--- Log opened Tue Dec 21 00:00:41 2004
00:09 < rxr> jsaw: feel free to apply the fixed version of the cmd-wrapper patch
00:11 < jsaw> okay. A bit later however - have to finish some important documents
00:11 < rxr> sure
00:11 < rxr> daja77: ouhm - your SECURITY updates seem to be rotting in submaster ... :-(
00:13 < rxr> first I tought this is a t2 copy:
00:13 < rxr> http://www.rocklinux.net/submaster/data//2004/12/1800292111471.patch
00:13 < rxr> but opening our files reviles we do not do it that hardcoded ...
00:14 < rxr> oh my is the stuff in the subjunk old ...
00:14 < mnemoc> ar $taropt $base/download/mirror/X/XFree86-*.tar.bz2 <--- god
00:15 < rxr> except the CVS checkouts they do for no real value ...
00:15 < rxr> and spamassasin I should udpate one day
00:15 * rxr does not yet want to test the update on the production server ...
00:27 < daja77> rxr: *shrug* i have them on my system ;)
00:28 < rxr> yeah M;-)
00:29 < mnemoc> mirroring rock' svn is damn slow
00:30 < mnemoc> mirroring t2's took ~1.5 per revision
00:30 < mnemoc> mirroring rock's is about +10 seconds per revision
00:31 < mnemoc> even 40-50 seconds on some revisions
00:31 < daja77> yep damn latency
00:33 < rxr> so much to tsa's comment on the rock mailing list that he never liked important stuff hosted on a private DSL line ...
00:33 < rxr> but rock's svn is always strangely slow ...
00:34 < rxr> but for some strange reason back-diffs in http://svn.exactcode.de SVN repositories is quite slow - I remeber it was faster in the past ...
00:34 < rxr> I need to review this ...
00:34 < mnemoc> rxr: what about moving it to fsfs?
00:35 < rxr> that should not make such a huge difference - but I review soon ...
00:35 < mnemoc> fsfs keeps every revision in a file
00:37 < rxr> yeah -> Compiler Cache Hits while pkg build: 4.71% (55 hits, 1113 misses)
00:37 < rxr> == 12/21/04 00:29:06 =[5]=> Finished building package gimp.
00:39 < mnemoc> 2.2.0?
00:39 < CIA-9> rene * r5158 /trunk/package/graphic/gimp/ (compile.patch gimp.desc): * updated gimp (2.0.6 -> 2.2.0)
00:39 < mnemoc> :D
00:42 < rxr> I have some more tiny updates in queue
00:46 < CIA-9> rene * r5159 /trunk/package/graphic/libsdl/ (5 files): * updated libsdl (1.2.7 -> 1.2.8)
00:59 < CIA-9> rene * r5160 /trunk/package/gnome2/gnumeric/gnumeric.desc: * updated gnumeric (1.4.0 -> 1.4.1)
01:28 < mnemoc> how is nexus theses days?
01:28 < rxr> why do you ask ?
01:28 < mnemoc> curious
01:28 < rxr> what do you wanna know ?
01:28 < rxr> bandwidth wise ?
01:28 < mnemoc> yes
01:30 < rxr> well - often shows up before gsmp, the amd64 cpu is of course doing the job compared to the 450Mhz PII in the gsmp ...
01:32 < rxr> so far we have 36 mirror cgi accesses for beta2
01:34 < mnemoc> =)
01:34 < mnemoc> and i'm not one of those :p
01:34 < rxr> and if I would find out how to grep for \t I could tell for -beta ...
01:35 < mnemoc> grep -e :)
01:35 < rxr> does not work for me ...
01:36 < mnemoc> -E ?
01:39 < rxr> grep is enough - and then [[:blank:]] does the trick
01:39 < rxr> :-(
01:39 < rxr> 40 for -beta
01:39 < rxr> 40 uniq
01:40 < rxr> 62 with doublicate (IP) accesses ...
01:41 < rxr> we need prebuilt t2 versions soon ...
01:41 < rxr> people like installing prebuilt stuff ...
01:42 < mnemoc> yes
01:43 < mnemoc> and cross-built bases
01:43 < mnemoc> they may at least be able to start their own builds on a friendly enviroment
01:44 < rxr> I try to fix cross builds over the xmas days ...
01:45 < mnemoc> fixing native and bootdisk we can beta2 before that, what do you think?
01:45 < mnemoc> that=xmas
01:45 < rxr> what do you mean with fixing native ?
01:45 < rxr> jsaw: could you apply the patch? I would like to restart the bootdisk build ... ;-)
01:46 < mnemoc> bootdisk is building for me
01:46 < rxr> here it failed
01:47 < rxr> [3] base/psmisc
01:47 < rxr> 162 builds total, 80 completed fine, 1 with errors.
01:48 < rxr> /usr/include/termcap.h:25: error: conflicting types for 'tgetnum'
01:48 < rxr> /usr/include/term.h:776: error: previous declaration of 'tgetnum' was here
01:48 < rxr> and many more ..
01:48 < rxr> has anyone seen this before ?
01:48 < CIA-9> mnemoc * r5161 /trunk/ (. package/base/dietlibc/parse-config-9): * set DIETHOME on parse-config if we use dietlibc
01:48 < mnemoc> a failing #include?
01:49 < rxr> nope
01:50 < mnemoc> 145 matches on google :)
01:54 < rxr> damn - f*cking gnomemeeting still not building ...
01:54 < rxr> now it wants the next package: howl ...
01:55 < mnemoc> i have been 17 hours mirroring rock and it is on rev 2276
01:56 < rxr> why do you do this ?
01:57 < rxr> (and I thought you gave up on svk ..)
