T2 IRC Log: 2006-01-01

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 Sun Jan 01 00:00:18 2006
00:02 < misl> happy new year to y' all.
00:05 < valentin> happy new year all
00:05 < valentin> :)
00:39 < morfoh> happy new year my friends :D
00:40 < karasz> anybody still on
00:40 < karasz> ?
00:40 < morfoh> I wish you all the best ... thanks for your support ;)
00:40 < karasz> yuheee morfoh
00:40 < morfoh> karasz: happy new year :D
00:41 < karasz> happy new year to u to morfoh
00:41 < morfoh> thankx :D
00:41 < karasz> it is amazing that i can spell ....
00:46 < karasz> at least i think so
00:47 < karasz> mnemoc, ensure me that i can spell....
00:47 < karasz> or at least morfoh do so...
00:48 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
01:01 * Stelz is back (gone 02:40:59)
01:02 < Stelz> happy new year :)
01:08 < karasz> 2 u 2 Stelz
01:08 < Stelz> :)
01:08 < karasz> wazup Stelz
01:08 < karasz> ?
01:08 < Stelz> none
01:08 < Stelz> all is ok
01:08 < Stelz> :)
01:08 < karasz> party at the firm here ...
01:09 < karasz> pretty boring i guess :(
01:10 < CIA-9> jsaw * r14372 /trunk/package/filesystem/parted/ (link-order-fix.patch parted.conf parted.desc):
01:10 < CIA-9> * A happy new year, and
01:10 < CIA-9> * fix linkorder of clearfat
01:10 < karasz> `glückliches neues Jahr` morfoh and rxr
01:25 < morfoh> thanks karasz ;)
01:26 < Stelz> hello, morfoh :)
01:26 < karasz> hohohoo what do you do online now?
01:26 < Stelz> happy new year
02:13 < rxr> re
02:13 < rxr> happy new year to anyone
02:13 < Stelz> happy new year, rxr
02:16 < rxr> hi Stelz
02:16 < rxr> in fact it is already 2:16 here, but just came back ... ,-)
02:17 < rxr> and soon to bed ...
02:17 < Stelz> gig
02:17 * Stelz Date/Time is now [ Sun Jan 1 04:15:28 ]
02:19 < CIA-9> rene * r14373 /trunk/package/shells/bash/bash.desc:
02:19 < CIA-9> * marked bash NOPARALLEL since on a dual-core ppc64 with PARALLEL_MAX=4
02:19 < CIA-9> it fails
02:21 < CIA-9> rene * r14374 /trunk/package/base/dietlibc/pkg_patch/pkg_bash.patch:
02:21 < CIA-9> * patch bash for dietlibc, include termios.h where needed and hack to
02:21 < CIA-9> make it cross compile with different libc's again (stupid Makefile
02:21 < CIA-9> setup)
02:21 < CIA-9> rene * r14375 /trunk/package/base/dietlibc/pkg_patch/pkg_curl.patch: * path away macro define sanity check on curl for dietlibc
02:35 < CIA-9> rene * r14376 /trunk/package/base/dietlibc/pkg_patch/pkg_ppp.patch:
02:35 < CIA-9> * fixed ppp cross build by actually removing a broken chunk of the
02:35 < CIA-9> current dietlibc ppp patch
02:35 < rxr> n8 all
02:47 < valentin> gn8 rxr and a happy new year to you
02:50 < Stelz> cu, rxr
02:57 < nzg> rehi and a happy new year
03:18 -!- VX [n=stealth@80.240.209.21] has joined #t2
03:19 -!- Stelz [n=stealth@80.240.209.21] has quit [Read error: 110 (Connection timed out)]
03:22 < karasz> morfoh, are you still online
03:22 < karasz> ?
03:23 < karasz> anybody still in 2005?
03:26 < karasz> guess not...
03:29 -!- nzg [n=tschmidt@p54AFD1D2.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)]
04:04 * VX is away: sleep
06:13 -!- sparc-kly [n=mubex@64.237.255.10] has joined #t2
06:31 -!- sparc-kly__ [n=mubex@64.237.242.138] has quit [Read error: 110 (Connection timed out)]
08:31 -!- morfoh [n=jeru@p54BEDAB5.dip.t-dialin.net] has left #t2 []
08:36 -!- mipe [n=mipe@dsl10040.japo.fi] has joined #t2
09:56 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
09:56 < misl> A happy new year on this brand new day to everybody :)
10:02 < mipe> yeah, happy new year... its starting really great... water pipes are broken... no water :)
10:05 < misl> mipe: It seems you are still smiling. :)
10:10 < mipe> well i'm getting used to it... water was out xmas day also...
10:10 < mipe> electricity blackout on pumping station
10:27 -!- mipe [n=mipe@dsl10040.japo.fi] has quit ["recon"]
10:28 -!- mipe [n=mipe@dsl10040.japo.fi] has joined #t2
10:33 -!- mipe [n=mipe@dsl10040.japo.fi] has quit ["re"]
10:36 -!- mipe [n=mipe@dsl10040.japo.fi] has joined #t2
10:38 < misl> mipe: seems like there is something wrong with your connection as well :)
10:40 < mipe> naah, needed to reconnect few times
10:45 -!- VX [n=stealth@80.240.209.21] has quit [Read error: 110 (Connection timed out)]
10:45 -!- Stelz [n=stealth@80.240.209.21] has joined #t2
10:58 -!- mipe_ [n=mipe@dsl10040.japo.fi] has joined #t2
10:58 < mipe_> damm, just had power cut
11:09 -!- mipe [n=mipe@dsl10040.japo.fi] has quit [Read error: 110 (Connection timed out)]
11:09 -!- mipe_ is now known as mipe
11:24 < valentin> moin
11:32 < rxr> moin
11:32 < rxr> happy new year valentin
11:33 -!- VX [n=stealth@80.240.209.21] has joined #t2
11:34 < misl> hi valentin, rxr happy new year and a prosperous t2 to all of us :)
11:34 -!- Stelz [n=stealth@80.240.209.21] has quit [Read error: 110 (Connection timed out)]
11:34 [Users #t2]
11:34 [@ChanServ] [ CIA-9 ] [ LMJ ] [ mtr ] [ valentin]
11:34 [ _Ragnar_] [ demian] [ mipe ] [ owl ] [ VX ]
11:34 [ axion ] [ jsaw ] [ misl ] [ rxr ] [ zod ]
11:34 [ Baldzius] [ karasz] [ mnemoc] [ sparc-kly]
11:34 -!- Irssi: #t2: Total of 19 nicks [1 ops, 0 halfops, 0 voices, 18 normal]
11:35 < valentin> hi misl
11:35 < valentin> hi rxr
11:35 < rxr> hi misl - happy new year to you, too ,-)
11:35 < valentin> happy new year to you (think i said this before at 0:00 CEWT)
11:35 < valentin> or at 0:05
11:35 < valentin> how were your celebrations ?
11:36 < rxr> gemuetlich - tiny group with board games and such
11:37 < valentin> we did a six person robo rally on a single board. Of course i am the winner :)
11:37 < rxr> "of course" ,-)
11:38 * VX is back (gone 00:05:25)
11:39 < rxr> hm
11:39 < valentin> kde stuff building
11:39 < valentin> should have disabled gnome - would have saved me a day or so
11:39 < rxr> the more of util-linux needs a.out.h - yet another ugly header shippd with glibc ...
11:39 < rxr> my oh my
11:40 < rxr> we really think about a ExactUn*x target where each file is reviewed and crap removed ...
11:40 * rxr patching a bare a.out into dietlibc to make latest util-linux/more heppy ...
11:40 < rxr> happy ....
