T2 IRC Log: 2007-03-07
This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.
--- Log opened Wed Mar 07 00:00:31 2007
01:24 -!- alfreedy [n=beastie@gprs-pool-1-027.eplus-online.de] has quit [Remote closed the connection]
01:48 -!- mtr_ [n=Michael@kobz-590cb054.pool.einsundeins.de] has joined #t2
01:55 -!- mtr [n=Michael@kobz-590caa0c.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)]
01:55 -!- mtr_ is now known as mtr
02:20 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Session timeout)"]
07:41 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
08:54 < CIA-8> susan * r22853 /trunk/package/kde/kalarm/kalarm.desc: * updated kalarm (1.4.8 -> 1.4.10)
08:55 < CIA-8> susan * r22854 /trunk/package/develop/re2c/re2c.desc: * updated re2c (0.11.0 -> 0.11.2)
08:55 < CIA-8> susan * r22855 /trunk/package/mail/dcc/dcc.desc: * updated dcc (1.3.52 -> 1.3.53)
08:56 < CIA-8> susan * r22856 /trunk/package/kde/digikamimageplugins/digikamimageplugins.desc: * updated digikamimageplugins (0.9.0 -> 0.9.1)
08:56 < CIA-8> susan * r22857 /trunk/package/emulators/dosbox/dosbox.desc: * updated dosbox (0.65 -> 0.70)
08:57 < rxr> re
09:10 < CIA-8> susan * r22858 /trunk/package/filesystem/reiser4progs/reiser4progs.desc: * updated reiser4progs (1.0.5 -> 1.0.6)
09:27 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
09:39 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
09:39 < LMJ_Work> Hello the chan
09:42 < LMJ_Work> Got some issue to get my custom desktop x64 done : http://pastebin.ca/384728 here is a part of my failed packages
09:45 < rxr> HI LMJ_Work
09:47 < LMJ_Work> moin moin rxr
09:51 < rxr> LMJ_Work: just some dependencies missing, no ?
09:52 < LMJ_Work> i guess so, but i'm wondering which ones
09:52 < LMJ_Work> libgnomeprintui deps are all compiled
09:52 < LMJ_Work> avifile, i wonder what is missing
09:53 < LMJ_Work> same for dvd+rwtools
09:54 < LMJ_Work> ardour : maybe liblrdf is missing ;)
09:54 < LMJ_Work> nfs-utils : what package provide libwrap ?
09:55 < LMJ_Work> tcp_wrappers ?
09:57 < LMJ_Work> portmap error, i don't know what's wrong
09:57 < LMJ_Work> vlc needs mozilla-config ?
10:00 < LMJ_Work> maybe seamonkey
10:07 < CIA-8> susan * r22859 /trunk/package/graphic/dcraw/dcraw.desc: * updated dcraw (8.61 -> 8.62)
10:12 < rxr> libwrap is in tcp-wrapper, yes
10:12 < rxr> vlc is not to now comile right not in trunk
10:12 < rxr> guess I finally should take a look .-(
10:13 < LMJ_Work> oki ;)
10:43 < rxr> ahh - what kind of crap have I written there above ...
10:55 < LMJ_Work> about vlc ? ;)
11:04 < rxr> yeah - the spelling and missing characters and such ..
11:04 < rxr> was a long day yesterday, yet again ...
11:14 < sepp_> i would enable fox, erlang-otp and esdl to build rezound and wings ;)
11:16 -!- sepp_ is now known as sepp
11:17 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Ping timeout)"]
11:25 < rxr> hi sepp
11:26 < sepp> hi rxr :)
11:28 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
11:28 [Users #t2]
11:28 [@ChanServ] [ idealm ] [ mtr ] [ sepp ] [ Stelz ]
11:28 [ Capey ] [ LMJ ] [ R4gnar0k] [ Shingoshi] [ valentin]
11:28 [ CIA-8 ] [ LMJ_Work] [ rxr ] [ sparc-kly]
11:28 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
11:30 < CIA-8> rene * r22860 /trunk/package/graphic/libjpeg/libjpeg.conf:
11:30 < CIA-8> * workarounded libjpeg cross compile hickup in a more obvious and
11:30 < CIA-8> reliable way
11:39 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
11:46 -!- idealm [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has quit [Remote closed the connection]
11:48 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
11:54 < CIA-8> rene * r22861 /trunk/package/graphic/fbv/ (cc.patch fbv.conf): * fixed fbv to cross-build (patch imported from Atmel)
12:03 < rxr> http://www.pro-linux.de/news/2007/10927.html
12:04 < rxr> ^- cool
12:17 -!- sparc-kly_ [n=mubex@adsl-64-237-249-210.prtc.net] has joined #t2
12:33 -!- sparc-kly [n=mubex@adsl-72-50-102-240.prtc.net] has quit [Read error: 110 (Connection timed out)]
12:48 < rxr> re
12:59 -!- AStorm [n=astralst@chello084010114027.chello.pl] has joined #t2
12:59 < AStorm> Hello.
12:59 < AStorm> I'm trying to build T2 trunk with tests enabled.
12:59 < AStorm> Of course, autoconf fails with some stupid meaningless failures.
12:59 < AStorm> Like line number off by 1
13:00 < AStorm> or "file not found" message being different from the one they predicted.
13:00 < AStorm> Is there any way to selectively skip tests?
13:00 < sepp> is avr32 some licensed arm or mips cpu with a silly name or something completely different?
13:01 < sepp> hi AStorm
13:01 < AStorm> AVR32 is not Mips.
13:01 < AStorm> It is ARM.
13:01 < AStorm> Don't know which version of ARM though.
13:03 < AStorm> Nice Atmel chip.
13:04 < AStorm> Ok, back to my problem. Any idea?
13:04 < AStorm> Or do I have to hack the scripts?
13:04 < sepp> make test?
13:05 < AStorm> Yes. Skip that for specified packages.
13:06 < AStorm> The package is just fine, it's that I'm building from a working Gentoo installation - a bit rough environment so to say.
13:07 < AStorm> E.g. autoconf 2.59 testsuite expects bash 2.x and not 3.x
13:07 < sepp> dont know, maybe just disable test cases
13:07 < AStorm> Heh.
13:07 < AStorm> No way to just skip make test for some pkgs?
13:08 < sepp> you could add some NO_CHECK flag to skip some packages
13:09 < AStorm> Where? .desc?
13:10 < sepp> scripts/functions.in and then flag thing like the other flags
13:10 < sepp> in the desc
13:10 < AStorm> Mhm.
13:10 < AStorm> Good, can do. Though I could fix the tests to be less stupid, even.
13:11 < sepp> guess there are too many stupid things to fix them all :)
13:12 < AStorm> That one I can fix :>
13:13 < AStorm> Gentoo devs, in their own retardedness, added unslotted autoconf 2.61
13:13 < AStorm> And that's incompatible with 2.5x
13:14 < AStorm> (AC_DEFINE_UNQUOTED et. al.)