01:57 < mnemoc> to keep my 2.0 branch good merged with my changes on t2
01:57 < mnemoc> angood merged with 2.0-stable
01:59 < jsaw> rxr: okay
02:04 < CIA-9> jsaw * r5162 /trunk/misc/tools-source/cmd_wrapper.c: (log message trimmed)
02:04 < CIA-9> * re-ordering cmd_wrapper:
02:04 < CIA-9> - different detection of executed wrappers: store these
02:04 < CIA-9> in the ENVPREFIX "_WRAPPER_OTHERS_DONE" env var and
02:04 < CIA-9> subtract it from ENVPREFIX "_WRAPPER_OTHERS" to avoid
02:04 < CIA-9> loops
02:04 < CIA-9> - move the ENVPREFIX "_WRAPPER_NOLOOP" after execution of others
02:04 < rxr> howl packaged ..
02:05 < rxr> let's see if this is now finally all that pleases gnomemeeting ...
02:05 < rxr> jsaw: thanks
02:05 < jsaw> CIA is not even finished...
02:05 < jsaw> rxr: de nada
02:05 < rxr> CIA: said "(log message trimmed)"
02:06 < jsaw> ic
02:06 < mnemoc> jsaw: studying spanish?
02:07 < rxr> -> Compiler Cache Hits while pkg build: 6.32% (6 hits, 89 misses)
02:07 < jsaw> mnemoc: sadly no, but our labrarian is from brazil iirc...
02:07 < rxr> ^- howl built the first time ...
02:07 < rxr> configure crap cached ... ;-)
02:07 < jsaw> mnemoc: so once in a while, a few short spanish sentence reside in my brain :)
02:08 < mnemoc> brazil speaks portuguese, not spanish :)
02:08 < jsaw> ups, list a few spanish south american countries please so I might remember correctly
02:09 < mnemoc> argentina, chile, uruguay, paraguay, peru, bolivia, ecuador, colombia, venezuela
02:10 < jsaw> still not sure... I think the last one.
02:10 < mnemoc> there are lot more on _central_ america :)
02:10 < jsaw> he's definitely from the south.
02:11 < mnemoc> so, those are.
02:12 < jsaw> okay. a colleague just told me, he's from argentinia.
02:13 < rxr> -> Compiler Cache Hits while pkg build: 1.80% (2 hits, 109 misses)
02:13 < rxr> == 12/21/04 02:10:35 =[5]=> Finished building package gnomemeeting.
02:13 < mnemoc> people from buenos aires (the capital) is _very_ pedantic, the rest are nice
02:14 < jsaw> mnemoc: then he is definitely not from buenos aires!!!
02:14 < mnemoc> s/is/are/
02:14 < jsaw> ;)
02:14 < mnemoc> :)
02:18 < rxr> <0 degree celcius ...
02:19 < jsaw> -11 expected in Stuttgart tomorrow night... they have snow, but until we arrive there for xmas, temperature is expected to be above 0...
02:19 < rxr> :-(
02:19 < rxr> we had snow sunday evening ...
02:19 < jsaw> nice.
02:19 < mnemoc> we have never had snow for xmas
02:20 < rxr> well, ... ;-)
02:20 < mnemoc> and i guess we will never have
02:20 < jsaw> Heidelberg like so often: snow on the hills, shitty (ice cold) rain in the city
02:23 < mnemoc> http://www.mail-archive.com/fink-users@lists.sourceforge.net/msg02586.html <--- uhm
02:27 < jsaw> hmm.
02:28 < rxr> do you see any value in the tiny script that does generate most of my commit messages nowadays ?
02:28 < CIA-9> rene * r5163 /trunk/package/ (4 files in 2 dirs):
02:28 < CIA-9> * updated gnomemeeting (1.0.2 -> 1.2.0)
02:28 < CIA-9> * added howl (0.9.8)
02:29 < rxr> so far 37 lines of Commit.sh ... ;-)
02:30 < CIA-9> rene * r5164 /trunk/target/desktop/config.in: * added gnomemeeting and dependences to the desktop target
02:30 < mnemoc> built?
02:32 < rxr> yep
02:32 < rxr> -> Compiler Cache Hits while pkg build: 1.80% (2 hits, 109 misses)
02:32 < rxr> == 12/21/04 02:18:34 =[5]=> Finished building package gnomemeeting.
02:33 < rxr> so - the next non-compiling thing lingering around
02:33 < rxr> gnucash
02:38 -!- rxr_ [~rene@p213.54.195.163.tisdip.tiscali.de] has joined #t2
02:38 -!- Topic for #t2: T2 | 2.1.0-beta RELEASED | The next generation of System Development Enviroments (SDE) | http://www.exactcode.de/t2
02:38 -!- Topic set by mnemoc [] [Sat Dec 18 15:39:34 2004]
02:38 [Users #t2]
02:38 [ _Ragnar_] [ daja77] [ mnemoc] [ rxr ] [ valentin]
02:38 [ CIA-9 ] [ jsaw ] [ nzg ] [ rxr_]
02:38 -!- Irssi: #t2: Total of 9 nicks [0 ops, 0 halfops, 0 voices, 9 normal]
02:38 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
02:38 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup
02:38 -!- rxr [~rene@p213.54.192.193.tisdip.tiscali.de] has quit [Nick collision from services.]
02:39 -!- Irssi: Join to #t2 was synced in 21 secs
02:39 -!- You're now known as rxr
02:39 < rxr> so
02:39 < rxr> re
02:39 < mnemoc> :)
02:39 < rxr> did you got all?
02:39 < rxr> 02:32 < rxr> -> Compiler Cache Hits while pkg build: 1.80% (2 hits, 109 misses)
02:39 < rxr> 02:32 < rxr> == 12/21/04 02:18:34 =[5]=> Finished building package gnomemeeting.
