Xenguy | There must be a protocol for troubleshooting, or decommissioning, mirrors | 02:30 |
---|---|---|
Xenguy | One would want to be able to yank a mirror if it is broken, for example | 02:31 |
golinux | There is and 99% of the time issues are user error. | 02:37 |
Xenguy | Good ol' PEBCAK | 02:49 |
EHeM | There seem to be a number of documentation failures too (out of date, not updated). | 02:51 |
Xenguy | In my experience that is Situation Normal for must documentation period | 02:53 |
Xenguy | *most | 02:53 |
golinux | EHeM: Fell free to correct/rewrite documentation errors or at a minimum report them | 04:08 |
golinux | Fell > feel | 04:08 |
bjb | How can I lock the screen and start a new login session? I used to use gdmflexiserver, but I can't seem to find it any more. | 05:26 |
ServiceRobot | hello peeps | 05:41 |
saptech | bjb, are you using gdm? I use lxdm or lightdm can switch screen sessions | 05:45 |
saptech | The Switch User feature | 05:46 |
bjb | saptech, xfce4 | 05:48 |
bjb | I will try those, thanks | 05:48 |
saptech | yw | 05:54 |
Unit193 | `dm-tool switch-to-greeter` for lightdm, but that won't lock the session. | 05:56 |
dundee | Where do I get the Beowulf .iso from? | 07:16 |
dundee | Do I install a package or run a command to migrate an ASCII installation to it? Is it just a matter of changing repos? | 07:18 |
Chanku | I don't think there is an .iso yet, although Beowulf is relatively unstable at this time. | 07:19 |
plasma41 | dundee: There isn't a Beowulf ISO. | 07:20 |
Chanku | If you wish to migrate from ASCII to it though, change stable to testing in the sources.list (or change ascii to beowulf although not sure if that would work). | 07:20 |
Chanku | Please note that things *WILL* probably break and be broken. | 07:20 |
dundee | How broken are we talking here? Individual package breaking or apt breaking? | 07:20 |
plasma41 | I am running an ASCII system, but I have selectively installed specific packages from Beowulf as needed. | 07:22 |
dundee | plasma41: I'll do that then. How do you selectively install packages from Beowulf? | 07:24 |
plasma41 | dundee: Add beowulf to your sources.list file temporarily, then apt-get update and install the package(s) you need, then remove or comment out beowulf in your sources.list and apt-get update again | 07:26 |
plasma41 | dundee: What package(s) do you want from Beowulf? | 07:28 |
dundee | plasma41: Updated versions of Emacs. My understanding is that ASCII is like Debian Stable, and so I wouldn't have versions of Emacs with recent features like multithreading. | 07:37 |
plasma41 | Which version of Emacs introduced that? | 07:43 |
gnarface | ascii has emacs24 and emacs25. you guys should try ascii-backports before resorting to beowulf though | 07:46 |
gnarface | last i heard, beowulf was pretty broken with regards to desktops | 07:47 |
gnarface | it might be fine for headless servers though | 07:48 |
plasma41 | dundee: AFAICT, the version of emacs you're describing is version 26 which is not yet even in Debian Sid yet. | 07:57 |
plasma41 | dundee: If you would like to use the latest version of emacs, I would recommend you either compile it from source or install the Guix package manager and install emacs from Guix. | 08:03 |
golinux | dundee: In Devuan ALWAYS use the <releaasename> in sources.list not stable, testing etc. | 08:14 |
* man_in_shack waves | 08:20 | |
plasma41 | o/ | 08:20 |
man_in_shack | persistent-net udev rules not working for me | 08:24 |
man_in_shack | something is trying to bring the network up before udev renames it :| | 08:30 |
man_in_shack | i'm getting "IP-Config" lines about my 2 network cards in boot before udev | 08:31 |
gnarface | is that on ascii? | 08:31 |
man_in_shack | ascii/ceres, ye | 08:32 |
man_in_shack | i want one card dhcp and the other left unconfigured. these "IP-Config" thingies both mention dhcp | 08:37 |
gnarface | hmmm, i just saw something going strange on an ascii box here | 08:39 |
gnarface | i suspect it might be related but i was about to attempt to solve it by uninstalling the dhcp client again | 08:39 |
gnarface | (not sure what pulled it back in anyway, wasn't using it) | 08:39 |
gnarface | there should be a config file somewhere that you set explicitly which ethernet devices you want dhcp activated for. i do remember that it will try them all if you don't change it, but i don't remember what the file was. something in /etc though i think... | 08:40 |
gnarface | maybe something to do with wpasupplicant | 08:40 |
man_in_shack | well this is all wired ethernet here | 08:41 |
gnarface | oh, then it wouldn't be wpasupplicant | 08:41 |
gnarface | actually, what i recall now i think is that it can be wpasupplicant too | 08:42 |
gnarface | if you're using it | 08:42 |
gnarface | but since you're not... i guess the first place i'd look is in /usr/share/doc/[package name] for your dhcp client | 08:42 |
man_in_shack | gonna have to murder wicd then :P | 08:43 |
man_in_shack | not that i have a wifi connection atm | 08:43 |
man_in_shack | will give it a try | 08:44 |
gnarface | oh | 08:44 |
gnarface | huh | 08:44 |
gnarface | wicd might have settings for it honestly | 08:44 |
gnarface | i'm not using that either | 08:44 |
gnarface | i don't use much of the default stuff really, i sorta had my own setup from prior to devuan | 08:45 |
man_in_shack | this is well before wicd starts though | 08:45 |
man_in_shack | hm | 08:55 |
man_in_shack | have tried redownloading initramfs-tools | 08:55 |
man_in_shack | there's some nfs stuff in there | 08:55 |
man_in_shack | wheeee i broke it :D | 08:56 |
man_in_shack | ok | 08:57 |
man_in_shack | so it's either an initrd thing or a kernel thing :| | 08:58 |
gnarface | do you have a custom kernel/ | 09:07 |
gnarface | ? | 09:07 |
man_in_shack | yah (: | 09:08 |
gnarface | hmmm, i recall that there is a in-kernel dhcp option possible now. perhaps it's activated accidentally | 09:10 |
gnarface | ? | 09:10 |
gnarface | you might be able to disable it with a kernel command-line option | 09:10 |
man_in_shack | could be | 09:14 |
man_in_shack | most of my problems stem from coming from debian unstable (: | 09:20 |
man_in_shack | ok i don't think it's kernel. CONFIG_IP_PNP_DHCP=n | 09:30 |
man_in_shack | ok | 09:34 |
man_in_shack | found networking stuff in initramfs-tools | 09:34 |
man_in_shack | for nbd and aoe | 09:34 |
man_in_shack | whheee | 09:41 |
man_in_shack | "ipconfig: no devices to configure" | 09:41 |
man_in_shack | nbd was the problem | 09:41 |