13:15 < AStorm> Don't do the same mistake - these have to be installed in parallel.
13:15 < AStorm> 2.1x, 2.5x and 2.6x
13:15 < rxr> hi
13:15 < AStorm> Hello.
13:15 < rxr> what architecture do you build for ?
13:16 < rxr> sepp: AVR32 is not ARM, it is designed from scratched
13:16 < rxr> AStorm: you mean the build fails due to make check | tes t?
13:17 < AStorm> Yes.
13:17 < AStorm> rxr, is it?
13:17 < AStorm> Well, I'm building x86_64
13:17 < AStorm> The failures are funny, different line numbers and different "404" message from bash.
13:17 < rxr> this automated make check | test thing never got too much attention
13:17 < rxr> just disable it if you do not want to see many stray build aborts due that
13:17 < AStorm> Guess it's the time to change it :P
13:18 < AStorm> I do want :P
13:18 < rxr> hehe - ok that is of course very welcome as well :-)
13:18 < AStorm> I just want to first see the failure, then disable the test for the stupid thing if I deem that safe.
13:18 < rxr> I guess it's best to fix the underlying package source then
13:18 < AStorm> Should be easy enough to just change run_check to be conditional.
13:18 < rxr> I also wonder if we run the check for native builds only
13:19 < rxr> I bet for cross builds most packages will fail to perform the checking anyway ...
13:19 < AStorm> For really cross ones, most likely.
13:19 < rxr> T2 always bootstraps a fresh cross compiler in the beginning
13:19 < rxr> stage 0 is "build cross compiler"
13:20 < rxr> stage 1 is "bootstrap fresh cross build stuff
13:20 < AStorm> Yes. And the autoconf failure is in stage0
13:20 < AStorm> I'd expect that :P
13:20 < rxr> and just stage 2 is "build in the fresh bootstraped sandbox"
13:20 < AStorm> Anyway, those autoconf tests have to be fixed a bit to be less stupid. Will do that.
13:21 < rxr> :-)
13:21 < AStorm> hah, it's impossible
13:21 < AStorm> They use diff on output, a braindead strategy.
13:21 < rxr> you can just drop a new patch in the t2 package/.../$pkg/ directory
13:21 < AStorm> I know.
13:21 < rxr> the patches will be picked up automatically ..
13:21 < rxr> AStorm: ah ok
13:22 < rxr> AStorm: most people expect more harcoding such as in other package formats ...
13:22 < rxr> just wanted to drop some basic-use information here so you have an easier life
13:22 < AStorm> rxr, well, I'm coming from Gentoo and have read the doc, even.
13:22 < AStorm> (poor English, but good enough)
13:22 < rxr> :-)
13:22 < rxr> http://www.linuxdevices.com/news/NS5501521223.html
13:23 < AStorm> Whoops, I mistook AVR32 for plain AVR.
13:23 < AStorm> That was ARM.
13:24 < AStorm> Highly interesting stuff...
13:30 < AStorm> BTW, I've some other ideas for T2 build system... eg. replacing chroot+loop with chroot+unionfs where available.
13:39 < AStorm> Yes, autoconf testsuite can't really be fixed.
13:41 < AStorm> The tests won't be enabled when building cross.
13:50 < AStorm> Hmm, one of the patches to vim contains "LIBTOOL-WUIRK"
13:50 < AStorm> Surely a tpyo.
13:53 < AStorm> Adding that NO-CHECK flag was very simple.
13:54 < AStorm> Will have to provide some variants maybe, like STAGE0-NO-CHECK
13:57 < AStorm> It's refreshing to change a distro after 6y with Gentoo.
14:05 < sepp> :)
14:07 < LMJ_Work> !> Binary file for rezound-0.12.3beta.tar.bz2 not present. Skipped in package database.
14:07 < LMJ_Work> !> Binary file for sun-jre-160-6.tar.bz2 not present. Skipped in package database.
14:07 < LMJ_Work> !> Binary file for vlc-0.8.6a.tar.bz2 not present. Skipped in package database.
14:08 < LMJ_Work> I stil have this warning even i removed those files from my custom pkgsel, how i can avoid this ?
14:09 < rxr> AStorm: you'r welcome :-)
14:09 < LMJ_Work> hello AStorm
14:09 < rxr> AStorm: for ideas you might also like to take a look at the GSoC site
14:09 < rxr> http://www.t2-project.org/developer/soc.html
14:10 < rxr> and some other roadmap bits are drafted here: http://www.t2-project.org/releases/
14:11 < AStorm> I've seen that...
14:18 < LMJ_Work> ping sepp
14:26 < CIA-8> rene * r22862 /trunk/architecture/avr32/package/binutils/binutils-fix-pico-in-out-regs.patch: * imported yet another avr32 binutils hotfix (this time for PiCo instructions)
14:42 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
14:56 [Users #t2]
14:56 [@ChanServ] [ CIA-8] [ R4gnar0k] [ Shingoshi ] [ valentin]
14:56 [ AStorm ] [ LMJ ] [ rxr ] [ sparc-kly_]
14:56 [ Capey ] [ mtr ] [ sepp ] [ Stelz ]
14:56 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
15:23 < CIA-8> rene * r22863 /trunk/package/multimedia/vlc/ (vlc.cache vlc.conf):
15:23 < CIA-8> * fixed vlc firefox/seamonkey support, but hard-disabled vlc flac
15:23 < CIA-8> support as it is broken right now
15:27 -!- LMJ_work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
15:28 < sepp> guess flac support is really broken atm
15:30 < rxr> in vlc or in general ?
15:30 < sepp> general
15:30 < rxr> oh
15:30 < rxr> maybe we have a too new flac version?
15:31 < sepp> maybe
15:31 < rxr> but there is also quite some stuff building with flac support
15:31 < sepp> :)
15:31 < LMJ_work> arg
15:31 < rxr> LMJ_work: ?
15:31 < LMJ_work> disconnected
15:32 < rxr> # grep flac *.err | cut -d : -f 1 | uniq
15:32 < rxr> 5-alsaplayer.err
15:32 < rxr> 5-audacity.err
15:32 < rxr> 5-gst-ffmpeg.err
15:32 < rxr> 5-libsndfile.err
15:32 < rxr> 5-mpd.err
15:32 < rxr> 5-sox.err
15:35 < sepp> 5-hydrogen.err
15:35 < sepp> :)
15:37 < sepp> http://www.xfce-look.org/CONTENT/content-pre1/51274-1.jpg <- so nice :p
15:37 < rxr> r22554 | rene | 2007-02-16 15:54:57 +0000 (Fri, 16 Feb 2007) | 2 lines
15:37 < rxr> * updated flac (1.1.2 -> 1.1.4)
15:37 < rxr> sepp: hm?
