#archlinux-ports | Logs for 2017-10-10

[00:50:05] -!- typikal has quit [Ping timeout: 240 seconds]
[00:51:24] -!- typikal has joined #archlinux-ports
[01:46:19] r00t^2 is now known as jth4n
[01:46:29] jth4n is now known as r00t^2
[01:47:01] -!- p71 has quit [Ping timeout: 240 seconds]
[01:52:06] -!- p71 has joined #archlinux-ports
[02:31:14] <tyzoid> fsckd: I didn't realise we were that close to implementing it
[02:31:32] <tyzoid> I thought that there was still a lot on the part of the software developers to allow that functionality
[02:33:08] <eschwartz[m]> tyzoid: nah, it is mostly really simple, and just a wrapper around `export SOURCE_DATE_EPOCH` everywhere
[02:33:18] <eschwartz[m]> reproducing it afterwards is slightly more involved
[03:42:57] -!- tyzoid has quit [Ping timeout: 260 seconds]
[03:44:48] -!- tyzoid has joined #archlinux-ports
[04:33:14] -!- typikal has quit [Ping timeout: 255 seconds]
[04:33:59] -!- typikal has joined #archlinux-ports
[06:16:06] -!- deep42thought has joined #archlinux-ports
[06:43:58] -!- tyzoid has quit [Ping timeout: 240 seconds]
[06:46:11] -!- tyzoid has joined #archlinux-ports
[06:46:58] <tyzoid> hey deep42thought
[06:58:11] <deep42thought> hi tyzoid
[07:06:42] <tyzoid> deep42thought: So I'm working on the i686 automated script
[07:06:57] <tyzoid> but it looks like the default arch32 releng build script puts the mainline mirrorlist in place
[07:07:15] <tyzoid> back when I made the october ISOs, I manually replaced this
[07:07:23] <tyzoid> but is this another build hook somewhere?
[07:07:31] <tyzoid> I mean - it should be installing our mirrorlist package
[07:08:38] <deep42thought> it extracts some information from the host system
[07:08:49] <deep42thought> I had to work around the cache dir already
[07:08:53] <deep42thought> maybe I missed the mirrorlist
[07:09:44] <tyzoid> okay. I haven't had time yet to sort through the hooks
[07:10:06] <tyzoid> though I had updated the mirrorlist in airootfs, but the working directory still generated the mainline mirrorlist
[07:10:39] <tyzoid> my workaround was to backtrack the iso build after it completed, replace the file, then rebuild the image
[07:11:37] <deep42thought> $ grep -n mirror build.sh
[07:11:37] <deep42thought> 98: curl -o ${work_dir}/${arch}/airootfs/etc/pacman.d/mirrorlist 'https://www.archlinux.org/mirrorlist/?country=all&protocol=http&use_mirror_status=on'
[07:11:47] <tyzoid> oh, right
[07:11:49] <deep42thought> this is the problem
[07:11:58] <deep42thought> I knew, I arrived there before
[07:12:03] <deep42thought> sorry for not fixing it then
[07:12:11] <tyzoid> Right when you pasted it, I realized I saw that code too
[07:12:12] <tyzoid> :/
[07:12:17] <tyzoid> It's no problem
[07:13:21] <tyzoid> deep42thought: Once this is done building, I'm going to throw up this i686 iso build up on my server if you wanted to test it/play around with it
[07:13:40] <tyzoid> It's getting a bit late for me, so I'll be off to bed in a min
[07:14:04] <deep42thought> ok, I'm off to work in a few minutes, too
[07:14:06] <deep42thought> good night!
[07:14:29] <tyzoid> figured
[07:15:00] <tyzoid> deep42thought: oh, btw, during the day, email is the best way to reach me
[07:15:19] <deep42thought> ok
[07:15:21] <tyzoid> I just keep this open in a screen, but I don't check it all that often
[07:15:32] <tyzoid> usually in the evenings
[07:15:34] <tyzoid> (my time)
[07:15:44] <deep42thought> your regular email?
[07:15:57] <tyzoid> yeah, or tyzoid@archlinux32.org
[07:16:01] <deep42thought> :-)
[07:16:01] <tyzoid> both go to the same place
[07:16:35] <deep42thought> ok (mine too), I'll leave now
[07:16:36] <deep42thought> cu
[07:16:40] <tyzoid> see ya
[07:16:41] -!- deep42thought has quit [Quit: Leaving.]