gnarface | ah | 09:43 |
gnarface | yea that would make sense | 09:43 |
man_in_shack | gah | 09:45 |
gnarface | can't you just like, not try to mount any network block devices until after the network is up? | 09:46 |
man_in_shack | well that would be the smart thing to do, wouldn't it, whoever it was who decided to put nbd in initrd :P | 09:46 |
gnarface | hmmm | 09:46 |
gnarface | i assumed it would be dormant unless something it needs is on the network | 09:47 |
gnarface | but i guess i don't know | 09:47 |
man_in_shack | very close to target now | 09:51 |
KatolaZ | EHeM: can you please articulate on your "breakage"? | 10:18 |
KatolaZ | I really would like to pin it down, but I can't reproduce your problem | 10:19 |
KatolaZ | EHeM: shouting and swearing won't have your problem solved :) | 10:20 |
KatolaZ | we have no other report of a similar breakage | 10:20 |
KatolaZ | and your is not the only installation using deb.devuan.org | 10:20 |
KatolaZ | the mirror in question serves all the packages and release files | 10:20 |
KatolaZ | for jessie, ascii, beowulf, and unstable | 10:20 |
KatolaZ | (including InRelease) | 10:21 |
KatolaZ | if this were not the case, almost anyody here would experience the very same issue | 10:21 |
KatolaZ | everyday | 10:21 |
KatolaZ | so there must be somthing wrong somewhere | 10:21 |
KatolaZ | if you are willing to help, we can try to pin it down | 10:21 |
KatolaZ | I have asked you to provide your sources.list, which is something you have refused to do so far | 10:21 |
KatolaZ | when you are ready to collaborate, just shout | 10:22 |
EHeM | KatolaZ: Check the logs of #devuan, there is at *least* one other. | 10:35 |
EHeM | KatolaZ: Quote the refusal. (hint, things could be misconstrued as a refusal, but no actual refusal happened) | 10:37 |
KatolaZ | EHeM: please paste your sources.list | 10:38 |
KatolaZ | your full sources.list, please | 10:38 |
KatolaZ | (that's an explicit request, can't be more explicit than that, I guess) | 10:39 |
EHeM | KatolaZ: I suspect most users are experiencing the issue every day, but since `apt-get`'s output is voluminous they don't notice the error buried in there, or they get an error from a GUI, but retry and the error disappears due to the magic of round-robin DNS. | 10:40 |
KatolaZ | EHeM: I can't reproduce your problem | 10:40 |
KatolaZ | and I am administering a couple dozens of devuan machines | 10:40 |
gnarface | i can't reproduce it reliably here, but it happens occasionally. i've been assuming that mirror is updating, and i try again. i have had to be careful to pay attention so i don't miss any | 10:41 |
KatolaZ | EHeM: would you please post your sources.list? | 10:41 |
gnarface | but i haven't paid close enough attention to figure out if it's the same one failing every time | 10:41 |
EHeM | KatolaZ: Do you have a cache for APT in front? Do they have special permissions or magic IP addresses which get special treatment? | 10:41 |
KatolaZ | nope | 10:41 |
KatolaZ | no apt cache | 10:42 |
KatolaZ | no special treatment | 10:42 |
KatolaZ | the same dns round-robin | 10:42 |
KatolaZ | from all around the world | 10:42 |
KatolaZ | now, please, EHeM would you be so kind to post your sources.list? | 10:42 |
EHeM | gnarface: The error likely disappears due to your second running of `apt-get` going to a different mirror which doesn't have the problem. | 10:42 |
KatolaZ | EHeM: see the request above, please | 10:43 |
EHeM | KatolaZ: Would you prefer a single line which reliably produces the error? | 10:43 |
KatolaZ | no | 10:43 |
KatolaZ | I would prefer to see your full sources.list, if you want me to help out | 10:43 |
EHeM | KatolaZ: I'm concerned about doing multi-line cut and paste to this channel, but if that is what you want. | 10:44 |
KatolaZ | use any pastebin, please | 10:45 |
KatolaZ | or you would be banned by a bot :\ | 10:45 |
KatolaZ | curl -L -F c=@- https://ptpb.pw <$FILEIN | 10:46 |
EHeM | https://ptpb.pw/dz7U | 10:47 |
KatolaZ | just remove the / after "merge" | 10:49 |
KatolaZ | "merged" | 10:49 |
KatolaZ | it shouldn't make a different, but some redirects might actually fail | 10:49 |
KatolaZ | EHeM: ^^^ | 10:50 |
EHeM | KatolaZ: "deb http://37.220.36.58/merged ascii-updates main" reliably fails "404 Not Found". | 10:52 |
KatolaZ | EHeM: which command are you using, please? | 10:53 |
EHeM | `apt-get update`. | 10:53 |
KatolaZ | (and why are you putting the mirror IP there?) | 10:53 |
KatolaZ | that mirror is not available at the named IP | 10:53 |
KatolaZ | it is available either behind deb.devuan.org | 10:53 |
KatolaZ | or as sledjhamr.org/devuan | 10:54 |
KatolaZ | not directly at the IP | 10:54 |
KatolaZ | please see https://pkgmaster.devuan.org/ | 10:54 |
EHeM | KatolaZ: Otherwise it will connect to any mirror and not reproduce. | 10:54 |
KatolaZ | sorry | 10:54 |
KatolaZ | https://pkgmaster.devuan.org/mirror_list.txt | 10:54 |
KatolaZ | then you should use http://sledjhamr.org/devuan/merged | 10:54 |
KatolaZ | not the IP | 10:54 |
KatolaZ | -_- | 10:54 |
EHeM | That I can do. | 10:54 |
EHeM | KatolaZ: Okay, that worked (with the '/' too). | 10:56 |
KatolaZ | ok | 10:57 |
KatolaZ | then use codenames | 10:57 |
KatolaZ | and everything should be working | 10:57 |
EHeM | KatolaZ: I'll have to wait a while before it can be called worked around, since it might work with that URL, but fail with the appropriate round-robin one (https://us.deb.devuan.org/merged/). | 10:57 |
KatolaZ | try the round-robin | 10:57 |
KatolaZ | we are chacking the round-robin *continuously* | 10:57 |
KatolaZ | and *automatically* | 10:57 |
KatolaZ | any fault is reported within 2 minutes | 10:58 |
KatolaZ | and acted upon, if needed | 10:58 |
KatolaZ | the fault you exprienced with the round-robin is most probably due to the usage of "stable" instead of "ascii" | 10:58 |
EHeM | KatolaZ: Yet this wasn't detected despite being present for 2+ months. | 10:58 |
man_in_shack | ok | 10:58 |
KatolaZ | what wasn't detected? | 10:58 |
man_in_shack | looks like aoe won't do what i want | 10:58 |
man_in_shack | and nbd is evil | 10:58 |
EHeM | KatolaZ: Tries that with "stable", no difference. | 10:58 |
KatolaZ | the fact that you used "stable" instead" of "ascii"? | 10:58 |
KatolaZ | o_O | 10:59 |
EHeM | ^Tries^Tried | 10:59 |
KatolaZ | EHeM: please test it | 10:59 |
KatolaZ | and report any misfunctioning | 10:59 |
EHeM | KatolaZ: Right now it *appears* functioning. | 11:00 |
KatolaZ | great | 11:00 |
EHeM | KatolaZ: I'll give good odds it will fail next time that particular mirror is hit via round-robin. | 11:01 |