15:37 < sepp> lol
15:37 < rxr> Martel Barby theme ?
15:38 < sepp> that is the worst thing ever
15:38 < rxr> barbi even
15:38 < rxr> hm - no idea how it is spelled
15:39 < sepp> they could even use it in guantanamo
15:40 < rxr> the flac 1.1 series goes rather way back
15:40 < rxr> and they claim 1.1.4 is way faster and such
15:40 < rxr> I guess we rather want to fix the compile errors currently seen
15:40 < rxr> than downdating something in the flac area ...
15:41 < sepp> oki
15:46 < LMJ_work> how to remove a package built with Build-Target (remove files inside build//*
15:46 < LMJ_work> )
15:48 < LMJ_work> i could chroot and do a mine -rf
15:50 < LMJ_work> ?
15:50 < sepp> that would work
15:51 < LMJ_work> nothing better ?
15:53 < sepp> i would not remove glibc or something everything is linked again ...
15:54 < rxr> LMJ_work: a single package ?
15:54 < rxr> mine -r $pkg -r build/...
15:54 < rxr> or
15:54 < rxr> ./scripts/Create-ErrList -cfg $cfg -delete $pkg
15:54 < rxr> ./scripts/Create-ErrList -cfg $cfg -remove $pkg
15:55 < rxr> IIRC -delete removes all files and -remove removes just the log
15:55 < LMJ_work> ok thanks
15:55 < rxr> (I think this option should be clarified=
15:55 < rxr> )
15:59 < sepp> -[full|new]delete ], ohh i was not aware that there is 1 delete, but 3 ...
16:06 < rxr> yep
16:14 < LMJ_work> sepp : what do you think about jahshaka : http://pastebin.ca/385022
16:16 < sepp> dont know, the real error is earlier ...
16:17 < sepp> guess it has not found some header file
16:18 < LMJ_work> sepp: here we go : http://pastebin.ca/385027
16:19 < sepp> it needs glew
16:19 < LMJ_work> opengl related error ?
16:20 < sepp> better build glew 1st :)
16:20 < LMJ_work> gonna add glew
16:20 < LMJ_work> in progress
16:20 < LMJ_work> done ;)
16:22 < LMJ_work> ok, still an error
16:22 < LMJ_work> http://pastebin.ca/385036
16:24 < sepp> glew needs FPIC-QUIRK
16:24 < LMJ_work> glew compiled fine
16:24 < LMJ_work> ok, let's try
16:27 < LMJ_work> same error
16:27 < LMJ_work> but stop for today, gtg ;)
16:28 < rxr> sepp: do you commit marking glew FPIC-QUIRK then ?
16:29 < LMJ_work> be back later at home
16:30 < sepp> still the same error?
16:33 < sepp> will try if it compiles later
16:35 < AStorm> Huh, tar failing its own testsuite was hilarious :P
16:35 < AStorm> multivolume and sparse.
16:36 < AStorm> 99
16:36 < AStorm> 99,9% sure it's related to the build system and loop file.
16:56 < CIA-8> rene * r22864 /trunk/package/graphic/fbv/bigendian.patch: * basically fixed fbv for a bigendian frame-buffer on big-endian CPUs
16:58 < AStorm> Weird. I can't do that NO-CACHE flag properly it seems.
16:58 < LMJ> moin moin back ;)
16:58 < AStorm> Testing with hasflag in the scripts/functions.in fails - is always true.
17:00 < AStorm> At least in the place I do it. I'll try moving the test.
17:02 < AStorm> Yes, the flags are not yet available where the hook is added... but they are available in the hook itself.
17:04 < AStorm> Yes, now they do really work.
17:09 < rxr> :-)
17:09 < rxr> you send patches of your work to the list or so then?
17:10 < sepp> jahshaka (+ needed libs) builds on my amd64 ...
17:13 < LMJ> ok
17:13 < LMJ> gonna remove it ;)
17:14 < AStorm> rxr, it's just a new flag
17:14 < AStorm> I still think it's not enough :>
17:14 < AStorm> Uh, glibc hates me? :P
17:14 < AStorm> !> checking for .preinit_array/.init_array/.fini_array support... no
17:14 < AStorm> !> configure: error: Need linker with .init_array/.fini_array support.
17:15 < AStorm> It's stage 1 AFAICT.
17:15 < CIA-8> rene * r22865 /trunk/package/graphic/fbv/bigendian.patch:
17:15 < CIA-8> * fixed fbv to handle 24 bit per pixel framebuffer
17:15 < CIA-8> (so far only the big-endian code-flow was tested)
17:15 < AStorm> I think this is caused by SSP.
17:16 < sepp> glibc has NO-SSP flag ...
17:17 < AStorm> Well, yes, SSP is disabled.
17:17 < AStorm> But what would cause the error? I'll better check the config.log
17:18 < rxr> do not remeber that I saw that recently ...
17:19 * rxr wonders if I fix mplayer on 16bpp big-endian framebuffers, next ...
17:22 < AStorm> Is T2 binutils built against uclibc or dietlibc?
17:23 < rxr> oeh
17:23 < rxr> depends on your build
17:23 < rxr> when you do a glibc build it is of course build against glibc
17:23 < rxr> when you do a uclibc or dietlibc based build this changes, of course
17:24 < AStorm> Well, it's a glibc build.
17:24 < AStorm> And the bintutils with dietlibc is unchecked.
17:27 < AStorm> Something is very much off then :P
17:27 < AStorm> Either the build system hates my Gentoo :P
17:32 < AStorm> or the compiler that's been built is broken.
17:41 < AStorm> In case I'm unable to figure it out, I'll try from a LiveCD chroot.
17:42 < AStorm> Info: Gentoo ~x86_64 with multilib
17:42 < AStorm> But that should be irrelevant, as T2 SDE is building its own binutils.
17:46 < rxr> yes it should
17:59 < CIA-8> rene * r22866 /trunk/package/graphic/fbv/bigendian.patch:
17:59 < CIA-8> * and finally fixed fbv for 32 bit per pixel framebuffer on
17:59 < CIA-8> big-endian
18:12 < AStorm> Unless I manage to run T2, I may end up with a personal distribution.
18:12 < AStorm> Got the dependency and package logging scripts already sorted out...
18:12 < AStorm> (package logging by the use of unionfs ro/rw split :>
18:13 < AStorm> dependency detection could possibly be done using unionfs logging if available, otherwise LD_PRELOAD or stracing.
18:14 < AStorm> Huh? WTF?
18:14 < AStorm> Now flex failed the tests...
18:14 < AStorm> Something is very much off here.
18:14 < AStorm> (previously it didn't)
18:16 < AStorm> Sensors do look ok, no weird log messages - hardware is probably fine.
