EHeM | Given the state of things, I'd almost go with unreliable installer ISOs. | 00:04 |
---|---|---|
* EHeM wonders how far he can let this disk go before declaring it failed. | 00:11 | |
xrogaan | I started getting theses: [xcb] Unknown sequence number while processing queue | 02:12 |
xrogaan | followed with the crash of a software | 02:12 |
xrogaan | I don't know what triggers it. | 02:12 |
golinux | You're on ceres iirc | 02:17 |
xrogaan | I'm supposed to be on beowulf | 02:22 |
tuxd3v | I am installing at least beowulf in a old laptop pentiumM :) | 02:27 |
tuxd3v | 1GB Ram | 02:27 |
tuxd3v | Any advice I should follow?I was thinking in going ascii, then upgrade.. | 02:27 |
yeti | ok | 02:29 |
yeti | only minimal install, then dist-upgrade | 02:29 |
yeti | I've beo on a T30 | 02:29 |
yeti | P4M@2000 ... started 1G RAM but could get it upgraded to 2G | 02:30 |
tuxd3v | nice I will try that :) | 02:30 |
yeti | and on a S10 | 02:30 |
yeti | 1.6GHz Atom "Netbook" | 02:30 |
yeti | 2G RAM too | 02:30 |
yeti | only on ARM I have systems with fewer RAM | 02:31 |
tuxd3v | downloading 'devuan_ascii_2.1_i386_cd-1.iso' | 02:31 |
yeti | but I'll dig out other 1G T30s and a 512M T23 someday... | 02:31 |
yeti | code density of x86 is better than on ARM | 02:32 |
yeti | so 512M there is more than 512M on e.g. Pi1 | 02:32 |
xrogaan | last that crashed for me is hexchat, any idea of what could cause a 'xcb_io.c:263: poll_for_event: Assertion `!xcb_xlib_threads_sequence_lost' failed.` | 02:32 |
yeti | i've no idea about hexchat and GUIstuff | 02:33 |
yeti | xcb_xlib <<< smells like a GUI problem | 02:34 |
fsmithred | looks like it's some freedesktop.org stuff | 02:39 |
fsmithred | I'm using hexchat in beowulf right now. | 02:39 |
xrogaan | It happens randomly, the software affected isn't always the same. | 02:41 |
xrogaan | I don't know how to reproduce :/ | 02:42 |
xrogaan | could it be due to the window manager? | 02:42 |
fsmithred | seems to be related to multithread | 02:43 |
fsmithred | lots of hits on those error messages | 02:43 |
fsmithred | related to different programs | 02:43 |
fsmithred | and answers that I don't know how to read | 02:44 |
xrogaan | same. | 02:44 |
xrogaan | It's almost always related to a programming error, but in my case it just doesn't make sense. | 02:45 |
xrogaan | I use compton as a compositor, it is threaded and old enough to have me wary. Have to use it because xfce's compositor has bugs with vsync. | 02:45 |
fsmithred | what wm or de? | 02:45 |
xrogaan | xfce | 02:45 |
fsmithred | same here, but I probably have compositing shut off | 02:46 |
fsmithred | not sure, I'll check | 02:46 |
xrogaan | how do you deal with non-transparent terminal window? :P | 02:46 |
xrogaan | also, how do you launch compton? | 02:46 |
fsmithred | I just have the default xfce settings for the compositor | 02:47 |
fsmithred | the only non-transparent terminals I use are xterms that pop up from scripts. | 02:48 |
xrogaan | compton usually vomit a ton of garbage in my logs, like | 02:50 |
xrogaan | [ 25952.66 ] error 140 BadRegion request 138 minor 17 serial 63633334: "BadRegion (invalid Region parameter)" | 02:50 |
fsmithred | what logs? | 02:50 |
xrogaan | xsession-errors | 02:50 |
xrogaan | or glx_bind_pixmap(0x01aad0c4): Failed to query Pixmap info. | 02:50 |
xrogaan | if you have a ~/.xsession-errors file, it'll be populated with the stderr output of all the software you run | 02:51 |
xrogaan | tied to the graphical session, I mean. | 02:51 |
fsmithred | this may take some time - it's over 8,000 lines | 02:51 |
fsmithred | no glx_bind_pixmap or BadRegion | 02:52 |
fsmithred | the xfce weather plugin has been hitting on a bad url. That what all my errors are. | 03:02 |
xrogaan | can slim exec task at user logout? | 03:17 |
xrogaan | or system shutdown? Doesn't seem to properly end my session when I shutdown. | 03:17 |
fsmithred | conlsolekit or elogind, and do you have the right libraries? | 03:20 |
xrogaan | what do you mean right libraries? | 03:31 |
xrogaan | I switched to consolekit | 03:32 |
fsmithred | it's in the release notes | 03:34 |
fsmithred | oh wait. There aren't any for beowulf yet. | 03:34 |
fsmithred | libpolkit-gobject... libpolkit-backend... | 03:35 |
fsmithred | they exit for elogind and consolekit | 03:35 |
fsmithred | you want a matching set | 03:35 |
fsmithred | also libpam-ck-connector for consolekit | 03:35 |
xrogaan | well yeah, have those things | 03:41 |
xrogaan | elogind or lightdm used to move ~/.xsession-errors to ~/.xsession-errors.old at either each login or logout. | 03:43 |
xrogaan | which help with not bloating the file. | 03:43 |
xrogaan | but with slim+consolekit, doesn't happen. | 03:43 |
djph | well, tongiht shouldbe fun ...starting my adventure of moving a webserver from jessie to beowulf ... on new hardware. Oh joy | 03:46 |
g4570n | \o/ | 03:54 |
pandakekok9 | \o\ | 03:55 |
DonkeyHotei | /o\ | 03:56 |
g4570n | /o/ | 03:58 |
_Ev | I confess to liking systemd for servers. Devuan surely seems like a more sane route however. | 03:59 |
fsmithred | xrogaan, did you look at /etc/slim.conf? I think you can add commands. | 04:08 |
xrogaan | indeed, thanks | 04:10 |
xrogaan | no documentation for the variables used in that config file htough | 04:12 |
fsmithred | what's the login_cmd? | 04:12 |
xrogaan | login_cmd exec /bin/bash -login /etc/X11/Xsession %session | 04:13 |
fsmithred | ok | 04:13 |
xrogaan | but I would need to do something like a mv $HOME/.xsession-errors | 04:13 |
xrogaan | no idea if $HOME is available | 04:13 |
xrogaan | there is a %user slim variable, no idea of its value | 04:14 |
xrogaan | sessionstart_cmd and sessionstop_cmd | 04:14 |
fsmithred | yeah, %user would the the user starting the session | 04:15 |
xrogaan | apparently ~/ is available | 04:15 |
xrogaan | but I believe it would be better to have a Xsession.d script to handle the xsession-errors file | 04:16 |
fsmithred | yeah, probably | 04:17 |
xrogaan | oh, wow, dwarf fortress is in debian's non-free | 05:33 |
DonkeyHotei | is that the game whose executable was simply "df" leading to the breakage of a lot of scripts? | 05:36 |
xrogaan | probably? | 05:36 |
xrogaan | df is a shell script | 05:36 |
xrogaan | or should be | 05:37 |