02:39 < rxr> 02:33 < rxr> so - the next non-compiling thing lingering around
02:39 < rxr> 02:33 < rxr> gnucash
02:40 < mnemoc> just until :32
02:40 < mnemoc> [22:34:02] :D
02:40 < mnemoc> [22:34:38] btw, don't forget ld.so.conf globbing on ld :)
02:41 < rxr> yeah - do you want/need it this night ?
02:41 < mnemoc> not really :)
02:41 < mnemoc> gnucash is a pain to build and a pain to use
02:42 < rxr> why do we have it then ?
02:42 < rxr> I mean I never tried to use it .. and most probably never will ..
02:46 < mnemoc> good question, and no answer cames to my mind
02:46 < rxr> hehe
02:46 < mnemoc> :)
02:46 < rxr> I try to update and fix it for 5 minutes if no end in sight I will leave it alone for now ..
02:46 < mnemoc> i think it's there just becasue 'it has to be'
02:47 < mnemoc> O it :)
02:47 * daja77 hands rxr an I hate gnucash banner
02:47 < jsaw> gnucash?
02:47 < jsaw> it should built already!
02:48 < jsaw> build I mean
02:48 < rxr> oehm - I think not ..
02:48 < rxr> didn't the bdb patch apply at all?
02:49 < jsaw> there's no bdb patch anymore
02:49 < jsaw> r4802r4802
02:49 < jsaw> ups
02:50 < rxr> ic ...
02:50 < rxr> it is gnome 1.4v ?
02:50 < rxr> we should move it then ...
02:50 < mnemoc> i think i'll add gambas :)
02:50 < jsaw> rxr: yes
02:51 < rxr> how can one maintain a gtk 1.x app nowadays ...
02:51 < jsaw> the maintainer knows....
02:52 < rxr> knows ?
02:52 < rxr> ah - yes ;-)
02:52 < mnemoc> tries?
02:52 < mnemoc> what happened to gimp-print?
02:52 < rxr> mnemoc: you mean this:
02:52 < rxr> http://gambas.sourceforge.net/2004-09-06.png
02:52 < mnemoc> :p
02:53 < mnemoc> i have mono, gtk#, _clip_, ... i need to add haskell and gambas :)
02:57 < rxr> :-(
02:57 < CIA-9> rene * r5165 /trunk/package/ (contrib/g-wrap/ develop/g-wrap/ develop/pwlib/ extra/pwlib/): * moved extra/pwlib and contrib/g-wrap -> develop
02:57 < daja77> clip ....
03:00 < rxr> hehe:
03:00 < rxr> >> > I was wondering, how do you pop a register (like EAX) into
03:00 < rxr> > a C variable?
03:00 < rxr> >
03:00 < rxr> > You don't. Messing with the stack like that behind the compiler's
03:00 < rxr> > back is strictly verboten.
03:00 < rxr> but it goes further, read this:
03:00 < rxr> Oh, come on, it's Christmas! Let's show him how to do it: use an inline asm
03:00 < rxr> to pop into a pre-reserved asm register variable like this:-
03:00 < rxr> #include
03:00 < rxr> int main (int argc, const char **argv)
03:00 < rxr> {
03:00 < rxr> register int Eax asm ("%eax");
03:00 < rxr> asm volatile ("pop %0" : "=r" (Eax) : );
03:00 < rxr> printf ("Eax was $%08x\n", Eax);
03:00 < rxr> return -1;
03:01 < rxr> }
03:01 < rxr> thats bizarre ...
03:02 < rxr> # ./pop
03:02 < rxr> Eax was $00000000
03:03 < mnemoc> :)
03:06 < rxr> for some packages ccache does still not scale
03:06 < rxr> examples are gnomemeeting and mozilla (or siblings)
03:07 < rxr> maybe it is due to libtool or so - we should analyze it one day ...
03:07 < rxr> subsequent builds of those give nearly no cache hits ...
03:13 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
03:14 < rxr> hi sparc-kly
03:14 < rxr> == 03:01:35 =[5]=> Building gnome2/gnucash [1.8.10 2.1.0-beta2].
03:14 < sparc-kly> hi
03:15 < rxr> -> Compiler Cache Hits while pkg build: 3.69% (14 hits, 365 misses)
03:15 < rxr> == 12/21/04 03:12:33 =[5]=> Finished building package gnucash.
03:15 * sparc-kly install now drock 2.0.1
03:15 < mnemoc> rxr: maybe discussing it with ccache people
03:15 < mnemoc> uhm?
03:15 < rxr> well - before discuss need to know what is going on ...
03:17 < CIA-9> rene * r5166 /trunk/package/gnome2/gnucash/ (configure-need-libstdc++-hotfix.patch gnucash.desc): * updated gnucash (1.8.9 -> 1.8.10)
03:21 < mnemoc> cc1: /usr/src/building/t2/test/build/boot-2.1.0-beta2-x86-pentium-mmx-32-bootdisk-expert/usr/src/building/t2/test/build/boot-2.1.0-beta2-x86-
03:21 < mnemoc> pentium-mmx-32-bootdisk-expert/usr/dietlibc/include/dietref.h: No such file or directory
03:21 < mnemoc> uhm
03:21 < mnemoc> that looks a bit long
03:21 < jsaw> indeed
03:22 * sparc-kly need hosting anyone have one for me ?
03:22 < CIA-9> rene * r5167 /trunk/package/base/lm_sensors/no_depmod.patch: * fixed lm_sensors my removing the obsolete no_depmod.patch
03:23 < rxr> sparc-kly: private, commercial? big small, ... ?