18:17 < AStorm> ah, previously, I probably had broken testing :P
18:17 < AStorm> CXX_not_available_in_stage_0 -c -o scanner.o -I. -I../.. -I../.. -g -O2 scanner.cpp
18:17 < AStorm> It seems I have to add even more flags.
18:18 < AStorm> Eg. NO-CHECK-STAGE0
18:25 < rxr> I hope you find T2 useful instead of starting your own scripts .-)
18:28 < AStorm> Maybe.
18:28 < AStorm> Except these are already started :-)
18:29 < rxr> yeah but once you sorted them out and fixed a few bugs you get 2600++ package descriptions and 17 or so CPU architectures for free :-)
18:29 < AStorm> make: Cross_CC_not_built_so_far_cc: Command not found
18:29 < AStorm> make: *** [/data/t2-trunk/src/Linux-x86_64/lua/lapi.o] Error 127
18:29 < AStorm> -> Error building the LUA bash accelerator.
18:29 < AStorm> Fun.
18:30 < AStorm> Shouldn't that try to build with system gcc then?
18:31 < AStorm> I'll probably use the package descriptions, but add more buildsystem support.
18:31 < AStorm> Maybe some option switches too (noticed none in T2)
18:32 < rxr> enable the expert options in the Config :-)
18:32 < AStorm> That's not exactly what I mean by options.
18:32 < AStorm> These are very lowlevel. I meant like per-package features.
18:32 < rxr> Config even has a few of those
18:32 < AStorm> E.g. you install, say, gnome, you can pick options.
18:33 < AStorm> E.g. no printer support- sure!
18:33 < AStorm> No accesibility, no problem at all.
18:33 < rxr> and the general stuff like jpeg or usb or not is controlled whether libjepg and co are selected in the target package selection
18:33 < AStorm> That's good.
18:34 < AStorm> Even good enough :-)
18:34 < rxr> and per package options exist too
18:34 < rxr> the whole Config stuff is build up dynamically by the various config.in files
18:34 < AStorm> I wonder how to test against something like ! (asstage toolchain && hasflag NO-CHECK-STAGE0)
18:35 < AStorm> *atstage
18:35 < rxr> e.g. vim has bloated, mediuam and tiny or so to 1:1 wire the assocaited vim feature set among other
18:35 < AStorm> Hmm... config.in... sounds... autotoolish, but could be nice.
18:35 < rxr> heh - no nothing todo with autotools
18:35 < rxr> (which I rather avoid like the plague ...)
18:36 < AStorm> rxr, good for you, I do too. M4 is evil.
18:36 < rxr> definetly ...
18:39 < AStorm> BTW, do you have any good description of the gem package format?
18:42 < rxr> the mine tarball includes some bites
18:42 < rxr> bits
18:42 < AStorm> Hm.
18:42 < rxr> but I hope you do not like that too much already, as the default is tarballs already and we plan to drop this .gem "nonsense" for future T2 releases
18:42 < rxr> did you already fall in love with it ?
18:42 < AStorm> "nonsense"?
18:42 < rxr> well - why "yet another package format" when vanilla tarballs are that handy ...
18:43 < AStorm> I'd store there 2 files more, maybe
18:43 < AStorm> E.g. dependencies
18:43 < AStorm> (in some form, not yet sure)
18:43 < AStorm> (have to be generic enough as to allow installing compatible packages)
18:44 < rxr> the tarballs has them
18:44 < AStorm> (maybe using library .version would be enough - with optional override too)
18:44 < rxr> in the usual T2 var/adm/* form
18:44 < AStorm> Well, that doesn't store any versioning info.
18:45 < AStorm> Just wait until someone tries to install an old tarball :D
18:45 < rxr> we sort those files to the front of in the tarball so the tar only needs to be partially decoded if just those are read
18:45 < AStorm> Tar is blocked anyway.
18:45 < rxr> the meta data is in it - jsut the install needs to use it
18:45 < AStorm> Shouldn't matter.
18:45 < AStorm> (unless the filesystem doesn't support seeks)
18:45 < rxr> and when you do not have the properitary installer around you still can use tar on any system from the last 20 years or so
18:45 < AStorm> 30 years.
18:46 < rxr> yeah
18:47 < AStorm> Optionally I'll also support .tar.7z
18:47 < rxr> but you can plug any other binary package creation between the build script lines as you wish of course as well
18:47 < rxr> yeah - just 3 lines of code to stick other compressors in :-)
18:49 < AStorm> Yep.
18:50 < AStorm> Still, the devel system has to fit on a pendrive.
18:50 < AStorm> With X and Xfce 4.4
18:50 < AStorm> And a lot of various interpreters/compilers.
18:50 < AStorm> Plus some libs.
18:50 < AStorm> Also other basic software.
18:51 < AStorm> T2 will be good _if_ I manage it to build.
18:51 < AStorm> I'll try the ISO in the chroot now.
18:52 < AStorm> Or maybe... I'll check the cross-compile switch.
18:53 < AStorm> plain LFS lookalike also failed to build here for weird reasons
19:03 < AStorm> I'm trying a cross build now.
19:08 < AStorm> If that'll work, Gentoo Glibc/GCC/Binutils get my personal IGNoble award.
19:27 < AStorm> Huh...
19:27 < AStorm> even the cross build failed like before.
19:27 < AStorm> So something is broken in T2 trunk.
19:28 < rxr> well - trunk works quite perfect on my box
19:29 < rxr> when your LFS attempts already failed maybe there really is some going on ...
19:29 < rxr> maybe just extract another system and chroot into it ...
19:30 < AStorm> I'll do that. Guess which system :P
19:33 * rxr out of office
19:33 < rxr> cu in the morning
19:33 < AStorm> Bye.