Ulrar | Hi, I'm trying to use your version of debootstrap but whatever I do it seems to select systemd anyway, even with --exclude systemd,libpamsystemd,systemd-sysv,libsystemd0 | 11:24 |
Ulrar | (And I can't manage to get it to work with fakechroot, but that's a separate issue) | 11:24 |
Ulrar | Isn't the whole point of maintaining your own version to avoid that forced dependency on systemd ? | 11:25 |
Ulrar | am I missing something ? | 11:25 |
EHeM | KatolaZ: I hope the test is done using `curl` for the normal URL, but using the "--resolve" option to force it to load from a particular mirror, otherwise the test isn't covering much. | 11:28 |
gnarface | Ulrar: systemd shouldn't be in there, but libsystemd is still inextricable. it's probably harmless though. | 11:32 |
Ulrar | I do see the actual systemd package | 11:33 |
Ulrar | (in the chroot error, but still) | 11:33 |
gnarface | you're using ascii? | 11:33 |
Ulrar | "stable" as indicated in the readme for your deboostrap | 11:34 |
Ulrar | should I replace that by ascii ? | 11:34 |
gnarface | yes but it should be the same thing | 11:34 |
Ulrar | Could it be pulled by something, maybe linux-image-amd64 ? | 11:35 |
gnarface | anything's possible, but i doubt it | 11:36 |
gnarface | what mirror are you using? | 11:36 |
Ulrar | In any case I do have that in the list : /var/cache/apt/archives/systemd_232-25+deb9u4_amd64.deb | 11:37 |
Ulrar | Not sure | 11:37 |
Ulrar | ./debootstrap --variant=fakechroot --components=main,contrib,non-free --arch amd64 stable /some/path | 11:37 |
gnarface | yea uh, i have no idea why that works at all | 11:37 |
gnarface | point it at a devuan mirror | 11:37 |
Ulrar | :/ I just copied the readme and added the variant | 11:38 |
Ulrar | Let's see | 11:38 |
gnarface | the readme didn't include a hostname in the command? really? | 11:38 |
gnarface | lots of the documentation hasn't been updated since debian, i don't know if this is any exception to that | 11:39 |
Ulrar | https://git.devuan.org/devuan-packages/debootstrap | 11:39 |
Ulrar | It does indeed reference debian | 11:39 |
Ulrar | Hadn't realised | 11:39 |
gnarface | it's worked for me but i've never read that readme and i always specified the mirror url on the command line | 11:40 |
Ulrar | ./debootstrap --variant=fakechroot --components=main,contrib,non-free --arch amd64 ascii /some/path http://deb.devuan.org/merged | 11:43 |
Ulrar | does that look correct ? | 11:43 |
Ulrar | I've never used deboostrap before, nor devuan | 11:44 |
gnarface | oh | 11:44 |
gnarface | i think so, lemme check my notes... | 11:44 |
Ulrar | but I figured it might be better this way than removing systemd from debian like we usually do | 11:44 |
gnarface | http://paste.debian.net/1046656/ | 11:47 |
gnarface | ^ that's the manpage from ceres, it's been updated for devuan | 11:48 |
gnarface | your commands looks right, i think | 11:49 |
gnarface | but i think you'll need fakechroot to actually be installed though too | 11:49 |
gnarface | i'm not sure if it uses fakeroot for that or what | 11:49 |
Ulrar | Ah great, thanks | 11:50 |
Ulrar | Yeah I have the fakechroot stuff | 11:50 |
Ulrar | it doesn't work, but it's there | 11:50 |
gnarface | it might complain about the keyring | 11:50 |
Ulrar | but maybe it's something to do with that repo, maybe now it'll work | 11:50 |
Ulrar | ah, good point | 11:50 |
gnarface | i've never used the "--components" | 11:51 |
gnarface | i add contrib and non-free later, and usually just temporarily | 11:51 |
gnarface | but i guess that might not make sense for all hardware | 11:51 |
Ulrar | Oh, I just noticed | 11:51 |
Ulrar | the repo URL in the readme for deboostrap is wrong too | 11:51 |
Ulrar | I'm using the debian deboostrap .. | 11:51 |
Ulrar | that won't help | 11:52 |
furrymcgee | you can debootstrap a netinstaller iso image | 11:56 |
Ulrar | Well I want a base system I can install stuff in with chroot. Then I'll turn that into an iso to make diskless web servers | 11:58 |
gnarface | you can use --include to add extra packages | 11:59 |
Ulrar | Yeah but we have our own repo, and as I understand it deboostrap can't use two repos | 12:02 |
Ulrar | so I'll just use chroot /blah /usr/bin/apt-get install .. after | 12:02 |
* man_in_shack flails | 12:04 | |
man_in_shack | nbd seems to work better than aoe but it screws with my network settings | 12:05 |
* man_in_shack flails | 12:45 | |
man_in_shack | ok | 12:55 |
man_in_shack | so current theory is neither aoe nor nbd pass ioctls. they just use open() to access the exported device | 12:56 |
man_in_shack | which works fine for regular devices, but not encrypted opticals | 12:56 |
man_in_shack | so ... i guess i can try iscsi next? | 12:58 |
gnarface | i still feel like you must be overlooking an easy way to mount after boot | 13:09 |
gnarface | i forget if there's another alternative to iscsi or not | 13:09 |
man_in_shack | mounting is the least of my problems (: | 13:14 |
man_in_shack | if it was just a matter of mounting, i could use nfs | 13:19 |
furrymcgee | whats wrong with nbd and network settings? | 13:24 |
man_in_shack | something about the way it works in initrd initialises network devices before udev gets a chance to rename them | 13:25 |
man_in_shack | iscsi looks like what i want. istgt package actually has examples for dvdrom | 13:25 |
HumanG33k | KatolaZ, Do you know where i can found a good tutorial doc for make a devuan repository (first a private mirror and after a public one) if it can mirroring throught tor it’s perfect :) | 14:54 |
KatolaZ | HumanG33k: is that for personal use? | 14:54 |
HumanG33k | first yes | 14:54 |
KatolaZ | then you should use apt-mirror or debmirror | 14:55 |
HumanG33k | and after a public one | 14:55 |
KatolaZ | but you still need to set the needed rewrites | 14:55 |
KatolaZ | as explained at: | 14:55 |
KatolaZ | https://pkgmaster.devuan.org/devuan_mirror_walkthrough.txt <- HumanG33k | 14:55 |
HumanG33k | thx | 14:55 |
KatolaZ | np | 14:56 |
* man_in_shack flails | 14:59 | |
KatolaZ | HumanG33k: on dev1galaxy there must be a (short) tutorial I wrote about using apt-mirror and/or debmirror | 15:00 |
KatolaZ | but I don't remember where :\ | 15:00 |
man_in_shack | iscsi is connecting but getting i/o error on read | 15:00 |
rrq | KatolaZ: "dh1" is searchable :) | 15:01 |
HumanG33k | rrq, KatolaZ i will search | 15:02 |
rrq | hmm hould be d1h | 15:02 |
rrq | s | 15:02 |
* rrq is lyxdexlic tonight | 15:03 | |