xrogaan | dwarf-fortress: /usr/lib/games/dwarf-fortress/df | 05:37 |
EHeM | xrogaan: Might you be running into hardware issues? | 06:24 |
xrogaan | I don't believe I do. | 06:29 |
xrogaan | EHeM: how would I make sure? | 06:42 |
EHeM | Find evidence something is failing. Try running a memory test program and see whether anything shows up. | 07:11 |
EHeM | Sounds like an intermittent failure with many programs? That could well be some flavor of hardware failure. | 07:12 |
xrogaan | memory heh? | 07:25 |
xrogaan | Doesn't happen often though. | 07:42 |
gour | morning | 07:44 |
gour | today i want to migrate to devuan - first my netbook (spare machine) and then the desktop | 07:45 |
gour | on netbook i want simple lvm2 install (single hd) and it seems that installer in manual mode does create separate /boot. my question is if it can be done without it (e.g. archlinux does work with /boot on root)? | 07:47 |
_Ev | My wifi card works on Ubuntu out of the box. I say to myself... "maybe I should just use ubuntu for a while." | 07:56 |
* _Ev googles vulnrabilies | 07:57 | |
_Ev | https://usn.ubuntu.com/releases/ubuntu-19.04/ | 07:57 |
EHeM | I'm pretty sure it is near-verboten if not actually verboten to ask, but do we have any ideas on a time-frame for beowulf? I'm under the impression /most/ of the pieces are ready, just a few (notably installer ISOs) are missing. | 07:57 |
_Ev | I need Beowulf now, haha. it's killing me. | 07:58 |
sedrosken | Yeah, I just dist-upgraded from a ASCII 2.1 fresh install myself. Minor hiccup in the OpenRC cyclic dependency checker (I use encrypted LVM and had a cyclic dependency because of that) in that it would segfault when apt-get dist-upgrade was trying to configure the new version but disabling that bit and doing a --fix-broken install and another pass of dist-upgrade had it going again | 07:59 |
sedrosken | the thing is, now it seems to ID as Debian 10 to sysinfo | 08:00 |
sedrosken | Client: HexChat 2.14.2 • OS: Debian 10.0 • CPU: Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz (1.60GHz) • Memory: Physical: 15.2 GiB Total (13.7 GiB Free) Swap: 16.4 GiB Total (16.4 GiB Free) • Storage: 1.5 TB / 8.0 TB (6.5 TB Free) • VGA: Advanced Micro Devices, Inc. [AMD/ATI] Ellesmere [Radeon RX 470/480] @ Intel Corporation Xeon E3-1200 v2/3rd Gen Core processor DRAM Controller • Uptime: 45m 5s | 08:00 |
sedrosken | which I can't decide if I find mildly frustrating or mildly amusing | 08:01 |
sedrosken | I'm going to attempt to migrate to openrc-init fully, I'm not the biggest fan of sysvinit and the way we normally have openrc riding on top of it | 08:02 |
sedrosken | plus, I'm mildly confused by the topmenu stuff being in ascii's repos but curiously not beowulf's | 08:03 |
sedrosken | I had wanted to do a little thing with that on my MATE install since I have a ton of open space on my top panel (I use plank) | 08:04 |
gour | sedrosken: what about runit which is used by e.g. voidlinux? | 08:42 |
_Ev | 2060 super will last 5+ years? | 08:47 |
_Ev | Sorry, wrong chan | 08:47 |
sedrosken | I do like runit, but for a machine that will serve as my main and home server I'd prefer some more robust service supervision | 08:51 |
sedrosken | as a matter of fact I *love* runit, it's dead simple to use and blindingly quick | 08:52 |
gour | ok | 09:04 |
gour | my netbook with / under lvm and no separate /boot can't boot. let me see if ceres can fix it | 09:05 |
gour | i'm also confused with https://wiki.debian.org/LVM#A.2Fboot since i believe that newer grub2 should support it | 09:08 |
gnarface | i thought someone told me you had to have bios level support to make that work | 09:25 |
gnarface | dunno about the grub1/grub2 thing | 09:26 |
gour | gnarface: well, with arclinux/manjaro it does work OotB | 09:28 |
gnarface | gour: did you say you tried it in ceres, too? ascii probably has a lot older versions of everything than current arch and manjaro | 09:32 |
gour | gnarface: now i'm updating to ceres...in rescue mode | 09:32 |
gour | even with ceres it does not work :( | 09:51 |
gour | will try uefi install | 09:56 |
gour | gnarface: how long before devuan release based on buster? | 10:10 |
sedrosken | beowulf from what I understand is supposed to hit release any day now | 10:12 |
gour | thumbs up | 10:12 |
sedrosken | you can dist-upgrade to it at any time if you like | 10:12 |
sedrosken | it's still technically "testing" so there isn't any isos available yet, afaict | 10:13 |
gour | doing uefi-install now | 10:13 |
sedrosken | but again, from what I understand, it's about as close to being released as it gets without actually having been so | 10:13 |
gour | bios/gpt with root under lvm didn't work | 10:14 |
gnarface | gour: no official date as far as i know, but people are already using it... it might be worth a try | 11:17 |
_Ev | What kernel will Beowulf be using? | 11:22 |
_Ev | s/what/which | 11:22 |
debdog | _Ev: https://packages.debian.org/search?keywords=linux-image&searchon=names&suite=stable§ion=all | 11:31 |
_Ev | Sigh. Guess I am done with Devuan for a year or so. | 11:32 |
* _Ev heart breaks | 11:32 | |
debdog | whyß | 11:32 |
debdog | s/ß/?/ | 11:32 |
debdog | I think there's some 5.something in backports | 11:33 |
debdog | there's always the option of building a kernel also | 11:33 |
_Ev | Too busy to attempt a kernel upgrade on this new Thinkpad. I'll figure it out. I need to travel ASAP and no internet is an issue. Almost had the external wireless card configured but then dhclient murderd me. | 11:34 |
_Ev | Just a lot of work for me right now | 11:34 |
debdog | oh | 11:34 |
debdog | that's never a good condition to set up a system | 11:34 |
debdog | haha, good ole days http://gunkies.org/wiki/Stacker | 11:43 |
xrogaan | _Ev you can use backports | 11:47 |
yeti | always the newest bugs? | 11:54 |
_Ev | I just ordered an rj45 converter. I will try a kernel upgrade in a few days | 12:01 |
_Ev | I require 5.1+ but will try 5.4.xyz | 12:01 |
_Ev | Actually no I wont. Sounds unstable to me. I'll just use another OS for a year or so. Hopefully les. | 12:25 |
xrogaan | From where do you see it be unstable? | 12:29 |
xrogaan | or is it just unstable for the wifi | 12:29 |
_Ev | I'm just assuming it would be unstable for the mainline packages | 12:30 |
