Dagmar> It's not a thing that's often done in Android because wifi 22:44 < phogg> Hmm. Since the last time I printed something I seem to have inadvertently broken my cups setup. What fun. 22:44 * phogg thinks 22:45 < phogg> It's been over two years, pretty sure. 22:45 < Dagmar> THe printer is out of ink 22:45 < phogg> no, the printer is maintained by someone else who uses it often 22:45 < Dagmar> If it's been sitting around for two years, count on it 22:45 < Dagmar> Ah 22:45 < Dagmar> So they used up all the ink then. ;) 22:46 < phogg> A good guess, but I'm more concerned with the fact that no printers appear in lowriter's print dialog. 22:48 < Lope> Dagmar, thanks! I'll try running Android 6 on the S3 and see if that helps! 22:49 < Lope> BTW I just checked out my 2 USB ethernet adapters to see what drivers they require. 23:06 < Smithe> Dagmar, for devices connected with eth* interfaces? 23:13 < Lope> Should I rather start buying USB-C peripherals, and get USB-C to USB-3 adapters for backwards compatibility with my old USB2/3.0 stuff? 23:13 < Lope> Is USB 3.0 going to disappear? 23:14 < nik> yes Lope 23:14 < nik> eventually 23:14 < nik> but there might be adapters i guess 23:17 < Pentode> heh. weird, the old fixed bitmap font appears _huge_ in fedora at the same point size compared to any other distro ive ever used. 23:18 < azaki> Lope: uhm, USB-C is just a different connector, it's still USB 2/3 signals going through the cable. 23:19 < azaki> The older connectors were called "Type A" (computer side) and "Type B" (for the device side) respectively. 23:19 < Lope> azaki, I see, but my question still stands, will USB-C replace A and B etc? 23:20 < azaki> Type A will probably last quite awhile I think, it doesn't seem anyone is in a hurry to replace it on desktops for instance. But on laptops it probably will happen relatively fast, especially on ultrabooks. 23:21 < azaki> Type B is already starting to disappear pretty quick, especially for USB 3.x devices, since the USB 3 'B' connector was pretty terrible. 23:21 < azaki> a lot of them basically now come with a type A on one end, and a type C on the other end instead of a B. 23:22 < Pentode> type b is falling out of popularity but i think type a is far worse.. 23:23 < azaki> yeah but type A is way more entrenched. also it's not that bad. there's kind of a trick to figuring out how to plug it in the right way 99% of the time 23:23 < Pentode> yeah, type b is just too clunky i think 23:24 < Lope> azaki, yeah, seems like a good thing 23:24 < azaki> basically, the pins inside a type A usually always face the PCB of the device (or the bottom, depending on the design of the product) 23:25 < cannabis_sativa> Pentode: I've noticed font size differences on the same distro changing from X11 to Wayland. In my case Wayland renders the fonts more readable in the few apps I notice that in. 23:25 < Lope> full size type B fits nicely, but micro-B takes a bit of wiggling 23:25 < Lope> But type C's omnidirectional nature is really cool! 23:25 < azaki> yeah, it's nice. 23:26 < azaki> also has support for alt modes and power delivery, which is cool. 23:26 < Pentode> nothing worse than realizing you have to flip the damn connector over to get it to fit 23:27 < afidegnum> hello, i m having an issue, my server is lock due to a malicious code being sent to an external server i don't know where this is coming from http://dpaste.com/1XV4J1P 23:27 < afidegnum> how do trace this out? 23:27 < afidegnum> and block all ports? 23:28 < azaki> Pentode: if it's a desktop, the usb connector is always oriented so the pins inside are 'facing' the motherboard. 23:28 < azaki> and on laptop, they'd be facing towards the bottom of the laptop 23:29 < cannabis_sativa> afidegnum: block all incoming, allow outgoing. At least until you've found the issue. 23:29 < Lope> azaki, USB-C's 100W power limit is a relief! 23:29 < darsie> Why won't 'scp -v compile s:' copy compile? It doesn't print an error. https://pastebin.com/Y02BUwkE 23:29 < Lope> finally, we can run a 3.5" Hard drive from a USB port. 23:30 < azaki> one thing i noticed is that the usb-c motherboard header that some companies have come up with doesn't give you two ports, only one. 23:30 < Lope> Pentode, first world problems. 23:30 < azaki> so they'd have to put two of those headers now 23:30 < azaki> for two ports 23:30 < azaki> all of the old usb headers gave you two ports for the front panel 23:31 < azaki> Lope: USB is so widespread that it's probably all over the third world too. =p 23:31 < Lope> azaki, yeah, except if you're holding a badly designed flash drive or cable, or in low-light scenarios, etc, directionality can be an issue. 23:31 < Lope> azaki, yeah, i realized. It was just a retort to "there's nothing worse" 23:31 < afidegnum> cannabis_sativa: i noticed there is a malicious code being drop to the server frequently at least once a day 23:32 < afidegnum> i had to remove it, but i want to trace from how it gets to the server 23:32 < azaki> Lope: i know, just being silly. =p 23:32 < azaki> in low light i'll often feel around the plug with a fingernail or something, to try and orient it right 23:34 < Dagmar> pfft. 23:34 < Dagmar> The things are n-dimensional 23:34 < Dagmar> Just flip them three times and they'll always plug right in 23:34 < oiaohm> azaki: usb C allows unique volate per port and lot more voltage and both directions of power flow. Makes things a lot more complex. 23:39 < Lope> azaki, up to 20v and 5A. I'm very impressed! 23:39 < Lope> night night 23:40 < darsie> I'll try in #debian ... 23:48 < afidegnum> i want to block all outgoing ports except to some specificed, is this correct ? ? http://dpaste.com/21TPPR4 23:54 < notmike> Oiaohmmm 23:54 < kraftb> Homer: I-phone or android ... flanders: i-phone or android ... vote! 23:55 < Psi-Jack> kraftb: Off-topic. 23:55 < kraftb> ^OT --- Log closed Wed Jul 04 00:00:16 2018