11:41 < VX> :)
11:42 < VX> happy ... rxr
11:44 < rxr> VX: thanks - you too
11:47 < Baldzius> moin & happy new year - valentin , rxr, VX , misl and * :)
11:48 < rxr> welcome Baldzius ,-) happy new year ,-!
11:48 < VX> :)
11:48 -!- VX is now known as Stelz
11:48 -!- Stelz [n=stealth@80.240.209.21] has quit ["There is intelligent life out there..."]
11:48 -!- Stelz [n=stealth@80.240.209.21] has joined #t2
11:49 < CIA-9> rene * r14377 /trunk/package/base/dietlibc/pkg_patch/pkg_util-linux.patch: * updated dietlibc util-linux patch
11:50 < valentin> hi Baldzius and a hny
12:00 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
12:06 < valentin> hehe: http://ars.userfriendly.org/cartoons/?id=20060101
12:10 < CIA-9> rene * r14378 /trunk/target/embedded/config.in:
12:10 < CIA-9> * moved not immediatly fixable for dietlibc packags into the uclibc
12:10 < CIA-9> section of the embedded target and added rocknet, mkinitrd and stone
12:18 < rxr> checking for growing stack pointer... configure: error: cannot run test program while cross compiling
12:18 < rxr> See `config.log' for more details.
12:18 < rxr> `- glib crap
12:29 < rxr> hm - libstdc++ does not compile with dietlibc out of the box
12:29 < rxr> however I think it is not far from compiling?
12:30 < jsaw> good morning
12:30 < rxr> anyone wanting to take a look? valentin ?
12:30 < rxr> mon jsaw
12:30 < rxr> moin even
12:30 < rxr> happy new year!
12:30 < jsaw> :)
12:30 < jsaw> too u 2
12:31 * rxr about testing a minimal as possible dietlibc static system booted from a usb stick
12:31 < jsaw> Chaarlotte on one arm, ..., I'm not used to type with one hand
12:31 < rxr> heh ,-) greetings to her 2
12:31 < jsaw> thx :)
12:33 < jsaw> arts tries to link to /usr/lib64/../lib/libstdc++.so and complains about file format... :/
12:34 < jsaw> on 2nd stage filesystem: etc/stone.d/default.sh is missing
12:34 < rxr> jsaw: hm - guess this is due to mnemoc adaptes ..
12:34 < rxr> package split
12:35 < rxr> jsaw: this ../lib is a new regressions
12:35 < rxr> not yet know what has caused it
12:35 < rxr> maybe libtool update or so
12:35 < rxr> it worked about a week ago
12:36 < jsaw> donÄ
12:36 < jsaw> don't know either...
12:36 < jsaw> investigating...
12:36 < CIA-9> rene * r14379 /trunk/package/base/dietlibc/a.out.h.patch: * added dietlibc/a.out.h - needed for util-linux/more, huh?
12:37 < jsaw> oh, libtool-quirk does not copy ltmain.sh
12:37 < rxr> == 01/01/06 11:36:39 =[1]=> Finished building package gcc.
12:37 < rxr> `- without c++/libstdc++ it builds fine
12:37 < rxr> dietlibc that is
12:37 < jsaw> ah, no, libtool.m4 versus ltmain.sh version mismatch in pcre configure...
12:38 < jsaw> I thought, libstdc++ relies on too many internals of glibc...
12:38 < rxr> btw. I did some research - decent flash memory is NAND nor NOR, and does survive more that 1.000 writes
12:38 < rxr> IIRC 100.000 up to 1.000.000 ...
12:38 < jsaw> rxr: where's the code that does the LIBTOOL-QUIRK?
12:38 < rxr> I think wikipedia had a quite nice flash summary ...
12:39 < rxr> jsaw: base/libtool/parse-config
12:40 < rxr> jsaw: do you mark arts LIBTOOL-QUIRK ?
12:43 < jsaw> no, pcre
12:43 < jsaw> but that does not work
12:43 < jsaw> configure script complains about version mismatch,...
12:44 < rxr> I think there is something more fundamental broken
12:44 < jsaw> "Gentoo sanity check failed!"
12:44 < jsaw> *g*
12:44 < rxr> it used to build just fine
12:44 < jsaw> you mean, x86-64 related ?
12:44 < jsaw> brb
12:45 < rxr> I mean some version change of s.th. else or rearrangement of some FS layout or whatever structure
13:04 -!- mtr_ [n=Michael@p54AFAA40.dip0.t-ipconnect.de] has joined #t2
13:04 < rxr> moin mtr_ and a happy 2006
13:04 < CIA-9> aldas * r14380 /trunk/package/graphic/asymptote/asymptote.desc: * updated asymptote (0.97 -> 0.99)
13:06 < CIA-9> aldas * r14381 /trunk/package/emulators/hercules/hercules.desc: * updated hercules (3.03 -> 3.03.1)
13:08 -!- sparc-kly [n=mubex@64.237.255.10] has quit [Read error: 104 (Connection reset by peer)]
13:08 -!- sparc-kly [n=mubex@64.237.255.10] has joined #t2
13:10 < jsaw> re
13:12 -!- mtr [n=Michael@p54AF8C67.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
13:18 < rxr> cool - after months of rotting embeded stil comes up with the login ...
13:18 < rxr> I feared more breakage ...
13:18 < jsaw> that`s always nice :)
13:18 < jsaw> bbl, cu
13:21 < CIA-9> rene * r14382 /trunk/package/base/minit/inits/mountall/run:
13:21 < CIA-9> * fixed minit/mountall to mount /sys and to create /dev/{pts,shm} mount points
13:21 < CIA-9> before mounting
13:29 < rxr> re
13:29 < rxr> hi from the Geode running from flash with dropbear to connect to the world ...
13:32 < CIA-9> jsaw * r14383 /trunk/package/base/pcre/pcre.conf:
13:32 < CIA-9> * libtool problems, LIBTOOL-QUIRK does not work,
13:32 < CIA-9> but passing "LIBTOOL=/usr/share/libtool" to make
13:32 < CIA-9> does it
13:41 < CIA-9> jsaw * r14384 /trunk/package/xorg/xorg-server/modular-x-conf.in: * fix fontconfig ingore pattern (must not start with "/")
13:41 < mipe> hmm, then i guess same patch as r14383 can be used to others too...
13:42 < valentin> jsaw: what is the current content of misc/lua/sde about ?
13:44 < jsaw> valentin: I try to get all parts for an installer in there
13:44 < jsaw> valentin: right now, I've started "lxio" (eXtensible IO for lua...)
13:44 < jsaw> so that we can hook any compression utility into it
13:45 < jsaw> then I'll continue with pkgdb.lua
13:45 < jsaw> I mean, with adapting it to lxio
13:46 < jsaw> luatar will also use this
13:46 < jsaw> once these components are done, an installer is not far away
13:46 < jsaw> brb
13:47 < jsaw> oh forgot:
13:47 < jsaw> mipe: could be...
13:48 < jsaw> but now, bias
13:53 < rxr> jsaw: mipe: the problem with r14383 is, it is most probably hiding some fundamental problem
13:54 < rxr> as I wrote all those packages built fine last week
13:54 < rxr> we really should review whats up instead of adding quirks to 100 packages
13:54 < rxr> and my build encourages the number in the region of 100 is correct
13:54 < rxr> valentin: at my localnet .3 a geode run on flash is up with dropbear ssh server
13:55 < rxr> all dietlibc
13:55 < mipe> rxr: yep, theres "lots" of stuff beeing done in gcc libstdc++, according to diff against snapshot...
