--- Log opened Sat May 07 00:00:46 2005
00:08 < rxr> screen has a lot of alignment cast warnings when build with -Wcast-align ...
00:08 < rxr> but I think I stop to investigate this crap
00:08 < rxr> no screen on sparc64 for now ...
00:19 < rxr> do you plan to add 8682 to trunk ?
00:23 < rxr> Clifford wolf: Fixed ash (crazy pointer magic: it didn't fail before)
00:24 < rxr> submaster - where even the name spellings bypass the extensive peer review ...
00:24 < rxr> mnemoc: do you apply that to trunk and remove the NOPARALLEL flag then ?
00:24 < rxr> thanks
00:28 < mnemoc> :)
00:28 < mnemoc> ok
00:28 < mnemoc> i have make mails flaged for merge
00:28 < mnemoc> s/make/many/
00:29 < mnemoc> i'll merge after work.... 2h
00:30 < rxr> ok - /me in bed ...
00:30 < mnemoc> gn8 rxr :)
00:30 < rxr> good n8
00:31 < valentin> gn8 rxr
00:32 < rxr> valentin: the new sparc box has .8
00:32 < rxr> I need to fix re-emerge ppp so the damn ppp-up script does not overwrite the name server on each reconect - so that the name resolving of internal addresses also works on the router itself ...
00:32 < valentin> ok
00:33 < rxr> it has an halfly working rootfs in 64bit mode and a more correct one in /home
00:33 < rxr> but it is busy building anyway ...
00:33 < rxr> n8 all
00:34 < rxr> hm - btw. I somehow doubt clifford's fix is a real fix
00:34 < rxr> the error was spurious, soo ...
00:35 < rxr> valentin: maybe you can review the revision I pointed out above mnemoc merged into 2.1 ...
00:35 < rxr> this is why our patches always should have an explaining header - such changes are just not self-explaining enough ...
00:35 < rxr> but now off - cu in the morning
00:35 -!- ojh [~omer@71-34-255-36.eugn.qwest.net] has joined #t2
00:37 < mnemoc> rxr: i tested ash here
00:51 < sparc-kly> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=5577795202 <-- LOOOOOOOOOOOOOOOOOL
00:53 < mnemoc> sparc-kly: you box keeps droping me
00:58 < valentin> US $15,500.00 ?? F*CK !!
00:58 * valentin going into his kitchen, starting his toaster ...
00:58 < sparc-kly> hahahahhaa
00:58 < sparc-kly> i win i make one 1st
00:59 < valentin> is this some kind of joke - i mean, who would bid a fortune for that shit ?
01:00 < sparc-kly> i dont have idea . is this some kind of joke . ebay droping auction in hours
01:29 < valentin> i think it is time to fall asleep
01:47 < valentin> gn8
03:05 < ojh> Does Gnome System Tools have a 'safe' profile that can be used with T2?
03:35 < mnemoc> ew
03:36 < mnemoc> ojh: no idea :\
03:37 < ojh> mnemoc: Oh well, I have T2 up and running on the laptop now. :)
03:38 < mnemoc> great! :)
03:38 < ojh> mnemoc: The one thing that did not work is that ROCK-Plug did not detect the NIC, so I used hwscan to configure and it worked.
03:39 < mnemoc> built by you, are one of the sample ISOs?
03:39 < mnemoc> ojh: now that the module is loaded run depmod, and set rockplug again
03:40 < ojh> mnemoc: Beta 3 ISOs.
03:40 < ojh> GNOME seems to come with no theme, lots of ugly filler icons. :-/
03:40 < ojh> And Special Nautilus :-/
03:41 < mnemoc> everything on default T2 looks... default :)
03:41 < ojh> mnemoc: IC.
03:42 < ojh> I was a bit surprised, no XFCE in the default ISOs.
03:42 < mnemoc> we are a system development enviroment (SDE, aka DBK), we minimalize T2 dna inside target builds
03:43 -!- Sparc-kly|G4 [~sparc@64.237.132.176] has joined #t2
03:43 < mnemoc> minimize*
03:43 < mnemoc> ojh: desktop target doesn't have xfce
03:43 < ojh> mnemoc: IC, well that's what a package/autobuilding system is for . . .
03:44 < mnemoc> get a checkout of 2.1 branch
03:44 < mnemoc> emerge there will emerge dependencies (one level) for you
03:44 < ojh> OK
03:47 -!- Sparc-kly|G4 [~sparc@64.237.132.176] has quit ["Leaving"]
03:48 -!- Sparc-kly|G4 [~sparc@64.237.132.176] has joined #t2
03:48 < mnemoc> Sparc-kly|G4: i'm at home now... looks ok
03:49 < Sparc-kly|G4> good
03:49 < Sparc-kly|G4> i have a problem in this laptop :S only 19kb/s in local network :S
03:50 < Sparc-kly|G4> 11kb/s wtf
03:51 < ojh> mnemoc: The REALLY cool thing about the laptop is that Xorg has full support for Radeon Mobility. Including GLX :D
03:51 < mnemoc> :)
03:53 < ojh> mnemoc: With a slightly more comfortable pointer (touchpads are great and all, but not for serous graphic work or games) it will be great for Blender and other graphics work.
03:53 < mnemoc> optical usb mice are not to expensive :)
03:55 < ojh> mnemoc: True, if I want expensive I would be shopping for a tablet . . . . well, think I just found the next item on my wish list . . . :)
03:56 < mnemoc> hehe
03:57 < sparc-kly> mnemoc in your house my SSH server run fine ?
03:58 < mnemoc> Sparc-kly: yep
03:58 < sparc-kly> tcp 0 0 amd:ssh pc-120-204-86-200:61357 ESTABLISHED
03:58 < sparc-kly>
03:58 < sparc-kly> hehehe
03:59 < mnemoc> :)
04:02 -!- Sparc-kly|G4 [~sparc@64.237.132.176] has left #t2 ["Leaving"]
04:21 < mnemoc> sparc-kly: i found your problem
04:21 < sparc-kly> really? ? ?
04:23 < mnemoc> i have to find how to solve it without reinstalling that crappy debian with another OS
04:23 < sparc-kly> LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL
04:29 < sparc-kly> ok you Win
04:31 < mnemoc> what?
04:31 < mnemoc> i have to solve it to let people get away of debian
04:33 < sparc-kly> hehehhe
04:34 < sparc-kly> damn it . DEBIAN sucks
04:35 < mnemoc> :)
04:58 -!- CIA-10 [~CIA@flapjack.navi.cx] has quit [Excess Flood]
04:58 < sparc-kly> q pongo
04:58 < sparc-kly> ?
04:58 -!- CIA-1 [~CIA@flapjack.navi.cx] has joined #t2
04:59 -!- keinek_ is now known as keinek
05:02 < mnemoc> sparc-kly: no me quites el debian!
