T2 IRC Log: 2005-01-04

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.

« prev | next »

--- Log opened Tue Jan 04 00:00:04 2005
--- Day changed Tue Jan 04 2005
00:00 < rxr> ehrm - of course:
00:00 < rxr> or s/\(match\)/\1\n/g
00:00 < mnemoc> ok... i'll to classic s///g with \n :p
00:05 < rxr> cool - did not know disown [-ar] [-h] [jobspec ...]
00:06 < mnemoc> is that what nohup does?
00:06 < CIA-9> jsaw * r5394 /trunk/scripts/functions: * revert r539[12] (bash workaround)
00:07 < rxr> mnemoc: only when -h is given .. ;-)
00:07 < mnemoc> :D
00:07 < rxr> mnemoc: and yes, this is why I find it interesting ... ;-)
00:07 < CIA-9> jsaw * r5395 /trunk/package/shells/bash/segfault-on-assign-dynamic.patch:
00:07 < CIA-9> * probably fix previous bash segfaults due to
00:07 < CIA-9> assigning certain variables
00:07 < mnemoc> rxr: bash3 only?
00:07 < rxr> mnemoc: AFA We K - yes
00:08 < rxr> it never was a problem ...
00:08 < jsaw> # ./src.wxpython.1104793593.24893.d9959e34/debug.sh
00:08 < jsaw> debug-wxpython:[TOOLCHAIN]# ls
00:08 < jsaw> . archdir chroot.sh debug.hooks dependencies.debug fl_wrapper.rlog install_wrapper.log xsrcdir.txt
00:08 < rxr> oh my hell ...
00:08 < jsaw> .. badfiles.txt cmd_wrapper.log debug.sh dependencies.txt fl_wrapper.wlog untar.txt
00:08 < jsaw> ERROR-LOG build.pid debug.buildenv debug_x.sh files.lst flist.txt wxPythonSrc-2.4.2.4
00:08 < jsaw> debug-wxpython:[TOOLCHAIN]#
00:08 < jsaw> so far this does not look bad.
00:08 < rxr> it seems I never executed a new bash since I updated the glibc and recompiled bash ...
00:09 < rxr> it now hangs as on my fresh build my athlon
00:09 < mnemoc> *G*
00:09 * rxr quickly reemerging as I already fixed it ...
00:10 < jsaw> ok, me now going into background mode
00:10 < rxr> jsaw: thanks for the work - and cu
00:10 < rxr> jsaw: maybe after the sleep cycle since I vanish soon
00:10 < jsaw> hehe, sleep well rxr.
00:10 < rxr> you too ;-)
00:13 < rxr> damn - straight way into bed or some food before ...
00:13 < rxr> grummel ...
00:15 < rxr> ./Update.sh cups 1.1.23
00:15 < rxr> anyone?
00:15 < rxr> Klear 0.1-alpha2
00:17 < rxr> PPCNerds has some news on the new images of the GNU-Darwin project now available, and includes a few screenshots.
00:17 < daja77> hm i could try the cups updatze
00:17 < rxr> daja77: using T2 :HEAD ?
00:17 < rxr> -> Compiler Cache Hits while pkg build: 99.20% (249 hits, 2 misses)
00:17 < rxr> ^ bash rebuild ... ,-)
00:17 < daja77> erh no
00:18 < rxr> haha
00:18 < rxr> welcome daja77 ;-)
00:19 * daja77 running fresh build, but in the beginning of stage 5, so cups update could be made ;)
00:20 < rxr> fresh t2 build then, huh?
00:21 < daja77> fresh crystal build )
00:21 < rxr> I know - I can be soooo annoying ,-)
00:22 < daja77> yeah you are afraid, that i moight steal all your code ...
00:22 * mnemoc moving home
00:22 < mnemoc> bbl
00:24 < rxr> daja77: nope, only that you do not specify the source correctly ...
00:24 < rxr> mnemoc: cu
00:24 < daja77> O_o
00:28 < daja77> .oO (ok 3d games are not running smoothly while building)
00:31 < rxr> *lol*
00:31 < rxr> nice the build ;-)
00:32 < rxr> or patch a idle scheduler into your kernel and priorize your build to only consume the idle cycles
00:32 < daja77> ^^
00:32 < daja77> the new briquolo levels are funny
00:33 < rxr> although the buld might take quite some time when you choose the later an run some heftly 3D shooter like Doom3 ...
00:34 < daja77> just a bit 3d breakout ;)
00:34 < rxr> ouh - even that is not smooth ?
00:35 < daja77> no
00:37 < rxr> it is a laptop?
00:37 < daja77> no
00:37 < daja77> wee cups dload finished
00:37 < CIA-9> rene * r5396 /trunk/package/printing/cups/cups.desc: * updated cups (1.1.22 -> 1.1.23)
00:38 < daja77> btw you could update briquolo as it breaks with gcc 3.4
00:39 < rxr> briquolo what *g* ? ;-)
00:39 < rxr> I take a look ;-)
00:39 < rxr> and thanks
00:39 < daja77> np
00:39 < daja77> on c3 i sent updates for most of my packages
00:39 < rxr> hm - without T2 announcements (which of course yield a high number of hits) my Avision stuff gets half of the hits T2 generate
00:39 < rxr> quite amazin how many linux Avision users are out there ...
