antranigv | hey all | 00:10 |
---|---|---|
antranigv | wanted to ask in #debian, but people here understand me better | 00:10 |
antranigv | I am installing a debian/devuan chroot. there's this new thing called usrmerge | 00:10 |
antranigv | but it kinda breaks the permissions, /bin/bash becomes rwx------ | 00:10 |
antranigv | then my user is not able to login | 00:10 |
antranigv | is this common or am I dumb? | 00:11 |
HimeHaieto | I think it's more that devuan seems to be a bit of a trigger word in that channel... | 00:29 |
HimeHaieto | you aren't the first to have issues with usrmerge though, I know that much | 00:30 |
HimeHaieto | maybe try providing the exact commands you're using to bootstrap the devuan chroot | 00:31 |
HimeHaieto | and the sources you're using | 00:31 |
HimeHaieto | there also appears to have been some discussion recently on usrmerge in the #devuan-dev channel, but I didn't really follow it too much | 00:33 |
HimeHaieto | seemed like it was providing hassle to a number of people | 00:33 |
fsmithred | antranigv, I played with usrmerge today. It broke my system when I installed it in ceres. It did not break when I installed it in daedalus and upgraded. | 00:42 |
fsmithred | Oh, and btw, I was unable to do a debootstrap install of ceres or excalibur, even with adding "--merged-usr" to the debootstrap command. | 00:43 |
HimeHaieto | ooh, there's a --merged-usr option now? | 00:43 |
fsmithred | *upgraded daedalus to ceres | 00:43 |
HimeHaieto | I'll have to make a note of that | 00:44 |
fsmithred | yeah, I was forced to look at the man page. | 00:44 |
HimeHaieto | last time I used debootstrap was I think maybe for jessie...but maybe even further back than that | 00:44 |
HimeHaieto | debootstrap is apparently what antranigv was using to try to set up the chroot | 00:44 |
HimeHaieto | antranigv: which release where you trying to bootstrap? | 00:45 |
onefang | mmdebstrap might work better. | 00:46 |
antranigv | HimeHaieto chimaera | 00:47 |
antranigv | I am actually running a Devuan chroot on a FreeBSD host | 00:47 |
HimeHaieto | antranigv: maybe try bootstrapping daedalus, and then upgrading that to chimaera? | 00:58 |
fsmithred | that's backward | 00:59 |
HimeHaieto | haha, meant beowulf XD | 00:59 |
fsmithred | we're going in alphabetical order | 01:00 |
HimeHaieto | been in the process of trying to upgrade to daedalus, so I guess I got switched around a bit lol | 01:00 |
fsmithred | debootstrap of daedalus should work. | 01:00 |
HimeHaieto | perhaps, but I assume there's a reason they wanted a fresh bootstrap of the *previous* release of devuan instead of current stable | 01:01 |
antranigv | :)) | 01:01 |
fsmithred | chimaera should work, too. You don't need usrmerge with chimaera or daedalus | 01:02 |
fsmithred | and the changed permissions on /bin/bash are weird | 01:03 |
fsmithred | but maybe usrmerge in chimaera is broken | 01:04 |
antranigv | մեյբի | 01:04 |
antranigv | maybe | 01:04 |
fsmithred | I added it in daedalus and ceres today. Permissions on bash are normal. | 01:04 |
antranigv | sorry, armenian | 01:04 |
fsmithred | your english seems pretty good | 01:05 |
fsmithred | oh, that. | 01:05 |
HimeHaieto | I still haven't installed usrmerge, and if anything what I've been hearing lately has made me ever more hesitant to | 01:05 |
HimeHaieto | is it required now by daedalus? | 01:06 |
fsmithred | no | 01:06 |
fsmithred | excalibur yes | 01:06 |
fsmithred | debian trixie yes | 01:06 |
HimeHaieto | oof...but that shouldn't affect me for a while so long as things can work without it | 01:06 |
HimeHaieto | I pretty much plan to go straight from daedalus to freia | 01:07 |
HimeHaieto | full system upgrade every two years is too frequent | 01:08 |
antranigv | HimeHaieto any chance I can bring you to FreeBSD? :P | 01:08 |