05:02 < mnemoc> keinek: hola.... trata de conseguir "mi mejor enemigo" la estrenaron hace dos dias... muy buena
05:02 -!- CIA-1 [~CIA@flapjack.navi.cx] has quit [Excess Flood]
05:02 -!- CIA-10 [~CIA@flapjack.navi.cx] has joined #t2
05:03 -!- CIA-10 [~CIA@flapjack.navi.cx] has quit [Excess Flood]
05:03 -!- CIA-10 [~CIA@flapjack.navi.cx] has joined #t2
05:03 -!- CIA-10 [~CIA@flapjack.navi.cx] has quit [Excess Flood]
05:04 -!- CIA-10 [~CIA@flapjack.navi.cx] has joined #t2
05:04 -!- CIA-10 [~CIA@flapjack.navi.cx] has quit [Excess Flood]
05:04 -!- CIA-10 [~CIA@flapjack.navi.cx] has joined #t2
05:36 < mnemoc> sparc-kly: ahora si que pasa
05:36 < mnemoc> oops
05:36 < mnemoc> sparc-kly: now i'm sure it will pass
05:36 < mnemoc> aaarg
05:37 < mnemoc> /bin/sh ./../ylwrap "flex" arlex.l lex.yy.c arlex.c --
05:37 < mnemoc> flex: fatal internal error, exec failed
05:43 < sparc-kly> qqqq
05:43 < sparc-kly> flex sucks
05:43 < sparc-kly> LOL
05:44 < mnemoc> you dont' have uname or bison neither
05:45 < mnemoc> that debian is very incomplete
05:49 < sparc-kly> ohh shit
05:54 < sparc-kly> you build binutils and build fine ?
05:54 < CIA-10> amery * r8683 /trunk/package/base/flex/flex.desc: * repriorized flex before binutils to have a safe flex whe building 0-binutils (debian)
05:55 < CIA-10> amery * r8684 /trunk/scripts/functions: * fixed --cache-file passing to use ./ because configure doesn't find it on debian if not
05:55 -!- ojh [~omer@71-34-255-36.eugn.qwest.net] has left #t2 []
05:56 < mnemoc> sparc-kly: not yet
05:56 * sparc-kly got o sleeep
05:56 < sparc-kly> go*
05:56 < mnemoc> cu sparc-kly
07:00 < CIA-10> amery * r8685 /trunk/package/shells/ash/mkinit.patch: * merged 8682 from 2.1: fixed ash init.c generator - from ROCKLinux
07:18 < mnemoc> rxr: debian's flex _and_ bison are not sane enough to build 0-binutils
07:18 < mnemoc> rxr: after building our bison on stage 0 between flex and binutils a weird headers problem apears when compiling the ld parser
07:19 < mnemoc> gcc -DHAVE_CONFIG_H -I. -I. -I. -D_GNU_SOURCE -I. -I. -I../bfd -I./../bfd -I./../include -I./../intl -I../intl -g -O2 -DLOCALEDIR="\"/home/mub
07:19 < mnemoc> ex/public_html/t2-trunk/build/test-2.2.0-alpha-x86-pentium-mmx-generic/TOOLCHAIN/tools.cross/usr/share/locale\"" -W -Wall -Wstrict-prototypes
07:19 < mnemoc> -Wmissing-prototypes -g -O2 -c ldlex.c
07:19 < mnemoc> lex.yy.c:10: error: parse error before '*' token
07:19 < mnemoc> lex.yy.c:10: warning: type defaults to `int' in declaration of `yyin'
07:19 < mnemoc> lex.yy.c:10: error: `FILE' undeclared here (not in a function)
07:30 < mnemoc> i'm very sleepy
07:30 < mnemoc> i don't understand how people use this env which is unable to build something as basic as binutils
07:38 < mnemoc> sparc-kly: too much for today... your debian is too weird.. i need to sleep
07:38 < mnemoc> how can your stdio.h don't define FILE?!
07:38 < mnemoc> gn8
07:49 -!- minto [~chatzilla@82-217-66-7.cable.quicknet.nl] has joined #t2
08:06 -!- keinek [~Keinek@201.254.12.85] has quit [Read error: 110 (Connection timed out)]
08:06 -!- keinek [~Keinek@201.254.8.228] has joined #t2
08:24 -!- mipe [~mika@dsl10040.japo.fi] has joined #t2
08:43 < rxr> re
09:39 -!- minto [~chatzilla@82-217-66-7.cable.quicknet.nl] has quit ["Chatzilla 0.9.66 [Mozilla rv:1.7.6/20050318]"]
09:47 < mipe> perl-libxml needs to be enabled for desktop
09:49 < mipe> /TOOLCHAIN/src.qt.1115386257.32335.7f0100/qt-x11-free-3.3.4/bin/uic: error while loading shared libraries: /usr/lib/libGL.so.1: cannot restore segment prot after reloc: Permission denied
09:50 < mipe> some wierd permission error during qt
09:51 < CIA-10> rene * r8686 /trunk/package/shells/sash/ (. sash.cache sash.conf sash.desc): * added sash (3.7)
09:52 < mipe> ah permission error comes from selinux
09:54 < CIA-10> rene * r8687 /trunk/package/shells/dash/dash.desc: * added another upstream URL to the dash package for reference
10:10 -!- Veki [~chatzilla@217.24.19.40] has joined #t2
10:12 < rxr> hi Veki
10:13 < Veki> hi
10:13 < Veki> I am willing to check t2
10:13 < rxr> Note: I just timed whether using dash as /bin/sh would speed up configure scripts - unfortunatly it does not really ...
10:13 < Veki> I will install it on on emachine and be probably borring to you with some questions
10:13 < rxr> Veki: what did you make change your mind ,-)?
10:14 < Veki> changeb mind,? that was my intention always, I am very busy and I do have negateive exoperience ehen people are not willng to help
10:14 < Veki> I think that you are willing to help and I will check it
10:15 < Veki> For example on RH8.0 there is no graphich fetchmailcponf
10:15 < Veki> I asked on fetchmail list how to configure it and I was replied with RTFM
10:15 < Veki> that is unkind
10:15 < rxr> yes - I also do not like such replied too much
10:15 < Veki> guy who replied in that way thought that there is graphic fetchmail conf, but it was not
10:16 < Veki> RH gave up from graphic config of fetchmail which I understand why
10:16 < Veki> I know that there is conf file anyway, but RTFM is not OK answer
10:17 < Veki> I understand that user should do his/her best to understand how something works, but we who know a lot should nt hurt them
10:18 < Veki> I am aware that we sometimes assume some things which usualk user do not know, but if we are going to give support than we have to be READY for that which means give support to someone who is not skilled as qwe are
10:18 < Veki> I taught kids in school on free software and they understood me and some of them are installiong BSDs and Linux everywhere
10:18 < rxr> nice!
10:19 < Veki> giving bad marks in any way is not good for the kids and adults too
10:19 < rxr> yes, I fully agree the tongue on many mailing lists is much too harsh
10:19 < Veki> Yes, some my former students are now working in labs worldwide, some my friends from 1995 when I started to use Linux are building free hardware with various processors
10:20 < rxr> another example is the glibc list - RedHat employee Ulrich Drepper automatically responses "you have no idea - leave me and the code alone, go away" ...
10:21 < Veki> yes, that is crazy, I understand that someone feels nice in his if I can say authistic world
10:22 < Veki> But, then please do not go on discussion lists, be on your own in your code paradise and that is fine
10:22 < Veki> we, who do want to communicate wouldliek to do so and we do nto like to be confused and disturbed by those who do not like to communicate
10:23 < Veki> For example, I am not crucial person in teh world of ffreee software but Richard Stallman always found time to reply on my E-mail
10:24 < rxr> nice
10:24 < Veki> That is reason more why I support his ideas. His behaviour is consewquent to what he says. ACtually, he practise what he says
10:24 < Veki> Thus, Richard may be someties wrong, but I am willing to communicate on some lists with him because he is friendly and the fact that someone is worng is not painful
10:25 < Veki> when someone is jsut geeting rid of you that is painful even when simple thing as configuring fetchmail is at question
10:28 < Veki> I think that writing good documentation for t2 is very important thing.