00:40 < rxr> http://www.exactcode.de/oss/avision/
00:40 < daja77> i have a canon scanner ^^
00:46 < rxr> just for the sake of it I should review is one can exploid this bash bug - just to proof it to the bash author ...
00:46 < rxr> too bad time is that short ...
00:46 < daja77> indeed
00:46 < daja77> i shoudn't be here atm
00:48 < rxr> *lol*
00:49 < daja77> have an exam on thursday
00:50 < rxr> ugh - then better start learning ...
00:50 < daja77> oh i started already ;)
00:52 < rxr> s/start/continue/
00:53 < rxr> ;-)
00:53 < daja77> ;-)
00:54 < daja77> just uploaded cups patch ;)
00:56 < rxr> ok - then it takes 1 week to show up in the ROCK trunk ...
00:57 < daja77> :D
00:58 < rxr> good that you can still smile about that ;-)
01:01 < rxr> cu /me in bed
01:02 < daja77> cu
01:03 < CIA-9> rene * r5397 /trunk/scripts/functions:
01:03 < CIA-9> * fixed scripts/functions to not dump PWD and thus yields a wrong
01:03 < CIA-9> prompt until the first chdir in the debug enfironment ...
01:04 * daja77 removes the subversion repo from rxr's bed
01:44 -!- LINX_Explorer [~yo@208.234.54.170] has joined #t2
01:51 < mnemoc> hehe
01:52 < mnemoc> re!
01:52 < sparc-kly> hi mnemoc
01:52 < mnemoc> hi sparc-kly
01:53 < sparc-kly> hey t2 is official name ?
01:55 < mnemoc> it seem to have become official
02:02 -!- LINX_Explorer [~yo@208.234.54.170] has quit ["Leaving"]
02:03 -!- ideal [~idealm@211.100.227.107] has joined #t2
02:34 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has joined #t2
02:49 -!- rxr_ [~rene@p213.54.193.122.tisdip.tiscali.de] has joined #t2
02:49 -!- Topic for #t2: T2 | 2.1.0-beta2 RELEASED | The next generation of System Development Enviroments (SDE) | http://www.exactcode.de/t2
02:49 -!- Topic set by mnemoc [] [Tue Dec 28 14:55:44 2004]
02:49 [Users #t2]
02:49 [ _Ragnar_] [ daja77] [ jsaw ] [ N0V4K] [ rxr_ ] [ valentin]
02:49 [ CIA-9 ] [ ideal ] [ mnemoc] [ rxr ] [ sparc-kly]
02:49 -!- Irssi: #t2: Total of 11 nicks [0 ops, 0 halfops, 0 voices, 11 normal]
02:49 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
02:50 -!- Irssi: Join to #t2 was synced in 11 secs
02:53 < jsaw> firing up email reader
02:55 -!- rxr [~rene@p213.54.198.163.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
02:55 < mnemoc> can't this ${guy:-person} give a $badword answer?
03:08 < jsaw> no he can't.
03:19 < jsaw> but I'm afraid to say he is right. see RANDOM in man bash....
03:20 < mnemoc> that's a point, but bash can't crash if you set 'by accident' FUNCNAME
03:25 < jsaw> any idea how a buffer overflow in the bash could be used?
03:26 < mnemoc> against yourself?
03:27 < jsaw> huh?
03:28 < mnemoc> *g*
03:28 < jsaw> I just have the feeling this guy wont believe until you present him a demo-exploit.
03:28 < mnemoc> yes
03:31 < CIA-9> mnemoc * r5398 /trunk/package/printing/cups/cups.desc: * fixed cups download location to use http:// instead of ftp:// on that server
03:31 < mnemoc> do you have space to add a 'key recorder'?
03:32 < jsaw> space?
03:32 < mnemoc> memory to modify without killing that shell
03:34 < mnemoc> what kind of thing do you want to do with that overflow?
03:44 < jsaw> more fun:
03:44 < jsaw> # ash
03:44 < jsaw> $ export FUNCNAME=HELLO
03:44 < jsaw> $ bash
03:45 < jsaw> # echo $FUNCNAME
03:45 < jsaw> HELLO
03:45 < jsaw> ...
03:45 < jsaw> the variable is not set correctly anymore to be a dynamic variable...
03:45 < mnemoc> oh
03:45 < jsaw> just a normal environment variable from this time on...
03:45 < jsaw> *gah*
03:46 < jsaw> can we start using a different shell?
03:48 < mnemoc> i think we can generate posix-sh scripts to use during buildings, but the main scripts..... i think not
03:48 < jsaw> we have to go back to bash 2.05...
03:48 < jsaw> then
03:48 < mnemoc> i vote for that
03:49 < mnemoc> but, what was wrong with your workaround?
03:49 < jsaw> nothing. Only that a fixed bash would have been better.
03:50 < jsaw> We can't distribute a bash that may contain the possibility to break into your system...
