gnu_srs | (11:09:04) gnu_srs: Problem is that dbus_1.14.10-3 now depends on usr-is-merged, which I don't want to install (or usrmerge): This is a problem also for devuan/linux. | 10:40 |
---|---|---|
gnu_srs | For all people not wanting usrmerge. Is this package modified by Devuan? | 10:41 |
brocashelm | doesn't look like it is: https://pkginfo.devuan.org/cgi-bin/package-query.html?c=package&q=usrmerge=38 | 10:42 |
brocashelm | anything that's a devuan package usually has "devuan" appended to the version | 10:42 |
brocashelm | example for dbus in daedalus: 1.14.10-1~deb12u1devuan1 | 10:43 |
brocashelm | i got usrmerge and usr-is-merged blacklisted on all my systems, and apt can still install ceres' dbus (1.14.10-3devuan1) | 10:46 |
brocashelm | >Depends: lsb-base, dbus-bin (= 1.14.10-3devuan1), dbus-daemon (= 1.14.10-3devuan1), dbus-system-bus-common (>= 1.14.10-3devuan1), libc6 (>= 2.34), libdbus-1-3 (= 1.14.10-3devuan1), libexpat1 (>= 2.1~beta3) | 10:46 |
brocashelm | so it's safe to upgrade | 10:46 |
gnu_srs | thanks: Do you have a link to the modifications made by Devuan for dbus (1.14.10-3devuan1)? | 10:51 |
brocashelm | all the information is provided here (feel free to search around): https://pkginfo.devuan.org/cgi-bin/package-query.html?c=package&q=dbus=1.14.10-3devuan1 | 10:53 |
brocashelm | https://git.devuan.org/devuan/dbus | 10:55 |
brocashelm | ^ this for the git | 10:55 |
brocashelm | https://git.devuan.org/devuan/dbus/src/branch/suites/unstable | 10:55 |
gnu_srs | Thanks again: Seems like LeePen modified 1.14.10-3: d/control: revert usr-is-merged dependency. | 10:57 |
gnu_srs | I think we really need to start up a Devuan/Hurd distro. | 10:58 |
gnu_srs | So far only a few Debian/Hurd packages depends on usrmerge/usr-is-merged: hurd, dbus, initscripts. But that list will probably grow. | 11:00 |
gnu_srs | LeePen: WDYT? I can provide a buildd if needed. | 11:00 |
gnu_srs | The only release needed for Hurd is sid/ceres. | 11:04 |
LeePen | gnu_srs: it may be short lived. I am not convinced Devuan as a whole will manage to avoid usrmerge in the medium term, even though we are at the moment. | 11:06 |
LeePen | (Much as I would like to) | 11:06 |
LeePen | Amprolla would also need work (I think) as Debian Hurd is a port rather than release (IIRC). | 11:09 |
LeePen | Maybe propose it at the meet this week? | 11:10 |
gnu_srs | Yes, please do! | 11:11 |
gnu_srs | Oh, you mean that I should joint the meeting? We'll see. | 11:12 |
LeePen | Yes, I meant you! I am not available, I a afraid. | 11:16 |
gnu_srs | LeePen: Which changes are needed for amprolla? | 11:21 |
LeePen | I don't know exactly, but Hurd isn't on deb.debian.org/debian, IIRC | 11:23 |
LeePen | WHich is where amprolla merged all the other archs. | 11:23 |
LeePen | meres | 11:23 |
LeePen | Even merges | 11:23 |
brocashelm | i'd be more interested in a devuan/bsd | 11:25 |
brocashelm | both are individually big projects, tbh | 11:25 |
gnu_srs | LeePen: Here is where Debian GNU/Hurd is: deb http://ftp.ports.debian.org/debian-ports sid main;deb http://ftp.ports.debian.org/debian-ports unreleased main | 11:29 |
onefang | I agree with brocashelm Devuan/BSD would be nice. | 11:30 |
onefang | And I still think it's just not worth the effort to fight usrmerge. Sure the dpkg mainainers think it's evil, so do lots of people. I haven't seen any actual problems, except for people constantly bitching about it. lol | 11:31 |
gnu_srs | onefang: I think IT IS worth fighting usrmerge! | 14:01 |
gnu_srs | BTW: I am now (and was in March) communicating with Guillem Jover about dpkg-fsys-usrunmess within dpkg! | 14:13 |
gnu_srs | onefang: Regarding BSD, Debian had kFreeBSD until very recently, but due to lack of interest it has now been removed completely. | 14:14 |
gnu_srs | (FTR: I was hosting the buildd kamp for Debian GNU/kFreeBSD many years but it broke due to a hardware fault. | 14:18 |
gnu_srs | Unfortunately nobody volunteered to help to recover the image!) | 14:18 |
onefang | gnu_srs: Well it's your effort to waste. shrugs | 16:57 |