man_in_shack | goddammit | 15:05 |
man_in_shack | read works on usb disk but not on dvd | 15:05 |
drisk | What's the difference between linux-image-4.9.0-6-amd64 and linux-image-amd64? | 15:34 |
telst4r | the other is simply just a metapackage pointing to the other. | 15:36 |
drisk | Thanks | 15:36 |
drisk | So none at all then. Curious. | 15:36 |
telst4r | 4.9... is a real package, the linux-image-amd64 just points to that (or some other version, if it's updated in future). | 15:37 |
telst4r | "real" as in 'complete', or something that actually includes a kernel. | 15:38 |
man_in_shack | ok now | 15:38 |
* man_in_shack flails wildly | 15:39 | |
telst4r | if you install the one without kernel number, it should forcibly install a versioned one | 15:39 |
telst4r | what? | 15:39 |
Ulrar | So that's strange, it works but as soon as I try to have ca-certificates installed in the chroot it hangs (on a grep in the post-inst for some reason) | 15:40 |
Ulrar | read the script, I don't understand what can hang there | 15:40 |
debdog | if you want a specific version, install that with the version number. if you want your system to automagically install the latest version use the meta package | 15:40 |
man_in_shack | i am out of ideas to get my dvd drives accessible over a network connection :( | 15:43 |
telst4r | My Jessie hangs on 3.16.0-6-amd64 but not -4-amd64, so it might be a worthy idea to still maintain an older, known-to-work version in handy incase the metapackage upgrades. | 15:44 |
man_in_shack | aoe and nbd would probably work if i wasn't trying to watch encrypted dvds | 15:44 |
man_in_shack | i haven't had luck with iscsi | 15:44 |
man_in_shack | oh hey, there's an in-kernel iscsi i didn't see before | 15:55 |
man_in_shack | kernel iscsi is for TYPE_DISK only :( | 16:08 |
man_in_shack | ooh | 16:10 |
man_in_shack | or maybe not | 16:10 |
golinux | HumanG33k: https://dev1galaxy.org/viewtopic.php?id=1933 | 16:13 |
HumanG33k | oh thx golinux | 16:14 |
golinux | I think that's what you want. | 16:14 |
* golinux is just catching up this morning | 16:15 | |
_stephen_ | what's the name of an off topic chat channel where devuan users congregate? | 16:40 |
debdog | _stephen_: #debianfork | 16:40 |
fsmithred | man_in_shack, pmount /dev/sr0 on one computer, then sshfs user@remote:/media/sr0 mountpoint | 16:44 |
fsmithred | on the other | 16:45 |
man_in_shack | fsmithred: doesn't work for encrypted dvds | 17:42 |
man_in_shack | libdvdcss needs to access the device node | 17:44 |
fsmithred | man_in_shack, sorry I never tried with encrypted dvd, but pmount does work with encrypted drives. | 18:38 |
fsmithred | oh, you want to watch movies from a remote machine? | 18:39 |
nemo | sooo ever since switch from ubuntu 14.04 to devuan ascii, I've had to remember to shut down my laptop entirely since it seems to fail to suspend | 19:38 |
nemo | even in 14.04 it was buggy, and about one time in, oh, twenty or so, I'd have to explicitly shut it down.. | 19:38 |
nemo | anyway. my inclination for something like this is to blame kernel or its params. I guess I should start by comparing an ubuntu 14.04 kernel config and boot params to a debian one | 19:39 |
nemo | unless anyone here more familiar with linux on laptops than I might know a simple obvious thing to try | 19:40 |
system16 | try shutdown -h (insert time here) | 19:41 |
nemo | system16: oooh yes | 19:42 |
nemo | system16: screw correct detection of lid | 19:42 |
nemo | so long as I can make it happen who cares | 19:42 |
system16 | :) | 19:42 |
nemo | well. assuming I can unsuspend | 19:42 |
system16 | i tried suspending my sftp server | 19:43 |
system16 | it didnt go well | 19:43 |
nemo | hah | 19:43 |
system16 | stick to shutdown -h | 19:43 |
system16 | unless u have a GUI | 19:43 |
nemo | oh. durn | 19:43 |
nemo | system16: I had to read the manpage for shutdown - I was hoping you were suggesting a suspend state, but it seems no | 19:44 |
nemo | system16: and yeah, it's a laptop, gui is pretty much a given | 19:44 |
system16 | my server is a netbook' | 19:44 |
system16 | it could not handle a GUI | 19:44 |
system16 | BRo. just go to the powersettings and sett a time to suspend after no use | 19:45 |
system16 | like debian | 19:45 |
system16 | or windows | 19:45 |
nemo | system16: that seems unlikely. I had a laptop that was a P166 (or a P233 I forget) with 96MiB of RAM and it handled a gui just fine ☺ | 19:45 |
system16 | i meant a DE | 19:45 |
nemo | ok. GUI vs DE is very different | 19:45 |
nemo | although there are lightweight DEs | 19:46 |
nemo | anyway | 19:46 |
system16 | i guess there is a setting like that in devuan ? | 19:46 |
nemo | basically I need to suspend immediately when I'm ready to haul this off somewhere. I guess I need to find a command that does that | 19:46 |
system16 | if windows has it | 19:46 |
system16 | in windows power>>sleep | 19:46 |
system16 | and in ubuntu | 19:46 |
system16 | and every other distro that i tried out. | 19:47 |
nemo | hm? not sure what you mean by power>>sleep | 19:48 |
system16 | in ubuntu there are 4 options | 19:48 |
system16 | shutdown | 19:48 |
system16 | restart | 19:48 |
system16 | sleep | 19:48 |
system16 | and the other one i cant remember' | 19:48 |
nemo | system16: just tested pm-suspend - it worked as expecte | 19:49 |
nemo | *d | 19:49 |
nemo | also woke up correctly on opening the lid | 19:49 |
nemo | that's good enough for me, I'll just do that | 19:49 |
* system16 will try that | 19:49 | |
nemo | it'd be nice if it was correctly firing on *closing* the lid, but whatever | 19:49 |
system16 | lets hope i dont brick my server | 19:50 |
system16 | like last time | 19:50 |
nemo | frankly it was none-too-reliable with ubuntu 14.04 and the one in twenty time that it'd fail would be a nasty surprise | 19:50 |
nemo | "oh look, my laptop back is making a weird sound" | 19:50 |
nemo | *bag | 19:50 |
system16 | see ? | 19:50 |
system16 | devuan hates suspending | 19:50 |
system16 | ether on or off' | 19:51 |
system16 | lol | 19:51 |
nemo | on worked fine | 19:51 |
nemo | and. eh. if it ever fails (which has happened with this old sucky laptop before - I'd literally get the "blinking capslock key of death") | 19:51 |
nemo | restarting is always an option | 19:51 |
KatolaZ | system16: have you tried to debug what happens during suspend? | 19:52 |
* nemo makes a quick desktop icon for gksudo pm-suspend | 19:52 | |
system16 | lol i can barely navigate the linux shell | 19:52 |
system16 | i cant do a debug | 19:52 |
