#archlinux32 | Logs for 2019-03-29

Back
[00:13:40] -!- oaken-source has quit [Ping timeout: 246 seconds]
[00:42:22] -!- guys has joined #archlinux32
[01:08:09] -!- ubuntuxp has joined #archlinux32
[01:46:45] -!- samantaz_ has quit [Ping timeout: 245 seconds]
[01:49:30] -!- yans has joined #archlinux32
[01:51:17] -!- thePiGrepper has quit [Ping timeout: 245 seconds]
[01:52:34] -!- ofara_ has quit [Ping timeout: 250 seconds]
[02:14:57] -!- ofara_ has joined #archlinux32
[02:28:50] -!- guys has quit [Ping timeout: 272 seconds]
[02:42:44] -!- guys has joined #archlinux32
[02:52:02] -!- ubuntuxp has quit [Quit: ERC (IRC client for Emacs 26.1)]
[04:29:10] -!- guys has quit [Ping timeout: 272 seconds]
[04:43:05] -!- guys has joined #archlinux32
[04:46:22] -!- yans has quit [Quit: chaos is the only true answer]
[05:11:01] <buildmaster> i686/haskell-comonad is broken (says eurobuild3) - I rescheduled: haskell-distributive.
[06:17:54] <elibrokeit> deep42thought: another way of thinking of what you do with [build-support], is that it's not a "real" repo, but only exists to build (unclean) packages that do go into the real repos
[06:18:15] <elibrokeit> hence, it should never actually be added to end-user systems, if I understand correctly
[06:18:34] <elibrokeit> and any case that needs a package from there, is probably broken
[06:19:05] <elibrokeit> I think I suggested that parabola disable the repo from being imported, and just add it by hand to the build chroots
[06:29:02] -!- guys has quit [Ping timeout: 250 seconds]
[06:42:59] -!- guys has joined #archlinux32
[06:50:14] <buildmaster> i686/c2hs are broken (says rechenknecht).
[07:39:16] -!- deep42thought has joined #archlinux32
[07:39:16] <buildmaster> Hi deep42thought!
[07:39:16] <buildmaster> !rq deep42thought
[07:39:17] <phrik> buildmaster: <deep42thought> it was too slow - they don't offer that speed anymore, so they rounded down to 0
[07:39:32] <deep42thought> elibrokeit: yeah, that sounds like the sanest thing to do
[07:40:05] <deep42thought> I would skip adding it to our mirror at all, too (to hide it from end users) if there was an easy, convenient way to still make it available for our build slaves ...
[07:51:34] <elibrokeit> add it to the build slaves as https://hidden.archlinux32.org
[07:52:06] <elibrokeit> I mean it's basically just another pacman-*.conf in /usr/share/devtools, right?
[07:52:09] <deep42thought> that part is easy, harder is the part of managing [build-support] by the build master
[07:52:25] <deep42thought> but otoh there's the question if we actually want that ...
[07:53:24] <elibrokeit> I guess you could host the master everything on some sekrit vhost, then also host it to the main mirrors but with exclusion rules
[07:53:33] <elibrokeit> OTOH you could also just say "pls don't do that"
[07:54:29] <deep42thought> yeah, conceptually, we also have other stuff on mirror.archlinux32.org which is not intended for end users
[07:54:33] <deep42thought> e.g. x86_64/releng
[08:06:19] <deep42thought> elibrokeit: I got something else: do you happen to know where the *.files.tar.gz database gets used ? (e.g. which commands need the additional (w.r.t. *.db.tar.gz) information)
[08:17:37] -!- oaken-source has joined #archlinux32
[08:29:00] -!- guys has quit [Ping timeout: 268 seconds]
[08:40:26] <elibrokeit> deep42thought: pacman -S pkgfile; pkgfile -u; pkgfile --verbose <file>
[08:40:55] <elibrokeit> That is the thing that powers phrik's "wtf" search tool, for example.
[08:40:56] <deep42thought> only the second and third part need the files file, right?