10:28 < Veki> Thus, we have to assume that someone is not expert in Linux and we have to write it in a way that intermediate user can use it
10:28 < rxr> yes - we need more documentation - altough I think the book is a good start
10:28 < Veki> that is important since t2 is in a way liberation tool
10:29 < Veki> yes, book is good start, but I think that procedure of creating one small , one middle size and one larger target is good exercise example which should be written in the book
10:31 < Veki> my experience when working with people who should use some tool that we hav to give them exercises and we have to do and predict some of common mistake in order to direct them on proper ways
10:32 -!- _jeru [~jeru@p54BFBA5A.dip.t-dialin.net] has joined #t2
10:32 -!- _jeru is now known as jeru
10:32 < jeru> moin :)
10:32 < Veki> hi
10:33 < jeru> hi Veki ... how are you ?
10:33 < Veki> rxr, I will let you know when I will be raedy and I will ask questions
10:33 < Veki> Oh, well, I just discussed writing documentation with rxr
10:33 < Veki> I am fine
10:33 < jeru> ah cool!
10:33 < Veki> yes
10:34 < jeru> what kind of documentation ?
10:34 < Veki> I am planning to write documentation for the user
10:35 < Veki> How to build target
10:36 < Veki> I think that his is very important
10:36 < jeru> I guess how to build an own one ... :)
10:37 < jeru> btw, hi "hiding" rxr :)
10:38 < jeru> Veki: do you make progress with your own target ?
10:38 < Veki> I iwll start in a coupe off days
10:40 < jeru> Veki: good luck
10:40 < Veki> I was writing book on audio in linux. I have some more things to do in other to complete that book and I will start on building target
10:41 < Veki> However, I will be borring to you asking some questions :-)
10:42 < rxr> jeru: hi
10:42 * rxr was writing a mail to the list - sorry ...
10:43 < rxr> Veki: asking questions is always a good thing (tm) here ,-)
10:43 < Veki> thanks
10:43 * rxr bakery - cu
10:44 < Veki> I have to leave now, but I will be oin line today again
10:45 < Veki> cu
10:45 -!- Veki [~chatzilla@217.24.19.40] has quit [Remote closed the connection]
11:01 < rxr> http://freshmeat.net/projects/xoverkill/?branch_id=57032&release_id=195528
11:06 < valentin> moin
11:06 < valentin> overkill ?
11:07 < jeru> moin monsieur valentin :)
11:07 < valentin> moin herr jeru
11:09 < jeru> monsieur sounds better ... :p
11:09 < valentin> sir jeru
11:12 < jeru> valentin: nah... sounds like a dusty old man working for the "East India Company"
11:14 < valentin> hehe
11:14 < valentin> i prefer a simple jeru
11:22 < jeru> valentin: ack
11:25 < valentin> during the last two years there were 499 accidents in Berlin with police cars involved. 376 of theese accidents were caused by the police...
11:26 < jeru> valentin: we should send the cops to the driving school
11:27 < valentin> yeah
11:27 < valentin> when we drove like they do, we would loose our license very soon...
11:28 < jeru> valentin: but it could be also a motoric problem while switiching from bat to wheel mode :)
11:29 < valentin> hehe
11:29 < valentin> within two years there will be helicopters only ...
11:29 -!- mipe [~mika@dsl10040.japo.fi] has quit ["Leaving"]
11:31 < jeru> valentin: is there a construction plan for a portable r*cket launcher published under the GPL ? :)
11:32 < valentin> jeru: beware ! this chat is logged ...
11:32 < jeru> valentin: ahhh! ok ok ...
11:32 < jeru> sorry
11:33 < valentin> btw: why do you need a construction plan ? i think it will not be too hard to buy one if you have the money ...
11:36 < jeru> valentin: ack ... but then they'll know that you have one
11:37 < jeru> btw, we should establish a secured chat for doing jokes ...
11:38 < valentin> r*cket l*unchers are soo destructive - i would prefer more elegant solutions like infrared flares and and artifical fog
11:39 < valentin> is there an irssi module that does en/decryption ?
11:39 < jeru> stunnel
11:39 < valentin> is the output 7 bit ascii ?
11:42 < jeru> I know some security groups are using stunnel with their irc servers
11:44 < rxr> valentin: what do you want to try ?
11:45 * jeru is phoning
11:47 < valentin> rxr: nothing - just talk to jeru ,)
11:47 < valentin> hi rxr
11:48 < rxr> hm - I think there is some stability problem with the 360Mhz SPARC ...
11:48 < valentin> it went down over night ?
11:48 < rxr> last night it stopped building glibc due to some "bla-blubs.h" that did exist - and on the rebuild it stopped with:
11:48 < rxr> sparc64-t2-linux-gnu-ranlib /t2-trunk/src.glibc.1115453532.1032.0/glibc-2.3.5/ob
11:48 < rxr> jdir/libc.a
11:48 < rxr> make[1]: *** [/t2-trunk/src.glibc.1115453532.1032.0/glibc-2.3.5/objdir/libc.a] S
11:48 < rxr> egmentation fault (core dumped)
11:48 < valentin> oha
11:49 < rxr> but this can not be my T2 system - that parts run stable on my other SPARCs ...
11:52 < valentin> one more helicopter ...
11:52 < valentin> i hear / see at least 3 per day
11:59 < daja77> rxr: have you tried glibc cvs for gcc4?
12:00 < valentin> moin daja77
12:00 < rxr> daja77: that works, yes - do you have a problem with it?
12:00 < rxr> btw. vanilla gcc-4.0 has 2 intrusive code miscompilation bugs
12:00 < rxr> use gcc 4.0 or pull the two pathes out of bugzilla. ...
12:01 < rxr> btw. gcc 4.0 is not yet faster than 3.4
12:01 < rxr> we (in T2) do not yet adapt to it due to:
12:01 < rxr> - genreated extremely defect code when not patched or CVS
12:01 < rxr> - is build slower code than 3.4
12:01 < rxr> - does not yet compile release versions of glibc or linux-kernel ...
12:01 < daja77> hm icm,that explains a lot
12:02 < rxr> so we can releax while redhat and suse patch the packages to build with gcc 4.0 and we jump on the bandwagon when the work is done - and thistime not be me ,-) *evil grin*
12:02 < rxr> and when 4.0 generated better code ...
12:03 < rxr> daja77: in the next issue of the linux magazine there is the usual gcc overview from me - as you will see gcc 4.0 does have regressions in nearly any benchmark
12:03 < valentin> rxr: which of the ppp module keywords and ppp-options should i support in the initial version ?
12:03 < daja77> hehe, it was fun to try it anyway
12:03 < rxr> valentin: only the ones joe user need ...
12:04 < rxr> user password inject the default and speed options, thats it for now ...
12:04 < daja77> atm I want to know why I can't format a 200GB partion on the drive I bought >_<
12:04 < rxr> ah - and of course the no. to dial for analog and cellphone use
12:05 < valentin> ppp-speed-defaults applies to analog modem only ?
12:05 < rxr> http://exactcode.de/rene/hidden/gcc-article/2005-gcc-4.0/stat2-rt.png
12:06 < rxr> valentin: no - better only to DSL
12:06 < valentin> what about chat-defaults ?
12:08 < valentin> and chat init for analog ?
12:08 < CIA-10> rene * r8688 /trunk/package/base/memtester/ (5 files): * added memtester (4.0.5)
12:08 < rxr> valentin: yes chat for analog and handy ...