xrogaan | I don't understand. | 12:30 |
_Ev | The Wi-Fi 6 card requires 5.1+ | 12:31 |
_Ev | xrogaan: I just am not educated enough. Merely assuming. | 12:31 |
fsmithred | There is a suite named unstable, and it is named that because the package versions keep changing, not because the software is flaky. | 12:32 |
_Ev | Devuan used 4.9 as the kernel only because Debian does? Not because if you upgrade it will break things? | 12:34 |
xrogaan | yes | 12:34 |
xrogaan | devuan's only purpose is to throw systemd away, everything else is debian. | 12:34 |
xrogaan | well, I should say main purpose instead of only. | 12:35 |
fsmithred | and one of our goals is to accomplish that purpose with the least amount of work | 12:35 |
xrogaan | Lazy people rise up! | 12:35 |
fsmithred | necessity may be the mother of invention, but laziness is its father | 12:36 |
yeti | ! | 12:36 |
_Ev | Haha. I can't argue with that. | 12:47 |
gour | how to get graphical splash screen in grub menu instead of boring textual one? | 12:47 |
_Ev | Other than just the systemd aspect, does devuan aim to be more secure in other areas? | 12:47 |
chillfan | gour try apt-get install desktop-base | 12:48 |
xrogaan | really, in desktop-base? | 12:50 |
hightower2 | Hey can something be done about xfce4 on ppc64le architecture... for quite some time now package xfce4 is not installable due to xfce4-session wanting libpolkit-gobject-1-0, which is not installable | 12:51 |
xrogaan | hightower2: it's a debian issue | 12:51 |
hightower2 | ouch. Any workarounds? | 12:52 |
hightower2 | well, I'll try package by package | 12:52 |
chillfan | _Ev There has been some (unofficial) talk about having a libre kernel or sub-distro. | 12:53 |
chillfan | We don't have a security specific focus, or any plans to add more features in that way that I know of. | 12:53 |
xrogaan | oh wait, libpolkit is forked by devuan | 12:54 |
_Ev | Debian itself is quite secure. Minus systemd of course -.- | 12:54 |
chillfan | Well, depends what you think of as secure. There's a good quality process for packages at least | 12:56 |
gour | chillfan: thanks! | 12:56 |
chillfan | no problem :) | 12:56 |
xrogaan | hightower2: try aptitude | 12:57 |
xrogaan | aptitude has a nice conflict resolution system | 12:57 |
hightower2 | xrogaan, on it | 12:57 |
chillfan | Has anyone managed to succesfully migrate a buster system to beowulf? | 12:57 |
gour | devuan boots with root under lvm2 on ceres here | 12:57 |
hightower2 | xrogaan, ah, no, there is no conflict. Some packages simply aren't installable | 12:58 |
xrogaan | libpolkit-gobject-1-0 should be installable regardless, if it can't then you have an issue with the dependencies (either elogind or consolekit) | 12:58 |
chillfan | I've been trying to document this but, it's a no go.. any attempt to migrate results in the kernel being removed | 12:58 |
hightower2 | xrogaan, for both of those (elogind and consolekit), the bottom line error is simply "no installation candidate" | 12:58 |
chillfan | Try specifying which one. e.g apt-get install pkgname consolekit | 12:59 |
xrogaan | ppc64le might not be supported by devuan? I don't know (ping fsmithred ) | 12:59 |
hightower2 | yes, yes, I did | 12:59 |
hightower2 | no luck | 12:59 |
fsmithred | I think somone is working on ppc | 13:00 |
fsmithred | so... not yet, but maybe later | 13:00 |
hightower2 | well yes but the installer images exist and they work, I have it installed | 13:00 |
fsmithred | not sure if it's the same. | 13:00 |
xrogaan | fsmithred: is there an easy way to get the list of stuff you forked and maintain? | 13:01 |
fsmithred | chillfan, there's a script for buster2beowulf | 13:02 |
chillfan | ah, where abouts? | 13:02 |
fsmithred | http://pkgmaster.devuan.org/bannedpackages.txt | 13:03 |
Evilham | a wild chillfan appeared! | 13:03 |
hightower2 | xrogaan, it was failing on some package which it lists as 'purely virtual', so I just added it as installed to /var/lib/dpkg/status. Now the install of xfce4 goes through. Will report if it actually also works afterwards | 13:03 |
fsmithred | was posted on DNG. I'll email it to you. | 13:03 |
fsmithred | assuming I have your address already | 13:03 |
chillfan | thanks. I guess they run into the same issue as I did | 13:03 |
fsmithred | there was also an earlier post on d1g with a path that worked | 13:04 |
fsmithred | I'll get link for that | 13:04 |
fsmithred | this is older and may need to be revised: https://dev1galaxy.org/viewtopic.php?pid=17716#p17716 | 13:05 |
xrogaan | hightower2: libpolkit-gobject-1-0 is virtual, you need either consolekit or elogind | 13:06 |
xrogaan | https://ci.devuan.org/job/policykit-1-binaries/architecture=ppc64el,label=ppc64el/ | 13:06 |
hightower2 | yes but it's already installed and it did not provide this | 13:06 |
chillfan | Thanks fsmithred I'll give it a try | 13:07 |
fsmithred | no, give the next one a try | 13:07 |
fsmithred | one minute... | 13:07 |
fsmithred | buster_migration.sh - https://termbin.com/2tuk | 13:08 |
fsmithred | chillfan, ^^^ | 13:08 |
chillfan | thanks | 13:08 |
xrogaan | hightower2: aptitude show libpolkit-gobject-1-0 | 13:09 |
xrogaan | shoud list: Provided by: libpolkit-gobject-consolekit-1-0 (0.105-25+devuan8), libpolkit-gobject-elogind-1-0 (0.105-25+devuan8) | 13:09 |
fsmithred | there's also a libpolkit-backend for each | 13:10 |
fsmithred | you want a matched set | 13:10 |
hightower2 | yes but I think you're missing the point... what I'm saying is that 'apt install xfce4' does not work | 13:15 |
DonkeyHotei | chillfan: yo | 13:15 |
hightower2 | and that's because ultimately something does not work... | 13:15 |
chillfan | hey DonkeyHotei | 13:15 |
xrogaan | hightower2: use aptitude, not apt | 13:17 |
hightower2 | yes I did, it's not helping | 13:18 |
xrogaan | it'll tell you why it can't install, and so you can tell us | 13:18 |
hightower2 | but ok anyway... | 13:18 |
xrogaan | We don't know why it fails, because you're not telling us. Most likely because you don't know yourself. | 13:19 |
yeti | had a problem with xfce in ceres a while ago... landed on mate because of that | 13:22 |
yeti | just a VM guinea pig... so not worth long trouble | 13:23 |
yeti | but it might have been related | 13:23 |