03:50 < mnemoc> almost every contains possible exploits
03:51 < mnemoc> you may publish the problem on full-disclosure and see how $mr_nice_guy acts
03:51 < jsaw> yes of course. But the fact that the memory is corrupted is a clear hint that it might be exploitable.
03:51 < mnemoc> yep
03:52 < jsaw> I think, yes, this should go to full-disclosure...
03:52 < mnemoc> let the wolfs eat $mr_nice_guy
03:52 < jsaw> okay. enough for know... have to go back to my books...
03:53 < mnemoc> have you tryed on non-t2 bashes?
03:53 < jsaw> ?
03:54 < mnemoc> fedora or bsd?
03:54 < jsaw> no.
03:54 < jsaw> bsd most probably does not have bash 3.0 yet... But I might check this later.
03:54 < mnemoc> if it fails only here our bashing may get against us
03:54 < mnemoc> sure, go to study
03:54 < mnemoc> see you
03:54 < jsaw> cu
03:55 < jsaw> (but a good proposal. I have to check other sources first, thanks? I should sleep over it anyway...)
03:55 < jsaw> s/?/!/
03:57 < mnemoc> *g*
04:07 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has quit [Remote closed the connection]
04:09 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has joined #t2
04:14 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has quit [Remote closed the connection]
04:15 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has joined #t2
04:44 < CIA-9> mnemoc * r5399 /trunk/package/filesystem/ntfsprogs/ntfsprogs.desc: * updated ntfsprogs (1.9.0 -> 1.9.4)
05:45 < CIA-9> jsaw * r5400 /trunk/package/shells/bash/ (comp-wordbreak.patch segfault-on-assign-dynamic.patch):
05:45 < CIA-9> * previous fix for the bash segfault was not correct
05:45 < CIA-9> * a new fix, but this time only affecting get_comp_wordbreaks:
05:45 < CIA-9> there's some out-defined code, when re-activated makes
05:45 < CIA-9> the segfault disappear... waiting for comment from chet
05:45 < CIA-9> * the FUNCNAME att_nounset fix is kept
06:02 < CIA-9> mnemoc * r5401 /trunk/package/base/uclibc/config-650.in: * fixed uclibc's config to disable glibc using it's current name
06:06 < CIA-9> mnemoc * r5402 /trunk/package/base/uclibc/parse-config-9:
06:06 < CIA-9> * fixed uclibc parse-config to handle new name of linux-headers
06:06 < CIA-9> * improved uclibc parse-config to handle .conf patches
06:07 < CIA-9> mnemoc * r5403 /trunk/package/base/uclibc/uclibc.conf: * fixed uclibc to eval preconf
06:09 < CIA-9> mnemoc * r5404 /trunk/package/base/uclibc/patches/ (8 files):
06:09 < CIA-9> * improved uclibc to patch binutils using sed
06:09 < CIA-9> * fixed uclibc binutils patches to current version
06:09 < mnemoc> gn8 *
06:09 < jsaw> gn8 ...
10:40 < rxr_> moin
10:40 -!- You're now known as rxr
10:52 < rxr> so - we have t2-project.org, soon
11:08 < rxr> GNU bash, version 3.00.16(1)-release (amd64-portbld-freebsd5.3)
11:08 < rxr> ^- does not crash on bashcrash.sh ...
11:08 < rxr> but this might be pure luck
11:08 < rxr> e.g. I saw no special patch in the ports collection that should prevent it ...
11:33 < rxr> whois t2-project.org
11:33 < rxr> does also already resolve here
11:48 * rxr shortly away
12:20 < CIA-9> rene * r5405 /trunk/package/gnome2/gedit-plugins/:
12:20 < CIA-9> * removed gnome2/gedit-plugins - it already was disabled and is
12:20 < CIA-9> abondone (or so)
13:15 -!- nzg [~tschmidt@p508EB359.dip.t-dialin.net] has joined #t2
13:17 < rxr> hi nzg
13:18 < nzg> moin rxr
13:44 < jsaw> re
13:48 < jsaw> 13:07 < praenti> bash is terrible. why they cannot use unified diffs...
13:48 < jsaw> hehe
14:02 -!- ideal [~idealm@211.100.227.107] has quit ["Leaving"]
14:08 < rxr> moin jsaw
14:13 < rxr> jsaw: are you in par with sparse files ?
14:14 < mnemoc> moin
14:15 < mnemoc> rxr: have you experienced the problem with SSP symbols?
14:16 < mnemoc> rxr: i think there is something on the new toolchain which catches host stuff on build
14:25 < rxr> no - none yet
14:25 < rxr> i have a strange problem
14:25 < rxr> I create a sparse file (using lseek ...)
14:25 < rxr> I then write and close open
14:25 < rxr> and later reads work fine - returning zeros ...
14:26 < rxr> but when I replace the close / open with a lseek(fd, 0L, SEEK_SET);
14:26 < rxr> I can not read the sparse zeros in the beginning - read return -1 ...
14:26 < rxr> does anyone seen this?
14:26 < rxr> how to work arouns this?
14:26 < mnemoc> -1??
14:29 < rxr> yes - I'm puzzled, too
14:36 < rxr> read: Bad file descriptor
14:36 < rxr> ...