fsmithred | with my current limited experience adding usrmerge, I think you'll need to add it before upgrading to excalibur. | 01:08 |
HimeHaieto | antranigv: eh? -.^ | 01:08 |
antranigv | oh never mind, we have upgrades every 2 years :D sometimes ever 1.5 years, depending on what it is | 01:08 |
HimeHaieto | my main use case here is for the server platform I manage | 01:10 |
HimeHaieto | if it were just some personal laptop or something, maybe pulling the rug out every couple years would be just fine | 01:10 |
HimeHaieto | I'm one of the reason things like long term releases/support exist | 01:10 |
HimeHaieto | also just so painful to have to go through the ordeal, that I wouldn't want to have to do it any more often than the every-other cycle | 01:11 |
golinux | Yes . . . behaving for the machine is painful . . . | 01:51 |
HimeHaieto | omg I just realised that as of bullseye even squid now has a dependency on systemd...why??? | 08:08 |
HimeHaieto | sometimes feels like at this rate, pretty soon hello world won't compile without linking to systemd... | 08:09 |
al1r4d | hello | 08:12 |
brocashelm | himehaieto: as debian trixie is "progressing", excalibur will require a lot more effort to de-systemd-ize than previous devuan releases; never mind usrmerge, which you can at least work around with careful symlinks | 08:43 |
HimeHaieto | I'm not sure if I've missed anything wrt usrmerge regarding it being much more than a technical issue, but how concerned should I be with what you appear to be implying regarding excalibur and beyond? | 08:44 |
HimeHaieto | what's going on that might be even more dramatically troublesome regarding everyone's favourite linux virus? | 08:47 |
brocashelm | very concerned; even a month or two after bookworm's release, libgudev was already breaking in testing/unstable for non-systemd distros: https://dev1galaxy.org/viewtopic.php?id=5822 | 08:48 |
brocashelm | https://github.com/eudev-project/eudev/issues/249#issuecomment-1624401542 | 08:50 |
brocashelm | which led to packaging issues, forcing users to stay on eudev 243 | 08:50 |
brocashelm | and systemd has that "bsod" and other "improvements" that will manifest itself in more packages than before | 08:51 |
HimeHaieto | bsod? | 08:51 |
brocashelm | debian is quickly turning into arch | 08:51 |
brocashelm | "blue screen of death" | 08:51 |
brocashelm | basically a glorified kernel panic | 08:51 |
HimeHaieto | I know that much | 08:51 |
brocashelm | best to take it to OT since this is for support, though | 08:51 |
brocashelm | join #devuan-offtopic | 08:51 |
HimeHaieto | but I generally don't try to let systemd infest my life enough to keep up on news of its awfulness | 08:51 |
brocashelm | well, a lot of work is put into blacklisting/forking packages that directly depend on systemd | 08:52 |
brocashelm | you're safe if you stick to daedalus or older releases | 08:52 |
brocashelm | but excalibur and beyond will present some challenges for a lot of non-sysd linux | 08:52 |
brocashelm | hence why it's encouraged to volunteer to test new builds | 08:53 |
tavasti | how do I update my up-to-date Daedalus to Ceres? edit apt sources and apt dist-upgrade? | 09:06 |
brocashelm | tavasti: first, make a backup of your system, then change daedalus to ceres (only one repo; disable the rest), run apt update && apt dist-upgrade | 09:42 |
tavasti | brocashelm, test system, so no worries if it will fail :-) | 09:43 |
brocashelm | have fun :) | 09:44 |
tavasti | thanks! | 09:44 |
devurandom | Hi, I just installed devuan(runit) on my system but it seems to hang on "Starting network connection manager: NetworkManager" on boot. I can still switch to other TTYs, what could the problem be? | 13:03 |
devurandom | I was using a netinstall for Daedalus | 13:04 |
fsmithred | devurandom, did you install runit-services? (I think that's the right name for the package.) | 13:06 |
