WebHosting Paid by #1Payday.Loans
[00:08] <owl> re [00:13] _BoS_ (n=BoS@dslb-084-059-203-012.pools.arcor-ip.net) left irc: Remote closed the connection [00:13] _BoS_ (n=BoS@dslb-084-059-201-236.pools.arcor-ip.net) joined #rocklinux. [07:51] <blindcoder> moin [07:54] kasc_ (n=kasc@dslb-084-060-124-014.pools.arcor-ip.net) joined #rocklinux. [07:56] <daja77> moin blindcoder [08:03] kasc (n=kasc@dslb-084-060-125-097.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [08:03] Nick change: kasc_ -> kasc [09:21] Freak (n=freak@erleuchtet.org) joined #rocklinux. [09:53] <netrunner> mnemoc: editing did not help [10:22] <netrunner> aha. libacl.la has hardcoded /lib/libattr.la [10:23] <netrunner> wooot [10:23] <netrunner> compiles :) [11:27] <daja77> netrunner: yeah I saw that too [11:27] <daja77> what have you done [11:29] <netrunner> daja77: edited libattr.la and libacl.la ... still building [11:30] <daja77> ic [11:30] <netrunner> in debug environment it helped. [11:30] <netrunner> I found it out by tracing libtool :) [11:42] <netrunner> daja77: success :) [11:42] <netrunner> kdelibs built. [11:42] <daja77> nice [11:44] <netrunner> now we need to find out where that line comes from, if it is really hardcoded or libtool's fault. [11:44] <netrunner> I first bring clothes to laundry. [12:00] blindcod1r (n=blindcod@dslb-084-059-055-033.pools.arcor-ip.net) joined #rocklinux. [12:00] blindcoder (n=blindcod@dslb-084-059-045-013.pools.arcor-ip.net) left irc: Read error: 104 (Connection reset by peer) [12:01] Nick change: blindcod1r -> blindcoder [12:04] <netrunner> hm, gnome has 2.13 already [12:04] Action: netrunner wants to test gnome ... [12:05] <blindcoder> re [12:44] <daja77> the kde update delayed my gnome update ... [12:45] <daja77> 2.12.1 for now, but 2.13 would be ok too [12:49] <SMP> daja77: I worked on that when Gnome 2.12 came out, but then ran out of time. how far gone are your efforts? [12:50] <daja77> SMP: i fixed gtk+ 2.8.6 to compile with gcc4, so some base packages build, other stuff builds after kde so I can't say if there are more errors to come [12:51] <SMP> daja77: do you have all platform + desktop packages up to date for 2.12.1 already? [12:52] <daja77> yes [12:52] <daja77> i used one afternoon to update them one by one ... [12:52] <daja77> and tested download so far [12:53] menomc (n=amery@200.75.27.27) joined #rocklinux. [12:53] Action: SMP used update-gnome.sh with heavy manual intervention [12:53] mnemoc (n=amery@200.75.27.60) left irc: Nick collision from services. [12:53] Nick change: menomc -> mnemoc [12:54] <daja77> the versions of the packages are too different, some were 2.12 some 2.12.1, annoying [12:54] <daja77> and I added 3 new ones [12:54] <SMP> I also found several occurences where the Gnome people are obviously just too damn stupid to ship working or even tested code [12:55] <SMP> and that doesn't make the job of updating Gnome more enjoyable at all ... [12:55] <daja77> the cairo dependency was confusing [12:55] <daja77> gtk depends on it, pango doesn't but gtk then needs a pango with cairo [12:57] <daja77> you have almost no chance to automate it [12:57] <SMP> automate what? the gnome update? [12:57] <daja77> i yes [12:58] <daja77> there even 2.12.0.1 packages out there [12:58] <netrunner> gnarf. kdebase already building 1.25h [12:58] <SMP> uh, well, update-gnome.sh helped me a lot [12:59] <daja77> I could try that too, but for 2.12.1 I already have all packages [13:01] <daja77> I can send you a diff for the gnome2 repo if you like [13:02] <daja77> but it is against trunk [13:02] <SMP> I'd be happy to have a look and maybe port some of the related changes I made back then, can't promise to do any useful testing, though [13:05] <daja77> https://www.tu-chemnitz.de/~jahre/gnome-2.12.1-update.patch [13:06] <SMP> thanks, bbl [13:07] <owl> moin [13:15] <netrunner> weha ... kdebase built. [13:15] <daja77> patch? [13:16] <netrunner> daja77: I fixed it