03:24 < sparc-kly> private . for configuration files and geek stuff hehehe
03:24 < rxr> hehe ;-)
03:25 < rxr> aren't there quite some sites offering those for free or so ?
03:27 < rxr> hahaha: regarding the poping stuff from above, here it goes on:
03:27 < rxr> >What is a stack? What does the compiler use it for? Why would
03:27 < rxr> >> messing with it be a bad idea? And no, I do not know
03:27 < rxr> >> assembly, and I haven't had computer science class yet.
03:27 < rxr> >>
03:27 < rxr> >> Samuel Lauber
03:27 < rxr> The "stack" is like a great big pile of junk where the compiler discards all
03:27 < rxr> the old data it doesn't need any more, deleted strings, values of dead
03:27 < rxr> variables, etc. etc., a bit like /dev/null only right inside the compiled
03:27 < rxr> program. It's a bad idea to go messing with it because all the stale old data
03:27 < rxr> can come flooding back out and fill your code with stuck bits, and they can take
03:27 < rxr> an awful long time to clean up again.
03:27 < sparc-kly> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=73321&item=5738581843&rd=1&ssPageName=WDVW \
03:28 < sparc-kly> micro webserver :S sound good for me
03:29 < rxr> are you sure you need an own wer server?
03:29 < rxr> I mean normally you just place some files on a remote box ...
03:30 < sparc-kly> my dsl is good . 90% of time
03:30 < sparc-kly> DSL sucks
03:35 < CIA-9> rene * r5168 /trunk/package/x11/freetype/freetype.desc: * updated freetype (2.1.5 -> 2.1.9)
03:35 < CIA-9> rene * r5169 /trunk/package/x11/freetype/freetype.conf: * added forgotten compile fix for the freetype update
03:35 < rxr> sorry
03:36 < mnemoc> weird, after some builds on stage 1, setting DIETHOME gets redundant
03:36 < mnemoc> i first needed it for 1-gzip, bug on 1-kiss it's redundant and for 1-gzip too
03:36 < rxr> hm
03:37 < rxr> looks like a more fundamental bug somewhere else ..
03:37 < rxr> ok - my athlon trunk is now modification less ..
03:37 < mnemoc> i'll comment DIETHOME= then
03:39 < rxr> I'm going to rebuild both desktop and install ....
03:45 < rxr> I do too much t2 stuff tonight ...
03:46 < CIA-9> rene * r5170 /trunk/package/filesystem/udev/ (config.hlp config.in udev.conf):
03:46 < CIA-9> * fixed udev to default to /dev (was /udev) - and added Config help
03:46 < CIA-9> entries
03:48 < rxr> oehm?: http://www.gnomemeeting.de/
03:49 < CIA-9> mnemoc * r5171 /trunk/ (. package/base/dietlibc/parse-config-9): * DIETHOME is not set in the right moment, commented until deeper analysis is performed
03:49 < CIA-9> mnemoc * r5172 /trunk/:
03:50 < rxr> http://www.webcamuser.de/forum/
03:51 < rxr> no - not that - this I meant:
03:51 < rxr> http://www.pitivi.org/
03:51 < mnemoc> you are confusing me :(
03:52 < mnemoc> at least this is in english :)
03:53 < rxr> yes
03:54 < rxr> the first one was just to note that now damn domain grabbers take open source project names to link to dubious talk whatever sites ...
03:54 < rxr> first I thought it links to some microsoft netmeeting site - but it is some strange sex-talklinie s.th. ...
03:54 < rxr> whatever ..
03:57 < mnemoc> oh
04:04 < jsaw> mnemoc: put DIETHOME in set_wrapper_and_misc ?
04:05 < mnemoc> fixed to $root/usr/dietlibc or some times to $tooldir?
04:05 < jsaw> hmmm... interesting question...
04:07 < jsaw> btw question, who knows what INA117 is?
04:09 < mnemoc> google?
04:09 < jsaw> no, that's unfair.
04:10 < jsaw> hint: I'm just soldering 2 of them.
04:10 < rxr> soldering?, hm
04:10 < mnemoc> chips?
04:11 < jsaw> instrumentation amplifiers, high voltage, Burr Brown...
04:11 < mnemoc> what for?
04:12 < jsaw> measuring current of a piezo bending element control
04:13 < rxr> seems you have fun over there ...
04:13 < rxr> reminds me I also have work to for this atmel based board stuff ..
04:14 < jsaw> my boss just asked me if I can't live without a computer... my laptop is right beside me in the lab...
04:14 < mnemoc> http://www-s.ti.com/sc/ds/ina117.pdf
04:16 < jsaw> :)
04:16 < jsaw> yeah, ti bought burr brown..
04:16 < mnemoc> whata re you constructing?
04:17 < jsaw> It's still my miniature microscope, which we will use to detect signals in nerve cells.
04:17 < rxr> mnemoc: why do you ask - he is constructing /something/ - this is already enough to know
04:18 < rxr> he is a terrorist ...
04:18 < mnemoc> :)
04:18 < jsaw> I know where you live, rxr ;)
04:18 < mnemoc> people like george is dangerous
04:19 < rxr> ouhm - damn /me pushing the atmega and co into the dark corner
04:20 < rxr> I do not believe this ....
04:20 < rxr> a quick test on my Athlon shows: it does not yield the same instability with the scanners ...
04:20 < rxr> if this is a OHCI USB whatever problem I debug on I I get angry ...
04:21 < rxr> then I need someone like george on my scanners ...
04:22 < rxr> damn - with x86 Via USB chip this damn scanner seems to be rock solid ...
04:22 < rxr> while true ; do
04:22 < rxr> replug
04:22 < rxr> scanimage
04:22 < rxr> replug
04:22 < rxr> batchscan
04:22 < rxr> done
04:22 < rxr> has not produced any lookup so far ...
