| golinux | <onefang> I'm still wondering what these right reasons for resisting usrmerge actually are. | 01:00 |
|---|---|---|
| Xenguy | More and more, resisting Red Hat/IBM et al. seems like the sensible default position, if you ask me | 01:04 |
| golinux | Technically there may not be one but limiting the freedom to 'do it your way' is a door that should not be permanently locked by those wanting to control and limit choice. Isn't that why we are here? | 01:05 |
| golinux | That is not saying that Devuan should do that but I hope that someone does . . . | 01:06 |
| golinux | And it wold be a good thing to do . . . | 01:06 |
| Xenguy | My impression is that somehow usrmerge will get rammed through, but you win some and you lose some | 01:06 |
| golinux | Xenguy: Exactly . . . nothing good comes from that camp . . . | 01:07 |
| golinux | We are just expendable pawns on the corporate chessboard of domination . . . | 01:08 |
| mason | onefang: So, the issue is that Debian's usrmerge breaks things. It'd be largely pointless aside from that, but as it breaks things it's a problem. | 02:28 |
| mason | onefang: The dpkg team has recommended better ways to do it but they've been ignored, resulting in a uniquely convoluted mess. | 02:28 |
| mason | Fedora, EL, SuSE, Ubuntu, Void... None of them *break* because of usrmerge. | 02:29 |
| onefang | mason: What things break? The dpkg team have provided a un-usrmerge script, so people still have choices. | 05:28 |
| onefang | Though I'm part way through converting my usual install script to daedalus, then there will be lots of testing. it starts with a mmdebstrap install of usrmerge, I might find out the hard way. lol | 05:47 |
| gnu_srs | init-system-helpers is not forked by Devuan, right? 1.64 is the latest OK version. 1.65 has:Depends: usrmerge | usr-is-merged | 13:54 |
| onefang | init-system-helpers is forked by Devuan. | 13:55 |
| mason | gnu_srs: It's not but it might be a good candidate. | 15:03 |
| mason | Oh, if it's forked then I guess locking the version wouldn't matter so much. | 15:03 |
| mason | onefang: The dpkg team lists a bunch of breakage: https://wiki.debian.org/Teams/Dpkg/MergedUsr For my personal case, the first time I did a test install of a usrmerged system, someone on IRC asked a support question and I got bad data back because I trusted "which", which of course was mangled by usrmerge. | 15:05 |
| mason | So, they asked if something was in the new stable, and I said "yeah, I've got it installed on this box," and then dpkg -S said "dpkg-query: no path found matching pattern foo" | 15:05 |
| mason | and then following symlinks I realized what had happened, and seeing no actual benefit to usrmerge, I figured out how to disable it. | 15:06 |
| mason | It was too late for that box, so I just reinstalled it. | 15:06 |
| onefang | I had read that already, and if I recall it boiled down to "we don't like it, we wont support it", so no wonder dpkg is broken. | 15:09 |
| gnu_srs | If your system has merged /usr by accident, you can always revert that with: dpkg-fsys-usrunmess | 15:51 |
| fsmithred | cool! unmerge un-merge ^^^ (now indexed for my log) | 15:56 |
| gnu_srs | I see at https://git.devuan.org/devuan/init-system-helpers/src/branch/suites/unstable the init-system-helpers-1.65.2 is forked. | 16:04 |
| gnu_srs | I don't see any branch for daedalus, chimaera, beowulf only: jessie, ascii, unstable, experimental. Only unstable is forked recently. | 16:04 |
| * golinux dreads the thought of moving to Daedalus | 16:04 | |
| LeePen | gnu_srs: that isn't the way branches work. It is forked in all suites. | 16:16 |
| LeePen | Packages migrate from unstable to testing without being rebuilt. There is only a suite branch if the src has been built for that specific suite. | 16:18 |
| LeePen | Try rmadison from the devscripts package to query dak on formed packages and versions in suites | 16:26 |
| LeePen | You need to specivy the dakweb url: | 16:26 |
| LeePen | https://api.pkginfo.devuan.org/madison | 16:27 |
| gnu_srs | Thanks LeePen, got it! | 19:42 |
| Centurion_Dan | Sorry I missed the meeting. | 23:26 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!