12:09 < valentin> can you paste me some analog config int /query ? i only have pppoe examples handy ...
12:09 < rxr> so far memtester has not caught an error on the sparc ...
12:33 < rxr> so - sparc cleaned ...
12:33 < rxr> somehow SPARCs loose the image of "iron boxes works for ages"
12:37 < rxr> I wonder if this output is quite right:
12:37 < rxr> # file core
12:37 < rxr> core: ELF 64-bit MSB core file, SPARC V9, version 1 (SYSV), SVR4-style, SVR4-style, SVR4-style, SVR4-style, SVR4-style
12:37 < rxr> quite some many SVR4-style and no program name ...
12:51 -!- mnemoc [~amery@200.75.27.18] has quit [Read error: 110 (Connection timed out)]
12:54 -!- mnemoc [~amery@200.75.27.94] has joined #t2
13:44 < rxr> # l */*/*.desc | wc -l
13:44 < rxr> 1644
13:51 < CIA-10> rene * r8689 /trunk/package/filesystem/udev/udev.desc: * updated udev (052 -> 057)
13:55 < rxr> building udev with dietlibc:
13:55 < rxr> make CC="diet gcc" LD="diet gcc"
13:58 < rxr> valentin: do you know if it is possible to transfer music files out of iTunes into my linux sytem if I decide to buy a few selected songs there ?
13:58 < rxr> otherwise I would need to plug a USB sound card with digial out into the box to stream the content either into the Delta 1010 digital in - or my old Tident 4D NX s.th. digital input jack ...
13:58 < rxr> that would be a bit time consuming ...
14:02 < valentin> rxr: no idea - i do not know much about apples drm
14:03 -!- keinek [~Keinek@201.254.8.228] has quit [Read error: 104 (Connection reset by peer)]
14:03 < valentin> rxr: there are some online shops without drm
14:05 < rxr> yeah - but you do not get much stuff there, do yo ?
14:07 < valentin> no idea - the question is: do they have your song or not
14:08 < rxr> what was the name of the vorbis selling shop ?
14:12 < valentin> no idea - google ?
14:13 -!- keinek [~Keinek@201.254.11.203] has joined #t2
14:13 < CIA-10> rene * r8690 /trunk/package/filesystem/udev/ (config.hlp config.in udev.conf):
14:13 < CIA-10> * added an option to udev to explicitly link it against dietlibc,
14:13 < CIA-10> defaulting to yes
14:17 < rxr> http://www.oggmusic.de
14:18 < rxr> http://www.OggStar.com
14:21 < rxr> "Hier k?nnen Sie sich eine Probe dieses Titels anh?renDieser Titel ist nicht einzeln verf?gbar" ... - Musicload.de
14:27 < rxr> lol - apple iTunes germany:
14:27 < rxr> TECHNISCHE SCHRITTE - 1-Click (R) 1-Click ist eine eingetragene und zu Gebrauch lizenzierte Marke von Amazon.com, Inc.
14:28 < rxr> ok - /me back to my initramfs work ,_)
14:29 < rxr> too bad the quality of the 128kBit mp3 radio streams is so poor - I would just dump them ...
14:29 < rxr> but with the artifacts I get headache listening to them too often ...
14:31 < rxr> what do you vote for, detecting the rootfs type to be mounted and loading only that fs module - or just load all filesystem modules available in the initrd?
14:36 < rxr> and I think we only want to include the major FSs ? that is not the toy amiga, beos and what else fluff lying around in the module dir, right ?
14:36 < rxr> I mean - does anyone boots from them *gr* ?
14:36 < valentin> rxr: depends a bit on the target
14:37 < valentin> for embedded, you would not want unecessary modules to be loaded
14:37 < rxr> for a real embeded solution used in products one will most probably carefully select the support needed and build all thos into the kernel ,-)
14:39 < rxr> so I'm going to hardcode the usually used mods for now ... and load all those available in the initramfs init ...
14:42 < valentin> ok ok
14:48 < rxr> 9.1M total
14:48 < daja77> thx rene ;)
14:48 < rxr> ^- the umcompressed size of what my find | grep selects for the initrd ...
14:48 < rxr> daja77: what for ? ;-)
14:49 < daja77> just found a t2 log on google about the ub driver, that causes pain here with that aldi drive too ^^
14:49 < rxr> hehe
14:49 < rxr> no problem ...
14:50 < daja77> low performance usb storage drive ts
14:50 < rxr> daja77: and, when do you start using T2 ,-)
14:51 < rxr> (just joking - not to affront you ... ,-)
14:51 < valentin> i have to buy some food for the weekend, bbl
14:51 < rxr> cu valentin
14:51 < daja77> for now a kernel rebuild would do, I hope ...
14:54 < daja77> just wondering where this option is
14:55 < daja77> ah ok found
15:04 < daja77> the question is, if that ub issue is still valid for 2.6.11
15:05 < rxr> yes, of course
15:06 < rxr> ub is a lightweight non-bug-workaround implementation - it will never perform as perfect as usb-storae and only work with a few devices ...
15:06 < rxr> but it is tiny and clean ... - but that's it ... ,-)
15:07 < rxr> 2.6M initramfs.gz
15:07 < rxr> oh - xfs is huge ...
15:08 < rxr> 557K fs/xfs
15:08 < rxr> who wants this beast ?
15:08 < rxr> even bigger than all the IDE drivers together :..
15:08 < rxr> 449K drivers/ide
15:09 < jeru> rxr: I like XFS for servers
15:09 < rxr> hi jeru
15:09 < rxr> what are the advantages you see ?
15:11 < jeru> rxr: It performs well, I had better experience than with reiserfs
15:14 < jeru> rxr: what do you use ?
15:15 < rxr> I use reiserfs for all requiring high performance since 3 years or so ...
15:15 < rxr> never had a single problem with it
15:15 < rxr> except the 2.4.10 s.th. kernel where some NFS inode mapping was broken - but it was not corrupting the FS or so - just preventing successfull I/O ...
15:16 < rxr> the most problem I ever experienced was this sh*ty Linux software RAID implementation
15:16 < jeru> rxr: ack
15:16 < rxr> I once hacked some bare code to recustruct data from a corrupted, due to ugly Miguel Icaza kernel code, in order to recover data ...
15:17 < rxr> since then I make a long bow around the software raid crap ...
15:17 < rxr> I wonder who signed off that code to make it into the kernel ...
15:17 < rxr> when s.th. from SCO is in the kernel it must be that brain deadness ..
15:18 < jeru> rxr: hehe :)
15:18 < rxr> so - next design decision - what shell to embed into the new initramfs ...
15:18 < rxr> kiss ?
15:19 < rxr> and I need the module tools ... statically linked ...
15:19 < rxr> hm - grummel ...
15:19 < rxr> is anyone screaming when I change the module packages to build against dietlibc by default ?
15:21 < rxr> ehrm
15:23 < jeru> rxr: I've no problems with statically linked module tools
15:23 < rxr> that would simplify the mkinitrs process - it could just use the system ones then ...
15:24 < rxr> otherwise we would neede an huge mkinitrd building them on the fly - or a 2nd set of in-system module tools that are linked with dietlibc ...