04:22 < rxr> damn
04:23 < rxr> me plugging the 2nd problematic scanner into the athlon
04:26 < rxr> oehm
04:27 < rxr> I do not believe this ...
04:27 < jsaw> and?
04:27 < rxr> scanner #2 works quite well on the athlon, as well ...
04:27 < rxr> arrrgggggghhhhhhhhh!!!!!
04:28 < mnemoc> better support to that chipset?
04:29 < jsaw> ohoh, somebody's gonna sue someone else soon
04:29 < rxr> well my driver is user-space - it does only do quite straightforward i/o over usb ...
04:29 < mnemoc> i blame kernel
04:29 < rxr> it happens under osx und linux ...
04:30 < mnemoc> GPL :)
04:30 < rxr> you mean to sue apple for cut'n pasting bugs out of the linux kernel into osx ?
04:31 < jsaw> yep...
04:31 < mnemoc> sue apple for having weak usb support
04:32 < rxr> damn - on the f*ck athlon with the ugly via chipset I can not stand and in gernally sort it behind OHCI devices in terms of stability ans support the scanner overlives serveral reconnections and scans ...
04:32 < rxr> what have I debugged on the last week ... :-(
04:36 < jsaw> nothing.
04:36 < mnemoc> :)
04:36 < jsaw> getting a coffee... then starting testing/debugging my electronics
04:42 < rxr> well - even leaving it idle for ten minutes does not freeze it ...
04:42 < rxr> well - seems I need to get a G5 rather sooner than later ...
04:47 < rxr> hm - but now it starts to freeze ...
04:48 < rxr> I somehow do not like those scanners ....
04:48 < mnemoc> nope, you love them but they make you suffer
04:51 < rxr> sure
04:54 < rxr> I have the feeling the scanner does not like the config requests as sent by libusb ...
05:02 < mnemoc> uhm, why termcap _and_ ncurses are enabled on bootdisk?
05:02 < mnemoc> shouldn't ncurses replace termcap?
05:03 < rxr> hm - s.th. broken, hm?
05:07 < mnemoc> target/bootdisk/config.in:ROCKCFGSET_PKG_TERMCAP_USEIT=1
05:07 < mnemoc> it's intentional
05:08 < rxr> them ncurses might not ...
05:08 < rxr> or whatever ...
05:08 < mnemoc> target/bootdisk/build_stage2.sh:-mine -bize -termcap +ncurse
05:09 < mnemoc> bootdisk is weird
05:14 < mnemoc> i'll try to build a bootdisk without termcap
05:15 < rxr> ok
05:16 < rxr> hm - loading a 46MB .pnm on my Athlon 2500XP does take quite some time ...
05:17 < mnemoc> it may be to not package termcap files, and only ncurses lib
05:18 < rxr> ouhm - now loading a 103MB scan into display ...
05:18 * rxr needs to Emerge gimp on the box ..
05:18 < rxr> It can't be as slow as display ...
05:19 < mnemoc> what about the eye of gnome?
05:20 < rxr> .oO
05:20 < rxr> do you think crappy example applications like such huge image data ?
05:20 < rxr> eog not installed at all ;-)
05:21 < mnemoc> =)
05:25 < rxr> I could try and report the result ...
05:25 < mnemoc> a segfault? :)
05:26 < mnemoc> "It can't be as slow as display ..."
05:26 < rxr> hehe
05:27 < rxr> yah - maybe "glib error: failed to allocate:
05:27 < rxr> "
05:28 < rxr> mnemoc: how comes you are awake at this time ?
05:28 < mnemoc> it would be nice to see how does it handle _huge_ files
05:28 < mnemoc> 1am here only
05:28 < mnemoc> 1:30
05:29 < mnemoc> but i guess i'll go to sleep :)
05:29 < rxr> you mean I should scan a A4 paper at 1200 or 2400 dpi ?
05:29 < mnemoc> i was playing with 7 builds of bootdisk in parallel with different hacks
05:29 < rxr> hey - I cold even scan 16bit per channel - just that nearly no open-source applications beside film-gimp does like 16bit per channel image data ...
05:29 < mnemoc> yes, huge enough
05:30 < rxr> and know you know why I packed film-gimp ,-)
05:30 < mnemoc> make it work for real people first :)
05:30 < rxr> cinepaint even ;-)
05:30 < mnemoc> oh
05:31 < rxr> I could add an annoying "ding-dong" exlusive focus dialog that pops up when one does select more than 400dpi
05:32 < rxr> "Please do not open the resulting file with display or eye-of-gnome. At least do not blame us for the result ...
05:32 < rxr> "
05:32 < mnemoc> :D
05:33 < rxr> btw. photoshp is dog slow ..
05:33 < rxr> I can not imagine why people invest 1.5k bucks into it ...
05:33 < mnemoc> http://cinepaint.sourceforge.net/pix/mac/MacFilmGimp.refuel.1440x900.png <-- i dream with a display like this
05:34 < mnemoc> becasue they are stupid
05:35 < mnemoc> ok, as you said i have to be sleeping now
05:35 < mnemoc> and you too
05:35 < rxr> yeah - soon ...
05:36 < mnemoc> you can't wait the sunrise when there is no sunrise
05:36 < rxr> need to complete some more test scans and send some quetions to device engeneers in taiwan
05:37 < rxr> +s
05:37 < rxr> sleep well mnemoc
05:37 < mnemoc> you too
06:40 < rxr> so bed ...
08:21 -!- mipe [~mipe@dsl10040.japo.fi] has joined #t2
09:00 < mipe> anyone awake?
09:00 < jsaw> more or less...