13:56 < mipe> and tryed with libtool cvs 1-5 branch and karasz tryed with .20 without fixing the problem, so atleast what i know only thing left is to check gcc
13:57 < rxr> mipe: one could build at svn up -r $data-of-last week
14:00 < mtr_> moin #T2, and a happy 2006 to you all
14:00 -!- mtr_ is now known as mtr
14:04 < jsaw> re
14:04 < jsaw> no, the problem is in libtool itself
14:05 < jsaw> if libtool wants "--tag", it cannot recognize the compile command
14:05 < jsaw> which most of the time happens, because it does not contain soupport for other compilers than C
14:06 < rxr> hm - I saw such an issue
14:06 < jsaw> the fact that libtool-quirk does not help here, is because the configure script internally checks the version information
14:06 < jsaw> so, alternative, a full auto* run would solve it
14:07 < jsaw> but passing LIBTOOL... to make is a cheaper solution
14:07 < jsaw> I remember that I fought with libtool half a year back to get to install it with C++ support...
14:08 < jsaw> rxr: when you say, it worked last week, what arch did you try it?
14:08 < rxr> x86-64 of course as well
14:09 < jsaw> with as-needed?
14:09 < rxr> nope
14:09 < jsaw> hmmm
14:12 < jsaw> oh oh, libstdc++.la
14:12 < jsaw> dependency_libs="... -lstdc++ ..." etal...
14:14 < jsaw> btw, /usr/lib contains 64bit stuff, or only 32?
14:14 < mipe> well should be just 32bit
14:14 < rxr> ack
14:14 < rxr> though some toy packages are not yet corrected to honor LIBDIR
14:15 < jsaw> like libstdc++ ...
14:15 < rxr> e.g. have hardcoded pathes, I have not yet bothered to fix them all
14:15 < rxr> the major stuff should be doing it right
14:15 < rxr> no
14:15 < rxr> libstdc++ is installed right
14:15 < jsaw> the 64bit libstdc++ is linked into /usr/lib
14:15 < rxr> jsaw: that should not be the case
14:15 < rxr> however if you cross build gcc Makefiles sometimes get those links wrong
14:16 < rxr> we already have a hack in the gcc.conf for libgcc_s* IIRC
14:16 < jsaw> no cross, native build!
14:18 < valentin> rxr: no route to host
14:20 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
14:21 -!- sparc-kly_ [n=mubex@64.237.255.10] has joined #t2
14:21 -!- sparc-kly [n=mubex@64.237.255.10] has quit [Read error: 104 (Connection reset by peer)]
14:23 < jsaw> rxr: wrong alarm, I was blind....
14:23 < valentin> rxr: not it works
14:23 < valentin> now
14:23 * jsaw is searching his glasses
14:25 < valentin> rxr: that machine has 750 MB ram ??
14:25 < valentin> oh - sorry confused
14:25 -!- nzg [n=tschmidt@p54AFD4D0.dip.t-dialin.net] has joined #t2
14:25 < valentin> i am on the rs6k still
14:26 * jsaw is glad that he's not the only one being confused
14:26 < rxr> valentin: shut it down right now
14:26 < rxr> wait 3 min or so
14:26 < valentin> ok
14:26 < rxr> have to copy some more file to the FS image
14:26 < valentin> it logged me off without me noticing it - and i wondered about the ibm processor :)
14:27 < valentin> back in some minutes...
14:38 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
14:39 < rxr> so
14:40 < rxr> geode booting up again
14:40 < rxr> hopefully able to compile binaries on its own
14:48 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
14:48 < jsaw> hmmm
14:48 < jsaw> rxr: I think you're right, there`s sth. fundamentally wrong...
14:48 < jsaw> rxr: taglib has the same problem as pcre does
14:49 < jsaw> adding LIBTOOL... to makeopt solves it...
14:49 < jsaw> .oO(?)
14:49 < rxr> haha
14:49 < rxr> gnu grep configure wants a grep and egrep already
14:49 < rxr> how I like auto*
14:50 < jsaw> o_O
14:50 < rxr> jsaw: yes - we have a problem
14:50 < rxr> oh - no make either
14:50 * rxr extending this embedded build ...
14:51 < jsaw> otoh, a minimal build (50-minimal-desktop) only had 12/519 errors
14:51 < jsaw> for x86-64 with as-needed
14:52 < rxr> it should not be as-needed
14:52 < rxr> I do not use it and get the same libtool trouble AFAIR
14:52 < rxr> valentin: btw. flashed geode up again
14:53 < jsaw> seems like all errors are libtool related, trying to link "/usr/lib64/../lib/libstdc++.so"
14:54 < jsaw> first pcre, then taglib, now libusb
14:54 < rxr> hm
14:55 < jsaw> rxr: do you think it would hurt to pass "LIBTOOL=..." globally via makeopt ???
14:55 < rxr> yes
14:55 < rxr> because we are hiding something
14:55 < jsaw> why?
14:55 < jsaw> why hide?
14:55 < rxr> something broke majorly in the last week or so and I want to know what it was
14:55 < rxr> jsaw: we did not had this errors in the past
14:55 < rxr> I built x86-64 like crazzy this all built fine
14:56 < jsaw> yes, we had, that
14:56 < jsaw> 's why we added the LIBTOOL-QUIRK
14:56 < rxr> that is only for 5-7 packages that have outdated libtool files included that need replaceing
14:56 < rxr> we had that in some .conf files and I wanted a unified place
14:56 < valentin> rxr: not bad...
14:56 < rxr> that is not the 100 pkg breakage
14:57 < jsaw> rxr: we have a lot more libtool hacks. libusb also has it.
14:57 < jsaw> rxr: sometimes the libtool quirk will not help, because autocrap is "intelligent" enough to recognize, that libtool has to be regenerated...
14:58 < jsaw> so copying libtool files over it might not help
14:58 < rxr> the LIBTOOL-HACK is for outdated libtool files where replacing them helps
14:58 < rxr> that are less then 10 packages right now
15:01 < jsaw> that
15:01 < jsaw> that's not true
15:01 < jsaw> only if you count the "LIBTOOL-QUIRK" flag
15:01 < rxr> jsaw: that was last week when all those just bult
15:02 < rxr> as I can only repeat - somthing broke seriously in the last 7 days and I'll not let it fix by adding LIBTOOL-QUIRK to 99 more packages
15:02 < rxr> jsaw: you can add some LIBTOOL in your build globally and I hunt the real issue here if you prever that
15:03 < rxr> but please do not commit a global LIBTOOL= or related quirk to hide the breakage we have
15:03 < jsaw> so, go for it
15:03 < jsaw> I think you're wrong...
15:04 < rxr> jsaw: I built a lot x86-64 builds and those enourmous bulid errors are new
15:04 < rxr> e.g. I pcre never used to fail and kde* built just fine as well
15:04 < valentin> one single gcc process for kdepim takes more ram than i have on my sisters old machine :/
15:05 < rxr> swap ?
15:05 < valentin> no wonder that box keeps swaping like hell
15:05 < rxr> ah
15:05 < valentin> yeah - it is kind of ... unresponsive
15:05 < rxr> with --enable-final (building all files in a sub-dir included in one source) some kde packages eat lots of ram
15:06 < valentin> brb
15:07 < CIA-9> jsaw * r14385 /trunk/package/base/pcre/pcre.conf: * reverse r14383 until real cause is found
15:10 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
15:10 < rxr> jsaw: in-system Build-Pkg of pcre works just fine on my turion x86-64 sub-notebook
15:12 < jsaw> ok, but how is it otherwise?
15:12 < jsaw> in-system was never a really reliable test
15:12 < rxr> jsaw: that jsut confirmed the system I have on my sub-notebook build ca. one week ago is fine
15:12 < rxr> jsaw: huh?
