#archlinux32 | Logs for 2024-09-23

Back
[07:39:16] -!- T`aZ has joined #archlinux32
[07:51:59] -!- morriset has quit [Quit: Leaving]
[08:36:29] -!- AtleoS has joined #archlinux32
[10:59:59] <KitsuWhooa> librsvg is fucked on i686
[10:59:59] <KitsuWhooa> *sigh*
[11:21:42] -!- AtleoS has quit [Ping timeout: 252 seconds]
[11:22:26] -!- AtleoS has joined #archlinux32
[11:31:45] <KitsuWhooa> fwiw I edited the apache config to tell supported clients to gunzip the build logs on the fly. This basically means that clicking on a build-log link will show the text in the browser without downloading it
[11:32:49] <KitsuWhooa> for example https://buildmaster.archlinux32.org
[13:43:52] -!- balrog has quit [Ping timeout: 252 seconds]
[13:44:01] -!- balrog_ has joined #archlinux32
[15:31:06] -!- AtleoS has quit [Ping timeout: 276 seconds]
[17:32:31] -!- abaumann has joined #archlinux32
[17:32:32] <buildmaster> Hi abaumann!
[17:32:32] <buildmaster> !rq abaumann
[17:32:33] <phrik> buildmaster: <abaumann> I'm all for an docdepends upstreams and a doc hook, in order to reduce all those silly dependencies just to get documentation nobody reads afterwards, because they don't know about the 'man' command.. (rant of the day)
[17:33:04] <abaumann> KitsuWhooa: oh, thanks for the implicit decompression of the logs.. I think I said once I'll do something about it, just to forget it again. :-)
[17:33:26] <abaumann> librsvg ist f*ed on all architectures, basically. :->
[17:33:35] <KitsuWhooa> the cool thing is that the browser is what decompresses it
[17:33:41] <KitsuWhooa> so it is transferred compressed
[17:34:00] <KitsuWhooa> it seems like librsvg built for pentium4
[17:34:02] <abaumann> this is cool indeed :-)
[17:34:07] <abaumann> ah?
[17:34:29] <abaumann> there is still librsvg-og.. could also be used on i686
[17:34:29] <KitsuWhooa> it was built on 2024-09-19 15:19:04
[17:35:00] <KitsuWhooa> I'll go complain first at some point
[17:35:36] <KitsuWhooa> well
[17:35:44] <KitsuWhooa> it seems to want SSE
[17:35:48] <KitsuWhooa> and I can't tell if it wants SSE or SSE2
[17:35:55] <KitsuWhooa> because if SSE is disabled then we can simply enable it on i686
[18:39:35] <abaumann> rsvg-css.c:866:20: error: assignment to 'xmlStructuredErrorFunc' {aka 'void (*)(void *, const struct _xmlError *)'} from incompatible pointer type 'void (*)(void *, xmlError *)' {aka 'void (*)(void *, struct _xmlError *)'} [-Wincompatible-pointer-types]
[18:39:35] <abaumann> 866 | handler.serror = rsvg_xml_noerror;
[18:39:35] <abaumann> | ^
[18:39:35] <abaumann> hum, also librsvg-og has some newer gcc issues..
[18:46:12] <abaumann> oh, libcroco has been dropped to the AUR, fine.
[18:46:23] <abaumann> and it needs autoconf-2.16, urg.
[18:47:32] <abaumann> we really have to think about adding an 'aur' repo
[18:48:08] <abaumann> /bin/sh: line 2: /usr/bin/glib-mkenums: No such file or directory
[18:48:26] <abaumann> oh, no. this hits me again. dropped glib scripts which have no to be added locally to the project..
[18:49:32] <abaumann> ah, 'rest' in AUR I patched in this way.