09:00 < jsaw> (means, I am almost falling from my chair .. :)
09:01 < jsaw> hi mipe
09:01 < jsaw> you were here yesterday and wanted to ask sth. ?
09:01 < mipe> hi jsaw, yeah. 00-dirtree is failing for some reason
09:02 < jsaw> a bit more detailed, please
09:02 < mipe> hmm. i mean Build-Pkg is failing on 00-dirtree package build. error log coming up
09:02 < mipe> mkdir: dl-close.c:134: _dl_close: Assertion `map->l_direct_opencount > 0' failed.
09:02 < mipe> ./scripts/Build-Pkg: line 260: 8953 Aborted mkdir -p $root/var/
09:02 < mipe> ups
09:03 < mipe> thats the error. i get same error for every mkdir -p call
09:03 < jsaw> you mean you even get it by executing it on the usual command line?
09:03 < mipe> nope. inside the build only
09:04 < jsaw> processor? optimization?
09:04 < mipe> amd64
09:04 < mipe> for athlon
09:04 < jsaw> both?
09:04 < jsaw> oh
09:04 < mipe> yeah
09:04 < jsaw> host system?
09:05 < mipe> i'm running arch. which isnt builded for amd64. my friend builded the iso for me and i thought he also optimized for amd64 but he just added needed packages so i wont need to download so much
09:06 < jsaw> 64 or 32 bit ?
09:06 < mipe> 32
09:07 < jsaw> hmm.
09:09 < mipe> hmm. it doesnt fail on 00-dirtree actually. fails on Build-Pkg on the lines where it adds /var/adm dirs
09:10 < mipe> but adds succefully adm/logs dir but the rest of fails
09:10 < jsaw> which stage is this?
09:13 < mipe> i'm new to t2. and dont know much about the stages. so... hmmm.... at the first stage since i call it /scripts/Build-Target -cfg system and it fails just after build-tools
09:14 < jsaw> the stage should appear at start and end/fail of building a package:
09:14 < jsaw> == 08:51:27 =[0]=> Building base/gcc [3.4.3 2.1.0-beta2].
09:14 < jsaw> == 12/21/04 09:02:44 =[0]=> Finished building package gcc.
09:15 < jsaw> the number in brackets gives you the stage, here 0
09:15 < mipe> oh. it doesnt get to that far. i mean 00-dirtree is the first package to be build but Build-Pkg fails before it starts to build the actual package... failing on creating the dirs for build dir /var/adm/*
09:16 < jsaw> oh
09:16 < mipe> unless this is a cross build checked on Config. should it build tools.cross to toolchain?
09:16 < mipe> i mean i havent checked it. but it builds tools.cross to toolchain
09:17 < jsaw> If you even do not reach the 0 stage, something else is wrong.
09:18 < jsaw> You are still on the host system. At stage 2, the system continues in a chroot.
09:18 < mipe> can i spam the entire log if it helps? ~10lines
09:18 < jsaw> yep
09:18 < mipe> Building tools in build/.../TOOLCHAIN/tools.cross:
09:18 < mipe> -> Building tools.cross/wrapper/cmd_wrapper.
09:18 < mipe> -> Building tools.cross/wrapper/strip_wrapper.
09:18 < mipe> -> Building tools.cross/wrapper/install_wrapper.
09:18 < mipe> -> Building tools.cross/bin/getdu.
09:18 < mipe> -> Building tools.cross/bin/getfiles.
09:18 < mipe> -> Building tools.cross/bin/fl_wrparse.
09:18 < mipe> -> Building tools.cross/bin/fl_stparse.
09:18 < mipe> -> Building tools.cross/lib/fl_wrapper.so.
09:19 < mipe> -> Testing tools.cross/lib/fl_wrapper.so.
09:19 < mipe> -> Creating tools.cross/lib/parse-config.
09:19 < mipe> -> Creating tools.cross/lib/pkg_*_{pre,post}.conf
09:19 < mipe> mkdir: dl-close.c:134: _dl_close: Assertion `map->l_direct_opencount > 0' failed.
09:19 < mipe> ./scripts/Build-Pkg: line 260: 9803 Aborted mkdir -p $root/var/adm/flists
09:20 < jsaw> My immediate guess is that there's something wrong with your host system
09:21 < jsaw> does sth. like mkdir /tmp/examplepath/{a,b,c,d,e,f} work?
09:21 < jsaw> -p
09:21 < mipe> yeah
09:25 < jsaw> actually I have no idea.
09:27 < jsaw> and this happens to both, the amd64 and the athlon cross build?
09:27 < mipe> yeah
09:29 < jsaw> I'm sorry but I can't help. Would you please post this information to the mailing list? (t2 at exactcode dot de)
09:30 < mipe> ok
10:01 < jsaw> gn8 all
10:36 -!- mipe [~mipe@dsl10040.japo.fi] has quit ["boot"]
11:18 -!- mipe [~mipe@dsl10040.japo.fi] has joined #t2
11:54 -!- grifters [~petrus@petrus.jmk.su.se] has joined #t2
11:54 -!- grifters [~petrus@petrus.jmk.su.se] has left #t2 []
11:55 < mipe> hmm. i think i solved my problem on Build-Target
12:03 < mipe> but now glibc doesnt compile :)
12:05 < daja77> hehe that's what I call fun
12:35 < mipe> hmm. i just dont understand this error and googling doesnt return anything usefull
12:35 < mipe> ../sysdeps/unix/x86_64/sysdep.S: Assembler messages:
12:35 < mipe> ../sysdeps/unix/x86_64/sysdep.S:59: Error: suffix or operands invalid for `push'../sysdeps/unix/x86_64/sysdep.S:64: Error: suffix or operands invalid for `pop'
12:35 < mipe> ../sysdeps/unix/x86_64/sysdep.S:66: Error: `(%rax)' is not a valid 32 bit base/index expression
12:41 < mipe> hmm. hmm... why does /Config have optimised for athlon and it passes the march value as athlon and not athlon64 or k8 as it should be
12:41 < mipe> me thinks
14:32 < mnemoc> moin
14:33 < mnemoc> mipe: what was your problem with mkdir?