20:08 -!- tntnet_ [n=tommi@dslb-084-058-140-155.pools.arcor-ip.net] has joined #t2
20:24 -!- d-marc [n=d-marc@HSI-KBW-091-089-002-247.hsi2.kabelbw.de] has joined #t2
20:25 -!- tntnet_ [n=tommi@dslb-084-058-140-155.pools.arcor-ip.net] has quit [Remote closed the connection]
20:58 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
20:59 < misl> hi folks
21:00 < Stelz> hello misl :-)
21:01 < misl> hi Stelz :-)
21:29 -!- LMJ_work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Session timeout)"]
21:53 -!- d-marc [n=d-marc@HSI-KBW-091-089-002-247.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
23:06 -!- sepp_ [n=sepp@p85.212.50.206.tisdip.tiscali.de] has joined #t2
23:06 -!- sepp [n=sepp@p213.54.174.43.tisdip.tiscali.de] has quit [Read error: 54 (Connection reset by peer)]
23:09 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
23:15 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
23:16 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Client Quit]
--- Log closed Thu Mar 08 00:00:32 2007
01:24 -!- alfreedy [n=beastie@gprs-pool-1-027.eplus-online.de] has quit [Remote closed the connection]
01:48 -!- mtr_ [n=Michael@kobz-590cb054.pool.einsundeins.de] has joined #t2
01:55 -!- mtr [n=Michael@kobz-590caa0c.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)]
01:55 -!- mtr_ is now known as mtr
02:20 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Session timeout)"]
07:41 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
08:54 < CIA-8> susan * r22853 /trunk/package/kde/kalarm/kalarm.desc: * updated kalarm (1.4.8 -> 1.4.10)
08:55 < CIA-8> susan * r22854 /trunk/package/develop/re2c/re2c.desc: * updated re2c (0.11.0 -> 0.11.2)
08:55 < CIA-8> susan * r22855 /trunk/package/mail/dcc/dcc.desc: * updated dcc (1.3.52 -> 1.3.53)
08:56 < CIA-8> susan * r22856 /trunk/package/kde/digikamimageplugins/digikamimageplugins.desc: * updated digikamimageplugins (0.9.0 -> 0.9.1)
08:56 < CIA-8> susan * r22857 /trunk/package/emulators/dosbox/dosbox.desc: * updated dosbox (0.65 -> 0.70)
08:57 < rxr> re
09:10 < CIA-8> susan * r22858 /trunk/package/filesystem/reiser4progs/reiser4progs.desc: * updated reiser4progs (1.0.5 -> 1.0.6)
09:27 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
09:39 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
09:39 < LMJ_Work> Hello the chan
09:42 < LMJ_Work> Got some issue to get my custom desktop x64 done : http://pastebin.ca/384728 here is a part of my failed packages
09:45 < rxr> HI LMJ_Work
09:47 < LMJ_Work> moin moin rxr
09:51 < rxr> LMJ_Work: just some dependencies missing, no ?
09:52 < LMJ_Work> i guess so, but i'm wondering which ones
09:52 < LMJ_Work> libgnomeprintui deps are all compiled
09:52 < LMJ_Work> avifile, i wonder what is missing
09:53 < LMJ_Work> same for dvd+rwtools
09:54 < LMJ_Work> ardour : maybe liblrdf is missing ;)
09:54 < LMJ_Work> nfs-utils : what package provide libwrap ?
09:55 < LMJ_Work> tcp_wrappers ?
09:57 < LMJ_Work> portmap error, i don't know what's wrong
09:57 < LMJ_Work> vlc needs mozilla-config ?
10:00 < LMJ_Work> maybe seamonkey
10:07 < CIA-8> susan * r22859 /trunk/package/graphic/dcraw/dcraw.desc: * updated dcraw (8.61 -> 8.62)
10:12 < rxr> libwrap is in tcp-wrapper, yes
10:12 < rxr> vlc is not to now comile right not in trunk
10:12 < rxr> guess I finally should take a look .-(
10:13 < LMJ_Work> oki ;)
10:43 < rxr> ahh - what kind of crap have I written there above ...
10:55 < LMJ_Work> about vlc ? ;)
11:04 < rxr> yeah - the spelling and missing characters and such ..
11:04 < rxr> was a long day yesterday, yet again ...
11:14 < sepp_> i would enable fox, erlang-otp and esdl to build rezound and wings ;)
11:16 -!- sepp_ is now known as sepp
11:17 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Ping timeout)"]
11:25 < rxr> hi sepp
11:26 < sepp> hi rxr :)
11:28 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
11:28 [Users #t2]
11:28 [@ChanServ] [ idealm ] [ mtr ] [ sepp ] [ Stelz ]
11:28 [ Capey ] [ LMJ ] [ R4gnar0k] [ Shingoshi] [ valentin]
11:28 [ CIA-8 ] [ LMJ_Work] [ rxr ] [ sparc-kly]
11:28 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
11:30 < CIA-8> rene * r22860 /trunk/package/graphic/libjpeg/libjpeg.conf:
11:30 < CIA-8> * workarounded libjpeg cross compile hickup in a more obvious and
11:30 < CIA-8> reliable way
11:39 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
11:46 -!- idealm [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has quit [Remote closed the connection]
11:48 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
11:54 < CIA-8> rene * r22861 /trunk/package/graphic/fbv/ (cc.patch fbv.conf): * fixed fbv to cross-build (patch imported from Atmel)
12:03 < rxr> http://www.pro-linux.de/news/2007/10927.html
12:04 < rxr> ^- cool
12:17 -!- sparc-kly_ [n=mubex@adsl-64-237-249-210.prtc.net] has joined #t2
12:33 -!- sparc-kly [n=mubex@adsl-72-50-102-240.prtc.net] has quit [Read error: 110 (Connection timed out)]
12:48 < rxr> re
12:59 -!- AStorm [n=astralst@chello084010114027.chello.pl] has joined #t2
12:59 < AStorm> Hello.
12:59 < AStorm> I'm trying to build T2 trunk with tests enabled.
12:59 < AStorm> Of course, autoconf fails with some stupid meaningless failures.
12:59 < AStorm> Like line number off by 1
13:00 < AStorm> or "file not found" message being different from the one they predicted.
13:00 < AStorm> Is there any way to selectively skip tests?
13:00 < sepp> is avr32 some licensed arm or mips cpu with a silly name or something completely different?
13:01 < sepp> hi AStorm
13:01 < AStorm> AVR32 is not Mips.
13:01 < AStorm> It is ARM.
13:01 < AStorm> Don't know which version of ARM though.
13:03 < AStorm> Nice Atmel chip.
13:04 < AStorm> Ok, back to my problem. Any idea?
13:04 < AStorm> Or do I have to hack the scripts?
13:04 < sepp> make test?
13:05 < AStorm> Yes. Skip that for specified packages.
13:06 < AStorm> The package is just fine, it's that I'm building from a working Gentoo installation - a bit rough environment so to say.
13:07 < AStorm> E.g. autoconf 2.59 testsuite expects bash 2.x and not 3.x
13:07 < sepp> dont know, maybe just disable test cases
13:07 < AStorm> Heh.
13:07 < AStorm> No way to just skip make test for some pkgs?
13:08 < sepp> you could add some NO_CHECK flag to skip some packages
13:09 < AStorm> Where? .desc?
13:10 < sepp> scripts/functions.in and then flag thing like the other flags
13:10 < sepp> in the desc
13:10 < AStorm> Mhm.
13:10 < AStorm> Good, can do. Though I could fix the tests to be less stupid, even.
13:11 < sepp> guess there are too many stupid things to fix them all :)
13:12 < AStorm> That one I can fix :>
13:13 < AStorm> Gentoo devs, in their own retardedness, added unslotted autoconf 2.61
13:13 < AStorm> And that's incompatible with 2.5x
13:14 < AStorm> (AC_DEFINE_UNQUOTED et. al.)