sgage | gour: one way to get a graphical grub menu is to simply put a | 13:30 |
sgage | graphics file in /boot/grub, and update-grub. that's how | 13:31 |
sgage | I've been doing it. | 13:31 |
chillfan | ah, I may have found a reasonable migration path for buster | 13:44 |
avbox | Trying to install falkon web browser, it does install elogind which seems to be incompatible with libsystemd. Looks like libsystemd is still needed to some other package. If setting libsystemd on hold, trying to install falkon says that it is not possible because of package libkf5wallet-bin. How to instal falkon without installing elogind? | 14:26 |
avbox | Looks like that libpam-systemd is the problem, want's to install elogind. | 14:29 |
fsmithred | try installing libelogind0 | 14:30 |
fsmithred | and libpam-elogind | 14:30 |
fsmithred | oh, you want without elogind | 14:30 |
avbox | @fsmithred: Yes, libelogind0 conflicts with libsystemd0. If I install elogind, I get serious troubles. Switching from normal user to root takes about 10 seconds, starting kodi takes very long time etc etc. | 14:32 |
djph | \o/ only a few headaches with moving from ascii -> beowulf for me server | 14:36 |
djph | *mailserver | 14:36 |
fsmithred | avbox, are you in ascii or beowulf? | 14:38 |
djph | ... and most of them "hey stupid, you forgot to move the perl scripts ... " | 14:38 |
avbox | @fsmithred: Beowulf, it really looks that install falkon without elogind does not work at all. | 14:38 |
fsmithred | I don't see any of that in the listed dependencies | 14:40 |
bilbo_b | I created myself an interesting problem when trying to migrate from buster to beowulf following the guide by nixer on dev1galaxy but forgetting to replace ascii by beowulf in the sources.list :/ | 14:41 |
avbox | fsmithred: So if you install falkon, elogind is not installed? And libsystemd0 still is there after it? | 14:42 |
bilbo_b | not sure if currently not being able to get an ip-adress on my ethernet interface is directly related | 14:42 |
fsmithred | avbox, I need to find an install without elogind to test | 14:43 |
fsmithred | I'm going by what apt-cache tells me | 14:43 |
fsmithred | try installing it with aptitude - it may give you other suggestions | 14:45 |
avbox | fsmithred: Do I understand it right, elogind is the "future" and "libsystemd" is only there because in ascii it was needed? | 14:47 |
fsmithred | elogind is one of our packages. It satisfies the dependency for systemd without being systemd. | 14:48 |
fsmithred | libelogind0 is the replacement for libsystemd0 | 14:48 |
bilbo_b | when booting eth0, eth1 and eth2 get recognised as interfaces and MAC-addresses listed, later they get renamed (by eudev?) to eth3, 4 and 5, but none show a link with ethtool (which they did previously) | 14:48 |
fsmithred | its job is to tell programs that systemd is not running | 14:48 |
enyc | hrrm, ooi, has anybody been looking for power management issues with devuan-beowulf? | 14:48 |
enyc | I note, mxlinux (systemd not used by default) having all sorts of issues with power consumption between 18 -> 19 (same debian base shift i believe) | 14:49 |
fsmithred | in ascii it's not possible to completely replace libsystemd0 with libelogind0 | 14:49 |
enyc | i wonder if debian relies more on systemd to turn on power-manglement matters etc | 14:49 |
sgage | but in beowulf it is possible to completely replace libsystemd0 with libelogind0. | 14:53 |
avbox | @sgage: Do I understand it right, that when replacing libsystemd0 with libelogind0, deamon eloingd must be started? | 14:55 |
sgage | my beowulf system is running elogind and libelogind0, along with libpam-elogind and the 2 appropriate libpolkit packages. | 14:57 |
fsmithred | I'm pretty sure you can have libelogind0 without elogind | 14:57 |
sgage | zero systemd anything. | 14:57 |
fsmithred | avbox, what desktop are you running? | 14:58 |
avbox | @fsmithred: Mate | 14:58 |
fsmithred | I think that should work with either consolekit or elogind | 14:59 |
fsmithred | I don't have a beowulf with consolekit right now to test falkon | 14:59 |
fsmithred | maybe I'll do an install today | 14:59 |
fsmithred | did you try it with aptitude? | 15:00 |
fsmithred | oh | 15:00 |
fsmithred | also try installing falkon without Recommends | 15:00 |
fsmithred | brb | 15:00 |
avbox | @fsmithred: If you like you can just take avmultimedia (installed in 10 seconds). Recommends are disabled, but I did not test it installing with aptitude. At this stage, I think pure slowness of the stytem is because I had not started 'elogind'. | 15:01 |
xrogaan | a lot of things depends on elogind | 15:10 |
* gour_ is doing backup prior to devuan install on main (desktop) machine | 15:10 | |
gour_ | wish me a good luck ;) | 15:11 |
xrogaan | apparently, elogind should be compatible with nearly everything, but it isn't the same with consolekit | 15:11 |
djph | o...kay; so either ascii was way better at being starved for RAM, or nextcloud is a cantankerous little bastard | 16:15 |
specing | the latter | 16:16 |
specing | PHP shit | 16:16 |
xrogaan | wait, nextcloud is written in PHP? | 16:17 |
specing | what else | 16:17 |
specing | afaik its a fork of owncloud | 16:17 |
djph | yes, PHP is the root problem | 17:31 |
djph | ... apparently nextcloud wants a big php heap. Guess what they don't put in the documentation (least that I saw -cue someone going "yeah, it's right there") and semi-hid the warning in one of the admin panel setup panes. | 17:32 |
tuxd3v | Is there any possibility to have Equinox Desktop Environment on Devuan? | 18:18 |
tuxd3v | there are packages for amd64 | 18:18 |
tuxd3v | for would ne nice to have for i386, due to memory limitation on this platforms | 18:19 |
tuxd3v | or even armhf,arm64 :) | 18:19 |
tuxd3v | https://www.youtube.com/watch?v=4QqNh_B5o4k | 18:24 |
wildjames | Last update of EDE was in 2014, why would you install that. It's probably to old to run with newer libraries or libc | 19:13 |
tuxd3v | hmm, could be, but in amd64 it uses 104Mb of ram after a could boot | 19:16 |
tuxd3v | so in 686 it will use less | 19:16 |
tuxd3v | also a good option to armhf and arm64 | 19:16 |
tuxd3v | due to Ram contraints.. | 19:16 |
tuxd3v | maybe it needs to be rebuilt against more recent packages.. | 19:17 |