14:34 < mipe> hmm. it went away when selecting flist detection tech to strace and not flist wrapper
14:35 < mipe> but now i get error on building kernel on stage 1
14:35 < mnemoc> 1-glibc?
14:36 < mipe> 1-linux26,error is
14:36 < mnemoc> oh
14:36 < mipe> In file included from include/linux/timex.h:188,
14:36 < mipe> from include/linux/sched.h:11,
14:36 < mipe> from include/linux/module.h:10,
14:36 < mipe> from mm/hugetlb.c:8:
14:36 < mipe> include/linux/time.h: In function `timeval_to_jiffies':
14:36 < mipe> include/linux/time.h:279: internal compiler error: Segmentation fault
14:37 < mnemoc> it's really bad to get that on a native build :\
14:52 < mnemoc> rxr: fake bypassed submaster an posted his patches here: http://f4k3.net/~fake/rock/HOTFIXES/ :)
15:07 < rxr> re
15:08 < mnemoc> re rxr
15:08 < rxr> mipe: what is the system you compile on ? what hardware and OS?
15:09 < rxr> mnemoc: interesting ...
15:09 < mnemoc> rxr: native on amd64 running arch-linux for 686 o_O
15:10 < mnemoc> i think he must rebuild his local kernel before doing anything else
15:11 < mipe> back,yep amd64 running arch,rebuilded kernel too
15:11 < rxr> but nothing is running in 64bit mode?
15:12 < mipe> btw. ignore the patch i send to ml
15:12 < mipe> it helps on cross but gives error on native
15:14 < rxr> so you run a i686 Linux ? and build t2 for x86 ?
15:15 < mnemoc> he builds t2 for amd64 on a 64bits machines running at 32 :)
15:15 < mipe> running i686 building for x86_64, the cross patch was test for cross building to friend of mine
15:15 < mnemoc> that's why i said we need to share cross-built bases :)
15:18 < mipe> btw. on -march=athlon doesnt use the 64bit optimizations?
15:18 < mipe> isnt that athlon64,k8 pr athlon-fx?
15:18 < mipe> pr=or
15:23 < rxr> your compiler will generate 64 bit code anyway ...
15:23 < rxr> I update them
15:23 < rxr> thanks for the note
15:30 < CIA-9> rene * r5173 /trunk/architecture/x86/config.in: * updated the x86-64 optimisation settings
15:36 < rxr> cool:
15:36 < rxr> http://www.ccc.de/hackabike/index_en.html
15:52 < rxr> GNU/Hurd: Support For Partitions Larger Than 2 GB
15:52 < rxr> http://kerneltrap.org/node/4429
15:53 < mnemoc> at last! :)
15:54 < mnemoc> over mach or l4?
15:57 < rxr> I read the patch is not yet upstream - but in debian/hurd
15:57 < rxr> and the texts indicates they still migrate to l4 - so most probably running on mach
15:57 < rxr> but I'm not a hurd expert ...
15:58 < rxr> ouhem - HEAD bootdisk build without any issue ?
15:58 < rxr> ah - no a few are left:
15:58 < rxr> Error logs from install-2.1.0-beta2-x86-bootdisk:
15:58 < rxr> [3] base/psmisc [5] network/minicom
15:58 < rxr> [5] base/dump
15:58 < rxr> 162 builds total, 159 completed fine, 3 with errors.
15:59 < mnemoc> termcap thing?
15:59 < rxr> cool - and the desktop is also still building ...
15:59 < rxr> nice
15:59 < rxr> File not found: download/mirror/m/minicom-2.1.tar.bz2
15:59 < rxr> Did you run ./scripts/Download for this package?
15:59 < rxr> *lol*
15:59 < mnemoc> i commented that USEIT on config.in and it built successfully
16:00 < mnemoc> # ./scripts/Download -mk-cksum download/mirror/m/minicom-2.1.tar.bz2
16:00 < mnemoc> download/mirror/m/minicom-2.1.tar.bz2: 2435954076
16:00 < rxr> In file included from pstree.c:21:
16:00 < rxr> /usr/include/termcap.h:25: error: conflicting types for 'tgetnum'
16:00 < rxr> /usr/include/term.h:776: error: previous declaration of 'tgetnum' was here
16:00 < rxr> ouhm - dump patch does not apply ...
16:00 < rxr> who update it ?
16:01 < rxr> i ...
16:01 < rxr> ok /me fixing ...
16:02 < mnemoc> :)
16:06 < rxr> maybe this is the time to push T2/Hurd ...
16:06 < mnemoc> let's finish T2/Linux first :p
16:07 < CIA-9> rene * r5174 /trunk/package/base/dump/dump-devfs.patch: * fixed dump to build again
16:09 < mnemoc> which is make's dry-run? --touch?
16:10 < mnemoc> -n :)
16:17 < mnemoc> rxr: DIETHOME is no longer needed after cmd_wrapper change.... this is _very_ odd
16:23 < rxr> INFO: Testing ... ok 70422.000
16:23 < rxr> INFO: Testing ... ok 69268.000
16:23 < rxr> INFO: Testing ... ok 17122.000
16:25 < mnemoc> gsmp took the head again :(
16:26 < rxr> yeah - but it is just moon rays and the current load that make the difference between the both here ...