15:12 < rxr> sure it is
15:12 < rxr> I have a plain t2 minimal on the box
15:13 < rxr> Build-Target with a toolchain built by HEAd does fail for pcre - in system works
15:13 < jsaw> rxr: see for example r14384
15:13 < rxr> thus something in the toolchain or files used by pcre (e.g. libtool whatever) got regressions in the last days
15:14 < jsaw> r14384 is a typical example of the difference with in-system builds
15:14 < jsaw> in-system builds are NOT reliable
15:14 < jsaw> my family bugs me to leave the computer...
15:15 < rxr> was that change to make in-system or build target work?
15:15 < rxr> i tested the pattern with Build-Target
15:15 < rxr> have a nice family day ,-)
15:17 < jsaw> hehe
15:17 < jsaw> I
15:17 < jsaw> damn new keyboard...
15:17 < jsaw> I'm trying to build a new minimal-desktop target
15:17 < jsaw> so, cu later
15:18 < rxr> cu - have a nice 1st day
15:18 < mnemoc> moin!
15:18 < rxr> moin moin mnemoc
15:18 < rxr> happy new year to you!
15:18 < mnemoc> happy 2006! :p
15:19 < mnemoc> (sorry... it was impossible to me to connect during night :'( )
15:19 < mnemoc> happy new year rxr
15:20 < mnemoc> valentin: happy new year!
15:22 < rxr> mnemoc: did you have some nice celebration ?
15:23 < mnemoc> yes, quite good. and you?
15:24 < rxr> some board game sitting together with some cooking ...
15:25 < rxr> what did you do ?
15:25 < mnemoc> i went to santiago as on xmas
15:26 < rxr> oh - the big family ?
15:26 < mnemoc> yep, but reduced
15:26 < mnemoc> about 80 tonight
15:27 < mnemoc> but go and return during night is exhausting
15:27 < rxr> I can imagine - how long was the bus trip? 1h?
15:27 < mnemoc> i rent a car
15:27 < mnemoc> 1:30 x 2
15:28 < mnemoc> getting bus tickets for this night is almost impossible
15:29 < rxr> my
15:30 < rxr> coreutils are also ugly coded
15:30 < rxr> we really need some diet / exact version of this stuff or one day I get a heart attack
15:32 < rxr> coreutils/sort does crash and burn wiredly when built with dietlibc ...
15:34 < mnemoc> o.o
15:41 < rxr> mnemoc: does busybox include an option to build the applets as single binaries ?
15:42 < mnemoc> nope
15:42 < rxr> wonderful
15:42 < rxr> I just need sort ...
15:42 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
15:42 < mnemoc> you can get symlinks, or 'wrapped'. but always a single binary
15:42 < mnemoc> happy 2006 misl
15:43 < rxr> and why i do not like busybox
15:45 -!- sparc-kly_ [n=mubex@64.237.255.10] has quit [Read error: 104 (Connection reset by peer)]
15:47 * mnemoc searching
15:48 -!- sparc-kly [n=mubex@64.237.255.10] has joined #t2
15:48 < mnemoc> rxr: http://heirloom.sourceforge.net/tools.html
15:53 * rxr disabing one thing after the other in busybox
15:54 < rxr> half of an OS like emacs ... :-(
15:56 < mnemoc> i guess you can change busybox's Makefile to produce single executables
15:57 < rxr> mnemoc: the sort yout pointed to comiles with a single gcc call
15:58 < rxr> I try that
16:01 < rxr> hm - it wants wchar stuff dietlibc does not ship
16:02 < mnemoc> :(
16:05 < rxr> I wonder if busbox includes a X and text to speach ...
16:06 < mnemoc> :)
16:06 < rxr> my name is busbox ...
16:06 < rxr> and if you do not keep an I on me I'll run away screaming ..
16:07 < rxr> an eye on me even ..
16:10 < rxr> /media/usr/src/t2-trunk/busybox-1.01/libbb/printf.c:123:3: error: #error dietlibc is currently not supported. Please see the commented source.
16:10 < rxr> /media/usr/src/t2-trunk/busybox-1.01/libbb/printf.c: In function 'bb_vfprintf':
16:10 < rxr> /media/usr/src/t2-trunk/busybox-1.01/libbb/printf.c:140: warning: implicit declaration of function 'SET_FERROR_UNLOCKED'
16:10 < rxr> make: *** [/media/usr/src/t2-trunk/busybox-1.
16:10 < rxr> that is one big ugly damn hell of a mess
16:10 < rxr> I stop this - I rather hack the wchar stuff out of the other sort ...
16:10 < rxr> my oh my
16:11 < valentin> re
16:11 < mnemoc> :)
16:11 < valentin> happy new year mnemoc
16:11 < mnemoc> happy new year valentin
16:12 < LMJ> happy new years mnemoc, valentin and rxr and all others users ;)
16:13 < mnemoc> happy new year LMJ
16:14 < valentin> hny LMJ
16:14 < LMJ> WOWOWOw
16:17 < Stelz> gigi
16:17 < Stelz> happy new year, mnemoc
16:17 < Stelz> :)
16:17 < mnemoc> hny Stelz :)
16:17 < valentin> hi Stelz
16:17 < Stelz> hello, valentin
16:18 < valentin> we need a hello-bot with hny feature
16:18 < Stelz> :)))
16:18 * Stelz pokes CIA-9
16:27 < rxr> 24881
16:27 < rxr> bytes - my new sort ...
16:27 < mnemoc> caldera based?
16:27 < rxr> yep
16:28 < mnemoc> how fat is coreutils'?
16:28 < rxr> coreutisl is crashing ...
16:28 < rxr> and I'm out of time to track this major corruption
16:28 < rxr> 46973 it is
16:29 -!- Stelz [n=stealth@80.240.209.21] has quit ["There is intelligent life out there..."]
16:29 < mnemoc> :)
16:32 < CIA-9> amery * r14386 /trunk/misc/lua/ (8 files in 3 dirs): * fixed and updated copyright notes on misc/lua
16:32 < rxr> mnemoc: btw
16:32 < rxr> mnemoc: let us not regen all copyrights with the new year
16:32 < rxr> only on the files we touch, ok ?
16:32 < mnemoc> :(
16:32 < mnemoc> svn revert -R .
16:32 < rxr> ouhm
16:33 < rxr> why would you inject the new year just for fun?
16:33 < rxr> also would you do that on 2.1 ?
16:33 < mnemoc> i wanted
16:34 < karasz> moin
16:34 < mnemoc> moin karasz
16:35 < rxr> mnemoc: other projects also do not do that
16:35 < rxr> mnemoc: when the file is updated the copright notice also covers the year of change
16:36 < rxr> we should mimize noise also
16:36 < rxr> especially seeing trunk worse and worse each day
16:36 < mnemoc> trunk worse and worse each day?
16:36 < rxr> e.g. I guess you read scrollback the libtool stuff ?
16:36 < mnemoc> revert it's update
16:36 < rxr> with copyright updates the diffs to review would include even more changes that make reviewing harder
16:37 < mnemoc> i commited misc/lua separated because dual licensing, but i wanted to commit all the rest as a single commit
16:37 < rxr> well just lua is okay
16:38 < rxr> but do we really have any advantage of regen them all ?
16:38 < mnemoc> but ok, i can revert my local changes
16:38 < mnemoc> rxr: just to keep legal crap consistent
16:38 < rxr> it is still legal crap listing the last change year
16:39 < mnemoc> some files still not copyrighted for 2005
16:39 < rxr> well, of course not updating the files keep the burden to remeber updating it
16:40 < mnemoc> i had to generate first for 2005, and then update to 2006
16:40 < rxr> archive/Commit could do ...