14:36 < mnemoc> uhm
14:36 < rxr> yeah
14:38 < jsaw> hi rxr
14:38 < jsaw> hi mnemoc
14:38 < rxr> ouhm - might be due to the creat in this damn excersie ...
14:38 < mnemoc> hi jsaw
14:38 < rxr> --typos ...
14:38 < rxr> hi jsaw
14:38 < mnemoc> :D
14:39 < jsaw> have no idea about sparse files...
14:39 < rxr> creat is equivalent to open with flags equal to
14:39 < rxr> O_CREAT|O_WRONLY|O_TRUNC.
14:39 < rxr> ok - this is the cause ..
14:41 < rxr> yep - that is was ...
14:44 < mnemoc> aha
14:52 < rxr> so - next exercise finished
14:52 < rxr> I now get lunch and then hope to finish the exercise due to in two weeks so I have it done ...
14:53 < rxr> (getting the annoying stuff from my desk ...)
14:57 < mnemoc> jsaw: last protector doesn't apply to our gcc :( why do we need every gcc spp patched?
15:08 < jsaw> brb
15:12 -!- mnemoc_ [~amery@200.75.27.19] has joined #t2
15:12 -!- mnemoc [~amery@200.75.27.1] has quit [Nick collision from services.]
15:16 -!- mnemoc_ is now known as mnemoc
15:34 < CIA-9> mnemoc * r5406 /trunk/package/base/uclibc/ (fix_shares_00dirtree.patch uclibc.conf): * fixed uclibc to really not share with 00-dirtree
15:47 < rxr> so
15:48 < mnemoc> not applying spp i'm on 1-linux26 of minimal-uclibc
15:54 < mnemoc> rxr: take a look into linuxrc.c:774, init is not finished!
15:56 -!- mtr [~michael@Ha5a3.h.pppool.de] has joined #t2
15:56 < mtr> hi
15:57 < mnemoc> hi mtr
15:57 < mtr> hi mnemoc
16:02 < mtr> mnemoc: seems you are working on the uclibc stuff again, that's great
16:03 < mnemoc> i want to do a bootdisk (uboot) based on it and busybox
16:03 < mnemoc> mtr: if you want to help you are very welcome :D
16:05 < mtr> mnemoc: uboot, :-) sounds good, if i can help to get the uboot running, i will do so
16:06 < mtr> afair there is still the uncommited fl_wrapper patch, without this you will not get beyond stage-1
16:06 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has quit ["FreeBSD N0V4K.Zero 5.3-RELEASE FreeBSD 5.3-RELEASE Chao"]
16:06 < rxr> mnemoc: I know uboot partially - how do you think it will improve our install target ?
16:09 < mnemoc> rxr: does uboot exist?
16:10 < mnemoc> rxr: currently stage1 is useless, it's almost imposible to do anything different than next next next
16:11 < rxr> yes
16:11 < rxr> it is a multi architecture boot loader ...
16:11 < rxr> I think created for PowerPC embedded boards and later exapnded to SPARC and x86 and many others ...
16:12 < mnemoc> but hardcoding C is not a easier/cleaner way to expand
16:12 < rxr> http://sourceforge.net/projects/u-boot
16:12 < rxr> stage is just to load stage2 ...
16:13 < rxr> it should handle module loading, swap setup, network config and then loding from some device or net ...
16:13 < mnemoc> yes but i have _never_ been able to do it
16:13 < rxr> I have always been ...
16:13 < rxr> http://u-boot.sourceforge.net/
16:14 < mnemoc> something cleaner may be easier to do it work on everyone
16:15 < rxr> actually the linuxrc.c was not the worst thing I "got in touch" so far - e.g. mine is far uglier ...
16:15 < mnemoc> i was thinking uboot target as bootdisk based on uclibc and busybox which would make the build much shorter and cleaner
16:15 < rxr> can't you just fix linuxrc in the way you need ?
16:15 < rxr> why cleaner ?
16:15 < mnemoc> mount doesn't mount my discs
16:16 < rxr> it does mount mine ...
16:16 < mnemoc> without fighting with diet
16:16 < rxr> e.g. I also patched a mount bug in iss ...
16:16 < mnemoc> i know it mount yours
16:16 < rxr> what hardware do you have?
16:16 < rxr> does this happen for all - or only for some ...
16:16 < rxr> btw. I do not want uclibc in some official target
16:16 < rxr> more likely is a dietlibc only install target ...
16:16 < mnemoc> i have done it in atleast 12 different machines and contexts
16:17 < mnemoc> why not?
16:17 < rxr> well - I never had a problem - on many machines tested ...
16:17 < rxr> strange
16:17 < rxr> because uclibc is a big ugly hack?
16:17 < rxr> pasted from dozens of other c libraries, unclear mixed license conditions
16:17 < rxr> and fat compared to dietlibc ... ?
16:18 < mnemoc> yes, but it builds without hacks and work on all machines
16:18 < mnemoc> i have never had problems with bg-rescue mounting and playing
16:19 < mnemoc> but bootdisk just doesn't work on the hardware i have tested
16:19 < rxr> is this all old hardware - or also stuff newer than pentium ?