[08:41:19] <elibrokeit> Right, you just need to install pkgfile first ;)
[08:41:24] <deep42thought> :-)
[08:41:37] <elibrokeit> Actually the second command extracts and builds an optimized uncompressed db
[08:42:02] <elibrokeit> Pacman itself recently supports -F to do searches, but it's not as efficient as pkgfile
[08:42:08] <elibrokeit> oaken-source: hi
[08:43:01] <deep42thought> ah, ok, so there is reason to manage "files" by repo-add and not by pkgfile itself
[08:43:15] <deep42thought> (in contrast to "links" which is never needed by pacman)
[08:43:23] <elibrokeit> deep42thought: right
[08:43:40] <elibrokeit> The files DB is an official part of the Pacman infra
[08:43:55] <elibrokeit> oaken-source: so what is the plan for build-support?
[08:44:04] <deep42thought> yeah, it just felt like "links" should be part of pacman infra, too
[08:44:26] <elibrokeit> I guess we could add links
[08:44:47] <deep42thought> would be nice - if it actually gets used by something
[08:44:53] <elibrokeit> But we'd need to have a well-thought-out implementation proposal to justify it
[08:44:57] <deep42thought> or: you could include the info in links into files
[08:45:07] <deep42thought> (and call it somewhat more appropriate, then)
[08:45:19] <deep42thought> yeah, no pushing :-)
[08:45:58] <elibrokeit> There's suggestions to have an "all" db and a files DB that does not have desc in it
[08:46:19] <elibrokeit> The files DB is also used by archweb to build the website index BTW
[08:46:38] <elibrokeit> That predates both Pacman -F and pkgfile
[08:49:51] <oaken-source> we are disabling build-support in our repo import
[08:50:23] <oaken-source> that seems to be the most obvious way forward for us
[09:01:58] <deep42thought> oaken-source: also note, that /all/ packages in [build-support] should be / should have been available in other repositories
[09:02:13] <deep42thought> e.g. there should be nothing which we only built for build-support
[09:49:23] <deep42thought> dig: error while loading shared libraries: libidn2.so.4: cannot open shared object file: No such file or directory
[09:49:23] <deep42thought> oops
[10:59:38] <buildmaster> i686/shellcheck is broken (says rechenknecht).
[11:02:30] <buildmaster> i686/idris are broken (says eurobuild3).
[11:54:41] <bill-auger> i think there is an insane situation in extra
[11:55:46] <deep42thought> what do you mean/refer to?
[11:55:49] <bill-auger> xorg-mkfontscale amd xorg-mkfontdir are both Required by ttf-dejavu
[11:56:23] <bill-auger> xorg-mkfontscale is Required by xorg-mkfontdir but also Conflict and Replaces xorg-mkfontdir
[11:56:46] <bill-auger> so installing ttf-dejavu is unsatifyable
[11:56:47] <deep42thought> that sounds indeed ... umm ... interesting
[11:57:19] <bill-auger> yea it has been "interesting" me for about an hour
[11:58:32] <deep42thought> looks like we should delete xorg-mkfontdir
[11:58:45] <deep42thought> because it is also /provided/ by xorg-mkfontscale
[11:59:43] <deep42thought> ah, only the latest does that ...
[11:59:46] <deep42thought> I'll move it :-)
[12:00:16] <bill-auger> no actually ttf-deja does install - im not sure what the blocker is yet
[12:00:37] <bill-auger> i noticed xorg-mkfontdir is not is x86_64
[12:00:48] <deep42thought> yeah
[12:01:01] <deep42thought> it's now provided by xorgg-mkfontscale, it seems
[12:03:52] <deep42thought> ok, moved and cleaned
[12:03:57] <deep42thought> does this resolve your issue?
[12:06:00] <bill-auger> simply installing ttf-dejavu seem to solve that
[12:06:15] <bill-auger> i actually have another problem now though
[12:06:21] <deep42thought> can you check if your original issue is also gone?
[12:06:39] <deep42thought> report whatever problem you like, first :-)
[12:06:46] <bill-auger> pacstrap keeps failing wiht a "corrupt package" but it doesnt say which one http://termbin.com
[12:07:37] <bill-auger> this is pacstrap from the LiveISO so i can try again in a few minutes
[12:08:22] <bill-auger> doesnt always say which package is corrupt ?