16:36 < mipe> hmm... gettext hangs on checking java compiler
16:54 < rxr> hm
17:10 < CIA-9> rene * r5175 /trunk/package/network/ppp/ip-up:
17:10 < CIA-9> * further fixed our ppp/ip-up to not leave stray comment lines in the
17:10 < CIA-9> /etc/resolv.conf ...
17:27 -!- rxr [~rene@p213.54.195.163.tisdip.tiscali.de] has joined #t2
17:27 -!- Topic for #t2: T2 | 2.1.0-beta RELEASED | The next generation of System Development Enviroments (SDE) | http://www.exactcode.de/t2
17:27 -!- Topic set by mnemoc [] [Sat Dec 18 15:39:34 2004]
17:27 [Users #t2]
17:27 [ _Ragnar_] [ daja77] [ mipe ] [ nzg] [ sparc-kly]
17:27 [ CIA-9 ] [ jsaw ] [ mnemoc] [ rxr] [ valentin ]
17:27 -!- Irssi: #t2: Total of 10 nicks [0 ops, 0 halfops, 0 voices, 10 normal]
17:27 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
17:27 -!- Irssi: Join to #t2 was synced in 13 secs
17:30 < rxr> glibc 2.3 stable branch
17:32 < mnemoc> uhm?
17:58 < jsaw> long live Poland...!
17:58 < mnemoc> o_O
17:58 < mipe> hmm. looks like stage 1 is done but i get following error chroot: cannot run command `bin/bash': Exec format error
17:58 < jsaw> mnemoc: software patents in EU ...
17:58 < mnemoc> kidding?
18:00 < daja77> mnemoc: poland blocks them and raised an opposition against them
18:00 < jsaw> hi all btw.
18:01 < jsaw> http://www.nosoftwarepatents.com/
18:01 < mnemoc> nothing on ffii
18:01 < jsaw> (this message is about 1h old)
18:02 < jsaw> But nosoftwarepatents has it already
18:04 < rxr> yeah - seems I should migrate some km into the east ...
18:05 < daja77> meeting esden there ^^
18:14 -!- mipe [~mipe@dsl10040.japo.fi] has quit ["Lost terminal"]
18:53 < rxr> mnemoc: trunk:HEAD looks quite ok right now
18:53 < rxr> mostly fixing this base/psmisc issue and testing we could tag -beta2, right ?
18:54 < rxr> Error logs from install-2.1.0-beta2-x86-bootdisk:
18:54 < rxr> [3] base/psmisc
18:54 < rxr> 162 builds total, 161 completed fine, 1 with errors.
18:54 < rxr> Error logs from default-2.1.0-beta2-x86-pentium-mmx-32-desktop-expert:
18:54 < rxr> 1019 builds total, 294 completed fine, 0 with errors.
18:57 < rxr> more t2 hits on our site due to beeing in this weekly news:
18:57 < rxr> http://distrowatch.com/weekly.php?issue=20041220
19:08 < CIA-9> rene * r5176 /trunk/package/base/psmisc/no-termcap.patch: * fixed psmic for the case ncurses and termcap are enabled
19:08 < rxr> ok - no bootdisk target errors anymore
20:52 < mnemoc> re
20:53 < mnemoc> ok, so i reenable termcap on my bootdisk
20:53 < mnemoc> generic is going good too
20:56 * rxr away - cu
20:58 < mnemoc> cu rxr
21:00 < rxr> I think, then, that we need to call 1.1.2 a "brown paper bag" release,
21:00 < rxr> since the perl bindings don't even compile.... and release 1.1.3 ASAP.
21:00 < rxr> ^ - subversion
21:00 < mnemoc> clkao (svk) is pushing 1.1.3 since yestarday
21:03 < rxr> well - I have a patch handy - but I have not committed it yet due to swig bindings failing ...
21:03 < rxr> so - cu
21:04 < mnemoc> cu
23:11 < rxr> re
23:16 < mnemoc> re rxr
23:16 < mnemoc> who the scanner thing went?
23:16 < mnemoc> how*
23:18 < jsaw> re
23:18 < mnemoc> re jsaw
23:20 < jsaw> while building gcc, look at this:
23:20 < jsaw> SHLIB_LINK='cmd_wrapper /scratch/src/t2/src.gcc.1103666536.18068.d9959e34/gcc-3.4.3/objdir/gcc/xgcc
23:20 < mnemoc> what?
23:20 < jsaw> the "cmd_wrapper" in front of xgcc
23:21 < mnemoc> yes, that's my "what?!"
23:22 < jsaw> there's a variable named "STAGE_CC_WRAPPER", which you can inject into the environment...
23:22 < mnemoc> tell us more
23:23 < jsaw> I'm creating the xgcc cmd_wrapper now.
23:23 < jsaw> export STAGE_CC_WRAPPER=cmd_wrapper
23:23 < jsaw> <- this line was enough
23:23 < rxr> to get rid of the ugly CFLAGS injection in the gcc.conf ?
23:24 < jsaw> no, actually I still have problems with the SSP thing.
23:24 < jsaw> But if that also solves the ugly CFLAGS injection, good...
23:25 < rxr> yes
23:26 < mnemoc> free bonus :)
23:26 < jsaw> :)
23:36 < rxr> re
23:36 < rxr> damn - just managed to drop a lot of juice onto my iBook ...
23:36 < rxr> had to power-down and clean it a bit ...
23:36 < rxr> still works, though ...
23:37 < jsaw> phew...
23:39 < mnemoc> =)
23:40 < mnemoc> once i found a used diaper over mine :\
23:41 < jsaw> :D
23:41 < daja77> shit happens!
23:41 < mnemoc> literally :)
23:41 < rxr> .oO
23:41 < daja77> yep
--- Log closed Wed Dec 22 00:00:41 2004