tuxd3v | I already tried, but failed in some regards.. | 19:17 |
tuxd3v | now if we at least can get dependencies straight.. | 19:18 |
tuxd3v | we can build it for more recent stacks :) | 19:18 |
tuxd3v | its installed in my amd64 machine, but due to dualgraphics, I can't use it.. | 19:18 |
cosurgi | Guys, iny idea how to disable this warning: | 20:35 |
cosurgi | *any idea? | 20:35 |
cosurgi | (EE) dbus-core: error connecting to system bus: org.freedesktop.DBus.Error.FileNotFound (Failed to connect to socket /var/run/dbus/system_bus_socket: No such file or directory) | 20:35 |
cosurgi | It is spamming my /var/log/Xorg.0.log | 20:35 |
cosurgi | the file is 40MB already. | 20:35 |
chillfan | how old is the log? I have ~0.25MB for one of mine | 20:37 |
debdog | and why is it in /var/log? shouldn't i be in ~ somewhere these days? | 20:38 |
chillfan | yeah, mine is somehow always in ~/.local/share/xorg | 20:38 |
chillfan | maybe the display manager puts it there | 20:39 |
cosurgi | yeah. My log files are in /var/log/Xorg.0.log | 20:40 |
cosurgi | I think that's because I startx from text virtual terminal with startx. | 20:41 |
debdog | so you're running xorg as root | 20:41 |
cosurgi | no. definitely not :) | 20:41 |
cosurgi | xorg-wrapper does something | 20:41 |
chillfan | do ls -l /var/log/Xorg* see what the creation time is | 20:42 |
cosurgi | I start Xorg as 6 different users. | 20:42 |
debdog | well, not as root but not as ordinary user either. what I meant is, mayhap that could be an issue here. | 20:42 |
debdog | depending on DE, too, of course | 20:42 |
cosurgi | few seconds ago. Ah it just updated in gvim. | 20:42 |
cosurgi | vim asked to reload file. | 20:43 |
cosurgi | few more same line appeared. | 20:43 |
chillfan | hm, 6 users.. that might be part of it. Bear in mind if dbus doesn't have the whole shebang it prints those warnings.. so disabling dbus can be as annoying as having it enabled | 20:43 |
cosurgi | I use the desktop without elogind consolekit dbus policykit. It works for over a month already. I just noticed it's spamming the log. | 20:43 |
systemdlete | disk space? | 20:43 |
cosurgi | nah ;) nvm. | 20:43 |
chillfan | I've had people say that the d-bus spamming was hurting their performance too, by printing too many logs | 20:44 |
cosurgi | actually these 40MB accumulated over the month of usage after I uninstalled elogind consolekit dbus policykit | 20:44 |
chillfan | a months usage without rotating the logs would do it | 20:45 |
enzosdfsdf | hello | 20:45 |
chillfan | do you have logrotate installed? | 20:45 |
cosurgi | not yet :) | 20:46 |
enzosdfsdf | devuan features a mirror which domain is "unetresgrossebite.com", which means in french "a very big c*ck". you probably wanna change that | 20:46 |
cosurgi | I still have 6GB free space on /var partition. | 20:46 |
systemdlete | ok. Just thought I'd ask. So many times, I've spent even hours(!) trying to figure out the source of some issue or another, only to find that it was low disk space all along. | 20:47 |
systemdlete | Of course, I am not the kind of geeks some of you all are. | 20:48 |
systemdlete | I've found that even fairly plentiful disk space can still lead to problems, even as little as half a gig or so. That seems like a lot, but some processes are voracious it seems. | 20:49 |
chillfan | yeah get logrotate, let it deal with that. Now and then I just check up on it and delete old gzipped files | 20:49 |
chillfan | you can probably tweak it to remove them sooner though | 20:50 |
cosurgi | I'm pondering to recompile Xorg just to remove this annoying message, | 20:53 |
chillfan | You can but things like firefox will still cause that to be printed in the xorg log | 20:54 |
chillfan | probably xorg isn't directly to blame | 20:54 |
cosurgi | ah, that may be right... | 20:54 |
cosurgi | ok. still 6GB free space, this can wait ;) | 20:54 |
debdog | or add something to startx to clear that log on startup | 20:56 |
debdog | *before startup | 20:56 |
debdog | well, dirty workaround | 20:57 |
Hum | Did you recognize https://fosdem.org/2020/schedule/event/riek_kubernetes/ | 20:58 |
Hum | How Containers and Kubernetes re-defined the GNU/Linux Operating System | 20:58 |
Hum | A Greybeard's Worst Nightmare | 20:58 |
chillfan | dbus is a major thorn, maybe worse than systemd for spreading through the system | 20:58 |
Hum | In the german article from heise.de he is cited with "greybeard's riot even worse then with systemd" | 20:59 |
chillfan | eh.. I don't think there is any connection with containers and people who don't like systemd | 21:02 |
chillfan | though lxc can probably be said to suck :p | 21:03 |
chillfan | I see no riots about this however | 21:05 |
Hum | Damn, I am more like a greybeard who wants to understand and control the technology. And that guy discriminates against me. Is Red Hat a company that wants me to consume without knowing how Soylent Green is made? | 21:07 |
chillfan | Well for sure you'll face a lot of abuse from that direction, your ways are old and outdated and you should get onboard without asking any questions | 21:08 |
chillfan | Making it personal is a good sign someone has an agenda | 21:09 |
Hum | the discussion about systemd and the fork from debian was a kind riot in the debian universe | 21:09 |
chillfan | a.k.a the "Straw Man Attack" | 21:09 |
chillfan | well we did make a fuss about it.. and I've probably been one of the loudest, having no beard at all and certainly not a grey one | 21:10 |
Hum | :D thx for straw man attack, I have to read about it | 21:10 |
chillfan | yeah you should read about these tactics, they are common amongst SJWs.. e.g go for the person, not their opinion. | 21:11 |
DarkUranium | chillfan, no u | 21:12 |
chillfan | Me what? | 21:13 |
Hum | chillfan: go for the person is imho not part of strawman attack | 21:14 |
chillfan | Well, the point is to misdirect people to something that's not at issue | 21:14 |
fluffywolf | ah, the republican strategy. | 21:15 |
chillfan | I mean to say there are different types of tactics, and sad to say most of the systemd crowd also seem to be sjws | 21:15 |
golinux | This is not a political channel folks | 21:15 |
Hum | ok | 21:16 |
golinux | Try #debianfork | 21:16 |