16:19 < mnemoc> stage2 is fine, but i can't get into that
16:19 < mnemoc> rxr: p2, p3, p4
16:19 < rxr> give me a download link to an ISO and I test it ...
16:20 < rxr> mnemoc: does it happen with my ISOs ?
16:20 < mnemoc> yes
16:20 < rxr> the former ROCK or beta2 one ?
16:20 < mnemoc> yes
16:20 < mnemoc> yes
16:20 < rxr> h
16:20 < rxr> m
16:20 < mnemoc> since 1.7
16:20 < rxr> I bet you do something ugly with your boxes
16:20 < mnemoc> kernel panics, mount problems, unable to symlink using kiss, etc etc
16:21 < rxr> maybe you pull the power-cord when you test the CD?
16:21 < rxr> maybe you do not hug the box often enough ?
16:21 < mnemoc> hehe
16:21 < rxr> a computer needs some love - otherwise it will not work reliable ...
16:21 < rxr> really ... - no joke intended
16:21 < mnemoc> i know they are female ;)
16:22 < mnemoc> my last iso mount and loads stage2 on laptop and home but kernel panics after linuxrc.c's main return 0
16:23 < mnemoc> and i coundn't mount it and play with kiss
16:25 < mnemoc> i'll rebuild bash and create a new iso for you
16:25 < rxr> well I have not that much free time really this week ...
16:26 < mnemoc> :\
16:26 < rxr> give it a low priority ...
16:26 < rxr> ()
16:26 < rxr> (since when are computers female ... - I mean - what do femals to please the box then ...)
16:26 < mnemoc> rxr: take a look into linuxrc.c:774 ?
16:26 < rxr> okok
16:26 < rxr> did I mentioned I need to hurry to uni ... ?
16:26 < mnemoc> :p
16:26 < mnemoc> cu rxr
16:26 < rxr> nah
16:27 < rxr> I look into linuxrc.c right now
16:27 < mnemoc> =)
16:27 < rxr> yeah - a TODO, and ?
16:27 < rxr> this is only used for the in-system bootup later ...
16:28 < rxr> as it is there it will work and let the kernel remount the root filesysmte
16:28 < mnemoc> btw. I do not want uclibc in some official target <--- this made me very sad, i dislike the need of using diet helper but my main goal for $uboot is busybox
16:28 < rxr> I already implemented the TODO two months ago
16:28 < rxr> but it is uncommited on my U30 ..
16:28 < mnemoc> grrrrr
16:28 < rxr> grrrrr what ?
16:28 < mnemoc> the uncommited stuff which makes me suffer
16:28 < rxr> why does it make you suffer ?
16:29 < rxr> in which use-case do you need that ?
16:29 < jsaw> re
16:29 < rxr> hi jsaw
16:29 < jsaw> hi rxr, mnemoc
16:29 < rxr> mnemoc: I mean to do you plan to use it for the in-system initrd or what ?
16:29 < mnemoc> installing on my laptop it kernel-panics there
16:29 < jsaw> mnemoc: do you need the fl_wrapper patch for uclibc?
16:29 < mnemoc> on next next next
16:29 < rxr> huh?
16:30 < rxr> if ( root && !strncmp("/dev/ram", root, 8) )
16:30 < rxr> stage1(); /* never returns */
16:30 < mnemoc> jsaw: i guess, but i'm on stage 1
16:30 < rxr> this is the code-path for the boot-cd bootup
16:31 < jsaw> mnemoc: it is a patch were you can pass "FLWRAPPER_LIBC" environment variable to tell fl_wrapper from which libc the symbols should be loaded
16:31 < mnemoc> jsaw: ok
16:31 < rxr> mnemoc: I can commit it this night if you really need it ..
16:32 < CIA-9> jsaw * r5407 /trunk/misc/tools-source/fl_wrapper.c:
16:32 < CIA-9> * add the possibilty to fl_wrapper which libc to dlopen
16:32 < CIA-9> by passing FLWRAPPER_LIBC env var
16:32 < mnemoc> rxr: i don't know why i get there, but it panics right after that message
16:32 < rxr> mnemoc: then most likely the argument parsing and /proc/cmdline parsing has some defect
16:32 < jsaw> I couldn't load stage2, gzip is killed due to oom
16:32 < mnemoc> jsaw: i'll add the export to uclibc's parseconfig now
16:32 < rxr> or you do not pass the right parameters from you boot loader
16:32 < mnemoc> rxr: just hit enter
16:32 < rxr> worked here ...
16:33 < mnemoc> i don't doubt it works for you anf for most people
16:33 < rxr> mnemoc: if you really think uclibc is that nice implement it - propose it - and/or add it as alternative bootdisk target
16:33 < rxr> bootdisk2 .
16:33 < rxr> ...
16:33 < rxr> mnemoc: instrument it with many printf's and follow the argument parsing ..
16:34 < mnemoc> i want a more customizable bootdisk, ash script maybe
16:34 < jsaw> I'd really go for dash.
16:34 < mnemoc> i have dash-0.5.1-3 as package here but doesn't build
16:34 < jsaw> I just hacked dynamic library loading into dash, just for fun.