sakrecoer | greetings everyone! Here's your weekly reminder of the weekly devuan meet! | 17:01 |
sakrecoer | pad is here: https://pad.dyne.org/code/#/2/code/edit/DtU8B5eo-3pG+q1grtJyYLW7/ | 17:01 |
sakrecoer | meet takes place November, Thursday 2nd @20:30UTC | 17:02 |
sakrecoer | November 2, Thursday, at 20:30 UTC to be more consistent :sweaty_smile: | 17:04 |
bb|hcb | ~. | 17:09 |
bb|hcb | onefang: I wouldn't say that usrmerge is evil. Moreover a pointless effort to please some corporate interest. The main problem are the bugs created during the transition. And what LeePen says is true - once Debian buildds become merged, we will no longer be able to avoid it. Like it or not | 17:14 |
bb|hcb | The alternative is to rebuild and maybe patch everything - not impossible but I do not see the people willing to and capable of doing that... | 17:15 |
onefang | I'm not seeing much in the way of bugs during the transition. None of them bit me. I think most of the reported bugs get fixed with "just do the symlinks thing and move on". | 17:19 |
onefang | Most of the pointless effort is in the wringing of hands about it. | 17:20 |
bb|hcb | Sadly, yes | 17:20 |
onefang | Sooo, don't worry, be happy. B-) | 17:22 |
DelTomix | backporting will take a hit at some point I think | 17:23 |
bgstack15 | I had one small problem in the systemctl-service-shim package but was able to update the code to get around it. | 17:30 |
bgstack15 | The systemctl script does some tricky /usr/bin versus /bin detection to avoid infinite recursion | 17:31 |
gnu_srs | I would still like to have a Devuan GNU/Hurd port. Preferably w/o usrmerge. Note GNU/Guix don't suffer from such problems. | 17:46 |
gnu_srs | bb|hcb: What is the problem for Devuan when Debian buildds becoming usrmerged? | 17:48 |
gnu_srs | (17:15:54) bb|hcb: The alternative is to rebuild and maybe patch everything - not impossible but I do not see the people willing to and capable of doing that...: Why? | 17:49 |
gnu_srs | What and why? | 17:49 |
gnu_srs | I'll try to participate somehow in tomorrows meeting. | 17:54 |
bgstack15 | I think we need to remember what Devuan is supposed to be: Debian with init diversity. Usrmerge is not a goal, unless we intend to make some official change to the plan. | 18:01 |
DelTomix | agree bgstack15 - usrmerge is an ill conceived and annoying decision - and creates some headache - but (afaict) its managable and should not hopefully compromise choice in the way that systemd does. Devuan's wisdom in containing its scope has no doubt contributed to its resilience (thus far). | 18:19 |
bgstack15 | I agree that it feels like change for the sake of change, but also that it barely affects me. I realize it can affect others more than me. | 18:20 |
gnu_srs | I can contribute more if Devuan continues to support non-usrmerge (and Hurd) | 18:32 |
golinux | Give to give not to get . . . | 19:28 |
gnu_srs | golinux: ? | 19:39 |
golinux | IMO, contributing should not be transactional. It has always been a labor of love, at least for me . . . To each his own, I guess . . . | 20:03 |
Centurion_Dan | usrmerge still is worth resisting as much as possible, but if they change the buildds to force it then it becomes impossible without going to a build from source model | 20:04 |
onefang | Why resist? | 20:04 |
Centurion_Dan | golinux: 100% | 20:04 |
Centurion_Dan | onefang: because a strong resistance makes them less likely to put lots of effort into such fruitless pursuits that are only to follow the corporate crowd. | 20:07 |
onefang | Are they following a corporate crowd? | 20:07 |
onefang | And I think most of their effort is done already. Most of the effort these days is in fruitless resisting. | 20:08 |
brocashelm | just worry about systemd for now | 20:14 |
brocashelm | the next debian release will have even more systemd parts that i'm not even sure how they're still going to claim to also "support" sysvinit | 20:14 |
brocashelm | elogind and eudev are just bandages placed on an open wound | 20:15 |