15:50 -!- rxr_ [~rene@e178129115.adsl.alicedsl.de] has joined #t2
15:50 -!- Topic for #t2: T2 | 2.1.0-beta4 RELEASED | The next generation of System Development Environments (SDE) | http://www.t2-project.org/
15:50 -!- Topic set by menomc [] [Sat Apr 16 00:33:34 2005]
15:50 [Users #t2]
15:50 [ _Ragnar__] [ jeru ] [ mnemoc] [ rxr ] [ valentin]
15:50 [ CIA-10 ] [ jsaw ] [ mtr_ ] [ rxr_ ]
15:50 [ daja77 ] [ keinek] [ nzg ] [ sparc-kly]
15:50 -!- Irssi: #t2: Total of 13 nicks [0 ops, 0 halfops, 0 voices, 13 normal]
15:50 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
15:50 < rxr_> re
15:50 < rxr_> hm - or I use the fedora nash with built in pivot_root and modprobe ...
15:50 -!- Irssi: Join to #t2 was synced in 18 secs
15:55 < rxr_> 2427 extents written (4 MB)
16:02 < valentin> re
16:06 < rxr_> hi valentin
16:06 -!- rxr [~rene@e178142170.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)]
16:06 -!- You're now known as rxr
16:07 < rxr> ok - only mod tools stuff missing
16:07 < rxr> the rest is working
16:07 < rxr> now ...
16:07 < rxr> it already initializes /dev with udev ...
16:07 < rxr> and does hardware scanning via hotplug++ ...
16:07 < rxr> 7.1M initramfs
16:07 < rxr> 2.7M initramfs.gz
16:07 < rxr> 2444 extents written (4 MB)
16:07 < rxr> GRUB boot CD -^
16:08 < rxr> after undusting the sparc glibc build ..
16:08 < rxr> == 05/07/05 13:10:12 =[1]=> Finished building package glibc.
16:09 < rxr> == 10:39:20 =[1]=> Building base/glibc [2.3.5 2.2.0-alpha].
16:09 < rxr> what times did I paste previously ?
16:09 < rxr> how much faster is that box ?
16:09 < rxr> == 13:10:38 =[1]=> Building base/glibc32 [2.3.5 2.2.0-alpha].
16:14 < rxr> grubby \- command line tool for configuring grub, lilo, and elilo
16:14 < rxr> nice ugly stuff fedora has in peto ...
16:16 < rxr> oh my god
16:16 < rxr> ok - I do not want nash ..
16:16 * rxr seting up to build the module tools statically with dietlibc - I think this will make me more happy ...
16:17 < jeru> rxr: hehe :)
16:17 < rxr> it does not even do modprobe itself ...
16:17 < rxr> Additionally, if nash is invoked as modprobe, it will immediately exit
16:17 < rxr> with a return code of zero. This is to allow initrd's to prevent some
16:17 < rxr> extraneous kernel error messages during startup.
16:17 < rxr> rm -rf ...
16:21 < valentin> rxr: nash invoked as modprobe ?
16:21 * valentin needs some tea
16:22 < valentin> sir jeru, please send me some from india ...
16:24 < jeru> valentin: I think the tea business wasn't the most profitable business from "East India Company" anyway ... :)
16:24 < CIA-10> rene * r8691 /trunk/package/filesystem/udev/ (config.hlp config.in udev.conf udev.desc):
16:24 < CIA-10> * removed dietlibc option from udev, decided to use the generic
16:24 < CIA-10> dietlibc support instread and thus marked it dietlibc
16:27 < mnemoc> hi
16:28 < rxr> hi
16:30 < jeru> hi! señor mnemoc :)
16:30 < mnemoc> don christian
16:31 < mnemoc> hi rxr ;)
16:31 < jeru> mnemoc: don christiano sounds better ;)
16:31 < mnemoc> cristiano in that case
16:31 < jeru> oh !
16:31 < mnemoc> but cristian exists
16:32 < mnemoc> jeru: i want to write an stone module to install, update and de-install zope modules from instances
16:32 < jeru> but don alejandro sounds nice too ... :)
16:32 < mnemoc> :)
16:32 < jeru> oh! ...
16:33 < mnemoc> jeru: i thought i adding a list of instances on /etc/opt/zope/ instancetab maybe
16:33 < valentin> hi mnemoc
16:34 < mnemoc> hi valentin
16:35 < jeru> mnemoc: hmmm ... we should have /etc/opt/zope/instance dir and placing all inctance configuration there
16:35 < rxr> the old 2.4 insmod is huge ... - damn beast ...
16:36 < jeru> we could even create the instances with the upcoming stone mod
16:36 < mnemoc> jeru: that would mean to patch mkzopeinstance
16:36 < jeru> why ?
16:37 < jeru> because of the zope.conf location ?
16:37 < mnemoc> yes
16:37 < jeru> we could make a link ?
16:37 < mnemoc> yes, but that would void the documentation
16:39 < daja77> the usb drive works now
16:39 < CIA-10> rene * r8692 /trunk/target/ (desktop/config.in generic/config.in):
16:39 < CIA-10> * enavled dietlibc for generic and desktop and marked the initrd tools
16:39 < CIA-10> to use dietlibc
16:40 < daja77> <- happy
16:40 < mnemoc> hi happy daja77
16:40 < CIA-10> rene * r8693 /trunk/package/ (base/kiss/ shells/kiss/): * moved base/kiss into shells
16:42 < mnemoc> .oO( uhm... 32 commits pending for review )o
16:45 < rxr> mnemoc: =
16:45 < rxr> ? even
16:45 < jeru> mnemoc: we should also consider the case where 2 or more instances of zope share the same products dir
16:46 < mnemoc> rxr: http://code.dogmap.org/lintsh/
16:46 < mnemoc> jeru: we don't need... symlinks will do the magic
16:48 < jeru> mnemoc: ok
16:48 < mnemoc> rxr: http://skarnet.org/software/execline/index.html
16:52 < rxr> mnemoc: thanks
16:52 < rxr> damn - I can not debug kiss ...
16:52 < rxr> break domount
16:52 < rxr> bkiss: mount got signalled with sig 5
16:55 < mnemoc> rxr: on skarnet.org you can find nice embutils complements
16:55 < rxr> good to know ,-)
17:13 < rxr> http://ec-outpost.dyndns.org/t2-initramfs-ng.iso
17:13 < rxr> if you have time test it on hardware and report feedback ,-)
17:13 < rxr> boots in qemu ...
17:14 < rxr> size: 4874KB#
17:16 < mnemoc> no cdr-s :(
17:16 < rxr> hehe ,-)
17:16 < rxr> then wait for v2 tomorrow - that will boot into X ,-)
17:16 < rxr> I'll have boxes to test here ,-)
17:16 < mnemoc> oh
17:16 * rxr soon off a bit
17:18 < mnemoc> checking whether db is at least 4.1... ./configure: line 9864: test: : integer expression expected
17:18 < mnemoc> no. you need at least db 4.1
17:19 < mnemoc> err....
17:19 < mnemoc> ooo-m100
17:20 < rxr> cdrecord is a hell of a software ...
17:21 < rxr> first it delays for 5 seconds or so due to the warning that tao or so might bot be the default in teh future
17:21 < rxr> and after the delay it tells me I placed the options wrong
17:21 < rxr> global ones after the track ...
17:21 < rxr> h*ll - that is when every crappy C monkey writes his own optoin parsing code ...
17:22 < rxr> good the sparc still building glib32 ...
17:22 < mnemoc> | if(DB_VERSION_MAJOR == 4 && DB_VERSION_MINOR == 2) return 0;
17:22 < mnemoc> configure:9869: result: no. you need at least db 4.1
17:22 < rxr> let's hope it finishes ...
17:22 < rxr> hahaha ...
