--- Log opened Sat Apr 30 00:00:47 2005
00:15 < jsaw> http://sourceforge.net/tracker/index.php?func=detail&aid=1186644&group_id=93438&atid=604308
00:16 < mnemoc> jsaw: we have pkgconfig 0.16.0 now
00:16 < mnemoc> jsaw: that fixed 7 packages on my ref
00:16 < jsaw> I know.
00:16 < jsaw> just wanted to say that the bug is on the packages, not pkgconfig
00:17 < jsaw> hi mnemoc :)
00:17 < mnemoc> hi jsaw :)
00:17 < jsaw> how's ooo?
00:17 < mnemoc> runnign :D
00:17 < mnemoc> it needs a patch for pkgconfig 0.17 also
00:17 < jsaw> running good/bad/soso ?
00:17 < mnemoc> good
00:17 < mnemoc> good speed, not crashing yet
00:17 < mnemoc> fat fonts on menu
00:18 < jsaw> looking forward to test it
00:18 < mnemoc> jsaw: can you reply to the list tell that we should keep 0.17 on trunk and patch the broken packages?
00:21 < jsaw> hmmm. no. I'd rather wait for the packages to be fixed. The fact that pkgconfig doesn't look for indirect dependencies is rather a removed feature - I don't think this makes sense. But... for an SDE... hmm. we might have to accept the new style
00:22 < jsaw> or maybe yes, I can raise my concerns
00:22 < jsaw> .oO(what does mnemoc mean with reply?)
00:22 < mnemoc> answer to a post :)
00:23 < jsaw> mnemoc: is there an email I should reply to, or just send one based this
00:23 < mnemoc> we have to accept on trunk
00:23 < mnemoc> [t2] pkgconfig 0.17
00:23 < jsaw> based on this chat, I mean
00:23 < jsaw> ah,
00:23 < jsaw> ic
00:23 < mnemoc> :)
00:30 < jsaw> the wiki was well spamed...were there any traces left?
00:31 < mnemoc> just ip
00:31 < mnemoc> but every time is a different one
00:32 < mnemoc> i want to drop that wiki asap
00:33 < mnemoc> jsaw: do you have account on www2?
00:33 < jsaw> I forgot ...
00:34 < mnemoc> jsaw, juergen or george?
00:36 < jsaw> jsaw
00:37 < jsaw> I think there was one when rxr tried d.* (I forgot the cms name too...), but since plone I don't have one.
00:37 < mnemoc> uhm... your mail server refused the connection
00:37 < jsaw> my mail server?
00:37 < mnemoc> drupal
00:37 < jsaw> ah
00:37 < mnemoc> sawinski.de
00:37 < jsaw> shrug
00:37 < mnemoc> uhm?
00:38 < mnemoc> or give me another address to set :)
00:38 < jsaw> why should it...?
00:38 < mnemoc> to create you the account
00:38 < jsaw> is this server blacklisted somewhere?
00:38 < mnemoc> "connection refused"
00:39 < jsaw> if I can change the email address, try george at mpimf-heidelberg dot mpg dot de
00:39 < jsaw> for now
00:39 < mnemoc> i forced the other
00:39 < mnemoc> try to log in, and change the password
00:40 < mnemoc> i want to do a section with T2 girls :)
00:41 < jsaw> hehe
00:41 < mnemoc> uhm... lars has a boy, i forgot
00:41 < mnemoc> :)
00:42 < mnemoc> .oO( where do i set the priviledges on this thing.... )o
00:44 < valentin> lars has no boy ?
00:44 < jsaw> okay, have to go back to work...
00:44 < valentin> cu jsaw
00:44 < jsaw> *click* shadow-mode
00:44 < jsaw> hi/cu valentin :)
00:44 < valentin> mnemoc: lars has 1.9 girls
00:44 < mnemoc> valentin: oh
00:44 < mnemoc> valentin: the boy is yours?
00:45 < mnemoc> valentin: who have the boy?
00:49 < sepp_> good night all :)
00:49 < valentin> nope, we get a girl, too
00:50 < mnemoc> so, we will have our T2 Girls section :)
00:54 < valentin> minto and mech have sons iirc
00:54 < mnemoc> .oO( T2 Family? ... uhm.... )o
01:04 < valentin> how many ppl on "the other side" have children ?
01:05 < mnemoc> i guess none
01:05 < CIA-9> amery * r8591 /branches/2.1/package/security/chkrootkit/ (chkrootkit.conf make_install.patch): * merged 8550 from trunk: fixed chkrootkit to also install chkrootkit script - chkrootkit still broken (bug #23)
01:06 < mnemoc> wb CIA-9
01:06 < mnemoc> btw, http://www.geeks.cl/~amery/dsc01479.jpg <--- my first picture in 8 years ;) .... the guy of the right, green sweater.
01:11 < CIA-9> amery * r8592 /branches/2.1/package/network/ktcpvs/make_install.patch: * merged 8582 from trunk: patch for ktcpvs adapted - ktcpvs still broken, needs postlinux.conf
01:15 < CIA-9> amery * r8593 /branches/2.1/package/security/snort/snort.conf: * merged 8583 from trunk: fixed snort to look for automake 1.9 - still broken due to prelude patch
01:20 < CIA-9> amery * r8594 /branches/2.1/package/gnome2/ (libbonobomm/gtkmm24.patch libbonobouimm/gtkmm24.patch): * merged 8540 from trunk: fixed libbonobo{,ui}mm to use gtkmm24 - they still broken.
02:52 < valentin> mnemoc: who are you ? the person on the right ?