16:35 < mnemoc> :D
16:35 < jsaw> mnemoc: you need the additional debian patch
16:35 < mnemoc> that's the -3 thing
16:35 < rxr> can you please commit it
16:35 < rxr> I did not found the dash thing IIRC and would like to review it ...
16:35 < rxr> so - /me off into uni - cu all later
16:35 < mnemoc> my broken dash or jsaw's ?
16:36 < jsaw> I don't have a package of dash, just compiled it by hand.
16:36 < jsaw> But I can create one.
16:36 < mnemoc> i'll commit and you fix
16:36 < mnemoc> that's faster
16:36 < jsaw> hehe :)
16:36 < mnemoc> dash is maintained by "my-beloved" runit's author
16:37 < jsaw> did I already mention that with dietlibc, stripped, it's 75k btw
16:37 < jsaw> ?
16:37 < mnemoc> :)
16:38 < mnemoc> i think we can easily use dash to do chrooted builds
16:38 < jsaw> only that we have to make our scripts POSIX compliant...
16:38 < jsaw> and that is, eh, will, ... don't know, is it feasible?
16:39 < mnemoc> not all the scripts
16:39 < jsaw> do we use arrays?
16:39 < mnemoc> hooks :\
16:39 < mnemoc> we can dump that into posix functions
16:40 < jsaw> only hooks? Then it should be easy to fix.
16:40 < CIA-9> mnemoc * r5408 /trunk/package/shells/dash/ (. dash.desc): * adds dash 0.5.1-3 (debian ash)
16:41 < mnemoc> jsaw: inside chroot bash blah i guess that's the only complex code
16:58 < jsaw> fixed.
16:59 < mnemoc> fixed what?
16:59 < jsaw> dash
16:59 < mnemoc> =)
16:59 < mnemoc> what was my mistake?
16:59 < jsaw> no
17:00 < mnemoc> no?
17:00 < jsaw> wait a sec
17:00 < jsaw> I meant you haven't done a mistake.
17:00 < CIA-9> jsaw * r5409 /trunk/package/shells/dash/tempfile-fix.patch: * we do not have the "tempfile" command, it is mktemp here
17:01 < mnemoc> i saw that thing.... but i forgot to patch it :\
17:01 < jsaw> it was crucial...
17:01 < mnemoc> -> Compiler Cache Hits while pkg build: 0.00% (0 hits, 2152 misses) <--- is ccache target dependant?
17:02 < jsaw> Compiler Cache Hits while pkg build: 24.32% (9 hits, 28 misses)
17:02 < jsaw> (dash)
17:02 < jsaw> hmmm.
17:02 < mnemoc> Compiler Cache Hits while pkg build: 100.00% (1020 hits, 0 misses) <--- 1-uclibc :D
17:03 < jsaw> !
17:04 < jsaw> btw, gnucash's port to gnome2 is on it's way... let's see how long it takes.
17:04 < jsaw> Should we bet what comes earlier? Gtk-3 or Gnucash for Gtk-2?
17:05 < mnemoc> btw, i had to disable spp apply to be able to build :'(
17:05 < mnemoc> gtk3 :D
17:05 < jsaw> :)
17:07 < jsaw> no news from chet... I hope my bash patch is correct... it's even in rock sm now.
17:09 < mnemoc> that #if 1 was a guess or you hunted the bug until there?
17:12 < jsaw> chet replied.
17:12 < jsaw> :)
17:12 < jsaw> :( but doesn't make me feel well.
17:13 < jsaw> and the answer is once more -eh- how should I say that: vague?
17:13 < mnemoc> full-disclosure?
17:14 < jsaw> obviously no way around that...
17:14 < mnemoc> they will find the way to make $fun with this bug
17:15 < jsaw> it is this one, or:
17:15 < jsaw> https://lists.netsys.com/mailman//listinfo/full-disclosure
17:17 < mnemoc> yep
17:18 -!- N0V4K [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has joined #t2
17:53 < rxr> 17:53 < jsaw> bbl
18:17 < rxr> I just did some t2 pr handing my exercies ...
18:17 < rxr> "what linux is that?"
18:17 < rxr> "erhm - one that works ... - e.g. no suse or debian ..."
18:17 < rxr> ;-)
18:59 < CIA-9> martin * r5410 /trunk/package/kde/amarok/ (. amarok.cache amarok.conf amarok.desc): * added amarok (1.2-beta2)
19:12 < CIA-9> martin * r5411 /trunk/package/textproc/kile/ (kile.conf kile.desc):
19:12 < CIA-9> * added kile.conf, including the kde-defaults
19:12 < CIA-9> * fixed kile.desc and took over maintainship
19:32 -!- sparc-kly [~sparc@64.237.130.154] has quit [Read error: 104 (Connection reset by peer)]
19:34 -!- nzg [~tschmidt@p508EB359.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)]
19:42 -!- praenti [~praenti@mail.obster.org] has joined #t2
19:51 < CIA-9> martin * r5412 /trunk/package/audio/mpd/ (mpd.conf mpd.init): * added mpd.init and the .conf installing it
19:53 < mnemoc> does unset 'unexports'?