system16 | if a thing is more complicated than a donut , i will come to freenode for welp | 19:53 |
nemo | my main fun at present is that we are apparently supposed to have "CylanceProtect Agent" on these laptops and they don't officially support Debian, and certainly not Devuan | 19:53 |
nemo | their installer supports upstart and systemd - so I wrote a little init wrapper | 19:53 |
system16 | btw i have tiny problem with my server | 19:54 |
system16 | i cant start it up when the lid is closed | 19:54 |
nemo | then it failed because their LinuxHook.h/LinuxHook.c which fires off if they don't have a prebuilt kernel module was doing a naughty redefinition that was freaking out gcc 6 - so fixed that | 19:54 |
nemo | then they were trying to do something that's IFDEF'd off by default in kernels but that I guess ubuntu 14.04/16.04 enables, so I changed *that* | 19:54 |
system16 | which means i have to unplug my netbook. pull it out . open the lid . start the damn thing. reverse. | 19:54 |
nemo | I now have a kernel module that loads, but no idea if it's working correctly - hope the workplace lets me slide | 19:55 |
nemo | I tried contacting their support but ofc zero interest in supporting non-"standard" linux | 19:55 |
system16 | are you talking to yourself ? | 19:55 |
nemo | eh. just kind of a warning to denizens of unexpected consequences of abandoning the Ubuntu Way Ⓒ on a corporate laptop | 19:56 |
nemo | I may end up just ditching this old crappy thing and buying myself my own damn laptop | 19:57 |
nemo | oddly enough they don't require "CylanceProtect" on personal equipment | 19:57 |
system16 | nemo, can i make my server power on when the lid is closed ? | 19:57 |
nemo | system16: WoL presumably | 19:58 |
nemo | or anything else it might be listening for. some systems have a button for that - mac keyboards for example | 19:58 |
nemo | system16: do you have a keyboard plugged in? | 19:58 |
nemo | system16: https://askubuntu.com/questions/848698/wake-up-from-suspend-using-wireless-usb-keyboard-or-mouse-for-any-linux-distro this seems relevant | 19:59 |
nemo | system16: not to me ofc, since have no particular need for that use case | 19:59 |
system16 | nemo no no keyboards | 20:00 |
nemo | system16: anything whatsoever plugged into it? | 20:00 |
nemo | system16: seems like any USB gadget would work, like one of those fun big red USB buttons | 20:01 |
system16 | a kan cable and a power adapter | 20:01 |
nemo | those are even wireless | 20:01 |
system16 | lan* | 20:01 |
nemo | could wake it up by smacking a big red button | 20:01 |
nemo | pretty satisfying | 20:01 |
system16 | wake on lan ? | 20:01 |
system16 | how can i do that | 20:01 |
system16 | i found a l3 option in its BIOs | 20:01 |
nemo | system16: why do you want to suspend your sftp server anyway? | 20:02 |
nemo | I generally like to keep my servers up 24/7 - that's kinda the point of a server | 20:02 |
system16 | nemo, because of arseholes trying to brake in | 20:02 |
system16 | i tried the ssh key thing' | 20:02 |
system16 | no luck | 20:02 |
nemo | uh... hopefully you're hardened enough that they are not going to be successful | 20:02 |
nemo | regardless of how much it's active | 20:02 |
nemo | system16: also, considered autoblocking them? | 20:03 |
system16 | (and its a netbook. its not built to be on all the time) | 20:03 |
system16 | i have fail2ban | 20:03 |
system16 | but idk how to configure it | 20:03 |
fsmithred | system16, try a non-standard port. All my break-in attempts stopped when I did that. (Unless someone is out to get you specifically.) | 20:06 |
unixman | Probably would work (caveat: I've not tried it) - https://www.atlantic.net/cloud-hosting/how-to-install-fail2ban-ubuntu-debian/ | 20:06 |
system16 | fsmithred, i failed that too | 20:06 |
fsmithred | what do you mean failed? | 20:06 |
system16 | didnt work | 20:06 |
fsmithred | couldn't set it up, or they tried anyway | 20:06 |
fsmithred | ? | 20:06 |
system16 | do i have to change something on the server ? | 20:06 |
unixman | fsmithred, I still have random hits on non-standard port @home. But much, much, much less than on port 22. | 20:06 |
system16 | could not set it up | 20:07 |
fsmithred | whatever router directs traffic to the server should be able to specify that it receives on a high port and then delivers on 22, so you don't need to change the server in that case, just the router | 20:07 |
unixman | system16, change the listening port for ssh and make sure that is open in the firewall. | 20:08 |
unixman | Oh, is this behind a router? | 20:08 |
fsmithred | if the router insists on using the same port for wan/lan, then change the port in /etc/ssh/ssh_config (or sshd_config) | 20:08 |
system16 | yes | 20:08 |
system16 | oh so i need to change the port on the router AND the server ? | 20:09 |
fsmithred | depends on what the router will do | 20:09 |
unixman | Yeah, system16, if behind a router just redirect a port, like 10222, to port 22 on your ssh host. | 20:09 |
fsmithred | if the router will accept outside traffic on a high port and then deliver it to 22 on the server, then you don't need to change the server | 20:10 |
* system16 is firing up the server. plz stand by | 20:10 | |
fsmithred | some will and some won't | 20:10 |
unixman | Depends on what the router can do. Some are less configurable than others. | 20:10 |
system16 | i can use whatever port i wan on the router | 20:11 |
fsmithred | well, any one that's not in use | 20:11 |
fsmithred | pick a number above 1024 | 20:11 |
fsmithred | I usually google a port number before I use it, just to make sure | 20:12 |
system16 | OK im in | 20:13 |
system16 | what was the command ? | 20:13 |
fsmithred | for which? | 20:13 |
system16 | changing the port | 20:13 |
fsmithred | are you in the router? | 20:13 |
system16 | no | 20:13 |
system16 | in the server | 20:13 |
system16 | (ssh) | 20:13 |
fsmithred | you have to change the router | 20:14 |
fsmithred | and then depending on how much you can change in the router, you may or may not need to change it in the server | 20:14 |
fsmithred | go to the port forwarding section in the router and see what you can do | 20:14 |
Acacia | btw IANA says ports 49152-65535 are custom ports to be specified by the user, so use those if anything, and it's the least likely port range to turn up a conflict | 20:15 |
fsmithred | brb | 20:15 |
system16 | # What ports, IPs and protocols we listen for | 20:15 |
system16 | Port 22 | 20:15 |
system16 | is that it ? | 20:15 |
system16 | (i changed the port on router) | 20:16 |