fsmithred | oh, that probably won't help. I don't see n-m listed in that package. | 13:08 |
devurandom | fsmithred no I did not install any additional packages. I also seem to have a network connection and my network seems correctly configured | 13:09 |
gnarface | devurandom: need a better description of the problem, if you can switch to other TTYs, boot completed, so it didn't "hang on boot" ... exactly where are you seeing this error? after the GUI starts? | 13:10 |
fsmithred | on startup. I see it sometimes, too. | 13:10 |
gnarface | hmm | 13:10 |
fsmithred | not necessarily with runit | 13:10 |
gnarface | well it could be a dhcp timeout or a dns issue... both could be as easily caused by a upstream misconfiguration as a local one | 13:12 |
devurandom | Oh nevermind I am dumb. Turns out there were just logging messages and I was put at the login screen. I did not see the "devuan login:" prompt | 13:12 |
gnarface | if it's wifi it could be missing firmware maybe... | 13:12 |
devurandom | I did select a desktop environment so I assumed I would get a login screen as I have had on debian | 13:13 |
fsmithred | lol | 13:13 |
gnarface | odd that you didn't, but maybe runit is missing the start script for it | 13:13 |
fsmithred | which desktop and which diplay manager | 13:13 |
fsmithred | which desktop and which diplay manager? | 13:13 |
fsmithred | oops | 13:14 |
devurandom | xfce4 | 13:14 |
fsmithred | shouldn't matter if there's a run script - it'll use the init script instead | 13:14 |
devurandom | When I run "startxfce4" I get errors from xorg about not being permitted to enable I/O ports "0000-03ff" | 13:14 |
devurandom | and for the root user I get "nvc0_screen_create:1072 - base screen init failed: -19" | 13:16 |
gnarface | is this a nvidia card? | 13:16 |
devurandom | Yes | 13:16 |
gnarface | looks like a nouveau error | 13:16 |
gnarface | you might need to install the nvidia drivers from non-free-firmware | 13:16 |
gnarface | or maybe you just need to be in the "video" group | 13:17 |
gnarface | not sure | 13:17 |
gnarface | try the video group first | 13:17 |
gnarface | if you're not in it by default | 13:17 |
devurandom | my user is in the video group but still gets the error | 13:19 |
devurandom | I guess I will try the nvidia driver | 13:19 |
gnarface | they've moved such packages from "non-free" to a new section called "non-free-firmware" as of daedalus | 13:19 |
devurandom | Turns out it was a problem with "nouveau" I installed nvidia-driver and booted up no problem, weird since I never had that problem any other time. Thanks for the help | 13:26 |
gnarface | it's possible if you installed another distro (or even this one using one of the live images) you would have gotten the nvidia drivers by default | 13:28 |
gnarface | it's also just as possible that the nouveau drivers broke recently; they're not very good | 13:28 |
gnarface | or if you used a different desktop environment, the other one might just not have needed whatever feature nouveu is choking on today | 13:29 |
rolfh | Somebody here who has usrmerge applied in Daedalus? Any side effects seen? | 15:29 |
gnarface | heh, yea there's side effects, they were just talking about it the other day | 15:35 |
gnarface | i forget the details | 15:35 |
fsmithred | rolfh, I added usrmerge in daedalus yesterday with no apparent problems. But that's in a small system with openbox and lxpanel and without dbus or policykit stuff. I then upgraded it to ceres without problems. | 16:09 |
fsmithred | Adding usrmerge in ceres itself did not work for me. | 16:09 |
onefang | I'm still experimenting in a VM with Daedalus and usrmerge. So far no problems. | 16:12 |
fsmithred | I'm heading out. Encrypted system didn't boot. Dropped to initramfs shell. I'll play with it later. | 16:13 |
fsmithred | hm, that should have been in the -dev channel | 16:14 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!