16:40 < mnemoc> auto copyright patching?
16:40 < mnemoc> uhm
16:40 < rxr> but well, if you would like regen then do so
16:40 < rxr> just wanted to note that it might be a bit too much noise
16:41 < mnemoc> i'll fix non-(C)2005 files, and wait for rc1 to update to 2006
16:42 < karasz> hmm mnemoc i waited till 5 o'clock to greet you...
16:42 < karasz> anyway
16:42 < karasz> i see thta there is something moving with libs
16:45 < mnemoc> happy new year karasz! :p
16:45 < karasz> same to u mnemoc :P
16:47 < rxr> mnemoc: hm
16:47 < rxr> your bz2filename does use a gnu extension of sed
16:48 < rxr> thus it fails when the system sed is minised
16:48 < rxr> `\?'
16:48 < rxr> As `*', but only matches zero or one. It is a GNU extension.
16:48 < mnemoc> {0,1} ?
16:49 -!- sparc-kly [n=mubex@64.237.255.10] has quit [Remote closed the connection]
16:49 < mipe> uhm, someone selling pixels from hes homepage --> www.milliondollarhomepage.com
16:49 < mipe> and has sold pixels enough to get $999k in cash
16:50 < mipe> hmm... will register www.onehunredbilliondollarhomepage.com and see if I get lucky
16:52 < karasz> hello rxr, mipe
16:52 < rxr> mnemoc: hm I think minised also does not supprt \|
16:52 < rxr> hm hm
16:52 < karasz> i see that pcre is a major issue...., i mean not th epackage but the error on which fails
16:52 * rxr building gnu sed ...
16:53 < rxr> lets leave the function as it is
16:53 < rxr> I improve minised some day
16:53 -!- sparc-kly [n=mubex@64.237.255.10] has joined #t2
16:54 < mnemoc> nah, let's find an standard expression
16:55 < mipe> hi karasz
16:55 < mnemoc> .oO( did i greet mipe? )o
16:56 < mipe> hi mnemoc :)
16:56 < mipe> i thought i greeted you... sorry :)
16:56 < mnemoc> happy new year mipe! ,-)
16:57 < mipe> thanks, you too and others too
16:57 < karasz> hmm hmm hmm i see that i stired up some waters with my fonts issue...
16:57 < karasz> Date: 2005-12-31 17:02:07 +0100 (Sat, 31 Dec 2005)
16:57 < karasz> New Revision: 14357
16:58 < karasz> Date: 2005-12-31 18:00:34 +0100 (Sat, 31 Dec 2005)
16:58 < karasz> New Revision: 14362
16:58 < karasz> Log:
16:58 < karasz> * reverted r14357
16:58 < rxr> sed: -e expression #1, char 8: Memory exhausted
16:58 < rxr> ouhm
16:59 < mnemoc> :(
17:01 < rxr> gnu sed with dietlibc
17:01 < mnemoc> oh
17:01 < rxr> mnemoc: let's sell add space over the t2 package matrix with this strategy ...
17:05 < mnemoc> rxr: i think we should try google adds
17:07 < rxr> yep
17:07 < rxr> do you know how they pay? view or click ?
17:07 < mnemoc> no idea
17:09 < zod> hello!
17:10 < mnemoc> hi zod :)
17:10 < karasz> hi zod
17:10 < karasz> hny 2u zod
17:10 < zod> happy new year 2u2 ;)
17:12 -!- morfoh [n=jeru@p54BEDE76.dip.t-dialin.net] has joined #t2
17:12 < karasz> hny senor morfoh
17:12 < morfoh> moin moin and happy new year again :)
17:12 < mnemoc> happy new year morfoh-chan
17:13 < karasz> mnemoc, morfoh-chan?
17:13 * karasz .oO
17:13 < morfoh> mnemoc: happy new year master mnemoc :)
17:13 < mnemoc> karasz: :)
17:13 < mnemoc> morfoh: :)
17:13 < karasz> errr...
17:14 < morfoh> karasz: I don't know what that "chan" means ...
17:14 < mnemoc> karasz: morfoh 'boy' on japanese
17:14 < mnemoc> little morfoh too
17:14 < karasz> mnemoc, are you sure :) ?
17:14 < mnemoc> nope
17:15 < morfoh> mnemoc: hmmmm ...
17:15 < mnemoc> let see http://en.wikipedia.org/wiki/Japanese_titles
17:16 < mnemoc> Chan is the informal, intimate, diminutive equivalent of san, used to refer to children and female family members, close friends and lovers. Chan is also used for adults as a title of affection.
17:16 < mnemoc> *G*
17:16 < morfoh> mnemoc: pfff :p
17:16 < karasz> thatt's why i was intrigued
17:16 < mnemoc> but i was close ,-)
17:16 < karasz> i thought that it applies only to females...
17:17 < mnemoc> it seems not
17:17 < karasz> yup
17:17 < karasz> it seems so
17:17 < morfoh> indeed ...
17:17 -!- karasz is now known as K
17:17 < morfoh> it seems so :p
17:18 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
17:18 -!- K is now known as Kohai
17:18 < mnemoc> "used to refer to children"
17:20 < rxr> "configure" is too big (>500000 bytes) [More...]
17:20 < mnemoc> but if you prefer, morfoh-kun
17:20 < rxr> `- elvis as vi ...
17:20 < Kohai> hai mnemoc sama ;)
17:21 * Kohai or should i say sensei mnemoc ??
17:21 < mnemoc> o.o
17:21 < mnemoc> san is ok ,-)
17:21 < morfoh> mnemoc: pffff ... ;)
17:21 * Kohai is searchin' the sensei's googles
17:24 * Kohai googles not found... me returnin' to polish it's sword...
17:24 < morfoh> mnemoc san, btw who won in the chilean election ? the "social-democratic" (neoliberal) women ?
17:25 -!- Kohai is now known as karasz
17:26 < mnemoc> morfoh: draw
17:26 < mnemoc> morfoh: jan 15 is the second round
17:26 < mnemoc> morfoh: and she is not neoliberal, her party is but she is anti-globalization
17:26 < morfoh> mnemoc: draw between ?
17:26 < morfoh> who ?
17:27 < mnemoc> mine
17:27 < morfoh> I feared so
17:28 < rxr> oh my god
17:28 < mnemoc> ?
17:28 < rxr> vi /usr/include/argp.h
17:28 < rxr> what crap does glibc ship ?
17:28 < rxr> rng-tools demands it ...
17:29 < rxr> this is definetly not my day
17:29 < rxr> should have done something else on the 1st
17:29 < mnemoc> rxr: what about injecting argp.h inside rng-tools?
17:30 < valentin> so - today is dependency eve
17:30 < rxr> rngtest.o: In function `parse_opt':
17:30 < rxr> rngtest.c:(.text+0x9a): undefined reference to `argp_usage'
17:30 < rxr> rngtest.o: In function `main':
17:30 < rxr> rngtest.c:(.text+0x3c4): undefined reference to `argp_parse'
17:30 < rxr> collect2: ld returned 1 exit status
17:30 < mnemoc> :)
17:30 * mnemoc waiting for 2.1.1 desktop isos from rene-sama
17:31 < valentin> have some question anyway - once we drop priority, if someone adds a package, he will have to specify the deps (cache file)
17:31 < mnemoc> oh,
17:31 < rxr> mnemoc: more than macro hackery ..
17:31 < rxr> valentin: of course
17:31 < rxr> valentin: new packages must be commited with build cache
17:31 < valentin> but what about deep insurgery operations like sorting out sysfiles into varios packages ?