fsmithred | yes, in sshd_config | 20:16 |
system16 | it was not in ssh_config | 20:16 |
fsmithred | so change it there and then restart ssh | 20:16 |
system16 | PermitRootLogin without-password | 20:17 |
system16 | is it off ? | 20:17 |
system16 | root login is bad af | 20:17 |
system16 | i thought i turned it off | 20:17 |
Acacia | ssh_config is the configuration file for the ssh client, sshd_config is the configuration file for the ssh daemon/server. Make sure you didn't use the wrong configuration file | 20:18 |
system16 | i used the sshd_config one | 20:19 |
system16 | changed the port to 2017 | 20:19 |
system16 | (because it was a good year for me) | 20:19 |
system16 | its restarting | 20:19 |
system16 | there was nothing related to ports in ssh_config afaik | 20:20 |
system16 | ssh: connect to host 192.168.0.107 port 22: Connection refused cmd is using port 22 | 20:20 |
system16 | how can i tell it to dont use port 22 ? | 20:21 |
system16 | oh gawd | 20:21 |
system16 | im gonna go back to 22 | 20:23 |
unixman | Well, 'man ssh' and 'man sftp' would tell you. But since you asked here, use 'ssh -p port# ...' or 'sftp -P port# ...' | 20:34 |
unixman | Where "port#" is just the number. Such as 'ssh -p 48222 foo@bar.baz.bat'. | 20:36 |
unixman | If you stay on port 22 then I strongly recommend getting fail2ban working. | 20:37 |
system16 | filezilla failed too | 20:38 |
system16 | rror:Cannot establish FTP connection to an SFTP server. Please select proper protocol. | 20:38 |
system16 | Error:Critical error: Could not connect to server | 20:38 |
system16 | it was proper protocol | 20:38 |
system16 | it did not accept port 2017 as a sftp port | 20:39 |
nemo | system16: what the heck client are you using? | 20:48 |
unixman | Well, trying to use FTP to connect to SFTP is going to fail. Too bad system16 is gone so I can't 'splain that. | 21:06 |
nemo | unixman: yeah I was kinda puzzled by that, which was why I was wondering what client he was using - my guess was, when he was on port 22 the client autoselected the correct protocol for him in whatever connection dialog he's using | 21:07 |
nemo | but it had no clue what to use for 2017 so left it at default and he didn't correct it | 21:07 |
nemo | buuut since he didn't answer | 21:07 |
unixman | nemo, Looks like his client was FileZilla. I use FileZilla often to connect to SFTP on altrernate ports. One just has to configure it correctly when setting up the connection. | 21:08 |
unixman | s/altrernate/alternate/ | 21:09 |
* unixman glares at his drunk fingers | 21:09 | |
nemo | unixman: I haven't used filezilla in a long while, but it was my goto SFTP client under windows once upon a time | 21:12 |
nemo | unfortunately their reputation has fallen recently | 21:12 |
unixman | nemo, fallen? You have a URL for that? | 21:12 |
nemo | unixman: you heard about the installer malware right | 21:12 |
nemo | I *think* that's windows only | 21:12 |
unixman | Ah, I don't run Microsoft systems. | 21:13 |
nemo | https://news.ycombinator.com/item?id=17381184 | 21:13 |
nemo | unixman: this is the 2nd iteration of bundled malware for them | 21:13 |
nemo | it's just scummy | 21:13 |
unixman | I use command line sftp and sftp in FileZilla on Linux, FreeBSD and OS X. | 21:13 |
nemo | if you're using distro bundled filezilla you should definitely be fine | 21:14 |
nemo | who knows what system16 is doing tho | 21:14 |
nemo | unixman: yeah. that's my usual approach, for my SO, I use a combination of automount and the MATE VFS | 21:14 |
unixman | Nice. Lately I have been playing with ssh tunnels and SFTP. It's quite cool to set up a tunnel through a jump box then SFTP to localhost to send/receive files. | 21:16 |
nemo | it's gotten easier to do since ssh supported that natively | 21:16 |
nemo | before you had to setup a netcat rule in your .config | 21:16 |
nemo | but yeah. it's pretty convenient to relay through gateway like that | 21:17 |
unixman | Yup | 21:17 |
nemo | you might also like sshuttle | 21:17 |
unixman | I'll check it out. | 21:17 |
unixman | Ah, a Python VPN thingie I see. | 21:19 |
nemo | unixman: yeeeah, although I'd say python is not the key aspect | 21:20 |
nemo | certainly it's not why I'd pick it | 21:20 |
nemo | their project page pretty much describes its advantages | 21:20 |
nemo | unixman: I find it pretty handy when on my laptop on a random remote network | 21:20 |
nemo | like, airport, starbucks whatever | 21:20 |
unixman | Well, for $COMPANY I am already provided a MacBook Pro with corporate VPN software installed. For home I just tunnel VNC over ssh. | 21:21 |
nemo | unixman: welp, depending on $COMPANY's TOS WRT ssh connections in or out, you might find sshuttle handy anyway. | 21:22 |
nemo | unixman: the use case when travelling would be to relay all or some of your traffic transparently through to your home machine/personal server host when on a nosy/untrusted public wifi | 21:23 |
nemo | unixman: it's also handy when travelling to access all of your home network's services, regardless of what they are and without special .config rules | 21:23 |
unixman | nemo, we have an official Linux based gateway that can be used and works fine for ssh + VNC. As for travel, I don't travel. I "hate" to travel. I'm a homebody type of dude. :D | 21:24 |
unixman | However, if I *have* to travel I will keep sshuttle in mind. | 21:25 |
nemo | heh. well, local coffeeshop still counts as travel | 21:26 |
nemo | if they have an official linux gateway at work, that probably implies they are blessing ssh tunnelling for trusted accounts | 21:26 |
nemo | unixman: in that case, you might find sshuttle handy for getting work done on something that's not the MBP | 21:27 |
nemo | anyway whatev. just brought it up 'cause you'd mentioned tunnelling | 21:27 |
unixman | I rarely use my laptop at coffee shops. I may play a game on my Samsung Android tablet, but that does not require connection to home/owkr. | 21:27 |
unixman | Damn! s/owkr/work/ | 21:28 |
unixman | nemo, like I said, I will keep sshuttle in mind. I may find a use for it yet. ;) | 21:28 |
unixman | nemo, reading the thread for FileZilla it seems to be a complaint about running untrusted executables on Microsoft more than anything else. AFAIK running random crapware downloaded from random sites is a "thing" on Microsoft systems. ;) | 21:39 |
nemo | unixman: random crapware downloaded from random sites BY the filezilla installer which the user trusts | 21:42 |