17:22 * rxr away - cu
17:22 < mnemoc> cu rxr
17:42 < valentin> cu rxr
18:26 < jeru> re
18:28 < mnemoc> hi www2^Wjeru
18:28 < jeru> mnemoc: I know ... that there is sth. pending ... I guess it will consume the rest of the weekend
18:30 < mnemoc> jeru: sorry to bother... but i want to drop that damn wiki
18:30 < jeru> mnemoc: no problem ... :)
18:31 < jeru> mnemoc: I've Zope3 in "shape" but have to extend it a bit.
18:32 < mnemoc> great
18:32 < jeru> mnemoc: I'll add a new instance this evening
18:33 < mnemoc> uhm? can you move old data?
18:34 < jeru> in the case of the bugtracker I'm not quite sure, but the rest will be migrated
18:35 < mnemoc> don't kill the old then... to be able to handcopy stuff from bugtracker
18:36 < jeru> mnemoc: now that you remember me, I will kill it :p
18:36 * jeru hides
18:36 < mnemoc> :(
18:36 < jeru> just joking
18:36 < jeru> of course I'll leave the "old" one
18:37 < jeru> don't cry don alejandro
18:37 < mnemoc> snif snif
18:38 < valentin> helicopter again ...
18:38 < mnemoc> uhm?
18:38 < valentin> i should sue otto vor noise polution
18:39 < valentin> *for*
18:40 < jeru> valentin: same here ... a lot of helicopters
18:41 < valentin> waste of money, polution with gas and noise ... only to "hunt down sprayers" ?
18:42 < jeru> valentin: as you pointed out some days ago, I guess the primary reason are not the sprayers anyway
19:05 < CIA-10> chris * r8694 /trunk/package/mail/qmail/qmail.desc: * updated qmail-ldap patch ( 20041201 -> 20050401a )
19:10 < CIA-10> amery * r8695 /branches/2.1/package/network/samba/fix_2686.patch: * fixed samba (samba bug #2686) due to problem using usemgr.exe managing LDAP groups
19:12 < mnemoc> usrmgr :(
19:13 < jeru> mnemoc: how is status of your samba-ldap work ?
19:15 < mnemoc> kind of working
19:15 < mnemoc> installed but client is trying to use it
19:16 < jeru> means it is working stable ?
19:17 < mnemoc> testing phase
19:17 < jeru> ok
19:18 < mnemoc> i wasn't able to add (ldap)users to groups on plone :\
19:18 < mnemoc> NotImplemented crap
19:19 < jeru> what ? really ? and "was" means you've implemented it ?
19:20 < mnemoc> no, "was" means i haven't continued
19:20 < jeru> hmmm ... :/
19:21 -!- minto [~chatzilla@82-217-66-7.cable.quicknet.nl] has joined #t2
19:22 < jeru> mnemoc: but as I rememeber you can map LDAP groups to Zope roles
19:22 < minto> hi all
19:22 < jeru> hi minto
19:22 < minto> hi jeru, how are your bunnies doing?
19:25 < jeru> minto: the small one is quite fine and very funny today, and the human one was replaced ...
19:26 < minto> jeru: So you do hava a substitute right now?
19:27 < jeru> yes ... :)
19:28 < jeru> I was out dancing salsa with her yesterday :)
19:30 < valentin> jeru: aha ...
19:30 < jeru> valentin: what aha ? :)
19:30 < valentin> i see why we have no www2 yet :)
19:30 < valentin> hi minto btw
19:31 < mnemoc> jeru: yes?
19:31 < minto> Hi valentin
19:31 < jeru> valentin: partially nope ...
19:31 < jeru> ;)
19:31 < jeru> mnemoc: yes what ?
19:31 * jeru hides
19:31 < minto> you're right valentin, he said partially
19:31 < mnemoc> jeru: mnemoc: but as I rememeber you can map LDAP groups to Zope roles
19:32 < jeru> mnemoc: ack
19:32 < mnemoc> jeru: i can only auth via LDAP by now
19:32 < mnemoc> jeru: i can't see names, emails, etc... or change anything
19:32 < minto> does libffi right a bell to anyone?
19:33 < jeru> mnemoc: mom phone
19:33 < valentin> ibffi is a library that really only provides the lowest, machine dependent layer of a fully featured foreign function interface.
19:33 < valentin> +l
19:34 < mnemoc> jeru: momment, i'm at phone or phoneing your mother?
19:34 < valentin> minto: why ?
19:34 < minto> valentin: that's what it said but it looks like it hasn't changed since 1998.
19:34 < minto> valentin: SableVM uses it.
19:34 < jeru> mnemoc: moment ;)
19:35 < minto> I read about it being part of gcc and then somewhere else this is denied.
19:35 < minto> Can't find it in any of my flists. So it probably is not part of T2.
19:35 < valentin> minto: gcc has foreign language call support ...
19:36 < minto> valentin: I wonder it that is enough for sableVM. Maybe I'll try.
19:36 < minto> thanks
19:38 < valentin> no package downloads libffi yet
19:40 < minto> I could add libffi, however http://sources.redhat.com/libffi/ says version 1.20 (1998). Looks ancient and out of date.
19:41 < minto> Especially if it is as low level as you say, it probably does not support any of the newer archs.
19:53 < valentin> think so, too..
20:10 < rxr> re
20:10 < rxr> isn't libffi part of gccc
20:11 < rxr> gcc even
20:11 < rxr> but by default not installed
20:12 < valentin> hi rxr
20:14 < rxr> noone tried the new livecd technology test ?
20:14 < rxr> at least it boots in qemu and susan's athlon ...
20:14 < rxr> slot-a that is
20:14 < minto> hi rxr
20:15 < valentin> nice
20:15 < valentin> what about X ?
20:16 < rxr> no X yet
20:16 < rxr> this will be done tonight and at least running early tomorrow ...
20:16 < rxr> I'm happy the hotplug++ early userspace with udev works so far ... ,-)
20:17 < rxr> major technology milestone
20:24 < valentin> sounds good
20:24 * valentin evening dinner
20:28 < rxr> damn - kiss has no "if" and "for" and such ...
20:28 < rxr> and it does not want to assign variables with lower case ... ?!?!
20:36 < minto> Uhmm, rxr: might gcc.conf be incorrect in respect to enabling shared libraries?
20:36 < minto> the conf file says: --enable-shared=yes
20:36 < minto> gcc configure manual says : --enable-shared[=package[,...]]
20:36 < rxr> I do not know off hand ...
20:37 < minto> I found this at: http://gcc.gnu.org/install/configure.html
20:37 < minto> To build libffi it needs: --enable-shared=libffi
20:38 < rxr> minto: then just replace yes with that an stry a new build ,-)
20:39 < minto> Maybe just remove =yes to build all as shared
20:39 < rxr> no idea ...
20:39 < rxr> I could test ,-9
20:39 < minto> I'll test too
20:42 < rxr> damn
20:42 < rxr> the dietlibc has not the tiny preparation for dynamic ctor / dtor handling ...
20:42 < rxr> grummel
20:42 * rxr hacking that ...
20:42 < rxr> as if my TODO would not already be long enough ...
20:46 < rxr> oh - I think that should be fixed now
20:50 < CIA-10> rene * r8696 /trunk/package/base/dietlibc/sparc64_dyn_start.patch: * fixed dietlibc with WANT_DYNAMIC for sparc64
20:50 < rxr> == 05/07/05 15:22:25 =[1]=> Finished building package glibc32.
20:50 < rxr> I hope the dust removal stability is permanent ...