17:31 < rxr> valentin: either tweak manually or do a reference build to get the new ones
17:32 < rxr> e.g. I also tweaked some manually during xorg (-modular)
17:32 < valentin> normal packages can be compiled into an existing build / existing system without problem, for such d.i.o.s we need a way the developer can specify build order
17:32 < rxr> valentin: don't worry - as long as you write the graph generaton and evaluation ,-)
17:33 < rxr> valentin: what do you mean ?
17:36 < mnemoc> valentin: developer can emerge, and get a .cache, or build inside an already grown sandbox and get a .cache
17:37 < mnemoc> valentin: or use [E] on .desc
17:39 < valentin> mnemoc: but emerge will only give you the deps for the package _itself_
17:39 < valentin> not the rest of the deps
17:39 < valentin> take a look at the mkinitrd issue we had on friday
17:40 < valentin> all packages depend on sysfiles. now you replace sysfiles with X,Y,Z and you need to tell the sde to build X,Y,Z before the other packages
17:41 < mnemoc> poink.desc:[E] foo, emerge will build foo and fail, so poink will not be build and person will 'know' to emerge foo
17:41 < valentin> there is no chance for the SDE to know that, except by telling it
17:41 < mnemoc> an stabilized tree wont have such splits
17:41 < rxr> guess I patch our rng-tools to include the needed lib functions ...
17:41 < valentin> mnemoc: yes, all i say is
17:42 < valentin> there has to be a way to tell the sde dep analyzer of such a thing
17:42 < mnemoc> ack
17:42 < valentin> like "manual priority"
17:42 < mnemoc> but we don't need more than [E] at. desc
17:42 < valentin> which you do _once_ after an d.i.o. until you did a new full reference build and have the cache
17:42 < rxr> http://www.lysator.liu.se/~nisse/misc/argp-standalone-1.4-test.tar.gz
17:42 < rxr> just note to self
17:42 < rxr> dinner
17:42 < rxr> cu
17:43 < valentin> cu
17:43 < mnemoc> after a package split, the harm-maker will have to loop over .cache files injecting or renaming the dep
17:43 < valentin> mnemoc: you want to do this manualy ?
17:43 < mnemoc> valentin: we could make a helper
17:43 < mnemoc> valentin: but don't bloat dep analizer
17:44 < valentin> all the thing the harm-maker has to do is to specify a build position for that one new package
17:44 < mnemoc> on a dinamic priorization context, [E] is like specifing a build position, but using a literal >= instead of a numeric ==
17:45 < mnemoc> which can be removed after .cache gets in
17:45 < morfoh> mnemoc: sorry, but what about "optional" deps ?
17:46 < valentin> morfoh: will not affect the build order
17:46 < mnemoc> {buildtime,runtime} {optional,mandatory}
17:46 < valentin> mnemoc: we do not need >= but <=
17:46 < mnemoc> valentin: yes, sorry
17:46 < mnemoc> morfoh: optional deps affect Emerge package list, not build order
17:46 < valentin> mnemoc: or can you think of some scenario where you would like to specify >= manually ? maybe when adding two new packages at once ?
17:47 < mnemoc> valentin: nah
17:48 < mnemoc> valentin: [E] can be set to solve any issue, beside staging
17:49 < valentin> ok, for now i just do the case "all .cache files exist and are consistent and true, because they are given by an oracle" and think about the special cases later
17:49 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
17:49 < mnemoc> yes, dep analizer has to _trust_ staging decitions and deps list
17:51 < valentin> mnemoc: as long as it is an external tool, that is ok. but i am thinking on to the future when static priority dies and it has to be integrated into the development process
17:51 < mnemoc> :)
17:52 < valentin> as i said, when adding a new package without .dep file, that package is build latest and we have the correct deps. but only works for non-insurgery stuff
17:52 < mnemoc> ack
17:52 < mnemoc> decitions related to deep changes have to be taken by a human, with a helper on the hand
17:59 < valentin> priority sorting for stage 0 (aka cross toolchain): not needed (?)
17:59 < valentin> priority sorting for stage x > 0: Let P be the set of all packages build in stage x for the current target. We split up this set into P_first, the set of packages first built in x and P_prior, the packages that have been build in a prior stage. Then, for each package in P, remove all dependencys to packages not in P_first from the list. Do a topological sorting of P with the reduced deps. In case of clean deps, there should be no loops wh
18:00 < valentin> what do you think about that ?
18:01 < mnemoc> i think we will have to split stage 0 to get this working.... but yes
18:02 < valentin> btw, with the current [P] flag, is it possible to specify different prioritys for different stages ?
18:03 < mnemoc> nope
18:04 < valentin> hm :( because i thought of let running the scheme described above and dump the result into the P flags and try a test build with that
18:05 < valentin> well, however. should be possible to combine the results
18:06 < mnemoc> we would need to use an auxiliary config/$config/package.$stage
18:06 < mnemoc> $stagelevel
18:06 < valentin> mnemoc: we have this feature ?
18:07 < mnemoc> nope, but not hard to implement
18:07 < rxr> re
18:07 < valentin> mnemoc: i wanted to have stage dependend dep lists anyway. our current deps are stage 9
18:07 < valentin> hi rxr
18:07 < mnemoc> valentin: deps are good _except_ python and screesaver
18:08 < valentin> how'd ya know ?
18:08 < mnemoc> python-programs, not python package
18:08 < mnemoc> lots of reviewing of all .cache files :)
18:08 < mnemoc> we don't understand how python work
18:09 < mnemoc> so all python-dependent packages depend on all python-dependent pacakges
18:09 -!- jsaw [n=jsaw@volans.mpimf-heidelberg.mpg.de] has quit [Read error: 104 (Connection reset by peer)]
18:09 -!- jsaw [n=jsaw@volans.mpimf-heidelberg.mpg.de] has joined #t2
18:09 < mnemoc> and we will have the same problem with java when we get more java packages
18:10 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
18:12 -!- nzg [n=tschmidt@p54AFD4D0.dip.t-dialin.net] has quit ["Leaving"]
18:12 < valentin> we will have problems with the rare cases of optional feature support for an optional feature
18:13 -!- sparc-kly_ [n=mubex@64.237.241.16] has joined #t2
18:13 < valentin> i.e X and Y compiled in stage 1 and Z in stage 2 and then X recompiled with support for Z and Y recompiled with support for X@Z
18:13 < rxr> valentin: we already said dozen times that we want to later mark optional deps in the E tag
18:13 < rxr> where is the problem ?
18:14 < valentin> that you need to mark that Y has to be build after Z in stage 2. but this should be doable with E ?
18:15 < mnemoc> staging
18:16 < valentin> mnemoc: when merging all stage depenencys down to one superposed build order, the problem would be solved.
18:16 < valentin> even without E
18:17 < valentin> because in stage 2, X and Y may be independend, but since the dep is in the cache file, they are not in the first stage they are build
18:18 < valentin> rxr: which language, c++ or bash or lua ?
18:18 < rxr> do you think bash will do ?
18:19 < valentin> rxr, well bash = bash + hell of other tiny script tools
18:19 < mnemoc> :)
18:19 < valentin> like tsort etc...
18:20 < rxr> valentin: I think not yet lua if we want to use it in trunk some time soon
18:20 < rxr> thus bash or C++
18:20 < rxr> I would vote for C++ for the first impl.