manually. Patch would contain a sed in acl.conf that edit's the hardcoded library location of attr. [13:16] <daja77> svn diff? ;) [13:16] <netrunner> daja77: second ... [13:25] <netrunner> daja77: 2005120713245513037 [13:25] Action: netrunner starting cluster mode again [13:31] <daja77> netrunner: what have you done to the kdebase joystick patch? [13:34] <netrunner> daja77: removed :) [13:34] <daja77> ah that's easy ^^ [13:34] <daja77> all patches or just that one [13:35] clifford (n=clifford@213-229-1-138.sdsl-line.inode.at) left irc: Remote closed the connection [13:35] <netrunner> only that one. [13:35] <netrunner> the others apply, so maybe they still make sense. [13:35] <daja77> k. [13:35] clifford (n=clifford@213-229-1-138.sdsl-line.inode.at) joined #rocklinux. [13:36] <netrunner> hi clifford [13:36] <netrunner> any gaim user here? [13:36] Action: daja77 party [13:36] <daja77> partly [13:36] <clifford> netrunner: what do you need pike for? [13:37] <netrunner> clifford: not that I'd remember .... :) [13:38] <netrunner> clifford: interesting. the .cache tells of an error, but here it built. [13:38] <clifford> You don't have ffmpeg on your system (or the older version of it) [13:38] <clifford> I want to disable the pike ffmpeg module. is that ok for you? [13:39] <netrunner> clifford: go ahead. iirc I was once pointed to it, and wanted to take a look. so I created the package and then forgot about it :) [13:40] <daja77> netrunner: what was you gaim question again? [13:40] <netrunner> daja77: can you make gaim to receive html? [13:40] <daja77> what for [13:41] <netrunner> daja77: I use sim, which sends html only. gaim users complain about the html tags. so gaim seems not to interpret it. [13:41] <netrunner> clifford: have you tried updating to 7.6.50? [13:41] <daja77> netrunner: no idea [13:41] <netrunner> daja77: ok. [13:43] <clifford> netrunner: yes, i du [13:43] <clifford> netrunner: yes, i did [13:51] <daja77> guess I finally messed it up, rerunning from scratch [13:55] <netrunner> superkaramba is integrated now ... [14:05] <blindcoder> re [14:06] <netrunner> daja77: disable syncml stuff in kdepim.conf and the syncml.patch there [14:07] <netrunner> stf^rocklinux: x264 fails to build right after extracting [14:08] <daja77> ok [14:10] <daja77> still in stage 0 now ;) [14:17] <mnemoc> netrunner: :'( [14:17] <netrunner> mnemoc: what? [14:18] <mnemoc> netrunner: libtool [14:19] <mnemoc> netrunner: uhm.... editing .la worked or not? [14:19] <netrunner> mnemoc: yes. patch is in submaster. [14:19] <mnemoc> :D [14:19] <mnemoc> great [14:29] Action: netrunner is eager to try out kde 3.5 :) [14:29] <netrunner> with updated hal and dbus suppoert :) [14:34] <blindcoder> yeah [14:34] <blindcoder> upload an ISO [14:37] demian (n=demian@201.206.44.210) joined #rocklinux. [14:37] <demian> hi [14:38] <blindcoder> moin demian [14:38] <demian> hi blindcoder [14:57] Action: netrunner want's also to have the gnome update in. now waiting for smp :) [15:04] markuman (n=supermar@p50927208.dip.t-dialin.net) joined #rocklinux. [15:13] <netrunner> tunimp now inclusdes /usr/bin/trm (shared with musicbrainz-trm) [16:06] <netrunner> daja77: kdepim needs the sasl2 patch to be reworked. [17:08] Action: clifford doing a test build with NPTL. [18:07] <daja77> oh i just remember that there were updates outside gnome2 repo for the gnome update [18:16] <netrunner> ts. koffice builds now for >2h, and kdepim almost 2h. [18:17] <blindcoder> sounds normal [18:18] <daja77> it is [18:19] <blindcoder> kde needs over a day to build whole [18:19] <mnemoc> anyone knows how to remove a file from a patch using sed or awk? [18:19] <blindcoder> humm, not me [18:20] <mnemoc> script master :( [18:20] <blindcoder> I'm not much into sed and awk [18:27] <mnemoc> bzcat download/mirror/q/q-s-1.25st-20050406.patch.bz2 | sed '/^--- .*qmail-scanner-queue.template.orig/,/^--- / {d;q}' <--- works :D [18:27] <netrunner> something like s,--- foo/bar.