02:53 < valentin> oh yes - overread that
02:53 < valentin> but i was correct :)
02:53 < valentin> who is the rest ?
02:55 < valentin> the leftmost guy has an open shoe ,)
03:01 < mnemoc> hehe
03:01 < mnemoc> re
03:02 < mnemoc> it's my group of "software engineering workshop"
03:02 < valentin> who is the boss ?
03:03 < mnemoc> the leftmost guy :)
03:03 < valentin> with the open shoe :)
03:03 < valentin> that is sympathic
03:03 < mnemoc> yeah :)
03:03 < mnemoc> it's a nice group
03:03 < valentin> where did you take this picture ?
03:04 < mnemoc> on my left it is the classroom
03:04 < valentin> looks so empty ?
03:05 < valentin> ah - that building
03:06 < mnemoc> yes :)
03:07 < mnemoc> http://www.uc.cl/campus/joaquin/index.html <--- this is my campus
03:09 < valentin> uh - i need flash ?
03:10 < mnemoc> i'm installing it :(
03:10 < mnemoc> we should keep that stuff as we do for kernel modules
03:10 < valentin> is there anything like an open source flash implementation ?
03:11 < mnemoc> to compile flash files, yes
03:12 < valentin> i just want to view ...
03:12 < valentin> my only flashplayer i can use is the one under osx
03:13 < valentin> or i could go to my gf's x86 box, she has a linux-flash binary installed
03:13 < mnemoc> ftp://ftp.sgi.com/sgi/graphics/grafica/flash/
03:13 < mnemoc> http://ming.sourceforge.net/
03:14 < mnemoc> but they seem to be just generators
03:14 < keinek_> hi
03:14 -!- keinek_ is now known as keinek
03:14 < mnemoc> http://www.swift-tools.com/Flash/
03:14 < mnemoc> hi keinek
03:15 < valentin> hi keinek
03:15 < mnemoc> keinek: do you know a GPL/BSD flash player?
03:15 < keinek> swift is a great program.
03:15 < keinek> no only flash player for firefox.
03:15 < keinek> firefox install automatic for you
03:15 < mnemoc> nice
03:16 < mnemoc> valentin: http://www.swift-tools.com/
03:17 < mnemoc> valentin: will you package it?
03:17 < valentin> i can try if i get it working
03:34 < valentin> f**k i get a thousand lines of compilation errors
03:35 < mnemoc> :(
03:35 < valentin> hm, some of them are due to paralles make screwing up
03:35 < valentin> at least one is a "real" problem
03:35 < valentin> what was the NOPARALLEL flag again ?
03:36 < valentin> [F] NOPARALLEL
03:36 < valentin> ?
03:36 < mnemoc> make -j 1
03:36 < mnemoc> instead of make -j 4
03:36 < mnemoc> yes, [F] NOPARALLEL
03:36 < mnemoc> bellow [C]
03:37 < valentin> this guy mixes shell code within his makefiles - no wonder something breaks
03:39 < mnemoc> outch
03:39 < valentin> whatever this guy coded, the makefile does not stop on first error
03:40 < valentin> it runs on until all failed
03:40 < valentin> however, first problem is an undeclared sqrt (should be easy to fix)
03:40 < valentin> then it does not find qt/kde stuff - depending on how ugly the config stuff is ...
03:43 < valentin> http://www.this-page-intentionally-left-blank.org/
03:43 < valentin> they have a "member list" for followers
03:43 < valentin> just to raise our google score :)
03:47 < mnemoc> :)
03:48 -!- sparc-kly [~sparc@65-23-205-193.prtc.net] has quit [Read error: 104 (Connection reset by peer)]
04:00 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: jsaw, valentin, keinek, mtr, nzg, CIA-9, sepp_, mnemoc
04:02 -!- Netsplit over, joins: CIA-9, mtr, sepp_, keinek, mnemoc, jsaw, nzg, valentin
04:02 -!- [freenode-info] why register and identify? your IRC nick is how people know you. http://freenode.net/faq.shtml#nicksetup
04:32 -!- keinek [~Keinek@201.254.13.197] has quit [Remote closed the connection]
05:18 -!- keinek [~Keinek@201.254.13.197] has joined #t2
05:18 < keinek> hi
05:18 < keinek> :)
05:19 < keinek> somebody knows why my kde menu dont have things like access to Kate... and other programs?
05:19 < keinek> i know this is my fault
05:19 < keinek> but i dont know why.
05:19 < mnemoc> wait two hours and ask kde users :)
05:20 < keinek> ok
05:20 < keinek> :)
05:20 < keinek> now alsamixer work fine in the init
05:20 < keinek> :)
05:20 < mnemoc> good
05:21 < mnemoc> how did you solved your vmware problem?
05:21 < keinek> every time that i reboot, i do alsamixer before...
05:21 < keinek> no way
05:21 < keinek> vmware 5 keinek 0
05:21 < keinek> :(
05:21 < mnemoc> outch
05:21 < keinek> i dont know to much to solve my problem with the headers.
05:22 < mnemoc> postlinux.conf
05:24 < keinek> yes i know...
05:24 < keinek> i have the log.
05:32 < mnemoc> the difficult part is removing interactivility
05:32 < mnemoc> not finding the sources ;)
05:33 < keinek> the difficult part is learn what is headers.
05:33 < keinek> and what need and how work vmware.
05:33 < keinek> when i just want use that.