bilbo_b | Hi there | 21:28 |
bilbo_b | after a migration from Debian buster to beowulf my interfaces get renamed from eth1 to eth3, from eth0 to eth4 and from eth2 to eth5 | 21:29 |
bilbo_b | any suggestion where I might look, why that is and how to change it? | 21:29 |
gnarface | bilbo_b: /etc/udev/rules.d/70-persistent-net.rules | 21:29 |
bilbo_b | still true after installation of eudev? | 21:29 |
gnarface | as far as i know, yes. if i'm wrong it's just some other file in /etc/udev/rules.d/ though probably | 21:30 |
bilbo_b | thank you, will have a look | 21:30 |
gnarface | it should be pretty easy to spot, because it's the one with all your ethernet devices in it | 21:30 |
bilbo_b | I find 3 entries with the original eth0, eth1, eth2 in there :/ | 21:31 |
bilbo_b | the above listed file is the only one in that dir | 21:32 |
gnarface | the idea is that when you add a new device or change the order of devices, or hotplug devices, it remembers which one is which. but the side effect of the fairly dumb logic doing that is that when you reinstall, it recognizes the old device as a new one, and just iterates the name | 21:32 |
bilbo_b | ifconfig -a gives only eth3, eth4, eth5 after boot into rescue-mode | 21:32 |
bilbo_b | so just deleting the file should help? | 21:33 |
gnarface | yea, though it's weird that you have devices not in there... | 21:33 |
bilbo_b | (or moving to a different name at first try) | 21:33 |
chillfan | It's kinda strange you had ethX at all in buster | 21:33 |
gnarface | there is also a /lib/udev/rules.d that as far as i know is read the same way, but should not be automatically populated by anything other than the package manager... still, maybe worth grepping that, too | 21:33 |
chillfan | usually they are ensX | 21:34 |
bilbo_b | could be because in buster I changed hw (mac-addresses) to same board-type but replacement, then the buster-upgrade changed them to en... and maybe it remembered my old mac-addres | 21:34 |
bilbo_b | it is expected that the grub menu items stay "Debian" when update-grub is run in Devuan? | 21:35 |
bilbo_b | that was it, after moving the rules file and reboot they are eth0, 1 and 2. thank you | 21:35 |
chillfan | yeah, usually it will still say Debian | 21:36 |
fsmithred | change /etc/os-release ID= | 21:37 |
fsmithred | if you want it to say Devuan | 21:37 |
bilbo_b | thank you, I guess I am done then with my conversion to Devuan Beowulf | 21:37 |
chillfan | Which instructions did you use to migrate btw? | 21:37 |
bilbo_b | https://dev1galaxy.org/viewtopic.php?id=2796 | 21:38 |
chillfan | ah thanks | 21:38 |
bilbo_b | but to keep it interesting I first had still ascii in sources.list and after reboot no ethernet interface as I mentioned this afternoon here, but noone replied | 21:39 |
bilbo_b | did then download all necessary packages on another machine and fed them through a usb-disk. now at least eth0 is again connected to the net | 21:39 |
bilbo_b | when not booting into rescue mode, it hangs after "starting openbsd sshd". maybe I need to configure a console login somewhere? | 21:40 |
bilbo_b | on serial? | 21:40 |
chillfan | So you came the long way round. Hah, fsmithred gave me some links earlier about this. I need to write some docs up for it, so.. just wondered which is working for people | 21:40 |
bilbo_b | apart from my mistake which took me some hours to fix it seems to have worked quite nicely. | 21:41 |
bilbo_b | maybe apart from no login on serial console | 21:41 |
fsmithred | bilbo_b, for the ssh hang, hit some keys to create entropy | 21:41 |
fsmithred | or install haveged (although I'm not certain that will fix it in a regular install) | 21:42 |
bilbo_b | ssh is running and I can login from another machine. I just dont get a login on serial console | 21:42 |
bilbo_b | what I have not tried yet is installing libelogind0 | 21:43 |
bilbo_b | maybe :) | 21:43 |
chillfan | Well libelogind0 isn't connected to sshd. But +1 for haveged. Usually it helps, provided it starts before the program you need to have entropy | 21:43 |
bilbo_b | the sshd server is running and accepting logins | 21:44 |
gnarface | you need an extra kernel command-line param for serial console, too | 21:44 |
chillfan | ah I read it wrong | 21:44 |
bilbo_b | I have in grub.cfg serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1 | 21:45 |
bilbo_b | terminal_input serial | 21:45 |
bilbo_b | terminal_output serial | 21:45 |
gnarface | i don't think those are them | 21:45 |
bilbo_b | it worked in buster, I don't see obvious changes in grub.cfg from buster | 21:45 |
chillfan | check if serial is setup in /etc/innittab .. I forget exactly how but you need the serial lines configured in there if i remember right | 21:46 |
gnarface | oh yea, there too | 21:46 |
chillfan | something like this | 21:47 |
gnarface | but fyi none of hotse command lines params you just pasted were the ones in question | 21:47 |
gnarface | of those* | 21:47 |
chillfan | # Example how to put a getty on a serial line (for a terminal)##T0:23:respawn:/sbin/getty -L ttyS0 9600 vt100#T1:23:respawn:/sbin/getty -L ttyS1 9600 vt100 | 21:47 |
bilbo_b | ok, thx | 21:47 |
chillfan | but eh, not misformatted like that. It's near the bottom somewhere | 21:47 |
bilbo_b | yay, works like a real Unix (tm) system, init q and the login appeared, thank you guys. | 21:48 |
chillfan | you're welcome | 21:49 |
chillfan | check also the docs for your device if there are any, to set the right baud rate | 21:49 |
bilbo_b | I already set 115200 and it works perfectly. | 21:49 |
chillfan | ah great | 21:49 |
bilbo_b | btw. when is Beowulf planned to be declared stable? | 21:49 |
fluffywolf | when it's stable. :) | 21:50 |
chillfan | when it's ready :P | 21:50 |
bilbo_b | you don't plan release dates (even though reality may interfere with plans)? | 21:50 |
fluffywolf | at some point you have to decide you really don't know, and saying a date before you know is just likely to cause issues. | 21:51 |
bilbo_b | ok, understand | 21:52 |
chillfan | Well same as in Debian we don't issue ETAs, but I would say an ISO release is a good milestone | 21:53 |
chillfan | just imho that is | 21:53 |
fsmithred | beowulf IS stable, but we can't have a release until we have isos to release. | 21:57 |