*---,---,g [18:27] <blindcoder> mnemoc: why remove the file? [18:28] <mnemoc> it's a file remove which fails to apply [18:28] <blindcoder> I see [18:28] <mnemoc> so i remove that from the patch and rm by hand ,-) [18:29] <netrunner> oh, koffice finished :) [18:29] <netrunner> daja77: should I commit the kde changes I had so far? [18:30] <netrunner> hm ... better wait. there are still some patches to come. [18:49] stf^rocklinux (n=user@heim-033-137.raab-heim.uni-linz.ac.at) left #rocklinux ("Kopete 0.10.3 : https://kopete.kde.org"). [18:57] stf^rocklinux (n=user@heim-033-137.raab-heim.uni-linz.ac.at) joined #rocklinux. [18:57] <SMP> clifford: NPTL-only glibc or an extra libc.so in /lib/i686/ ? [19:37] demian (n=demian@201.206.44.210) left irc: Read error: 110 (Connection timed out) [20:28] netrunner (n=andreas@anvame.net) left irc: Remote closed the connection [20:34] netrunner (n=andreas@anvame.net) joined #rocklinux. [21:25] markuman (n=supermar@p50927208.dip.t-dialin.net) left irc: "https://www.linwiki.org - wiki around os linux!! #linwiki" [21:52] [raphael] (n=raphael@host-84-235-196-162.satlynx.net) joined #rocklinux. [21:53] [raphael] (n=raphael@host-84-235-196-162.satlynx.net) left #rocklinux. [21:53] [raphael] (n=raphael@host-84-235-196-162.satlynx.net) joined #rocklinux. [22:07] Freak (n=freak@erleuchtet.org) left irc: Read error: 110 (Connection timed out) [22:33] <netrunner> daja77: more bproblems to come. [22:33] <daja77> yeah! [22:42] Freak (n=nfreak@erleuchtet.org) joined #rocklinux. [22:59] <daja77> netrunner: you mean ftp://ftp.kde.org/pub/kde/security_patches ;) [23:00] <netrunner> daja77: kde installs .la files in /opt/kde3/libs, which also contain hardcoded links to /lib/libattr [23:01] <daja77> omfg [23:01] <netrunner> daja77: we should find a more generic solution. [23:01] <daja77> --without-acl ... [23:01] <netrunner> no. [23:01] <netrunner> If my throat hurts, I eat medicine, and don't chop my head off. [23:02] <mnemoc> *G* [23:02] <daja77> it is broken ... [23:03] <mnemoc> you could run a single sed call after relocating changing every /lib/*.a to /usr/lib/ on .la files [23:03] <netrunner> mnemoc: the problem does not occur at movement time. [23:03] <netrunner> hm. I have an idea. [23:04] <netrunner> I think I have missed sth. [23:04] <mnemoc> netrunner: it occurs on ranlib time, but .la gets inconsistent after relocation, so it should be fixed on that moment [23:04] <mnemoc> uhm, no ranlib. [23:05] <mnemoc> when ever .la is generated :p [23:16] <SMP> the only root of this evil is the acl package, because libacl.la references /lib/libattr.la. all other packages are merely just innocent victims of the attr/acl build system mess [23:18] <mnemoc> so no need to genericly solve this issue? [23:18] <SMP> no. [23:18] <mnemoc> :) [23:19] <netrunner> SMP: no. [23:19] <mnemoc> fight fight fight [23:19] <netrunner> second, am currently discussing. [23:20] <SMP> I think libattr.la's libdir must be fixed [23:20] Action: SMP testing [23:21] <mnemoc> netrunner tried that yesterday [23:21] <SMP> and? [23:21] <daja77> SMP: i forgot some packages in base and public repos in that gome update patch [23:22] <mnemoc> no results, then he changed depend_list... with success. [23:22] <SMP> daja77: can you update the patch? [23:22] <daja77> mom [23:22] <netrunner> SMP: I have fixed that acl thing. [23:23] <netrunner> SMP: now some kde .la's are referencing the libacl.la hardcoded to /lib/libacl.la [23:23] <SMP> bingo [23:23] <SMP> netrunner: of course [23:23] <SMP> libacl.la needs the same treatment [23:24] <mnemoc> they reference each other?? [23:24] <SMP> you can't move the .la without reflecting that in the libdir definition [23:24] <mnemoc> so, the generic fix [23:24] <SMP> mnemoc: no, the problem occurs with every .la that gets moved [23:25] Action: netrunner forgot to change the libdir. only changed the deplist, which helped in the first run. [23:25] <netrunner> trying to change the libdir also now. [23:25] <SMP> yeah, well, no. it's actually a bit more complicated [23:25] <SMP> netrunner: it worked for me, and it did so for good reasons [23:26] <mnemoc> .la global sanitization by replacing every /lib/*.