05:33 < keinek> :)
05:33 < mnemoc> :)
05:50 < CIA-9> amery * r8595 /branches/2.1/package/network/pilot-link/pilot-link.conf: * fixed pilot-link to not build java support, at least it doesn't build with kaffe/jikes
06:01 -!- ojh [~omer@67-42-199-46.eugn.qwest.net] has joined #t2
06:10 -!- keinek [~Keinek@201.254.13.197] has quit [Remote closed the connection]
06:13 -!- keinek [~Keinek@201.254.13.197] has joined #t2
07:06 -!- ojh [~omer@67-42-199-46.eugn.qwest.net] has left #t2 []
07:25 -!- mipe [~mika@dsl10040.japo.fi] has joined #t2
07:30 < CIA-9> amery * r8596 /branches/2.1/package/network/openldap/openldap.desc: * updated openldap (2.2.24 -> 2.2.26)
08:24 < mipe> damm, bootdisk failed after all
08:24 < mipe> http://koti.japo.fi/~pmika/error.log.bz2 for the error
08:24 < mipe> http://koti.japo.fi/~pmika/2-gcc.log.bz2
08:24 < rxr> moin
08:24 < mipe> http://koti.japo.fi/~pmika/1-gcc.log.bz2
08:24 < mipe> hi rxr
08:25 < mipe> had to compress them since i dont have much webspace there
08:25 < rxr> uhm
08:25 < mipe> 2 and 1 gcc is build with xtrace, i didnt see any noticable errors that could lead to this error.log.bz2
08:25 < rxr> damn - my U30 idled the night due to perl missing from the cross build and thus 0-autoconf was not built ..
08:26 < mipe> for short error is that i dont have stddef.h
08:27 < rxr> keinek: I track the kde menu empty issue
08:27 < rxr> mnemoc: do you keep your 2.1 branch updates and fixes in sync with trunk ?
08:27 < mipe> but have to go,back later. rxr if you are not busy, could you take a look at this?
08:29 < rxr> ouhm - I'm already quite busy working over the weekend
08:29 < rxr> but I can try
08:55 < mipe> prepare for little spam
08:55 < mipe> tar -cf - .; exit 0) | (cd /usr/lib64/gcc/x86_64-t2-linux-gnu/3.4.3/include; tar xpf - )
08:55 < mipe> tar: ./stdbool.h: file is unchanged; not dumped
08:55 < mipe> tar: ./stddef.h: file is unchanged; not dumped
08:55 < mipe> tar: ./pmmintrin.h: file is unchanged; not dumped
08:55 < mipe> tar: ./varargs.h: file is unchanged; not dumped
08:55 < mipe> tar: ./iso646.h: file is unchanged; not dumped
08:55 < mipe> tar: ./README: file is unchanged; not dumped
08:55 < mipe> tar: ./syslimits.h: file is unchanged; not dumped
08:55 < mipe> tar: ./stdarg.h: file is unchanged; not dumped
08:55 < mipe> tar: ./emmintrin.h: file is unchanged; not dumped
08:56 < mipe> tar: ./float.h: file is unchanged; not dumped
08:56 < mipe> tar: ./mmintrin.h: file is unchanged; not dumped
08:56 < mipe> tar: ./xmmintrin.h: file is unchanged; not dumped
08:56 < mipe> tar: ./unwind.h: file is unchanged; not dumped
08:56 < mipe> tar: ./limits.h: file is unchanged; not dumped
08:56 < mipe> on my system target this never happen for gcc,only on bootdisk target, this is the cause?
08:56 < mipe> all of the above files are missing
08:57 < mipe> oh damm,wrong paste
08:57 < mipe> rm -f /usr/lib64/gcc/x86_64-t2-linux-gnu/3.4.3/include/$i; \
09:02 < rxr> mipe: never noticed those unchanged; not dumped stuff
09:03 < rxr> in which state is that ?
09:03 < rxr> can you cd src.gcc......
09:03 < rxr> ./debug.sh
09:03 < rxr> which tar ?
09:05 < mipe> rxr:in a moment, i need to break gcc. gcc itself doesnt fail, it builds succefully. its just that after 2-gcc it rebuilds tools,which fails
09:23 < mipe> so that is during make install and tar is generated from objdir/gcc/include and it should dump them to /usr/lib....
09:27 < mipe> well actually that is during make install-headers
10:10 < rxr> mipe: still - wich tar is used?
10:11 < mipe> uhm, you mean the tar version?
10:13 < rxr> cd src.gcc* or so
10:13 < rxr> ./debug.sh
10:13 < rxr> which tar
10:14 < mipe> debug-gcc:[src.gcc.1114847853.21689.7f0100]# which tar
10:14 < mipe> /bin/tar
10:14 < mipe> debug-gcc:[src.gcc.1114847853.21689.7f0100]# tar --version
10:14 < mipe> tar (GNU tar) 1.15.1
10:14 < mipe> so system tar,not the one t2 builded
10:15 < mipe> have to try if its simple as changing install-headers method to cp and not tar, but that isnt the real solution
10:20 < rxr> which base system do you use ?
10:20 < rxr> btw. for me it is:
10:20 < rxr> which tar
10:20 < rxr> /t2-trunk/build/default-2.2.0-alpha-sparc64-v9vis-desktop/TOOLCHAIN/tools.cross/bin/tar
10:21 < mipe> hmm
10:21 < mipe> i have fc4test2
10:22 < mipe> but not using gcc4
10:22 < rxr> where did you run debug.sh ? in a stage2 or stage1 package build ?
10:23 < mipe> stage 2
10:32 < rxr> ah - then it is all right - that is in the sandbox ...