20:53 < rxr> when I call for it:
20:53 < rxr> !> FATAL: can't create bin-sparc64/start.o: No such file or ..
20:58 < CIA-10> rene * r8697 /trunk/package/shells/dash/ (dash.conf dash.desc):
20:58 < CIA-10> * marked dash DIETLIBC compatible and fixed it to install into /bin/
20:58 < CIA-10> by default
20:59 < CIA-10> rene * r8698 /trunk/target/ (desktop/config.in generic/config.in):
20:59 < CIA-10> * made desktop and generic build dash instead of kiss with dietlibc
20:59 < CIA-10> for use in the initramfs - kiss is too limitted ...
20:59 < rxr> ah - my god
20:59 < rxr> dash is also too brain dead to expand s.th. as simple as this ...
21:00 < rxr> /lib/modules/*/kernel/fs/{*/,}*.{ko,o}
21:01 < valentin> rxr: screems for an exactsh ?
21:01 < minto> new build with modifie gcc is started
21:01 < rxr> yes - in
21:01 < rxr> C++
21:01 < rxr> esh or sh++ =
21:01 < rxr> ? even
21:01 < valentin> or xsh++
21:01 < rxr> x for ?
21:01 < valentin> eXact
21:02 < rxr> # echo {a,b}
21:02 < rxr> {a,b}
21:02 < rxr> ...
21:02 < rxr> f*cking shell ...
21:04 < rxr> oh myyyyyyyyyyyyyyyyyyy
21:07 < rxr> hey!
21:07 < rxr> sash is able to expand {a,b}
21:08 < _Ragnar__> kiss isn't meant to be used as a scripting shell
21:08 < rxr> _Ragnar__: I need s.th. tiny for the bootup
21:08 < rxr> I did not expected people leave out s.th. as simle as {a,b} expansion ...
21:08 < _Ragnar__> ash or sash would be better for that
21:09 < rxr> I already used dash, the slithly fixuped ash - but it doesn't do that neither ...
21:09 < rxr> sash does it ...
21:09 < _Ragnar__> ah
21:09 < rxr> it is just a tiny script for the initramfs ...
21:09 < rxr> no monster script
21:09 < _Ragnar__> :)
21:09 < rxr> I can leave out much - but not s.th. as simple as {,} expansion ...
21:11 < valentin> how many *sh are there ?
21:11 < rxr> hey! sash even is able to:
21:11 < rxr> echo ${PATH//:/ }
21:12 < rxr> but how to set a var ...
21:12 < rxr> > a=123
21:12 < rxr> a=123: No such file or directory
21:12 < rxr> > a="bc"
21:12 < rxr> a=bc: No such file or directory
21:12 < rxr> > and no for also ...
21:12 < _Ragnar__> set or export?
21:12 < valentin> rxr: so we will have a xsh++ soon ?
21:12 < rxr> rxr: already tried seet ...
21:12 < rxr> valentin: I need a running live cd until monday ..
21:12 < rxr> I do not start an own shell now ...
21:12 < valentin> i know
21:13 < rxr> I just wrote the base of an STL and a hotplug in C++ ...
21:13 < valentin> i said soon, not tonight
21:13 < rxr> this need to stop soon ...
21:13 < rxr> yeah - maybe soon ...
21:13 < rxr> unbelievable ...
21:13 < rxr> rxr: neither export nor set works in sash
21:14 < rxr> setenv
21:15 < _Ragnar__> *lol* oh, csh clone
21:16 < rxr> but sash has no for ...
21:16 < rxr> damn
21:17 < valentin> is there another thing than tcp/udp for accept/reject/drop rules ?
21:17 < rxr> I think udp/tcp is optional
21:18 < rxr> so that it does not matter what goes to e.g. port 80 - tcp and udp are dropped ...
21:18 < rxr> but I would also need to relookup the details of that rules
21:18 < rxr> .oO(should I really use pdksh for the initramfs ...)
21:19 < _Ragnar__> port 80 makes only sense with tcp or udp
21:19 < _Ragnar__> icmp doesn't have a port field for example
21:20 < rxr> _Ragnar__: I only said specifing rcp or udp is optional ...
21:20 < rxr> drop 80 will drop both, tcp and udp IIRC
21:20 < _Ragnar__> yea I was adding to your statement;)
21:20 < rxr> ah - ok
21:22 < valentin> i will look it up in the module
21:23 < rxr> sash also builds with dietlibc without a tweak ...
21:24 < CIA-10> rene * r8699 /trunk/package/shells/sash/sash.desc: * marked sash DIETLIBC
21:25 < CIA-10> rene * r8700 /trunk/package/shells/sash/sash.desc: * added the sash' authors mail address to the sash.desc
21:25 < rxr> damn - what do I do with my startup code ...
21:25 < rxr> wither for or xargs would be nice ...
21:25 < rxr> neither has sash ...
21:25 < valentin> no for ??
21:25 < rxr> no for ...
21:25 < rxr> dash and ash have neither ...
21:26 < valentin> while ?
21:26 < rxr> while
21:26 < rxr> while: No such file or directory
21:26 < valentin> repeat until ? loop ?
21:26 < valentin> there must be sth like that ...
21:29 < valentin> isn't there a manual for sash ?
21:29 < rxr> valentin: there must nothing
21:29 < rxr> sash is a stand alone rescue shell
21:29 < rxr> if the author did not need loops there are none ...
21:29 < daja77> goto ^^
21:30 < valentin> how will you rescue sth if you have no possibility to iterate over stuff ?
21:30 < valentin> even the DirtyOS shell had a for statement
21:30 < valentin> iirc
21:30 < valentin> rxr: how about recursive function calls ?
21:31 < rxr> no functions in sash ...
21:32 < rxr> I hardcode the filesystem modules to load for now ...
21:39 < rxr> 184040 May 7 21:38 /bin/pdksh
21:39 < rxr> damn - a 184kB shell just for loops ... and conditionals ...
21:43 -!- Veki [~chatzilla@217.24.19.40] has joined #t2
21:44 < rxr> oh my
21:44 < rxr> and pdksh can not ${var//pat/subst}
21:44 < rxr> ...
21:45 < rxr> hi Veki
21:45 < Veki> hi rxr
21:45 < Veki> :-)
21:46 < Veki> how are you doing
21:46 < rxr> oh - shells get on my nerves
21:46 < rxr> aside from that fine ,-)
21:46 < valentin> hi Veki
21:46 < Veki> hi
21:47 < Veki> I am planning to build target However, before all I want to make a plan which packages I do need
21:47 < Veki> which shells are getting on your nerves
21:47 < rxr> all except bash
21:47 < rxr> and bash is also ugly enough ... ,-)
21:47 < Veki> ah, that is sometimes really frustrating thing
21:48 < Veki> o_O
21:49 < Veki> sinec ethis would be my first project with t2, i am beginner really in this
21:49 < Veki> I owuld liek to ask from time to time some questions I need to be at leats someone who tried a couple of times to do something in order to overcome the first barrier
21:50 < Veki> how to build target and what are preconditions for that
21:50 < Veki> in addition this will help me to write manual fro th eneginners since I can see what beginner may to do mistakenly when planning or starting to build target
21:51 < rxr> yes, sure, just ask here
21:51 < rxr> or on the list if noone is awake
21:54 < Veki> ok, thanks
21:54 < Veki> for example, If I need to build educational software target with KDE should I consider packages by their name like Kchem which is opackage for chemistry or I should know all package's dependencies
21:56 < rxr> so far we do not have a way to allow target to specify "add kdeedu and all its dependencies" - we wll have this some time soon in 2.2 or 2.3
21:56 < rxr> so you have to do the selection like e.g. "qt, arts, kdelibs, kdebase, kdeedu" manually ...