golinux | onefang: Doing the right thing for the right reasons is never ethically fruitless which is all that matters in the end . . . | 20:21 |
onefang | I'm still wondering what these right reasons for resisting usrmerge actually are. | 20:32 |
golinux | onefang: Good question . . . in the end, everything turns to dust . . . so why bother with any of it? It is all just froth and foam without substance . . . | 20:52 |
golinux | Seeking security and stability is a fool's venture because the sands are constantly shifting . . . | 20:54 |
golinux | I am exhausted with it . . . | 20:54 |
DelTomix | I would guess that everyone here wants to do the right things for the right reasons, it seems the questions is to establish or understand what is "right", and more specifically, what constitutes "resistance" - what can be done and to what effect? Also worth considering is where should that originate? Devuan is essentially downstream of it, Debian's maintainers are ones upstream. | 21:03 |
DelTomix | ^are >the< ones upstream | 21:05 |
bgstack15 | I'm amused that Debian gets stuck taking the ride that RHEL is shoving. More systemd, more freedesktop.org, more pulseaudio, dbus, etc. | 21:22 |
bgstack15 | I remember reading that Debian had a menu system before the freedesktop.org one.... | 21:22 |
onefang | My desktop has both the FDO menu and the Debian menu. | 21:23 |
brocashelm | debian has gone far and away from continuing unix philosophy; the closest thing to "unix philosophy" is slackware, and that already has elogind, avahi, pulseaudio, and grub2 shoved in by default | 21:30 |
brocashelm | that's why i say just focus on systemd | 21:31 |
brocashelm | even just replacing with a freebsd kernel will be a lot of work; hyperbola has been stuck trying to combine gnu with freebsd kernel for years already, never mind the licensing conflicts | 21:32 |
DelTomix | I don't care much about defaults - I see it mostly just as a subjective decision - as long as they can be overriden and don't block choice of altertnatives. | 21:33 |
brocashelm | for now, i've blacklisted usrmerge and usr-is-merged | 21:33 |
brocashelm | i blacklist subversive software that still happen to be available on the repo | 21:34 |
brocashelm | the future of linux is more rust, wayland, pipewire, and other such things | 21:35 |
onefang | I'm using usrmerge with no problems. | 21:36 |
DelTomix | tactically its a bit like blowing up a bridge between future and past, with scripts for example (such as sysvinit and all their bits and pieces), a legacy of zillions with #!/bin/thing will have to be rewritten or rely on symlinks, and then eventually new scripts with #!/usr/bin/bash won't work without modification or reverse symlinking on pre-usrmerge installations. One can still get back and | 21:48 |
DelTomix | forth through the valley but not as easily as when the bridge was there. | 21:48 |
mason | brocashelm: Are you blacklisting via pins? That ends up being fairly complete. | 21:54 |
mason | brocashelm: One thing I had to do in addition was pin init-system-helpers:any to a local repository. Half a sec and I'll have the version for you. | 21:55 |
mason | Got it. 1.64 appears to be the last safe version. | 21:57 |
mason | Anyway, you're in good company blacklisting it. There are a number of Debian developers right there with you. | 21:57 |
brocashelm | mason: apt pinning, yes | 21:59 |
brocashelm | i recently did this with any newer kernels than 5.19, because i've been having some swap issues with 6.4 and 6.5 won't even start my session (black screen) | 22:00 |
bgstack15 | mason: maybe those guys will end up being Devuan developers, if Debian because too unpalatable? | 22:08 |
mason | bgstack15: Yeah, I think about that somewhat often. The level of dismay there seems appropriate. | 22:10 |
DelTomix | Thats exactly what I've also been thinking | 22:10 |
brocashelm | i seem to recall one debian developer who was ousted from the project go straight into arch | 22:14 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!