[07:56:06] -!- isacdaavid has quit [Quit: isacdaavid]
[08:21:36] -!- deep42thought has joined #archlinux-ports
[08:58:11] -!- deep42thought has quit [Quit: Leaving.]
[09:52:37] -!- deep42thought has joined #archlinux-ports
[11:46:39] <deep42thought> brtln: Should we expect situations like in bug https://bugs.archlinux.org to stay for ever / appear again (on purpose) or did felixonmars simply not yet fix that one?
[11:46:39] <phrik> Title: FS#54696 : [perl-list-moreutils] spurious old PKGBUILDs for extra-x86_64 and extra-i686 (at bugs.archlinux.org)
[12:08:52] <brtln> I forgot about it so I would expect that anybody could
[12:11:53] <deep42thought> my question aims more to: Is it (sometimes) desired to have multiple PKGBUILDs in repo/*/ even if the corresponding package does not exist in the repositories?
[12:17:47] <brtln> it's not, it's a mistake when architecture specific package becomes any or the other way around
[12:17:57] <deep42thought> ok
[12:18:05] <deep42thought> thanks for the clarification
[12:18:06] <brtln> it's not handled by archrelease at all
[12:18:24] <brtln> thus these issues
[12:19:04] <deep42thought> I have similar issues with my own repositories
[12:19:17] <deep42thought> once I change architecture fields, the old packages stay behind
[12:19:34] <deep42thought> well, it's rather packages than pkgbuilds - but it's similar
[12:27:33] <tyzoid> deep42thought: So update on the auto-build script
[12:27:53] <tyzoid> It seems to work as intended (in that, after running it several times for debugging - it didn't crash halfway through)
[12:28:00] <tyzoid> which is something that it tended to do before
[12:28:12] <tyzoid> but archiso32 by default uses the system mirrorlist
[12:28:33] <tyzoid> which would make it a mainline i686 iso
[12:28:51] <tyzoid> so we should package a custom pacman.conf in archiso32 which has our mirrors instead
[12:29:06] <tyzoid> that said - we still haven't resolved the -dual cds needing mirror changes
[12:29:25] <tyzoid> should I have a script that downloads the packages from the appropriate sources, and compiles a local repo via repo-add?
[13:09:07] <deep42thought> maybe, compiling a combined cache (without *.db.tar.gz and *.files.tar.gz) is enough
[13:09:22] <deep42thought> on the other hand, I really don't see, why we should not provide a combined mirror
[13:24:59] -!- Faalagorn has joined #archlinux-ports
[13:30:36] -!- Faalagorn has quit [Quit: Leaving.]
[13:31:20] -!- Faalagorn has joined #archlinux-ports
[13:31:32] <deep42thought> tyzoid: I added an x86_64 directory to the master mirror with symlinks for core, community and extra
[13:31:43] <deep42thought> can you try, if the iso builds from that?
[13:35:28] <deep42thought> also, please try this patch https://pastebin.com to create our own mirror list
[13:35:30] <phrik> Title: diff --git a/configs/releng/build.sh b/configs/releng/build.sh index d602d78..1 - Pastebin.com (at pastebin.com)
[13:36:37] -!- Faalagorn has quit [Remote host closed the connection]
[13:50:06] <deep42thought> I just noticed, that the create-mirrorlist-command is run separately for each $arch, so it's trivial to create different mirrorlists for different architectures
[14:26:57] -!- Faalagorn has joined #archlinux-ports
[14:43:41] -!- deep42thought has quit [Quit: Leaving.]
[15:33:48] -!- Faalagorn has quit [Quit: Leaving.]
[15:39:35] -!- Faalagorn has joined #archlinux-ports
[15:43:21] -!- Faalagorn has quit [Remote host closed the connection]
[16:10:55] -!- shentey has joined #archlinux-ports
[16:20:38] -!- shentey has quit [Ping timeout: 255 seconds]
[17:05:51] -!- p71 has quit [Read error: Connection reset by peer]
[18:36:40] -!- p71 has joined #archlinux-ports
[19:23:08] fsckd is now known as hald
[19:25:58] hald is now known as fsckd
[20:17:35] -!- p71 has quit [Remote host closed the connection]
[20:20:44] -!- p71 has joined #archlinux-ports