nemo | the only thing that protects linux users, at present is the distro | 21:44 |
nemo | but unfortunately alternate less-reviewed paths are becoming popular | 21:44 |
nemo | flatpak crap... | 21:44 |
nemo | unixman: there's also the sad state of affairs for chrome users | 21:44 |
nemo | unixman: the other thing that helps protect ofc is fact that it's not quite worth trying to compromise the smaller linux audience with crapware | 21:45 |
nemo | otherwise a filezilla flatpak could do the 'sact same thing | 21:45 |
unixman | Well, Linux kernel is used in billions of internet connected devices. It's not as small a target as some folks think. | 21:46 |
golinux | <nemo> flatpak crap... | 21:47 |
nemo | unixman: ofc. but vast majority of those are smartphones or extremely limited use stripped down servers and commodity devices | 21:47 |
* golinux agrees 100% | 21:47 | |
nemo | golinux: did you see the link trending on hacker news today? | 21:47 |
golinux | No | 21:47 |
nemo | golinux: WRT flatpak | 21:47 |
nemo | http://flatkill.org/ | 21:47 |
golinux | But I've seen signs here and elsewgere. | 21:48 |
nemo | I ranted about it a bit in #hedgewars | 21:48 |
golinux | Oh yeah, I did see that | 21:48 |
golinux | flatpak, snapd all rubbish | 21:48 |
nemo | security... that's one aspect of it - the other that bothers me is it feels like it's an end run around distros and open source ecosystem. once people accept flatpak, linux is just a kernel for executing these self-contained massive blobs - which weakens open source quite a lot. I would actually be less bothered if it was kind of a 3rd class alternative for the closed source world. but spreading it into | 21:49 |
nemo | major FOSS packages, ew... | 21:49 |
diskp | How do you install and update an individual package from beowulf instead of ascii? | 21:49 |
nemo | turning linux desktop into android/osx | 21:49 |
nemo | diskp: dpkg -i ? | 21:49 |
nemo | diskp: which package? | 21:50 |
nemo | speaking of that sort of thing... I'd like to have libreoffice 6.1 which is in ascii stable backports | 21:51 |
nemo | normally I'd force version, but libreoffice has quite a large chain of deps so doing that causes broken packages | 21:51 |
nemo | I was wondering what the most elegant way to do this would be | 21:51 |
diskp | nemo: Emacs. | 21:51 |
diskp | I was told something involving sources.list, but I only see instructions for all packages. | 21:52 |
nemo | huh... 25.1 vs 25.2 - that's a pretty small version jump - I'm surprised it matters, not that I've used emacs in, oh, 20 years | 21:53 |
diskp | nemo: Using Emacs on Arch, they're on 26.1, which has a lot of really useful features. Emacs in ASCII tries to install v24 (.5, I think). | 21:54 |
gnarface | diskp: the trick is you add backports to sources.list, just upgrade that one package, then remove it again | 21:54 |
nemo | diskp: ascii has 25.1 | 21:54 |
nemo | diskp: beowulf has 25.2 | 21:55 |
nemo | diskp: https://pkginfo.devuan.org/cgi-bin/d1pkgweb-query?search=emacs25&release=ascii | 21:55 |
gnarface | diskp: if the package has dependencies that are also in backports, you actually probably want them too | 21:55 |
fsmithred | I see 46.1 in ascii and 1:25.2+1-11 in beowulf | 21:55 |
nemo | diskp: frankly I don't think people pick debian for latest hotness ☺ | 21:56 |
nemo | diskp: I went to the trouble of adding a new sources.list to get a usable gimp tho - I can't believe gimp 2.10 was only just released this spring. yeesh | 21:56 |
fsmithred | older = better tested | 21:56 |
nemo | fsmithred: yeah. sure, makes sense most of time, is just, well, super irritating in a few minor domains like games | 21:57 |
nemo | fsmithred: Debian/Ubuntu/Mint users are bane of our hedgewars dev team ☺ | 21:57 |
nemo | fsmithred: similar issues with Spring | 21:57 |
nemo | although they do a frankly insane end-run around the distro by having the lobby client fetch builds from latest git | 21:57 |
fsmithred | I've run into the problem other places. Not a big gamer here. | 21:57 |
nemo | fsmithred: issue w/ games is people usually want to play against other humans and deterministic lockstep is a popular model in FOSS to reduce cheating and network load | 21:58 |
nemo | fsmithred: and that instantly means debian/ubuntu/mint users who are almost never on latest are constantly wandering around the lobby complaining about why no one wants to play with them ☹ | 21:58 |
nemo | fsmithred: I feel debian's policy could use more nuance, based upon, you know, the trivalness of the app ☺ | 21:58 |
Artemis3 | well there are rolling distros if you want to live in the edge... rest would be making your own repository, compile directly, distribute static binaries... The ubuntu camp has PPAs tho. | 21:59 |
nemo | Artemis3: yeah, for ubuntu, the maintainer for debian/ubuntu/mint (LocutusOfBorg) etc maintains a PPA - that solves their situation so long as we can communicate this to users | 21:59 |
diskp | Someone suggested either installing from source or using Guix. I don't see a Guix package, but the website package search utility has been acting funny. | 21:59 |
nemo | Artemis3: there's also playdeb for ubuntu | 21:59 |
nemo | Artemis3: for debian users, we usually tell them to dpkg -i | 21:59 |
Artemis3 | nemo, i'd try gdebi | 22:00 |
nemo | Artemis3: I just wish their policy could draw distinction between a game vs a critical system package - steam can update games without problems, why not debian | 22:00 |
nemo | Artemis3: I totally get security fixes only for, you know, important stuff | 22:00 |
nemo | Artemis3: but right now their policy makes Hedgewars local-play-only on debian the majority of the time | 22:00 |
nemo | even if we just do one or two releases a year | 22:00 |
fsmithred | flatpak! | 22:01 |
Artemis3 | An official repository would be best then? | 22:01 |
* nemo glares at fsmithred | 22:01 | |
Artemis3 | lol | 22:01 |
nemo | Artemis3: that's an interesting idea | 22:01 |
* fsmithred ducks | 22:01 | |
nemo | Artemis3: hm | 22:01 |
nemo | Artemis3: I'll ask LocutusOfBorg about that | 22:01 |
nemo | since he knows the intricacies of all that | 22:01 |
nemo | maybe we could add that to our download page in the Debian section | 22:01 |
nemo | Artemis3: what's the functional difference between adding a sources.list official repo and a PPA? apart from nicer management interface? | 22:02 |
nemo | Artemis3: still, it requires quite a lot of interaction on user's part, instead of just promoting anything that's a game to a special dynamic updates section | 22:02 |