20:20 -!- N0V4K is now known as N0V4K|AT-04
20:20 * N0V4K|AT-04 is away: Estoy ocupado
20:23 * N0V4K|AT-04 is away: Estoy ocupado
20:23 * N0V4K|AT-04 is away: off
20:23 * N0V4K|AT-04 is back (gone 00:00:10)
20:23 * N0V4K|AT-04 is away: Estoy ocupado
20:23 * N0V4K|AT-04 is back (gone 00:00:01)
20:24 -!- nzg [~tschmidt@p508EB359.dip.t-dialin.net] has joined #t2
20:29 -!- N0V4K|AT-04 [~N0V4K@246.Red-80-33-149.pooles.rima-tde.net] has left #t2 ["Adios"]
20:55 < rxr> re
20:55 < rxr> mnemoc: do not know off hand
21:09 < rxr> this chart: http://svn.exactcode.de/svn.png from http://svn.exactcode.de/ is pretty cool ... ,-
21:12 < mnemoc> i think t2 and rock must be in one graph and the rest in another
21:13 < rxr> okok
21:13 < rxr> but our line is quite cool ... ;-)
21:15 < mnemoc> yeah :D
21:42 -!- maddin [~martin@pD956B345.dip.t-dialin.net] has joined #t2
21:42 < maddin> re hi
21:43 * maddin like Yogi tea ... :)
21:43 < mnemoc> hi maddin
21:50 < maddin> hi mnemoc
21:50 < rxr> hi maddin
21:50 < rxr> my stomarch likes to get any food at all ...
21:51 < rxr> mnemoc: I now improve the svn commit graphs ...
21:51 < maddin> hmmm, is "maddin lige yogi tea" right? Or it's like "maddin ist wie yogi tee"? :)
21:51 < rxr> a bit ...
21:51 < rxr> you most probably just forgot the trailing "s" on "like" ...
21:52 < rxr> when you wanted to express that you like jogi tea ...
21:52 < rxr> since it is simple present ...
21:52 < rxr> this "he, she, it - s muss mit" thing ...
21:53 < maddin> ups, yes. I know. And I have think so.
21:56 < mnemoc> rxr: what do you use for those graphs?
21:57 < rxr> ploticus
21:57 < rxr> why ?
21:57 < mnemoc> curious
21:57 < rxr> ploticus matured to my favourite tool to plot (non 3d) stuff
21:58 < rxr> but it is ugly hacked C - and I once debugged an annoying background fill bug where the author did not wanted to apply my fix ...
21:58 < mnemoc> i don't plot stuff but i good to have a 'selection' for the moment the need comes
21:58 < rxr> we have it as t2 package
21:58 < rxr> but I patch it to install the binary as ploticus rather than pl conflicting with prolog ...
21:58 < mnemoc> i supposed :)
21:59 < mnemoc> oh
21:59 < rxr> so keep care when you hit ploticus or pl on some foraign system
21:59 < mnemoc> some systems have perl as pl too :\
21:59 < rxr> urgs ..
22:03 < jsaw> re
22:03 < rxr> hi
22:04 < jsaw> hi usual suspects! :)
22:05 < rxr> uhm?
22:05 * rxr was not at CCC if you mean that ...
22:06 < jsaw> but mnemoc and rxr are the "usual suspects" in _this_ channel. (ok ok, me too)
22:07 < rxr> aha ;-) *lol*
22:07 < mnemoc> :(
22:07 < rxr> yeah - you too
22:07 < jsaw> valentin is more like a shadow warrior...
22:07 < mnemoc> hehe
22:08 * jsaw ate a clown for breakfast -sorry-
22:09 < rxr> someones birthday? yours, doughter ... ?
22:10 < mnemoc> nope
22:10 < jsaw> "eating a clown for breakfast" is some term I learned from my wife's family , meaning something like being in a "funny/strange" mood
22:11 < mnemoc> err, minimal doesn't have readline
22:11 < jsaw> huh
22:11 < mnemoc> that's a bit 'unconfortable'
22:14 < mnemoc> rxr: why your svn.png doesn't label Y completely?
22:14 < rxr> because I somehow messed the y scaliing when I first hacked it ...
22:14 < rxr> or maybe due to a bug in ploticus ...
22:15 < rxr> I just hack on the graph anyways ...
22:16 < CIA-9> mnemoc * r5413 /trunk/package/network/openldap/openldap.desc: * updated openldap to the new "most stable release" (2.2.19 -> 2.2.20)
22:17 < mnemoc> rxr: do you have Download -all on cron?
22:17 < rxr> yep
22:17 < mnemoc> rxr: how did you solved progress-bar/progress-meter crap?
22:18 < rxr> I have a cron-work script that does pipe that into a log (I nearly never look into ...)
22:18 < mnemoc> rxr: i want to receive output, but not e^n lines of progress
22:19 < mnemoc> :(}
22:19 < rxr> pipe into log and grep it away ...
22:19 < mnemoc> s/\}//
22:29 < jsaw> hmm. my bash patch at least hides the actual error. With it, valgrind reports no error, while it does if it is not applied...
22:29 < mnemoc> "coincidence"
22:29 < jsaw> :)
22:30 < jsaw> there's some hidden code path that screws it up...