fsmithred | stable != Stable | 21:57 |
bilbo_b | sounds fair :) | 21:57 |
fsmithred | even though it is | 21:57 |
AntoFox | \o/ | 21:57 |
bilbo_b | is docker supposed to run the same as in buster? | 21:58 |
fsmithred | I think there are only parts of it in the repo | 21:58 |
bilbo_b | for buster I had a sources.list entry for docker-ce | 21:59 |
fsmithred | if it does not require systemd, it should work | 21:59 |
fsmithred | beowulf will be recognized as debian by third-party software | 21:59 |
chillfan | Are we able to import the devuan keyring (e.g package devuan-keyring) using gpg somehow? | 22:00 |
bilbo_b | it is the "if" I am not sure of | 22:00 |
bilbo_b | I did it with dpkg -i as described in the above url to dev1galaxy | 22:01 |
bilbo_b | @chillfan | 22:01 |
fsmithred | chillfan, I think so | 22:01 |
chillfan | try apt-cache depends <pkg_name> | 22:02 |
chillfan | I guess we need to download the package or wget it if importing with gpg though. I tried just gpg --recv-key but it wouldn't import it | 22:03 |
fsmithred | what key id did you use? | 22:04 |
chillfan | Ah, I've forgotton by now.. apt prints the message "could not find key <key_id>" | 22:04 |
fsmithred | wget the deb package and install it | 22:05 |
chillfan | alright, will try it | 22:05 |
chillfan | the other way it can be done: apt-get update --allow-insecure-repositories then; apt-get install devuan-keyring --allow-unauthenticated | 22:06 |
chillfan | but the security warnings are pretty annoying | 22:06 |
bilbo_b | is there anything necessary to access an ip-port from a remote machine? one of my docker-containers opens port 8080, the software in the container seems to run fine and reach to the internet, but I can not reach its port 8080 from other machines on my net | 22:06 |
bilbo_b | I mean in addition to what I would do on buster with systemd? | 22:07 |
fsmithred | I don't think so | 22:08 |
bilbo_b | probably a layer8 problem to debug | 22:08 |
buZz | heh | 22:08 |
fsmithred | no firewall on that machine? | 22:08 |
chillfan | IDK about docker but some software doesn't have transport layer and needs xinetd | 22:08 |
chillfan | but it would be unusual in something as modern as docker | 22:09 |
bilbo_b | not that I remember configuring one and up until this afternoon things worked. so likely something that happend when migration to devuan. | 22:10 |
bilbo_b | would it hurt to install xinetd and try? | 22:11 |
chillfan | nope, you can just purge it after if you want | 22:11 |
bilbo_b | docker is not very modern wrt ip, e.g. ipv6 is not very automatic within docker yet | 22:11 |
chillfan | worth remember, systemd replaces many services in your system | 22:12 |
chillfan | so when you move over, you need to find what to use instead | 22:13 |
chillfan | I would also check if you have static ip setup now | 22:13 |
bilbo_b | i know I do | 22:13 |
bilbo_b | thx | 22:13 |
bilbo_b | which other daemons or subsystems are commonly missing when removing systemd? | 22:14 |
chillfan | the good thing is you can use any inetd implementation you want | 22:14 |
bilbo_b | is the a doc on what to check? | 22:14 |
chillfan | that I don't know, I'm blissfully ignorant of systemd. But it would be handy to know mind you | 22:14 |
gnarface | bilbo_b: dunno about a doc on what to check, but make sure you have lm_sensors | 22:15 |
chillfan | dhcp client is one thing | 22:15 |
gnarface | bilbo_b: oh it's actually lm-sensors. also make sure you have acpid and acpi-support-base | 22:15 |
gnarface | bilbo_b: oh, and yea a dhcp client probably (though your network setup might have included that already) | 22:16 |
chillfan | and also net-tools as in the link you posted | 22:16 |
bilbo_b | net-tools was already installed in debian | 22:16 |
chillfan | isc-dhcp-client, inetutils-ping (maybe?) | 22:16 |
chillfan | ah, maybe it's netbase that's missing but you probably got that when migrating anyway | 22:17 |
chillfan | does systemd do sensors now? | 22:18 |
bilbo_b | inetutils-ping would remove iputils-ping, is that what I want? | 22:18 |
bilbo_b | dont need dhcp-client, lm-sensors was already installed | 22:18 |
chillfan | well you can choose between them, iputils-ping is smaller if i rememeber | 22:18 |
gnarface | yes, systemd axed sensors right away | 22:18 |
chillfan | Oh yeah, I heard they had indeed done an inetd replacement | 22:18 |
chillfan | so probably you should try xinetd or inetd.. or whatever else you can find that does the same | 22:19 |
gnarface | bilbo_b: mostly what's different if you weren't using all the window-manager features that have been integrated into systemd is just the permissions setup | 22:19 |
bilbo_b | i installed xinetd | 22:19 |
bilbo_b | it was a console-only server, no GUI | 22:19 |
chillfan | you'll have to configure it somewhere in /etc/xinetd/ I think | 22:19 |
chillfan | which port you need | 22:20 |
bilbo_b | i have docker expose port 8080 from one container to the host | 22:20 |
bilbo_b | so ip-of-host:8080 should be reachable from other machines, but is not | 22:20 |
chillfan | restarted xinetd? | 22:20 |
bilbo_b | connection refused | 22:20 |
bilbo_b | I rebooted the system entirely | 22:21 |
bilbo_b | after xinetd-install (without any manual configuration) | 22:21 |
chillfan | ah. Well this is only an educated guess. You could probably get a better answer from the docker folks | 22:21 |
chillfan | I've never used it | 22:22 |
bilbo_b | wasn't the services stuff so that a server would be started, when a port got accessed? here docker wants to provide a port to the outsideworld without anyone asking for it | 22:22 |
* gnarface doesn't know wtf is wrong with docker, either | 22:24 | |
bilbo_b | don't find much on the interwebz regarding devuan and docker :( | 22:25 |
specing | bilbo_b: what do you need docker for? | 22:26 |
fluffywolf | Containers are a horrible idea in general. | 22:26 |
gnarface | bilbo_b: there is something wrong with docker. there is a fix but i don't know it. you have to use an older version, or get a patch or something... | 22:26 |
specing | they aren't, fluffywolf | 22:26 |
gnarface | bilbo_b: i just remember this comes up every time | 22:27 |
bilbo_b | well "need" is a strong word :) | 22:27 |