10:57 < CIA-9> rene * r8597 /trunk/ (5 files in 3 dirs):
10:57 < CIA-9> * added mdnsresponder (107), the Apple Zeroconf implementation
10:57 < CIA-9> used by KDE 3.4
10:57 < CIA-9> * registed and added APSL - Apple Public Source License
10:59 < mipe> looks like it works when install-headers method is cp and not tar
11:00 < mipe> atleast the header files are in place
11:02 < mipe> this is wierd problem, maybe because of tar? on system target tar is build against glibc and on bootdisk target tar is build against dietlibc?
11:02 < mipe> thats the only relevant diffrence i can come up with
11:02 < mipe> bbl
11:11 < rxr> it only happens in your bootdisk build ?
11:12 < mipe> rxr:yep
11:13 < rxr> mipe: hm
11:13 < rxr> intersting how long a bdb can build ...
11:13 < mipe> rxr:and it worked by changing install-header method,now bootdisk build is pass stage 2
11:13 < rxr> on my UltraSPARC
11:13 < rxr> hm
11:13 < rxr> maybe a bug in dietlibc on x86-64 ...
11:13 < rxr> (if tar is linkged against it)
11:14 < rxr> export ROCKCFG_DIETLIBC_tar='1'
11:14 < rxr> ouhm
11:14 < rxr> I think it is a dietlibc bug then ...
11:15 < mipe> would there be any other problems if its linked against glibc?
11:15 < rxr> well - I think the initrd extraction is using it
11:16 < rxr> so it needs to be tiny and statically linked ...
11:16 < rxr> we should fix the bug in dietlibc if there is on
11:16 < rxr> since soon dietlibc is used for any bootup when we switch to full udev support with a fully nodular kernel to boot on any hardware available in 2.2 soon
11:17 < mipe> ncurses is failing too,No fmod yet.make[1]: *** [unctrl.c] Error 1
11:18 < rxr> I suspect s.th. with stat ...
11:18 < rxr> or - hm
11:19 < rxr> maybe your box is too fast and the time does not differ ...
11:19 < rxr> mipe: we could timply fix it, by not passing the incremental option to tar
11:20 < rxr> hm - strange there is no incremental option passed ...
11:20 < rxr> mipe: do you want to debug why tar behaves that way?
11:20 < rxr> it is in tar src/create.c: WARN ((0, 0, _("%s: file is unchanged; not dumped"),
11:21 < rxr> I suspect if there is a bug with dietlibc, it will be in the stat code-path
11:23 < mipe> i just fetched dietlibc from cvs,i could try using that and if it works,i'll diff it more to get just the fix,unless you want the latest dietlibc from cvs,which might cause other troubles
11:24 < rxr> if you think there might be a fix in it ,-)
11:25 < mipe> according to changelog,it has quite many fixed on 64bit platforms,havent yet looked over what kind of changes
11:25 < mipe> fixed=fixes
11:41 < rxr> ouhm - just C-c'ed a hours building glibc build on sparc64 ...
11:41 < rxr> dam
11:41 < rxr> n
11:45 < valentin> moin
11:46 < rxr> hi
11:54 < valentin> 2.6.11.8 released
12:03 < rxr> yep
12:07 < rxr> == 09:46:52 =[9]=> Building develop/apr [0.9.6 2.2.0-alpha].
12:07 < rxr> == 04/30/05 09:55:12 =[9]=> Finished building package apr.
12:07 < rxr> == 09:55:17 =[9]=> Building develop/apr-util [0.9.6 2.2.0-alpha].
12:07 < rxr> == 04/30/05 10:01:13 =[9]=> Finished building package apr-util.
12:07 < rxr> == 10:01:18 =[9]=> Building textproc/libxml [2.6.17 2.2.0-alpha].
12:07 < rxr> ...
12:07 < rxr> I wonder if that is not a bit slow even for this U30 ...
12:12 < rxr> top - 12:18:22 up 16:42, 5 users, load average: 6.19, 5.12, 3.94
12:12 * rxr stresstesting the box a bit ,-)
12:12 < rxr> top - 12:18:47 up 16:42, 5 users, load average: 6.50, 5.28, 4.02
12:13 < rxr> top - 12:19:15 up 16:42, 5 users, load average: 6.78, 5.46, 4.12
12:29 < rxr> ok - ping works on sparc64 ,-
12:29 < rxr> )
12:30 < rxr> mnemoc: xmule 1.10.0b
12:34 < CIA-9> rene * r8598 /trunk/package/kde/kxdocker/kxdocker.desc: * updated kxdocker (0.29 -> 0.31)
12:54 < rxr> hey
12:54 < rxr> my native sparc64 svn build finished ...
13:30 < valentin> nice
13:31 < rxr> yep - when this continous, we would have gcc-3.4 sparc and sparc64 soon
13:31 < rxr> it only depends on how good the non basic stuff will build for the later
13:32 < rxr> so far it looks ok
13:32 < rxr> but the last time e.g. X did not start up in 64 bit mode ...
13:32 < rxr> I also do not throw CPU time on the 32bit build right now - because I do not want to have - let me count - 5 boxes powered up here ...
13:40 < rxr> screen does still refuce to work on sparc64
13:47 < valentin> will the fixes be pulled into stable ?
13:47 < rxr> sparc build fix ?
13:47 < rxr> I mean so far I did not fix much - the only basic linux-ehader problem was immediatly backported by myself
13:53 < CIA-9> rene * r8599 /trunk/package/ (92 files in 92 dirs): * removed dpkg from all .cache files - they are annoying ...
14:00 < CIA-9> rene * r8600 /branches/2.1/package/ (92 files in 92 dirs): * merged r5999 from trunk: removed dpkg from all .cache files
14:02 < rxr> it seems openbsd has also no functional sparc64 gdb ...