13:15 < AStorm> Don't do the same mistake - these have to be installed in parallel.
13:15 < AStorm> 2.1x, 2.5x and 2.6x
13:15 < rxr> hi
13:15 < AStorm> Hello.
13:15 < rxr> what architecture do you build for ?
13:16 < rxr> sepp: AVR32 is not ARM, it is designed from scratched
13:16 < rxr> AStorm: you mean the build fails due to make check | tes t?
13:17 < AStorm> Yes.
13:17 < AStorm> rxr, is it?
13:17 < AStorm> Well, I'm building x86_64
13:17 < AStorm> The failures are funny, different line numbers and different "404" message from bash.
13:17 < rxr> this automated make check | test thing never got too much attention
13:17 < rxr> just disable it if you do not want to see many stray build aborts due that
13:17 < AStorm> Guess it's the time to change it :P
13:18 < AStorm> I do want :P
13:18 < rxr> hehe - ok that is of course very welcome as well :-)
13:18 < AStorm> I just want to first see the failure, then disable the test for the stupid thing if I deem that safe.
13:18 < rxr> I guess it's best to fix the underlying package source then
13:18 < AStorm> Should be easy enough to just change run_check to be conditional.
13:18 < rxr> I also wonder if we run the check for native builds only
13:19 < rxr> I bet for cross builds most packages will fail to perform the checking anyway ...
13:19 < AStorm> For really cross ones, most likely.
13:19 < rxr> T2 always bootstraps a fresh cross compiler in the beginning
13:19 < rxr> stage 0 is "build cross compiler"
13:20 < rxr> stage 1 is "bootstrap fresh cross build stuff
13:20 < AStorm> Yes. And the autoconf failure is in stage0
13:20 < AStorm> I'd expect that :P
13:20 < rxr> and just stage 2 is "build in the fresh bootstraped sandbox"
13:20 < AStorm> Anyway, those autoconf tests have to be fixed a bit to be less stupid. Will do that.
13:21 < rxr> :-)
13:21 < AStorm> hah, it's impossible
13:21 < AStorm> They use diff on output, a braindead strategy.
13:21 < rxr> you can just drop a new patch in the t2 package/.../$pkg/ directory
13:21 < AStorm> I know.
13:21 < rxr> the patches will be picked up automatically ..
13:21 < rxr> AStorm: ah ok
13:22 < rxr> AStorm: most people expect more harcoding such as in other package formats ...
13:22 < rxr> just wanted to drop some basic-use information here so you have an easier life
13:22 < AStorm> rxr, well, I'm coming from Gentoo and have read the doc, even.
13:22 < AStorm> (poor English, but good enough)
13:22 < rxr> :-)
13:22 < rxr> http://www.linuxdevices.com/news/NS5501521223.html
13:23 < AStorm> Whoops, I mistook AVR32 for plain AVR.
13:23 < AStorm> That was ARM.
13:24 < AStorm> Highly interesting stuff...
13:30 < AStorm> BTW, I've some other ideas for T2 build system... eg. replacing chroot+loop with chroot+unionfs where available.
13:39 < AStorm> Yes, autoconf testsuite can't really be fixed.
13:41 < AStorm> The tests won't be enabled when building cross.
13:50 < AStorm> Hmm, one of the patches to vim contains "LIBTOOL-WUIRK"
13:50 < AStorm> Surely a tpyo.
13:53 < AStorm> Adding that NO-CHECK flag was very simple.
13:54 < AStorm> Will have to provide some variants maybe, like STAGE0-NO-CHECK
13:57 < AStorm> It's refreshing to change a distro after 6y with Gentoo.
14:05 < sepp> :)
14:07 < LMJ_Work> !> Binary file for rezound-0.12.3beta.tar.bz2 not present. Skipped in package database.
14:07 < LMJ_Work> !> Binary file for sun-jre-160-6.tar.bz2 not present. Skipped in package database.
14:07 < LMJ_Work> !> Binary file for vlc-0.8.6a.tar.bz2 not present. Skipped in package database.
14:08 < LMJ_Work> I stil have this warning even i removed those files from my custom pkgsel, how i can avoid this ?
14:09 < rxr> AStorm: you'r welcome :-)
14:09 < LMJ_Work> hello AStorm
14:09 < rxr> AStorm: for ideas you might also like to take a look at the GSoC site
14:09 < rxr> http://www.t2-project.org/developer/soc.html
14:10 < rxr> and some other roadmap bits are drafted here: http://www.t2-project.org/releases/
14:11 < AStorm> I've seen that...
14:18 < LMJ_Work> ping sepp
14:26 < CIA-8> rene * r22862 /trunk/architecture/avr32/package/binutils/binutils-fix-pico-in-out-regs.patch: * imported yet another avr32 binutils hotfix (this time for PiCo instructions)
14:42 -!- LMJ_Work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (EOF)"]
14:56 [Users #t2]
14:56 [@ChanServ] [ CIA-8] [ R4gnar0k] [ Shingoshi ] [ valentin]
14:56 [ AStorm ] [ LMJ ] [ rxr ] [ sparc-kly_]
14:56 [ Capey ] [ mtr ] [ sepp ] [ Stelz ]
14:56 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
15:23 < CIA-8> rene * r22863 /trunk/package/multimedia/vlc/ (vlc.cache vlc.conf):
15:23 < CIA-8> * fixed vlc firefox/seamonkey support, but hard-disabled vlc flac
15:23 < CIA-8> support as it is broken right now
15:27 -!- LMJ_work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
15:28 < sepp> guess flac support is really broken atm
15:30 < rxr> in vlc or in general ?
15:30 < sepp> general
15:30 < rxr> oh
15:30 < rxr> maybe we have a too new flac version?
15:31 < sepp> maybe
15:31 < rxr> but there is also quite some stuff building with flac support
15:31 < sepp> :)
15:31 < LMJ_work> arg
15:31 < rxr> LMJ_work: ?
15:31 < LMJ_work> disconnected
15:32 < rxr> # grep flac *.err | cut -d : -f 1 | uniq
15:32 < rxr> 5-alsaplayer.err
15:32 < rxr> 5-audacity.err
15:32 < rxr> 5-gst-ffmpeg.err
15:32 < rxr> 5-libsndfile.err
15:32 < rxr> 5-mpd.err
15:32 < rxr> 5-sox.err
15:35 < sepp> 5-hydrogen.err
15:35 < sepp> :)
15:37 < sepp> http://www.xfce-look.org/CONTENT/content-pre1/51274-1.jpg <- so nice :p
15:37 < rxr> r22554 | rene | 2007-02-16 15:54:57 +0000 (Fri, 16 Feb 2007) | 2 lines
15:37 < rxr> * updated flac (1.1.2 -> 1.1.4)
15:37 < rxr> sepp: hm?