18:20 < valentin> ok, i do bash now and migrate to the next higher level language we'll definitly use for t2
18:20 < mnemoc> we can't start on C++ if we are no clear on what to really do
18:20 < mnemoc> valentin: ack
18:21 < rxr> #include
18:21 < rxr> int daemon(int nochdir, int noclose);
18:21 < mnemoc> bash+friends is good for prototyping this imo
18:21 < valentin> rxr: i do bash - it is not very hard thing. sorting out P and P_first is not that hard task
18:21 < rxr> functions exist ... my oh myu ..
18:21 < valentin> the rest is a grep -f and a tsort thing :)
18:21 * mnemoc hate self-demonizing apps
18:22 < mnemoc> valentin: yep
18:22 < valentin> however merging down to priority numbers could be tricky. for now i just could check the [P]s we have against the sorting
18:22 < rxr> ah - in dietlibc -lcompat ...
18:23 < mnemoc> valentin: dump a sorted list, then grep -n '^' dump :p and awk to generate packages.$stagelevel ,-)
18:24 < mnemoc> or simply while read
18:25 < rxr> there was or is parallel build sorting code from rock in t2
18:25 < rxr> no idea if I already removed it
18:25 < rxr> maybe it is worth a look?
18:25 < rxr> might be over cryptic maybe
18:26 < valentin> rxr: no, i do not want to rely on the old cluster code
18:27 < rxr> valentin: ok
18:27 < mnemoc> tsort and grep are enough
18:27 < rxr> mnemoc: I remove this old cluster code crap from t2 trunk, ok ?
18:27 < mnemoc> sure
18:27 < rxr> ok - so today will end as happy day
18:27 < rxr> despite soo much ill code today ...
18:27 < mnemoc> 2.1.1 building on your ppc? *g*
18:27 < valentin> mnemoc: i will try to tsort a full list over all stages but with clever removal of edges
18:27 < rxr> should not have tried to convert his 24x7x... router to geode on flash ...
18:28 < mnemoc> valentin: i believe you will find lot of loops
18:28 < valentin> mnemoc: i _know_ i will
18:28 < valentin> but this will give me some feedback about troubles there are and then i may have an idea how to do better
18:29 < mnemoc> ack
18:29 < rxr> or what packages we should fix urgently
18:29 < zod> anyone wants to assist me building my first t2 build? ;)
18:29 < valentin> hi zod
18:29 < rxr> I think we should have a clean parser / sorter and fix the really troubling packages instead adding too much inteligent into the generator
18:29 < mnemoc> rxr: python deps need urgent atention
18:30 < mnemoc> bbl
18:30 < mnemoc> (1h)
18:30 < valentin> zod: is there any specific problem you came across or just wanting a general guidline ?
18:30 -!- sparc-kly [n=mubex@64.237.255.10] has quit [Read error: 110 (Connection timed out)]
18:30 < zod> well i tried once but i got some errors while building the kernel
18:31 < zod> but this was the embedded uclibc target
18:31 < karasz> zod for starters i think it is better to stick with general target
18:32 < karasz> it is more tested that embeded
18:32 < karasz> which mnemoc told you that wasn't touched in 8 month...
18:32 < zod> yeah i know
18:33 < karasz> if you use trunk, general with minimal or minimal + xorg will build fine
18:34 < karasz> guys gottago
18:34 < karasz> cu tommorow
18:34 < rxr> cu karasz and mnemoc
18:34 < valentin> cu karasz and mnemoc :)
18:35 < mipe> zod:mtr(or someone) did some uclibc changes last night(?), dont know thought if it fixes your problem
18:37 < rxr> valentin: I'll wait ordering geode until we next order ones anyway - that is not ordering a single unig
18:37 < rxr> t
18:37 < rxr> do you need it /urgently/ and want to have one of mine or can it wait some weeks ?
18:37 < mipe> rxr: you got the price tag for me yet?
18:37 < rxr> mipe: ah yes - called them
18:37 < rxr> mipe: the employee I had on phone told me there are no transmeta anymore
18:38 < rxr> no idea if one of them was irritated, since the former told me weeks ago they have still a lot of them or so ...
18:38 < valentin> rxr: well it is not toooooo urgent, but i would like to have a babyphone replacement :)
18:38 < mipe> :(
18:38 < rxr> mipe: the price we have to paz for the geode is 3xx EUR in quantities, single end user prise on their homepage ix 5xx EUR
18:39 < mipe> have to search other low power pc then, only need it as firewall+router and such
18:39 < valentin> mipe: with the geode you get a full featured multimedia station (without the media drives ;)
18:39 < mipe> rxr:hmm, what was the their homepage again
18:39 < rxr> mipe: if you need an AP the linksys one with linux is mod-able ...
18:40 < rxr> http://www.cool-shop.de or something as silliy as that name ,-)
18:43 < CIA-9> rene * r14387 /trunk/package/base/dietlibc/pkg_patch/pkg_cpio.patch: * removed dietlibc/pkg_cpio.patch, doesn't apply and compiles without
18:58 < CIA-9> aldas * r14388 /trunk/package/kde/kbarcode/ (compile-fix.patch kbarcode.desc):
18:58 < CIA-9> * updated kbarcode (1.9.8 -> 1.9.9)
18:58 < CIA-9> * removed compile-fix.patch - obsolete
19:17 -!- rxr_ [n=rene@e178140090.adsl.alicedsl.de] has joined #t2
19:17 -!- Topic for #t2: T2 wish you all a very happy 2006! | 2.1.1 TAGGED and 2.2.0-epsilon RELEASED! | The System Development Environment (SDE) | http://www.t2-project.org/ | Give us a greeting and don't hesitate to ask any questions you have :-)
19:17 -!- Topic set by mnemoc [] [Sat Dec 31 18:55:11 2005]
19:17 [Users #t2]
19:17 [@ChanServ] [ CIA-9 ] [ LMJ ] [ mtr ] [ sparc-kly_]
19:17 [ _Ragnar_] [ demian] [ mipe ] [ owl ] [ valentin ]
19:17 [ axion ] [ jsaw ] [ mnemoc] [ rxr ] [ zod ]
19:17 [ Baldzius] [ karasz] [ morfoh] [ rxr_]
19:17 -!- Irssi: #t2: Total of 19 nicks [1 ops, 0 halfops, 0 voices, 18 normal]
19:17 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
19:17 -!- [freenode-info] if you need to send private messages, please register: http://freenode.net/faq.shtml#privmsg
19:17 -!- Irssi: Join to #t2 was synced in 30 secs
19:18 < mnemoc> hny Baldzius
19:24 -!- rxr [n=rene@e178139187.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)]
19:24 -!- You're now known as rxr
19:40 < CIA-9> jsaw * r14391 /trunk/package/base/vlock/link-fix.patch: * -l... should appear at the end of argument list
19:41 < jsaw> re
19:41 < mipe> http://koti.japo.fi/~pmika/libstdc++.la
19:41 < mipe> thats the installed file from t2 minimal trunk iso
19:41 < mipe> some wierd values there... -L/BACKUP/
19:42 < mipe> same -L/BACKUP values are generated current trunk, desktop build i'm running
19:43 < jsaw> mnemoc: you did split up stone, right? Was there something else?
19:43 < rxr> rocknet mkinitrd
19:43 < rxr> those are most problably not causing this
19:44 < jsaw> where are the stone.d files ?
19:44 < jsaw> should they be in the stone package ?
19:44 < rxr> the individial packages mostly
19:44 < jsaw> oh my
19:44 < jsaw> and which?
19:44 < rxr> I think
19:44 < rxr> not checked ...
19:45 < rxr> jsaw: well the basic modules in stone and the other stuff as it used to be in e.g. grub, openssh, ...
19:45 < jsaw> grep stone.d .../flists/*
19:45 < jsaw> -> nothing
19:45 < rxr> hehe
19:45 < rxr> that is a but then I would propose ..