14:02 < rxr> well - let's hope the one I'
14:02 < rxr> m building right now does "s.th." ...
14:03 < rxr> damn - start gnome-control-center once and all your X resources are messed up
14:03 < rxr> damn crap ..
14:04 < rxr> now my frely started xemacs's are white ...
14:04 < rxr> *grmbl*
14:04 < rxr> and I do not really want to restart X right now - some builds running in not-screened xterms ...
14:09 -!- sparc-kly [~sunultra@65-23-205-193.prtc.net] has joined #t2
14:11 < rxr> == 04/30/05 12:17:02 =[9]=> Finished building package gdb.
14:12 < rxr> This GDB was configured as "sparc64-t2-linux-gnu".
14:12 < rxr> (gdb)
14:12 < rxr> ok - so far so good ...
14:12 < sparc-kly> :D
14:12 < rxr> (gdb) bt
14:12 < rxr> #0 0xfffff800004b1ac0 in sigsuspend () from /lib64/libc.so.6
14:12 < rxr> #1 0xfffff800005174f0 in pause () from /lib64/libc.so.6
14:12 < rxr> #2 0x000000000012a964 in getlogin ()
14:12 < rxr> #3 0x00000000001087d0 in ?? ()
14:12 < rxr> #4 0x00000000001087d0 in ?? ()
14:12 < rxr> Previous frame identical to this frame (corrupt stack?)
14:13 < rxr> hm - at least no crash ...
14:13 * sparc-kly reboot
14:13 -!- sparc-kly [~sunultra@65-23-205-193.prtc.net] has quit [Client Quit]
14:17 < mipe> damm,i guess i have to start debugging tar
14:19 < rxr> :-(
14:19 < rxr> mipe: I would setup a tiny testbed
14:20 < rxr> tar some test files and reproduce it ...
14:20 < rxr> if you managed to do so - just use gdb and tset a breakpoint at the place of the error message and examine that the stat structs contain ...
14:20 < rxr> sparc frooze ..
14:20 < rxr> damn
14:20 < rxr> it is starting again
14:21 < valentin> your fridge again ?
14:22 < rxr> no idea
14:22 < rxr> I somehow doubt it is the rdige
14:22 < rxr> fridge
14:22 < rxr> it seems to happen under high disk load ...
14:22 < sepp_> moin *
14:23 < rxr> maybe either a real kernel bug - or the fat IBM drives draw too much power ...
14:23 < rxr> moin sepp_
14:25 < mipe> hmm, how can i run it through gdb inside chroot? since outside t2 it works but inside it doesnt
14:25 < rxr> well - there is no gdb inside the bootdisk target, is there?
14:25 < mipe> nope
14:26 < rxr> do you have a more normal target of T2 handy?
14:27 < mipe> minimal
14:28 < rxr> if minimal has dietlibc and gdb I would charoot into it
14:28 < rxr> build a dietlibc linked tar in it and debug it there ...
14:30 < CIA-9> rene * r8601 /branches/2.1/package/base/screen/screen.desc: * changed screen to use the original gnu download URL
14:42 < CIA-9> rene * r8602 /trunk/package/base/screen/screen.desc:
14:42 < CIA-9> * merged 8601 from branches/2.1: moved screen download to official
14:42 < CIA-9> GNU server
14:46 < rxr> valentin: a C++ iterator style glob wrapper can be implemented in less than 25 lines *g*
14:46 < rxr> and it worked on first try ,-)
14:47 < rxr> a few more lines if it should be const friendly ,-)
14:55 < rxr> ach - damn fstream
14:55 < rxr> dosn't take a std::string as fname .. arghs
15:00 -!- sparc-kly [~sunultra@65-23-205-193.prtc.net] has joined #t2
15:02 < rxr> class ifstreamwstr : public std::ifstream
15:02 < rxr> {
15:02 < rxr> public:
15:02 < rxr> ifstreamwstr (const std::string& str)
15:02 < rxr> : std::ifstream (str.c_str()) {}
15:02 < rxr> ~ifstreamwstr() {};
15:02 < rxr> };
15:02 < rxr> the STL can be soo annoying ...
15:06 < mipe> hmm,how can i get symbol table in to binary. or how do i know where to set breakpoint?
15:06 < rxr> cool - works out
15:06 < rxr> mipe: you need to copmile with -g
15:06 < rxr> best is '-ggdb'
15:06 < rxr> and -O0
15:07 < rxr> with higher optimization debugging can get complex
15:07 < mipe> oh, what did -pg do?
15:07 < rxr> profiling
15:07 < rxr> and do not forget to _not_ use '-s' or 'strip' to strip the debug info off of the binary again ;-)
15:07 * rxr preparing lunch
15:12 < mipe> st=0x7fffffd7d130
15:21 < valentin> rxr: all those stream stuff uses too much char*
15:25 < mipe> is there a way to debug what a function call returns? without iterating it myself
15:26 < mipe> oh damm, its a define clause
16:24 < keinek> hi
16:24 < keinek> morning
16:25 < keinek> rxr thanks
16:25 < keinek> remember kde panel and menu change.