21:59 < Veki> OK, but if that is enough that is OK. ACtually, it does not need to be Kde , it can be whatever, but idea is that I do not need to know all dependnecies by hearst since some packages really need a lot dependencies etc.
21:59 < valentin> rxr: what is that "calculating gnome dependencies" stuff that is executed during scripts config ? can't you code sth similar for a target
21:59 < valentin> ?
22:00 < rxr> that is a custom "hack"
22:00 < rxr> well - we have the dependies in the cache file - it is just that we do not yet have the graph generator
22:00 < valentin> and we had too mutch trash in the dep's in the past ....
22:01 < rxr> the desktop target has an own "hack" that does try to get the basic gnome overview ,-)
22:01 < rxr> yep - after the next build the deps should be quite clean
22:01 < rxr> as soon as they arrive I plan to take a look into the remaining noise
22:02 < rxr> oh
22:02 < rxr> I accidently noticed
22:02 < rxr> /usr/usr/bin/gocr
22:02 < rxr> if anyone could take a lok into gocr some time ,-)
22:03 < valentin> is this an OCR software ?
22:03 < rxr> yes
22:04 < rxr> we have two OCR packages
22:04 < rxr> both untested - I only packed them ...
22:04 < rxr> packaged
22:08 < rxr> I'll need some tea soon ...
22:12 < rxr> http://ec-outpost.dyndns.org/t2-initramfs-ng-v2.iso
22:12 < rxr> still no X ,-)
22:16 < valentin> rxr: i could test it on K's box
22:16 < rxr> not yet
22:17 < valentin> :)
22:17 < rxr> as soon as it detects the CDROM and mounts the sqash and unionfs you can give it a try
22:17 < rxr> I'm preparing the test quash with /bin /sbin /lib and such only ...
22:19 < rxr> 53755904 May 7 22:18 live.squash
22:19 < rxr> # du -csh bin etc lib sbin
22:19 < rxr> 149M total
22:32 < rxr> I wonder why mkdir in my test iso does not work ...
22:32 < rxr> sh complains no such file or diretory
22:33 < rxr> but when I execute the same static linked dietlibc binary on my system it works just fine ...
22:33 < rxr> strange
22:33 < rxr> ah - it is dyn linked
22:50 < rxr> .oO
22:52 < rxr> I think I once had loop device setup code in some initrd
22:52 < rxr> I'm afraid I think I lost that code ...
23:01 -!- mtr [~Michael@p54AFB0A5.dip0.t-ipconnect.de] has joined #t2
23:03 < mnemoc> re
23:05 < mnemoc> checking for /usr/share/java/db-4.3.jar... no
23:05 < mnemoc> checking for /usr/share/java/libdb4.3-java.jar... no
23:05 < mnemoc> configure: error: db Java bindings not found
23:05 < mnemoc> i'm hating this ooo
23:09 < rxr> 36641 May 7 23:08 losetup
23:09 < rxr> ok - huge and not pretty ...
23:09 < rxr> util-linux marathon vi'ed to compile stand alone and with dietlibc ... *g*
23:09 < rxr> later I patch it into embutils - but first I need this livecd going ...
23:10 < mnemoc> rxr: does ooo-m100 build there?
23:10 < rxr> no idea
23:10 < rxr> mtr or so updated it ...
23:10 < rxr> I have no spare CPU cycles for it in the next 2 days
23:11 < mnemoc> mtr: does ooo-m100 build there?
23:13 < mnemoc> rxr: how is your project going?
23:13 < mnemoc> go open
23:13 < mnemoc> err
23:15 < rxr> ok - quick hack losetup works ... ,-)
23:15 < rxr> working in a todays commercial world gets more eXTreme programming every week ...
23:15 < valentin> rxr: are there option-only ppp options ?
23:15 < valentin> i.e there is no value behind the option name ?
23:16 < rxr> yes, there are value less ppp options ...
23:16 < rxr> usepeerdns and so on ...
23:16 < valentin> :(
23:16 < rxr> why do you ask ?
23:16 < valentin> usepeerdns is not toggle-able anyway
23:16 < rxr> mnemoc: stressing but we are going forward ...
23:16 < valentin> because i did not code it that way ....
23:16 < rxr> the deadline is a bit strong - as always
23:17 < valentin> ok, i fix it
23:17 < rxr> fix what ?
23:18 < rxr> ah - ...
23:19 < valentin> for the gui the option stuff is quite hardcoded in the first version
23:19 -!- mtr_ [~Michael@p54AF96DE.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
23:19 < rxr> valentin: I have no problem with that ...
23:19 < rxr> see what I struggle with ...
23:20 < valentin> yep
23:20 < valentin> for the little money they pay they get a 'little brain' product ...
23:21 < valentin> just kidding ,)
23:23 -!- minto [~chatzilla@82-217-66-7.cable.quicknet.nl] has quit [Read error: 145 (Connection timed out)]
23:23 < mnemoc> rxr: what should i do with bdb-java-bindings needed by openoffice now? move jdk-s before bdb?
23:24 < rxr> no idea - I do not want the ooo build depend on any binray only java crap
23:24 < rxr> if ooo does not build anymore in an opensource setup I'll be forced to svn revert to the last know good revision ...
23:24 < mnemoc> m90
23:25 < mnemoc> i hope this bindings can be built with jikes
23:26 < rxr> why is a m90 m10x change so intrusive ?
23:27 < mnemoc> src680 still highly volatile
23:37 < valentin> i am confused
23:38 < valentin> is it ppp_option or ppp-option ?
23:38 < valentin> same for all other ppp keywords ?
23:38 < rxr> - is translated to _ IIRC
23:38 < valentin> i have different versions in different working configs
23:38 < rxr> so both work
23:38 < valentin> :((
23:38 < rxr> why ?
23:38 < valentin> so _ is the official version ?
23:38 < valentin> i just support _ now
23:39 < valentin> is this translation done for all keywords ?
23:40 < rxr> yep
23:41 < valentin> this is $1 only ?
23:41 < rxr> yes, IIRC only the keyword should be substituted ...
23:44 < mnemoc> valentin: bash doesn't support '-' on names
23:44 < mnemoc> valentin: ppp wants '-' on options
23:48 < rxr> eiiiiiiiiiiii
23:48 < rxr> does this embutil mount not support read-only mounty my oh my
23:48 < rxr> after ths work I'll have a lot to merge upstream ...
23:48 < rxr> and properly in t2 that is first...
23:49 < rxr> ah - maybe -r only ...
23:49 < rxr> ahhh!
23:49 < rxr> works ,-)
23:50 < rxr> and I think this line is wrong:
23:50 < rxr> if(result==-1 && errno==EROFS) {
23:51 < rxr> embutils mount line 285
23:51 < rxr> it should be EPERM or so ..
23:51 < rxr> s/so ../both/
23:52 < rxr> ok - not it boots up, detects the cdrom, and loopback mounts the squashfs ..
23:53 < rxr> finally time for a tea and a tiny night snack ...
23:53 < valentin> snack - remembers me about the ice i have in my fridge ...
23:56 < rxr> my snack are rolls with meat salat
23:56 < rxr> no coco gadgets ... ,-)
--- Log closed Sun May 08 00:00:14 2005