[12:08:25] <deep42thought> sry, I cannot see this termbin
[12:08:45] <bill-auger> http://termbin.com
[12:09:37] <deep42thought> there should be something above, I think
[12:10:00] <deep42thought> ah, wait
[12:10:25] <deep42thought> is you gpg broken in some way?
[12:10:46] <deep42thought> because usually, I see an error from gpg before that line
[12:14:08] <bill-auger> here a full run http://termbin.com
[12:15:04] <bill-auger> i have been repeating, removing one package at a time to see if i can find the pesky one
[12:15:17] <deep42thought> bisect it - it's faster
[12:16:29] <bill-auger> yea i know - but ya know flukes and stuff
[12:17:07] <deep42thought> anyway, you'll get faster to the one thing which makes everything fail
[12:17:23] <deep42thought> (assuming that the failing is reproducible)
[12:18:57] <bill-auger> ok its one of ours - either icedove or iceweasel
[12:19:33] <bill-auger> but why doesnt the error mention that, i assumed it always did
[12:20:13] <deep42thought> the errors were always somewhat misleading, in my opinion, but I thought, too, that there was at least /some/ error
[12:25:04] -!- abaumann has joined #archlinux32
[12:25:05] <buildmaster> Hi abaumann!
[12:25:05] <buildmaster> !rq abaumann
[12:25:06] <phrik> buildmaster: <abaumann> Why not copy postfix into systemd?
[12:25:15] <deep42thought> Hi abaumann!
[12:25:20] <abaumann> hi deep42thought
[12:26:05] <abaumann> dependency troubles?
[12:26:14] <deep42thought> only on parabola, it seems
[12:26:20] <abaumann> ah, ok.
[12:26:28] <deep42thought> ok for us ;-)
[12:26:50] <abaumann> :-)
[12:38:05] <bill-auger> ok i found the source of that error - it was a package in the /var/cache/pacman that was an incomplete .part file
[12:38:27] <deep42thought> shouldn't that have a different ending and not be relevant at all?
[12:46:29] <bill-auger> that i dont know but it was the source of the error -- when i deleted that file everything installed
[13:44:32] <bill-auger> the xorg-mkfontdir /xorg-mkfontscale problem is solved BTW
[13:44:56] <deep42thought> thx, good to know :-)
[13:47:19] <bill-auger> it did result in xorg-mkfontdir still being in the initial package list and it was removed automatically as befoe, but this time without incident
[13:50:50] -!- john__ has quit [Remote host closed the connection]
[13:52:52] -!- ubuntuxp has joined #archlinux32
[13:53:23] -!- thePiGrepper has joined #archlinux32
[14:05:52] -!- abaumann has quit [Ping timeout: 245 seconds]
[14:21:16] -!- guys has joined #archlinux32
[14:28:19] <elibrokeit> bill-auger, deep42thought: immediately when I saw "Pacman does not provide the error message" I was going to ask about .part files. There's a fix on pacman-git for exactly this.
[14:30:07] -!- guys has quit [Ping timeout: 246 seconds]
[14:34:14] -!- thePiGrepper has quit [Ping timeout: 250 seconds]
[14:43:08] -!- guys has joined #archlinux32
[14:56:57] -!- oaken-source has quit [Ping timeout: 246 seconds]
[15:24:02] -!- guys has quit [Ping timeout: 246 seconds]
[15:29:07] -!- oaken-source has joined #archlinux32
[15:38:29] -!- guys has joined #archlinux32
[15:48:45] -!- thePiGrepper has joined #archlinux32
[15:58:32] -!- thePiGre1per has joined #archlinux32
[15:58:52] -!- thePiGrepper has quit [Remote host closed the connection]
[16:12:07] -!- thePiGre1per has quit [Ping timeout: 245 seconds]
[16:13:59] -!- thePiGrepper has joined #archlinux32
[16:37:15] -!- quequotion has joined #archlinux32
[16:45:01] -!- oaken-source has quit [Ping timeout: 255 seconds]
[17:23:43] -!- guys has quit [Ping timeout: 255 seconds]
[17:34:20] -!- quequotion has quit [Quit: quequotion]
[17:38:07] -!- guys has joined #archlinux32
[19:23:28] -!- DepositePirate has quit [Remote host closed the connection]
[19:23:52] -!- DepositePirate has joined #archlinux32
[19:24:19] -!- guys has quit [Ping timeout: 255 seconds]
[19:38:28] -!- samantaz_ has joined #archlinux32
[19:38:46] -!- guys has joined #archlinux32
[19:40:57] -!- samantaz__ has joined #archlinux32
[19:43:38] -!- samantaz_ has quit [Ping timeout: 250 seconds]
[19:59:01] <buildmaster> i686/hedgewars are broken (says nlopc46).