nemo | could even have it be optional ☺ | 22:03 |
nemo | enable trivial-game-unstable-non-critical-backports | 22:03 |
Artemis3 | oh PPAs are maintained by Canonical, and they removed Debian support sometime ago. Its their place to go for personal repository where the keys and compiling is handled by them | 22:03 |
nemo | lame | 22:03 |
nemo | (the debian support part) | 22:04 |
nemo | Artemis3: I know I've added a PPA to devuan a couple of years ago though, so it must still work soemtimes | 22:04 |
Artemis3 | its good for Ubuntus but Debian needs another solution, tho i have seen people ocasionally using ubuntu PPAs in debian, but thats ugly | 22:04 |
Artemis3 | yes depending what the package links against | 22:05 |
nemo | hm. playdeb is ubuntu only... | 22:05 |
nemo | Artemis3: well... the frustrating thing for a game with a deterministic lockstep model (like hedgewars or spring) is we can change NOTHING about the libs we depend on - which has been true for 90% of the releases, such that debian could trivially compile it... but still be stuck with debian stable being years out of date | 22:06 |
nemo | Artemis3: actually, we were recently dragged kicking and screaming into the Qt5 world by gentoo threatening to remove us from the distro | 22:06 |
Artemis3 | there is also.. compile in the lowest common denominator, and try with debian/ubuntu stable to see if everything is ok, that is, binary distribution | 22:06 |
nemo | Artemis3: frustrating as all hell since qt5 removed features we depended on and also introduced fun surprise unicode bugs | 22:06 |
nemo | Artemis3: hmmmm | 22:07 |
nemo | Artemis3: yeah, I guess I'd just like to avoid having to explain all this to users. it'd be nice if they could find it in their distro tools somehow. I really feel backports lacks nuance | 22:07 |
Artemis3 | i remember some project using some old fedora to compile everything in, and they distributed those binaries for the major distros. Not sure if its still feasible, could take long testing | 22:07 |
Artemis3 | oh they would just download the binary and run it (when things work, that is) | 22:08 |
nemo | yeah, we really have been trying to avoid doing our own binary distribution. we've barely been able to keep windows functioning | 22:08 |
nemo | we're a very small team | 22:08 |
nemo | Artemis3: actually Windows 10 1809 just broke play for all users due to a bug on part of Microsoft. yay. | 22:08 |
Artemis3 | lol | 22:08 |
nemo | (causing physfs that we depend on to malfunction) | 22:08 |
Artemis3 | why i'm not surprised | 22:09 |
nemo | apparently microsoft pulled 1809 'cause it was deleting a ton of user files, but their rereleased version still doesn't function it seems | 22:09 |
nemo | Artemis3: we were Windows XP compatible up until the forced Qt5 upgrade - we would have maintained XP but we were having trouble finding the Qt5 long term support binaries that were XP compatible | 22:10 |
nemo | Artemis3: amusingly had a few users complain about this | 22:10 |
Artemis3 | or do you acually want to go the steam way? Tho they basically compile for debian/ubuntu, there was a low common denominator. They also bundle a bunch of base libs that you may use if you want | 22:10 |
nemo | Artemis3: we reaaaaaaally want to leave it up to the distro | 22:10 |
nemo | does Steam handle the builds for you? | 22:10 |
nemo | Artemis3: at present we only have one windows build "machine" which is a VM on unc0rr's desktop ☺ | 22:11 |
Artemis3 | i don't think so, but with the client you can see the bundled libs (which are usually older than the distro itself) | 22:11 |
nemo | Artemis3: we lost our OSX "build machine" 'cause koda has been too busy working for vimeo to do builds | 22:11 |
nemo | Artemis3: ew. vulnerabilties and all no doubt | 22:11 |
nemo | Artemis3: just crying out for DLC exploits | 22:12 |
Artemis3 | suppose it was ubuntu 14.04 (it might be the Debian from SteamOS now days) you see the lib versions are bundled with the client, even if the distro doesn't have those the client do | 22:12 |
Artemis3 | so you compile for those specific versions and it "should" be covered, thats their way i think | 22:12 |
nemo | Artemis3: anyway. speaking of backports ... do you know the "correct" way to get libreoffice 6 from backports into ascii? I'm trying to avoid screwing up my package tree | 22:12 |
nemo | Artemis3: yeah, I dunno. Locutus does maintain 14.04 for now, since it is in ubuntu LTS until next april - no idea what happens then. maybe steam will bump theirs | 22:13 |
nemo | or drop linux. who knows | 22:13 |
Artemis3 | if its in backports there should be no problem, it should be lower priority anyway, so you have to tell apt to specifically pull the backports version with -t iirc | 22:13 |
nemo | Artemis3: so I guess in theory could maybe do some sort of autowrapper to fetch the ubuntu 14.04 .deb for now and feed it into steam | 22:13 |
nemo | Artemis3: oh.. | 22:14 |
nemo | hm | 22:14 |
* nemo RTFMS | 22:14 | |
fsmithred | add backports repo to sources.list, update, then 'apt-get -t ascii-backports install $package' | 22:14 |
nemo | yay | 22:14 |
Artemis3 | nemo, for that i'd need to update myself with steam... i think they have a page somewhere explaining the developing of linux | 22:14 |
Artemis3 | yes that | 22:14 |
Artemis3 | for | 22:14 |
nemo | that's muuuuch better | 22:14 |
fsmithred | you shouldn't need to disable backports repo after you install what you want, but it's a good idea to do that anyway | 22:15 |
nemo | Artemis3: up until now my gui clueless way to do it was "force version" in synaptic which tended to break things with large complex webs of packages like this one | 22:15 |
fsmithred | just in case the priority gets screwed up in the repo | 22:15 |
nemo | latest libreoffice and gimp is pretty important ☺ | 22:15 |
Artemis3 | unless you to -t ascii-backports update/upgrade (don't lol) it will only pick the required package and its dependencies | 22:15 |
Artemis3 | you can safely leave backports in sources.list, won't harm you in anyway | 22:16 |
nemo | Artemis3: yeah. I already had it enabled, it was using it correctly that was the problem | 22:17 |
Artemis3 | the fun is when the package in backports isn't the one you want, but if it is in backports, thats no brainer xD | 22:17 |
nemo | Artemis3: far riskier was doing this https://forums.bunsenlabs.org/viewtopic.php?id=4770 | 22:17 |
nemo | but seems fine so far | 22:17 |
Artemis3 | yes, thats an example | 22:18 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!