19:45 < mipe> so could it be that these libstdc++ troubles (for example pcre) has been there longer than expected, since minimal doesnt have pcre for example
19:45 < rxr> mipe: I modified to build a cross compiler - maybe that was it
19:45 < rxr> or the tweaks for cross compiling due to libtool uglyness
19:45 < jsaw> mipe: what libstdc++ troubles?
19:46 < rxr> I modified that last a week ago or so
19:46 < rxr> maybe those are it
19:46 < rxr> but I just let the wrapper do some stuff behind libtools back though ...
19:46 < mipe> jsaw:http://koti.japo.fi/~pmika/ERROR-LOG
19:48 < CIA-9> rene * r14392 /trunk/package/network/gatling/gatling.conf:
19:48 < CIA-9> * fixed gatling for cross compilation (prefix to libowfat headers and
19:48 < CIA-9> doubled DESTDIR due falsely added to prefix)
19:49 < jsaw> mipe: libtool problem... yes... the new libtool does not have this problem...
19:49 < mnemoc> what version is good?
19:49 < mipe> none
19:49 < jsaw> ?
19:50 < mipe> tryed with .20 .22 .23a none works right atm
19:50 < jsaw> it compiles when passing LIBTOOL=/usr/bin/libtool to makeopt
19:50 < jsaw> but rxr objects this fix
19:50 < mnemoc> hack :)
19:50 < jsaw> (see discussion around 15:00)
19:51 < mipe> hmm, since its only libstdc++, is gcc libtool much differ than the real libtool stuff?
19:51 < mnemoc> # grep stone.d build/minimal-2.2.0-alpha-x86-pentium-mmx-generic/var/adm/flists/* | wc -l
19:51 < mnemoc> 14
19:51 < jsaw> nothing here, absolutely NOTHING
19:51 < mnemoc> jsaw: you may need to rebuild the tools to get stone/parse-config considered
19:51 < jsaw> (generic/minimal)
19:52 < mnemoc> jsaw: do you have stone package enabled?
19:52 < jsaw> it was not... :(
19:52 < mnemoc> :(
19:52 < CIA-9> jsaw * r14393 /trunk/target/generic/pkgsel/00-minimal.in: * add "stone" to the minimal pkgsel
19:53 < mnemoc> uhm
19:53 < mnemoc> i did that
19:53 < jsaw> ?
19:54 < jsaw> lemma see
19:54 < mnemoc> # svn blame 00-minimal.in | grep stone 14271 amery X stone
19:54 < jsaw> ??????
19:54 < jsaw> I had no stone
19:54 < jsaw> lemma check $confdir/packages
19:55 < mnemoc> confdir=package/$repo/$pkg :p
19:55 < CIA-9> jsaw * r14394 /trunk/target/generic/pkgsel/00-minimal.in: * revert r14393, stone was already there (?)
19:56 < mnemoc> *G*
19:56 < rxr> ...
19:56 < jsaw> I don't get it...
19:56 < jsaw> why did stone disappear if it should be there?
19:57 < mnemoc> i think Build-Target didn't notice the change and Config -oldconfig wasn't run
19:57 < jsaw> I always do oldconfig
19:57 < jsaw> It was a fresh build _after_ these changes
19:57 < mnemoc> o.o
19:57 < mnemoc> can you reproduce?
19:58 < jsaw> trying to...
19:58 < mnemoc> i tried hard to make the changes as harmless as posible
19:59 < jsaw> no question about that... :)
20:00 < mnemoc> as posible :)
20:00 < mnemoc> but the split was needed
--- Log opened Sun Jan 01 21:57:34 2006
21:57 < rxr> bye
21:58 < CIA-9> jsaw * r14399 /trunk/misc/isomd5sum/ (libcheckisomd5.c libimplantisomd5.c):
21:58 < CIA-9> * compiler complains about signedness of char/unsigned char usage
21:58 < CIA-9> -> cast to (char *) where appropriate
22:01 < jsaw> testing build, cu later
22:34 -!- sparc-kly_ [n=mubex@64.237.241.16] has quit [Read error: 104 (Connection reset by peer)]
22:37 -!- sparc-kly [n=mubex@64.237.241.16] has joined #t2
22:56 < rxr> re
22:56 < rxr> australian bar is closed today ...
22:56 < rxr> how boring
23:07 < rxr> mnemoc: do you have a clean cron to propose ?
23:14 -!- nzg [n=tschmidt@p54AFEB50.dip.t-dialin.net] has joined #t2
23:18 < CIA-9> aldas * r14400 /trunk/package/multimedia/xvid/xvid.desc: * updated xvid (1.0.3 -> 1.1.0)
23:30 < rxr> jsaw: the problem is due to invalid junk in the libstdc++.la files
23:31 < rxr> this mostly happens due to our xgcc-wrapper I assume - and/or the related hackery I have in gcc.conf due to that
23:31 < rxr> I can take a look tomorrow how we could tweak the gcc build to not leak those strange deps into the libstdc++.la file ...
23:34 < rxr> alternatively not instaling the file for now might be a solution for now - earlier gcc versions did not install such a file
23:42 < jsaw> re
23:43 < jsaw> installation did not work out
23:43 < jsaw> 1. problem rmdir missing in initramfs
23:43 < jsaw> 2. hotplug++: "No such file or directory"
23:44 < jsaw> 1. is solved, 2. ... investigating
23:46 < rxr> is there a diff for 1 ?
23:48 < rxr> jsaw: work around for your build - remove all the junk from the libstdc++.la files in the dependecy line ...
23:48 < mnemoc> re
23:48 < mnemoc> [19:09:40] mnemoc: do you have a clean cron to propose ? <-- what?
23:49 < rxr> mnemoc: some clean and tiny ..
23:49 < mnemoc> bcron
23:49 < rxr> mnemoc: do you use it ?
23:49 < mnemoc> yes
23:49 < rxr> ok - will try ASAP
23:49 < rxr> sleep and uni first :-(
23:49 < mnemoc> uni?
23:50 < mnemoc> i thought you had 'vacations'
23:50 < rxr> till today ...
23:50 < jsaw> rxr: my workaround is still using var_append makeopt " " "LIBTOOL=/usr/bin/libtool" (and I have removed the LIBTOOL-QUIRK locally)...
23:50 < rxr> tomorrow I should apply ..
23:50 < jsaw> rxr: diff follows
23:51 < rxr> jsaw: please do not commit this workaround, ok ?
23:52 < CIA-9> jsaw * r14401 /trunk/target/install/build_initrd.sh: * rmdir seems to be missing in initramfs -> copy from embdutils
23:52 < jsaw> rxr: I won
23:52 < jsaw> `t
23:52 < mnemoc> *G*
23:52 < mnemoc> jsaw: foodirlist change tested?
23:52 < jsaw> but I keep it, because I think it is correct and saves more troubles than it creates
23:53 < jsaw> mnemoc: haven't seen any troubles so far :)
23:53 < mnemoc> good :)
23:53 < jsaw> but that hotplug++ does not work bugs me
23:54 < rxr> does not work or is not in the initrd ?
23:54 < jsaw> does not work
23:54 < rxr> huh?
23:54 < jsaw> yep, huh
23:55 < rxr> hm - what does it do ?
23:57 < jsaw> nothing, the output is sth. like "/sbin/hotplug++: No such file or directory"
23:57 < jsaw> same if executed manually
23:58 < jsaw> s/if/when/ ?
23:59 < rxr> hm
23:59 < rxr> file ?
23:59 < rxr> exacutable bit ?
--- Log closed Mon Jan 02 00:00:00 2006