[20:13:40] <thePiGrepper> hi, does anyone know what changes do I need to make to a rootfs originally booted with a gpt+efi scheme, to be able to boot it from a legacy bios one? I already changed the fstab file to remove mention of the efi partition, but the kernel is still panicking. any suggestion?
[20:14:46] <deep42thought> thePiGrepper: you installed a bootloader, I assume?
[20:15:11] <thePiGrepper> yeah, sorry, grub
[20:15:21] <thePiGrepper> it's trying to boot
[20:15:28] <thePiGrepper> it just makes the kernel panic
[20:16:17] <thePiGrepper> maybe some part of the output from grub-mkconfig is not right
[20:19:10] <deep42thought> yeah, I'd guess it might produce an invalid config if generated while booted in efi mode
[20:20:12] <deep42thought> I guess, it's a matter of getting the kernel options / initramfs location / root partition path right
[20:20:16] <deep42thought> in the grub.cfg^
[20:21:14] <thePiGrepper> no, I didnt do that. It was originally in a different disk. I copied the whole partition to a different disk which already had free space and a tiny arch install. so now, the new disk has two partitions(the tiny arch one, and this newly copied one), I mounted the newly copied one and rerun grub-mkconfig, the new output include /dev/sda3 which is the partition, however something is making the
[20:21:20] <thePiGrepper> kernel panic
[20:23:31] <thePiGrepper> I think I know what it is.
[20:25:03] <thePiGrepper> yeah, it worked. I knew it was something dumb. it always is
[20:25:46] <thePiGrepper> the generated output from grub-mkconfig didnt add /boot/initramfs-linux.img to the 'initrd' line after intel-ucode.img
[20:26:30] <thePiGrepper> lightdm/Xorg is still dead though.. :-)
[20:29:12] <thePiGrepper> when you change your linux rootfs from one machine to a different one, do you need to do something in particular? besides external drivers like Nvidia's I mean. for instance, udev or systemd save some configuration from your previous hardware?
[20:38:54] <deep42thought> nothing beyond the "usual" manual crafting of udev rules for unrecognized devies, etc :-)
[20:38:56] <deep42thought> I thionk
[20:38:58] <deep42thought> *think
[21:14:34] <buildmaster> i686/qcad is broken (says rechenknecht).
[21:15:33] -!- oaken-source has joined #archlinux32
[21:17:24] <elibrokeit> thePiGrepper: I suggest not using grub-mkconfig :p
[21:17:40] <deep42thought> I lived without it for years
[21:17:49] <deep42thought> but: why do you suggest that, eli?
[21:24:37] -!- guys has quit [Ping timeout: 245 seconds]
[21:26:23] -!- deep42thought has quit [Quit: Leaving.]
[21:29:48] <buildmaster> i686/jupyter-notebook is broken (says rechenknecht).
[21:33:52] <buildmaster> i686/gpgme is broken (says eurobuild3).
[21:34:45] <elibrokeit> deep42thought: because I reimplemented the whole thing in 40 lines of bash: https://paste.xinu.at
[21:34:57] <elibrokeit> Which produces this: https://paste.xinu.at
[21:37:37] <elibrokeit> now you tell me whether it is sane to use grub-mkconfig
[21:39:09] -!- guys has joined #archlinux32
[21:46:48] -!- oaken-source has quit [Ping timeout: 246 seconds]
[21:58:47] -!- thePiGrepper has quit [Ping timeout: 245 seconds]
[22:52:04] <buildmaster> i686/devtools are broken (says rechenknecht).
[23:15:20] -!- thePiGrepper has joined #archlinux32
[23:25:04] -!- guys has quit [Ping timeout: 255 seconds]
[23:39:31] -!- guys has joined #archlinux32
[23:52:48] -!- thePiGrepper has quit [Ping timeout: 246 seconds]