16:26 < keinek> kate, and another programs
16:26 < keinek> work
16:26 < keinek> but run call from shell
16:41 -!- rxr_ [~rene@85.178.174.126] has joined #t2
16:41 -!- Topic for #t2: T2 | 2.1.0-beta4 RELEASED | The next generation of System Development Environments (SDE) | http://www.t2-project.org/
16:41 -!- Topic set by menomc [] [Sat Apr 16 00:33:34 2005]
16:41 [Users #t2]
16:41 [ _Ragnar_] [ keinek] [ mtr] [ rxr_ ] [ valentin]
16:41 [ CIA-9 ] [ mipe ] [ nzg] [ sepp_ ]
16:41 [ jsaw ] [ mnemoc] [ rxr] [ sparc-kly]
16:41 -!- Irssi: #t2: Total of 13 nicks [0 ops, 0 halfops, 0 voices, 13 normal]
16:41 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
16:41 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup
16:41 -!- Irssi: Join to #t2 was synced in 11 secs
16:46 < mipe> hmm,it looks like it is stat stuff from dietlibc
16:50 -!- rxr [~rene@e178191252.adsl.alicedsl.de] has quit [Read error: 60 (Operation timed out)]
17:00 -!- You're now known as rxr
17:02 -!- veki [~veki@bunuel-cpe-54.nat-pool.bgd.sbb.co.yu] has joined #t2
17:09 -!- keinek [~Keinek@201.254.13.197] has quit [Read error: 54 (Connection reset by peer)]
17:12 < mipe> does bootdisk target just uses mine to install packages,no chrooting once base packages have been installed? just wondering if i should do x86 bootdisk for x86-64 system
17:16 < rxr> yes, that could work
17:16 < rxr> but it would be nice if you could fix dietlibc or tar or whatever ...
17:17 < mipe> i think its dietlibc, i dont even see stat64 for it
17:17 < mipe> on x86-64
17:17 < valentin> http://www.heise.de/newsticker/meldung/59176
17:24 -!- keinek [~Keinek@201.254.13.188] has joined #t2
17:34 < rxr> ok
17:34 < rxr> first C++ PCI probing drafted ...
17:34 < rxr> even mathces more accurate and more than e.g. the orig. ROCK awk script
17:35 < rxr> valentin: ack
17:36 < rxr> awk need:
17:36 < rxr> user 0m0.205s
17:36 < rxr> C++ needs:
17:36 < rxr> user 0m0.022s
17:37 < rxr> ROCK Plug needs:
17:38 < rxr> user 0m29.845s
17:38 < rxr> *grrr*
17:38 < rxr> on a 2500+ Athlon-XP ...
17:39 < mipe> rxr:any pointers where to start looking,i belive its a stat problem
17:40 < valentin> rxr: matches _more_ ? .oO
17:40 < valentin> bbl
17:41 < rxr> mipe: well directy in the stat syscall implementation
17:42 < rxr> I would antar dietlibc
17:42 < rxr> and tar within - buildit together with debug info and set a breakpoint at the dietlibc stat implementation and take a look what is done there ...
17:42 < rxr> mipe: have you looked what stat returns? pure junk more mostly reasonable values?
17:43 < rxr> for a start it would be nice to know what value is wrong
17:43 < rxr> valentin: hwscan:
17:43 < rxr> modprobe bttv
17:43 < rxr> modprobe snd-bt87x #no-initrd
17:43 < rxr> modprobe snd-ice1712 #no-initrd
17:43 < rxr> modprobe i2c-viapro
17:43 < rxr> modprobe via-agp
17:43 < rxr> modprobe snd-via82xx #no-initrd
17:43 < rxr> modprobe via-rhine
17:43 < rxr> modprobe budget-av
17:43 < rxr> my C++:
17:44 < rxr> Module: via-agp
17:44 < rxr> Module: shpchp
17:44 < rxr> Module: budget-ci
17:44 < rxr> Module: bttv
17:44 < rxr> Module: bt878
17:44 < rxr> Module: snd-ice1712
17:44 < rxr> Module: via-ircc
17:44 < rxr> Module: i2c-viapro
17:44 < rxr> Module: snd-via82xx
17:44 < rxr> Module: via-rhine
17:44 < rxr> as you can see the hwscan does not match some modules by class correctly
17:44 < rxr> e.g. the shpchp - pci hotplug module ...
17:45 < rxr> however now that I look at the IRC pasting, my implemetation seems to kac some too
17:45 < sepp_> how much free space do i need for a ooo build?
17:47 < jsaw> rxr: were's the code?
17:48 < rxr> jsaw: under development
17:49 < rxr> sepp_: a lot - over 2GB - I think nearly or more than 3 ...
17:49 < valentin> so hwscan sucks :)
17:49 < valentin> cu
17:49 < rxr> cu valentin
17:49 < sepp_> rxr: uhm - thanks
17:49 < rxr> jsaw: to be released as soon as I implemented more ...
17:49 < rxr> I hope tomorrow
17:50 < rxr> and I hope to finally have a udev + hotplug++ + embeddedSTL due to monday
17:51 < rxr> jsaw: I'm sorry, I'll quite reimplement most
17:51 < rxr> especially since you invented some streams and container that have an own API
17:51 < rxr> I want it STL compatible - so the base library is interchangable ...
18:03 -!- veki [~veki@bunuel-cpe-54.nat-pool.bgd.sbb.co.yu] has left #t2 ["Leaving"]
18:14 < mnemoc> rxr: i do syncs 2.1-trunk
18:14 < mnemoc> rxr: but most is done first on trunk
18:15 < rxr> mnemoc: sorry that I commit the screen change
18:15 < rxr> was in the wrong PWD ...
18:16 < jsaw> rxr: why sorry? I told you from the beginning..
18:18 < rxr> jsaw: because it so sad to delete all the work ... :-(
18:18 < mnemoc> uhm?
18:18 < jsaw> mnemoc: rxr's rewriting the hotplug++ code
18:19 < mnemoc> jsaw: all your work is being flushed?