DonkeyHotei | like any idea, containers can be abused. abusing them is what's horrible | 22:27 |
fluffywolf | specing: yes, they are. Shared libraries are one of the best and most important ideas in modern software. Ignoring them and bundling per-application libraries is an utterly idiotic idea that should die. | 22:28 |
specing | fluffywolf: you don't have to bundle libraries | 22:28 |
specing | thats a bastardisation of containers | 22:28 |
bilbo_b | I just have 2 machines with 1 running 12 containers, the other 4 and would not mind, if I could keep it that way and not find other ways to fulfill the potential not interchangeable dependencies otherwise | 22:28 |
fluffywolf | and how many contains do you know that do not bundle libraries? | 22:28 |
fluffywolf | containers | 22:28 |
specing | fluffywolf: all the ones I make | 22:29 |
DonkeyHotei | specing: that's macOS in a nutshell, actually | 22:29 |
specing | but I use LXC, not docker | 22:29 |
bilbo_b | if you have the same shared libraries in different containers they are only there once | 22:29 |
chillfan | if docker has an older supported release I would try that | 22:29 |
fluffywolf | bundled libraries are how you end up with an insecure maintenance nightmare, unless you start rebuilding containers yourself... | 22:30 |
specing | docker started off as a good idea, as single-program isolation | 22:30 |
bilbo_b | the container seems to run fine and communicates with the outside world, I just can not access it from my laptop with a browsser | 22:30 |
specing | but docker was taken over as a distribution method | 22:30 |
specing | fluffywolf: so don't use docker | 22:30 |
specing | or rather | 22:30 |
fluffywolf | I don't. :P | 22:30 |
specing | Or rather | 22:31 |
specing | use it, but only make your containers | 22:31 |
bilbo_b | I don't want to convince anyone of docker, would just like for convenience to keep using it | 22:31 |
bilbo_b | most of my containers are selfmade | 22:31 |
chillfan | you might try on #docker to see if they have a solution for running without systemd | 22:32 |
gnarface | people know the answer to this, they might be resistant to give it out though because they're systemd advocates | 22:33 |
bilbo_b | thanks | 22:34 |
chillfan | that's likely. Normally I would say remove the offending software and try something else, but.. there's enough of that going around :p | 22:34 |
bilbo_b | ah, devuan brings its own docker package, named docker.io | 22:37 |
bilbo_b | will try removing docker-ce and installing docker.io from beowulf source | 22:38 |
bilbo_b | now i have the devuan-docker.io package running and still the same, container running smoothly, getting outside internet access, but it's port is not reachable :/ giving up for tonight I guess | 22:43 |
chillfan | Try again without xinetd since you're not sure if it's needed | 22:45 |
gour | eveningi | 22:45 |
gour | *evening | 22:46 |
chillfan | let us know if you solve it, we'll be likely to tell others the solution | 22:46 |
bilbo_b | will do, thank you | 22:50 |
gour | just to say 'hello' to everyone and big "thank you" for all devuan devs!! my main machine is upgraded from manjaro/gnome to devuan/xfcer running ceres. in last few days i had hard time booting and doing shutdown on my machine which was today stuck in some kind of kernel loop not enabling me to even launch devuan install. now everything is fine - both boot and shutdown and there is no more bloody systemd. hope to be able to contribute back somehow to help | 22:53 |
gour | keep devuan rolling for a long time ;) | 22:53 |
gnu_srs | tom-mediapc: Hi, can you please try the proposed solutions in #261. Anyway, can you please explain how to set up one partition to be encrypted, as in: https://paste.debian.net/1128429/ | 22:58 |
chillfan | Oh we definitely intend to | 22:59 |
bilbo_b | ah, seems the application I was trying to connect to was not started, because previously in docker I got error-messages that modprobe could not find e.g. iptable_mangle :/ | 23:07 |
bilbo_b | manually modprobing (outside the container) worked | 23:09 |
chillfan | ah, drop iptable_mangle in /etc/modules | 23:09 |
chillfan | that will make it persist over reboots | 23:09 |
bilbo_b | surprised that was not necessary previously | 23:10 |
bilbo_b | now my application works, no problem with docker (at the moment) | 23:10 |
bilbo_b | chillfan: regarding your question why I had eth0 in buster, it was upgraded from debian 7 to 8 to 9 to 10 ;) so they kept the names, only when i replaced hw and got new mac-addresses, i got the new en... names | 23:20 |
chillfan | ah, normally you would replace it with maintainers new version when uprading | 23:21 |
chillfan | unless you had manual changes | 23:21 |
fsmithred | bilbo_b, if you want to boot devuan with the "predictable" names, add 'net.ifnames=1' to the boot command | 23:22 |
bilbo_b | thank you, prefer the eth0, e.g. because etherwake has that interfacename hardcoded as default | 23:23 |
bilbo_b | chillfan: I believe the change to predictable names happened from debian 8 to 9 for new interfaces, previously configured interfaces were kept. and the upgrade to 10 also did not change previously noted interfaces (with recorded mac-adresses in the /etc/udev/rules.d/ file | 23:26 |
bilbo_b | I hated predictable but hard to remember device names when Solaris started this stuff 20 years ago :( | 23:27 |
chillfan | I don't think much of the long and weird naming | 23:31 |
* bilbo_b is not a fan | 23:32 | |
chillfan | Also of ipv6 hex.. with enough address space for the solar system.. | 23:32 |
fsmithred | I don't know what solaris had, but to predict the new names, I think you have to open the box and count slots | 23:32 |
bilbo_b | ;) | 23:32 |
chillfan | it's ok to be technical, but it should be slightly intuitive too | 23:33 |
bilbo_b | it was similar in solaris, they had one mapping file for everything (disk device, all sbus or later pci cards) and something when wrong and you had to edit this file, chances were good you could reinstall your operating system afterwards if you did not know 150% what you were doing | 23:34 |
bilbo_b | sorry, should have read "when something went wrong". its late... | 23:34 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!