15:37 < sepp> lol
15:37 < rxr> Martel Barby theme ?
15:38 < sepp> that is the worst thing ever
15:38 < rxr> barbi even
15:38 < rxr> hm - no idea how it is spelled
15:39 < sepp> they could even use it in guantanamo
15:40 < rxr> the flac 1.1 series goes rather way back
15:40 < rxr> and they claim 1.1.4 is way faster and such
15:40 < rxr> I guess we rather want to fix the compile errors currently seen
15:40 < rxr> than downdating something in the flac area ...
15:41 < sepp> oki
15:46 < LMJ_work> how to remove a package built with Build-Target (remove files inside build/
15:46 < LMJ_work> )
15:48 < LMJ_work> i could chroot and do a mine -rf
15:50 < LMJ_work> ?
15:50 < sepp> that would work
15:51 < LMJ_work> nothing better ?
15:53 < sepp> i would not remove glibc or something everything is linked again ...
15:54 < rxr> LMJ_work: a single package ?
15:54 < rxr> mine -r $pkg -r build/...
15:54 < rxr> or
15:54 < rxr> ./scripts/Create-ErrList -cfg $cfg -delete $pkg
15:54 < rxr> ./scripts/Create-ErrList -cfg $cfg -remove $pkg
15:55 < rxr> IIRC -delete removes all files and -remove removes just the log
15:55 < LMJ_work> ok thanks
15:55 < rxr> (I think this option should be clarified=
15:55 < rxr> )
15:59 < sepp> -[full|new]delete ], ohh i was not aware that there is 1 delete, but 3 ...
16:06 < rxr> yep
16:14 < LMJ_work> sepp : what do you think about jahshaka : http://pastebin.ca/385022
16:16 < sepp> dont know, the real error is earlier ...
16:17 < sepp> guess it has not found some header file
16:18 < LMJ_work> sepp: here we go : http://pastebin.ca/385027
16:19 < sepp> it needs glew
16:19 < LMJ_work> opengl related error ?
16:20 < sepp> better build glew 1st :)
16:20 < LMJ_work> gonna add glew
16:20 < LMJ_work> in progress
16:20 < LMJ_work> done ;)
16:22 < LMJ_work> ok, still an error
16:22 < LMJ_work> http://pastebin.ca/385036
16:24 < sepp> glew needs FPIC-QUIRK
16:24 < LMJ_work> glew compiled fine
16:24 < LMJ_work> ok, let's try
16:27 < LMJ_work> same error
16:27 < LMJ_work> but stop for today, gtg ;)
16:28 < rxr> sepp: do you commit marking glew FPIC-QUIRK then ?
16:29 < LMJ_work> be back later at home
16:30 < sepp> still the same error?
16:33 < sepp> will try if it compiles later
16:35 < AStorm> Huh, tar failing its own testsuite was hilarious :P
16:35 < AStorm> multivolume and sparse.
16:36 < AStorm> 99
16:36 < AStorm> 99,9% sure it's related to the build system and loop file.
16:56 < CIA-8> rene * r22864 /trunk/package/graphic/fbv/bigendian.patch: * basically fixed fbv for a bigendian frame-buffer on big-endian CPUs
16:58 < AStorm> Weird. I can't do that NO-CACHE flag properly it seems.
16:58 < LMJ> moin moin back ;)
16:58 < AStorm> Testing with hasflag in the scripts/functions.in fails - is always true.
17:00 < AStorm> At least in the place I do it. I'll try moving the test.
17:02 < AStorm> Yes, the flags are not yet available where the hook is added... but they are available in the hook itself.
17:04 < AStorm> Yes, now they do really work.
17:09 < rxr> :-)
17:09 < rxr> you send patches of your work to the list or so then?
17:10 < sepp> jahshaka (+ needed libs) builds on my amd64 ...
17:13 < LMJ> ok
17:13 < LMJ> gonna remove it ;)
17:14 < AStorm> rxr, it's just a new flag
17:14 < AStorm> I still think it's not enough :>
17:14 < AStorm> Uh, glibc hates me? :P
17:14 < AStorm> !> checking for .preinit_array/.init_array/.fini_array support... no
17:14 < AStorm> !> configure: error: Need linker with .init_array/.fini_array support.
17:15 < AStorm> It's stage 1 AFAICT.
17:15 < CIA-8> rene * r22865 /trunk/package/graphic/fbv/bigendian.patch:
17:15 < CIA-8> * fixed fbv to handle 24 bit per pixel framebuffer
17:15 < CIA-8> (so far only the big-endian code-flow was tested)
17:15 < AStorm> I think this is caused by SSP.
17:16 < sepp> glibc has NO-SSP flag ...
17:17 < AStorm> Well, yes, SSP is disabled.
17:17 < AStorm> But what would cause the error? I'll better check the config.log
17:18 < rxr> do not remeber that I saw that recently ...
17:19 * rxr wonders if I fix mplayer on 16bpp big-endian framebuffers, next ...
17:22 < AStorm> Is T2 binutils built against uclibc or dietlibc?
17:23 < rxr> oeh
17:23 < rxr> depends on your build
17:23 < rxr> when you do a glibc build it is of course build against glibc
17:23 < rxr> when you do a uclibc or dietlibc based build this changes, of course
17:24 < AStorm> Well, it's a glibc build.
17:24 < AStorm> And the bintutils with dietlibc is unchecked.
17:27 < AStorm> Something is very much off then :P
17:27 < AStorm> Either the build system hates my Gentoo :P
17:32 < AStorm> or the compiler that's been built is broken.
17:41 < AStorm> In case I'm unable to figure it out, I'll try from a LiveCD chroot.
17:42 < AStorm> Info: Gentoo ~x86_64 with multilib
17:42 < AStorm> But that should be irrelevant, as T2 SDE is building its own binutils.
17:46 < rxr> yes it should
17:59 < CIA-8> rene * r22866 /trunk/package/graphic/fbv/bigendian.patch:
17:59 < CIA-8> * and finally fixed fbv for 32 bit per pixel framebuffer on
17:59 < CIA-8> big-endian
18:12 < AStorm> Unless I manage to run T2, I may end up with a personal distribution.
18:12 < AStorm> Got the dependency and package logging scripts already sorted out...
18:12 < AStorm> (package logging by the use of unionfs ro/rw split :>
18:13 < AStorm> dependency detection could possibly be done using unionfs logging if available, otherwise LD_PRELOAD or stracing.
18:14 < AStorm> Huh? WTF?
18:14 < AStorm> Now flex failed the tests...
18:14 < AStorm> Something is very much off here.
18:14 < AStorm> (previously it didn't)
18:16 < AStorm> Sensors do look ok, no weird log messages - hardware is probably fine.