[18:51:40] <abaumann> pkg-config --variable=glib_mkenums glib-2.0
[18:51:50] <abaumann> still results /usr/bin/glib-mkenums
[18:51:54] <abaumann> -bash: /usr/bin/glib-mkenums: No such file or directory
[18:52:11] <abaumann> sorry, I don't know what the Gnome guys are trying to do here. :-)
[18:57:23] <KitsuWhooa> you need to build glib2 without introspection first, then rebuild glib2 with introspection, then rebuild gtk
[18:57:35] <abaumann> and then I get glib-mkenums?
[18:57:38] <KitsuWhooa> I believe so
[18:57:44] <abaumann> oh my.
[18:57:52] <abaumann> this is sooo good for bootstrapping..
[18:57:58] <KitsuWhooa> I ended up making packages on build support for i686 and pentium4 to automate it
[18:58:13] <KitsuWhooa> I can't remember if they are in manual or not
[18:59:44] <KitsuWhooa> ah yeah it's in manual, but I left the pkgbuild under disabled
[19:00:04] <abaumann> but this means, glib2 is also broken on 64-bit upstream in that regard, because I'm working on Arch64
[19:00:05] <KitsuWhooa> it's -D introspection=disabled the first build
[19:00:12] <KitsuWhooa> oh
[19:00:14] <KitsuWhooa> that's weird
[19:00:24] <KitsuWhooa> there's a new glib devel package
[19:00:26] <KitsuWhooa> are you installing it>?
[19:00:30] <KitsuWhooa> I thought this was i486
[19:00:49] <abaumann> ah.
[19:00:52] <abaumann> thanks, that's it'
[19:00:57] <KitsuWhooa> on a32 I added a dependency on it because so many things were broken
[19:01:05] <KitsuWhooa> as in, glib2 depending on glib2-devel
[19:01:10] <abaumann> ah, good.
[19:01:11] <KitsuWhooa> but upstream it doesn't
[19:01:21] <abaumann> glib2-devel is part of base-devel or so?
[19:01:41] <KitsuWhooa> https://git.archlinux32.org
[19:01:42] <phrik> Title: PKGBUILD « glib2 « core - packages - Archlinux32 package modifications (at git.archlinux32.org)
[19:01:58] <KitsuWhooa> I don't remember what the problems were, honestly
[19:02:00] <KitsuWhooa> it's been a while
[19:02:23] <abaumann> no worries. I'll find out.
[19:02:32] <abaumann> ..and I have to fix 'rest' too now.
[19:02:40] <KitsuWhooa> but if it uses mkenums, it needs to depend on glib2-devel
[19:02:41] <abaumann> adding a glib2-devel there
[19:05:21] <abaumann> AC_MSG_CHECKING([for glib-mkenums script])
[19:05:21] <abaumann> GLIB_MKENUMS=`$PKG_CONFIG --variable=glib_mkenums glib-2.0`
[19:05:21] <abaumann> AS_IF([test "x$GLIB_MKENUMS" = "x"],
[19:05:21] <abaumann> [AC_MSG_ERROR([glib-mkenums not listed in glib-2.0 pkg-config file])],
[19:05:21] <abaumann> [AC_SUBST([GLIB_MKENUMS])
[19:05:23] <abaumann> AC_MSG_RESULT([$GLIB_MKENUMS])])
[19:05:27] <abaumann> from easytag. This look saner
[19:22:23] <abaumann> ok. for now I'll add libcroco to extra..
[19:32:12] <abaumann> so, hope, it rebuilds. :-)
[19:32:15] <abaumann> also fixed 'rest'
[19:32:17] -!- abaumann has quit [Quit: leaving]
[19:38:34] -!- T`aZ has quit [Remote host closed the connection]
[19:39:10] -!- T`aZ has joined #archlinux32
[19:48:05] <KitsuWhooa> abaumann: Any updates on getting the key signed by the master keys (if needed?) Also, girls, your master key is also expiring
[21:14:45] mavicaway is now known as mavica
[23:36:54] mavica is now known as mavicaway