18:19 < jsaw> I guess the outline will be left...but almost no code :)
18:20 < mnemoc> rxr: removing dpkg from .cache considering i will update the all was weird
18:20 < mnemoc> jsaw: oh
18:21 < mnemoc> jsaw: OT: we will have to change our CopyPatch to support your copyright on some headers
18:22 < jsaw> ?
18:22 < rxr> mnemoc: the intention was, that we always diff .cache changes
18:22 < rxr> mnemoc: and only commit if they make sence
18:22 < rxr> sense
18:23 < rxr> mnemoc: otherwise we would constantly break people's Emerge-Pkg
18:23 < rxr> and since dpkg has nothing to do on all the .caches (and it was scheduled to build on many Emerges this morning here) I deleted it, so we will noticy when it still tries to slip in after reference builds ..
18:23 < jsaw> bbl
18:23 < rxr> cu jsaw
18:27 < mnemoc> cleaning [DEP]s on .cache do reduce emerging can be a problem because on automatized priorization will ignore them
18:27 < mnemoc> [DEP] -foo
18:27 < mnemoc> [DEP] blah
18:27 < mnemoc> [DEP] +bar
18:27 < mnemoc> uhm....
18:27 < mnemoc> we need to do something
18:29 < mnemoc> -foo: foo can be a dependency, but we don't want to emerge them
18:30 < mnemoc> blah: blah is a good dependency, but we can build without it
18:30 < mnemoc> +bar: bar is a mandatory dependency
18:31 < mnemoc> and maybe... =poof: poof is mandatory but only needed at build time
18:31 < mnemoc> .oO
18:32 < mnemoc> -foo, -blah, poof, and +bar can be better
18:32 < mnemoc> .oO
18:34 < mnemoc> xmule-1.10.0b has the same wx25 problem than a
18:35 < rxr> mnemoc: I wanted to implement an optional specified in a 2nd column ...
18:35 < rxr> like
18:35 < rxr> [DEP] foo >= 1.2
18:35 < rxr> [DEP] foo optional
18:36 < rxr> where those additions are merged from manual specifications in the .desc
18:36 < rxr> [E] foo optional
18:36 < rxr> [E] foo >= 1.2
18:36 < rxr> or so ...
18:36 < mnemoc> looks like autoconf....
18:37 < rxr> where we of course only have to mark the optional ones ..
18:37 < rxr> or where we see space for important optimizations
18:37 < rxr> most packages will not have those ...
18:37 < mnemoc> i was looking for samething less intrusive and more automatizable
18:37 < rxr> well - you can not really automate much more in this area
18:37 < mnemoc> with .cache reporting you can
18:37 < rxr> the only think is to parse configure files with some regex and hope you match s.th. for many packages
18:38 < rxr> and you can track pkg-config invocations for new-style packages ...
18:38 < rxr> and apply some marks automatically
18:38 < mnemoc> "send us your .cache today!"
18:39 < mnemoc> .oO
18:40 < mnemoc> how do you plan to discriminate runtime/buildtime dependencies?
18:40 < mnemoc> only a few need gcc/binutils at runtime (example)
18:42 < rxr> what do you expect will need binutils ?
18:43 < mnemoc> gcc
18:43 < mnemoc> if i _install_ gcc i need to install binutils, bison, etc...
18:45 < rxr> well, ok
18:45 < rxr> (although no bison)
18:45 < mnemoc> :)
18:45 < rxr> well - basically we could differentiate if the dependency is a program
18:46 < rxr> and create a 2nd dependency mark (runtime-dep) where those are not used - normally ...
18:46 < rxr> that would only leave stuff like .h .so and so on in the runtime dep list
18:46 < mnemoc> you are making it more complex that what it deserves
18:46 < rxr> what do you propose ?
18:47 < mnemoc> a cgi to process .cache submittions with an overlay handled by us
18:47 < mnemoc> which is the only that alter official .cache files
18:48 < mnemoc> flagging [DEP]s with something less intrusive
18:48 < mnemoc> cheap to parse
18:48 < mnemoc> overlay on $confdir of course
18:48 < rxr> mnemoc: you asked about runtime deps - what has your cgi submitted todo with runtime deps ?
18:49 < mnemoc> having the 'overlay' at .desc is ugly
18:49 < rxr> ah - yes - of course we do not need a 2nd set - but could just use my "header/library or program" conditional to make them build-only ..
18:50 < mnemoc> i don't see that safe...
18:50 < rxr> ?
18:50 < rxr> It will be a quite good start - and I see [E] as an quite obvious mark to tweak them permanently ...
18:51 < sepp_> why not ldd and check the flists to get the runtime deps?
18:51 < mnemoc> sepp_: what about mozilla? ;)
18:52 < sepp_> hmm no idea what is with mozilla?
18:52 < mnemoc> not all apps use hardcoded .so dependencies
18:53 < mnemoc> some load them by code
18:53 < sepp_> hmm yes :(
18:53 < mnemoc> ldd tell things, but not everything
19:08 < mnemoc> uhm... my last 2.1->trunk sync was 8569....
19:08 < CIA-9> amery * r8603 /branches/2.1/package/network/xmule/ (install.patch xmule.desc): * updated xmule (1.10.0a -> 1.10.0b)
19:30 < rxr> ok - proc read out for PCI implemented, too
19:34 < mnemoc> rxr: i haven't get 8603 notification
19:35 < CIA-9> amery * r8604 /trunk/package/ (16 files in 9 dirs): (log message trimmed)
19:35 < CIA-9> * merged 8585,8586 from 2.1: moved wxpython24 to attic and disabled the only package which depends on it, gnue-designer.