18:17 < AStorm> ah, previously, I probably had broken testing :P
18:17 < AStorm> CXX_not_available_in_stage_0 -c -o scanner.o -I. -I../.. -I../.. -g -O2 scanner.cpp
18:17 < AStorm> It seems I have to add even more flags.
18:18 < AStorm> Eg. NO-CHECK-STAGE0
18:25 < rxr> I hope you find T2 useful instead of starting your own scripts .-)
18:28 < AStorm> Maybe.
18:28 < AStorm> Except these are already started :-)
18:29 < rxr> yeah but once you sorted them out and fixed a few bugs you get 2600++ package descriptions and 17 or so CPU architectures for free :-)
18:29 < AStorm> make: Cross_CC_not_built_so_far_cc: Command not found
18:29 < AStorm> make: *** [/data/t2-trunk/src/Linux-x86_64/lua/lapi.o] Error 127
18:29 < AStorm> -> Error building the LUA bash accelerator.
18:29 < AStorm> Fun.
18:30 < AStorm> Shouldn't that try to build with system gcc then?
18:31 < AStorm> I'll probably use the package descriptions, but add more buildsystem support.
18:31 < AStorm> Maybe some option switches too (noticed none in T2)
18:32 < rxr> enable the expert options in the Config :-)
18:32 < AStorm> That's not exactly what I mean by options.
18:32 < AStorm> These are very lowlevel. I meant like per-package features.
18:32 < rxr> Config even has a few of those
18:32 < AStorm> E.g. you install, say, gnome, you can pick options.
18:33 < AStorm> E.g. no printer support- sure!
18:33 < AStorm> No accesibility, no problem at all.
18:33 < rxr> and the general stuff like jpeg or usb or not is controlled whether libjepg and co are selected in the target package selection
18:33 < AStorm> That's good.
18:34 < AStorm> Even good enough :-)
18:34 < rxr> and per package options exist too
18:34 < rxr> the whole Config stuff is build up dynamically by the various config.in files
18:34 < AStorm> I wonder how to test against something like ! (asstage toolchain && hasflag NO-CHECK-STAGE0)
18:35 < AStorm> *atstage
18:35 < rxr> e.g. vim has bloated, mediuam and tiny or so to 1:1 wire the assocaited vim feature set among other
18:35 < AStorm> Hmm... config.in... sounds... autotoolish, but could be nice.
18:35 < rxr> heh - no nothing todo with autotools
18:35 < rxr> (which I rather avoid like the plague ...)
18:36 < AStorm> rxr, good for you, I do too. M4 is evil.
18:36 < rxr> definetly ...
18:39 < AStorm> BTW, do you have any good description of the gem package format?
18:42 < rxr> the mine tarball includes some bites
18:42 < rxr> bits
18:42 < AStorm> Hm.
18:42 < rxr> but I hope you do not like that too much already, as the default is tarballs already and we plan to drop this .gem "nonsense" for future T2 releases
18:42 < rxr> did you already fall in love with it ?
18:42 < AStorm> "nonsense"?
18:42 < rxr> well - why "yet another package format" when vanilla tarballs are that handy ...
18:43 < AStorm> I'd store there 2 files more, maybe
18:43 < AStorm> E.g. dependencies
18:43 < AStorm> (in some form, not yet sure)
18:43 < AStorm> (have to be generic enough as to allow installing compatible packages)
18:44 < rxr> the tarballs has them
18:44 < AStorm> (maybe using library .version would be enough - with optional override too)
18:44 < rxr> in the usual T2 var/adm/* form
18:44 < AStorm> Well, that doesn't store any versioning info.
18:45 < AStorm> Just wait until someone tries to install an old tarball :D
18:45 < rxr> we sort those files to the front of in the tarball so the tar only needs to be partially decoded if just those are read
18:45 < AStorm> Tar is blocked anyway.
18:45 < rxr> the meta data is in it - jsut the install needs to use it
18:45 < AStorm> Shouldn't matter.
18:45 < AStorm> (unless the filesystem doesn't support seeks)
18:45 < rxr> and when you do not have the properitary installer around you still can use tar on any system from the last 20 years or so
18:45 < AStorm> 30 years.
18:46 < rxr> yeah
18:47 < AStorm> Optionally I'll also support .tar.7z
18:47 < rxr> but you can plug any other binary package creation between the build script lines as you wish of course as well
18:47 < rxr> yeah - just 3 lines of code to stick other compressors in :-)
18:49 < AStorm> Yep.
18:50 < AStorm> Still, the devel system has to fit on a pendrive.
18:50 < AStorm> With X and Xfce 4.4
18:50 < AStorm> And a lot of various interpreters/compilers.
18:50 < AStorm> Plus some libs.
18:50 < AStorm> Also other basic software.
18:51 < AStorm> T2 will be good _if_ I manage it to build.
18:51 < AStorm> I'll try the ISO in the chroot now.
18:52 < AStorm> Or maybe... I'll check the cross-compile switch.
18:53 < AStorm> plain LFS lookalike also failed to build here for weird reasons
19:03 < AStorm> I'm trying a cross build now.
19:08 < AStorm> If that'll work, Gentoo Glibc/GCC/Binutils get my personal IGNoble award.
19:27 < AStorm> Huh...
19:27 < AStorm> even the cross build failed like before.
19:27 < AStorm> So something is broken in T2 trunk.
19:28 < rxr> well - trunk works quite perfect on my box
19:29 < rxr> when your LFS attempts already failed maybe there really is some going on ...
19:29 < rxr> maybe just extract another system and chroot into it ...
19:30 < AStorm> I'll do that. Guess which system :P
19:33 * rxr out of office
19:33 < rxr> cu in the morning
19:33 < AStorm> Bye.
20:08 -!- tntnet_ [n=tommi@dslb-084-058-140-155.pools.arcor-ip.net] has joined #t2
20:24 -!- d-marc [n=d-marc@HSI-KBW-091-089-002-247.hsi2.kabelbw.de] has joined #t2
20:25 -!- tntnet_ [n=tommi@dslb-084-058-140-155.pools.arcor-ip.net] has quit [Remote closed the connection]
20:58 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
20:59 < misl> hi folks
21:00 < Stelz> hello misl :-)
21:01 < misl> hi Stelz :-)
21:29 -!- LMJ_work [n=c3196415@laf31-2-82-224-107-105.fbx.proxad.net] has quit ["CGI:IRC (Session timeout)"]
21:53 -!- d-marc [n=d-marc@HSI-KBW-091-089-002-247.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
23:06 -!- sepp_ [n=sepp@p85.212.50.206.tisdip.tiscali.de] has joined #t2
23:06 -!- sepp [n=sepp@p213.54.174.43.tisdip.tiscali.de] has quit [Read error: 54 (Connection reset by peer)]
23:09 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
23:15 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
23:16 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Client Quit]
--- Log closed Thu Mar 08 00:00:32 2007