{a,la} and libdir to /usr/lib ? [23:26] <mnemoc> why would it be more complex than that? [23:28] <daja77> SMP: updated patch at the same location [23:29] <netrunner> there is already some sed work done??? [23:29] <netrunner> in scripts/functions line 329 [23:29] <mnemoc> iirc chris did it [23:30] <netrunner> oh, but it expects a space before the = if I see that correctly :) [23:30] <SMP> mnemoc: attr and acl internally (rightly) make a difference between where the .so should go (--libdir=) and where .so, .la and .a for development should go (--libexecdir=). the problem is that libtool cannot reflect this. libtool only knows about libdir. attr and acl then go and install .so, .la and .a into PKG_DEVLIB_DIR (/usr/lib), but the libtool generated .la reflects libdir=$PKG_LIB_DIR, and so they work around that by symlinking the dev-libs in [23:31] <SMP> so actually attr and acl do the right thing, but libtool is just too braindead to get it right [23:32] <mnemoc> let's teach libtool about libexecdir :) [23:32] <daja77> mnemoc: some packages using their own libtool ... [23:32] <mnemoc> daja77: which can be replaced on postpatch [23:33] <SMP> you gotta be kidding. libtool is such a piece of crap code [23:33] <mnemoc> SMP: yes, i ranted about it's code yesterday [23:33] <mnemoc> SMP: but you have the gift :) [23:35] <SMP> mnemoc: the funny thing about the sentence of yours is, that the German word "Gift" means "poison". so, yes, taking poison would be preferable to hacking libtool ;-) [23:35] <netrunner> mnemoc: all kde packages use their own libtool# [23:35] <daja77> netrunner: yes ... [23:36] <SMP> netrunner: I'm still looking at that regexp ... [23:36] <mnemoc> SMP: :-) [23:38] <mnemoc> netrunner: are they significativly different than the one from libtool? [23:41] <netrunner> mnemoc: no [23:41] <netrunner> SMP: do you agree that this is the problem? [23:42] <netrunner> SMP: it seems to work without the secodnd backreference [23:43] <netrunner> no. weird. [23:43] <SMP> netrunner: the sed expression works for me just as it stands [23:44] <SMP> which means we shouldn't be seeing this problem in the first place [23:44] <SMP> which is obviously not the case [23:45] <SMP> oh! i've got it [23:46] <SMP> line 325, [ -e $root/$fn -o -L $root/$fn ] || continue [23:46] <SMP> the loop takes /lib/*.la, but skips symlinks [23:46] <SMP> what attr and acl install: [23:47] <SMP> /lib/libattr.la -> /usr/lib/libattr.la [23:47] <mnemoc> oh [23:47] <netrunner> SMP: just have seen it :) [23:47] <mnemoc> you have the gift ,-) [23:47] <SMP> so we're back to square one [23:48] <SMP> libtool produces a broken .la and our magic doesn't pick it up because of the wacky way it gets installed [23:49] <netrunner> SMP: we could apply the sed script also on links. [23:49] <SMP> so we should add a sed -i 'bla..bla' on the /usr/lib/lib{attr,acl}.la and we're done [23:51] <netrunner> if [ -L $fn -a [[ $fn = *.la ]] ]; then sed ... [23:52] <SMP> netrunner: that's not necessary and it will break with the current code, because the input and output files for sed would be identical [23:52] <mnemoc> netrunner $fn != "${fn%.la}" ,-) [23:53] <SMP> I don't feel like doing some global magic when only attr, acl and possibly dmapi are affected [23:54] <SMP> ok, xfsprogs, too [23:54] <mnemoc> what if it's not only attr, acl and possibly dmapi are affected? [23:55] <SMP> mnemoc: I've never seen another package with a build system that allows for this, and we don't seem to have patches for other packages like we need for those we know are affected [23:56] <mnemoc> yet :) [23:56] <netrunner> SMP: inline sed? (sed -i) [23:57] <SMP> fuck, my laptop's background light just died [00:00] --- Thu Dec 8 2005