22:34 -!- mtr_ [~michael@H803c.h.pppool.de] has joined #t2
22:34 -!- mtr [~michael@Ha5a3.h.pppool.de] has quit [Read error: 110 (Connection timed out)]
22:44 < rxr> could someone commit s.th. I need to test my tuned log parser ...
22:44 < mnemoc> what kind of shit?
22:44 < mnemoc> i have two adds pending
22:44 < rxr> yeah - s.th. like that ..
22:45 < rxr> well - ok - I could also much in the files to construct a testcase ...
22:46 < mnemoc> i'm open to test your script
22:47 < rxr> just commit when you want
22:51 < mnemoc> one on road
22:51 < CIA-9> mnemoc * r5414 /trunk/package/gnue/ (5 files in 2 dirs):
22:51 < CIA-9> * introduced gnue repository for GNUe packages
22:51 < CIA-9> * adds gnu-common (0.5.12)
22:52 < rxr> ok - works as expected - thanks
22:52 < mnemoc> :)
22:52 < rxr> in some minutes we will have split up graphs
22:53 < rxr> but I did a major performance / network bandwith tuning to make the log obtaining less expensive ...
22:54 < mnemoc> :)
23:05 < rxr> ouhm - need to update our ploticus package to utilize a new feature ...
23:10 < maddin> gn8 all
23:10 < jsaw> gn8 maddin
23:10 < rxr> cu maddin
23:10 -!- nzg [~tschmidt@p508EB359.dip.t-dialin.net] has quit [Nick collision from services.]
23:11 -!- tschmidt__ [~tschmidt@p508EB359.dip.t-dialin.net] has joined #t2
23:11 -!- nzg [~nazgul@xchangecenter.de] has joined #t2
23:11 -!- maddin [~martin@pD956B345.dip.t-dialin.net] has left #t2 ["Verlassend"]
23:28 < CIA-9> rene * r5415 /trunk/package/graphic/ploticus/ploticus.desc: * updated ploticus (221 -> 230)
23:34 < rxr> urgs bash 3.0 does parse
23:35 < rxr> lrev ++
23:35 < rxr> where bash 2 does not ...
23:36 < CIA-9> mnemoc * r5416 /trunk/package/gnue/gnue-common/gnue-common.conf: * fixed gnue-common typo sourcing file
23:36 < mnemoc> uhm?
23:36 < rxr> don't ask me ...
23:37 < rxr> ./generate.sh: line 11: lrev ++ : syntax error: operand expected (error token is " ")
23:37 < mnemoc> (( lrev++ )) ?
23:37 < rxr> http://svn.exactcode.de/
23:37 < rxr> yep ..
23:39 < rxr> now the graphs are also properly limitted - this is a new feature in ploticus 230 - perviously limitting was not available for curvefit plots ...
23:39 < rxr> only e.g. for normal non-interpolated line plots ...
23:39 < mnemoc> i see it as always
23:40 < rxr> reload button ?
23:40 < rxr> there are two graphs now ...
23:40 < mnemoc> without proxy it worked
23:40 < rxr> ic - agressive caching ...
23:41 < rxr> any further complains ?
23:41 < mnemoc> no compains ;)
23:42 < mnemoc> 'average speed' statistics between rock and t2? ;)
23:42 < rxr> speed ?
23:42 < rxr> you mean svn bandwidth or security updates ?
23:42 < mnemoc> commits per $timerange
23:43 < rxr> this is already slightly averaged to remove to noisy parts ...
23:43 < mnemoc> time svn log can't be accurate considering one is on your machine
23:43 < rxr> do you want some real big average like e.g. one week ?
23:43 < mnemoc> that's ok to me
23:43 < rxr> who cares - at least the t2 graph would rock then ...
23:43 < mnemoc> :D
23:43 < rxr> 23:43 < mnemoc> that's ok to me
23:43 < rxr> you mean I sou
23:44 < rxr> should add week averaging ?
23:44 < mnemoc> rock's patch bombs make graph lie
23:45 < mnemoc> week average as a third and forth line sounds atractive to me
23:45 < rxr> ic - you want me to keep me away from my real work ...
23:45 < mnemoc> :p
23:45 < rxr> btw - I got two old 1995 or so PowerPC boards ...
23:46 < rxr> 133Mhz or so ...
23:46 < mnemoc> i want to port my payed work to gnuenterprise :p
23:46 < rxr> I think IBM - at lesat they are PReP conforming and one SCSI disk carries an AIX label ...
23:46 < rxr> mnemoc: payed what, where ?
23:47 < mnemoc> here, i have to eat too ;) and i don't know to write articles
23:48 < rxr> nah - I meant just to know some details ;-) what gnuenterprise what and where and so ...
23:48 < mnemoc> gnue.org as platform to my payed development
23:57 < rxr> http://www.gnuenterprise.org/
23:58 < rxr> just to make the irc log parser happy converting google trackable a href's out of it ...
23:58 < rxr> what do you mean using it as platform ?
23:58 < mnemoc> :D
23:58 < mnemoc> yes
--- Log closed Wed Jan 05 00:00:40 2005