19:35 < CIA-9> * merged 8587,8603 from 2.1: updated xmule (1.10.0a -> 1.10.0b) and fixed to use wxwidgets24
19:35 < CIA-9> * merged 8588 from 2.1: updated bogofilter (0.92.8 -> 0.94.6) - stable to beta because stable doesn't support bdb43
19:35 < CIA-9> * merged 8589 from 2.1: fixed fyre by not running update-{desktop,mime}-database
19:35 < CIA-9> * merged 8590 from 2.1: repriorized ooo after apache-ant
19:35 < CIA-9> * merged 8595 from 2.1: fixed pilot-link to not build java support
19:48 < rxr> mnemoc: notification still missing ?
19:48 -!- sepp_ [~sepp@p213.54.199.88.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)]
19:48 < rxr> we do not need ldd - we already have that all in the dep list
19:48 < mnemoc> rxr: yes
19:54 < mipe> 20:58:50 up 12:30, 4 users, load average: 5.06, 5.07, 5.15
19:54 < mnemoc> hi mipe
19:54 < mipe> hi mnemoc
19:55 < mnemoc> rxr: no notification yet
19:58 < rxr> hm
20:11 < mnemoc> rxr: i got 8604 but not 8603 yet
20:11 < rxr> hm
20:12 < mnemoc> but i can live with that
20:47 < mnemoc> finally 8603 got here
20:47 < rxr> hm
21:01 -!- mnemoc_ [~amery@200.75.27.67] has joined #t2
21:03 < rxr> ouhm
21:03 < mnemoc_> rxr: OT: after rebuild pilot-link after build was finished, evolution found it and build fine... but nothing significative new on .cache
21:04 < mnemoc_> rxr: did i miss anything?
21:05 < mnemoc_> ^--- this was about the 'ouhm'
21:05 < rxr> nope
21:05 < rxr> I just ouhmed about C cruft junk ...
21:05 < mnemoc_> :)
21:07 < rxr> it seems this crappy C junk - glob this time
21:07 < rxr> is overly dump
21:07 < mnemoc_> :)
21:07 < mnemoc_> # du -sh t2-download/{,*}
21:07 < mnemoc_> 6.1G t2-download
21:07 < mnemoc_> 2.7G t2-download/2.1.0-beta3
21:07 < mnemoc_> 3.7G t2-download/2.1.0-beta4
21:07 < mnemoc_> 512 t2-download/2.1.0-current
21:07 < mnemoc_> 512 t2-download/2.1.0-rc1
21:07 < mnemoc_> 512 t2-download/2.1.0-rc2
21:07 < mnemoc_> 512 t2-download/2.2.0-alpha
21:07 < mnemoc_> 512 t2-download/2.2.0-current
21:07 < mnemoc_> 4.1G t2-download/T2-current
21:07 < mnemoc_> 4.0K t2-download/unify.sh
21:12 -!- mnemoc [~amery@200.75.27.9] has quit [Read error: 110 (Connection timed out)]
21:16 < rxr> ok - was my fault
21:18 < mnemoc_> the crappy C junk?
21:25 -!- mnemoc_ is now known as mnemoc
21:25 < rxr> yeah - one can easily misuse ...
21:25 < rxr> good that I have a tiny C++ wrapper ... ,-)
21:25 < rxr> so I'll never fall into that pit again lll
21:26 < mnemoc> :)
22:26 < rxr> re
22:28 < rxr> mnemoc: do we want to relocate Documentation/Developer/* to misc/share for 2.1, too ?
22:28 < mnemoc> sure
22:28 < rxr> ok - I migrate it in trunk soon
22:29 < mnemoc> and i'll merge after you test it
22:29 < rxr> ack
22:29 < mnemoc> merging it as an atomic commit i don't break my rules
22:33 < CIA-9> rene * r8605 /trunk/architecture/share/kernel-block.conf.m4:
22:33 < CIA-9> * marked CONFIG_USB_STORAGE as modul (so it does accidently get
22:33 < CIA-9> built in ..)
22:34 < mnemoc> :)
22:34 < rxr> +e
22:39 < mipe> hmm,looks like dietlibc problem is fixed now
22:39 < mipe> but ncurses is failing now :(
22:39 < rxr> mipe: how did you fix dietlibc ?
22:40 < mipe> changing stat struct, {m,c,a}time from unsinged to signed
22:41 < mnemoc> uhm
22:42 < mipe> since on glibc those are time_t which is 16bit signed long, and dietlibc had unsigned long
22:45 < mipe> but /me going bed,i'll post dietlibc diff tomorrow,after i do some more testing
22:45 -!- mipe [~mika@dsl10040.japo.fi] has quit ["sleep"]
22:45 < mnemoc> gn8 mi
22:45 < mnemoc> pe
22:54 -!- mtr_ [~michael@p54AFA12C.dip0.t-ipconnect.de] has joined #t2
22:55 < rxr> ok - usb /proc parsing also hacked ...
22:55 < rxr> user 0m0.046s
22:56 < rxr> parsing PCI, matchin pci modules and parsing USB ...
22:56 < rxr> I'll hack USB and PCI into a releaseable state
22:57 < rxr> and after my first "all this new code" + udev + modular kernel boots hopefully tomorrow evening I'll implement ISAPNP and IEEE1394 thereafter ...
22:57 < rxr> and hack it into a releaseable form until monday I think ...
22:57 < rxr> n8 all
22:57 < rxr> cu early then ...
23:10 -!- mtr [~michael@p54AF9D5F.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
--- Log closed Sun May 01 00:00:48 2005