--- Log opened Sat Oct 29 00:00:30 2005
00:51 -!- sparc-kly [n=Mubex@64.237.253.126] has quit [Read error: 110 (Connection timed out)]
01:34 < karasz> bed time for me too
01:35 < karasz> gn8 all
01:35 -!- karasz is now known as karasz[Away]
02:44 -!- karasz[Away] [n=karasz@80.97.102.202] has quit [Read error: 104 (Connection reset by peer)]
04:02 -!- sparc-kly [n=Mubex@64.237.253.126] has joined #t2
06:39 -!- Baldzius_ [n=mx@85.206.103.43] has joined #t2
06:39 -!- Baldzius [n=mx@85.206.141.205] has quit [Read error: 110 (Connection timed out)]
08:34 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
08:36 -!- mtr_ [n=Michael@p54AF95EA.dip0.t-ipconnect.de] has joined #t2
08:45 -!- mtr [n=Michael@p54AF8F33.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
09:00 < misl> good morning to all. :)
09:09 < rxr> moin misl
09:09 < misl> moin rxr
09:09 < rxr> mnemoc: powerpc desktop and boot builds are in stage 5 now
09:11 < misl> rxr: Do you have some time in the near future to move noisy to a full 64 bits installation?
09:13 < rxr> misl: yes - however we should do this when you are around to select the old grub entry for the case of a to be fixed typo or so
09:13 < rxr> misl: do you want to come over the the linux world expo in frankfurt am main?
09:14 < misl> I'd love to but my current job does not allow it. There is too much presure on that project to skip a day.
10:08 -!- SerWou [n=serwou@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
10:10 < misl> I am off for some desperately needed family quality time :)
10:10 < misl> bbl
10:22 -!- rxr [n=rene@e178187068.adsl.alicedsl.de] has joined #t2
10:22 -!- Topic for #t2: T2 | 2.1.0 TAGGED! | The System Development Environment (SDE) | http://www.t2-project.org/
10:22 -!- Topic set by mnemoc [] [Wed Oct 26 05:11:03 2005]
10:22 [Users #t2]
10:22 [ _Lewellyn] [ CIA-15 ] [ misl ] [ owl ] [ SerWou ]
10:22 [ _Ragnar_ ] [ jsaw ] [ mnemoc] [ rxr ] [ sparc-kly]
10:22 [ Baldzius_] [ lewellyn] [ mtr_ ] [ sepp] [ valentin ]
10:22 -!- Irssi: #t2: Total of 15 nicks [0 ops, 0 halfops, 0 voices, 15 normal]
10:22 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
10:22 -!- [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup
10:22 -!- Irssi: Join to #t2 was synced in 6 secs
10:30 < SerWou> good morning
11:00 -!- sepp [n=sepp@p213.54.208.191.tisdip.tiscali.de] has quit [Read error: 104 (Connection reset by peer)]
11:02 -!- sepp [n=sepp@p213.54.208.191.tisdip.tiscali.de] has joined #t2
11:02 < sepp> moin
11:02 < rxr> moin sepp
11:03 < sepp> hi rxr
11:08 < rxr> 974 builds total, 927 completed fine, 47 with errors.
11:08 < rxr> Error logs from default-2.2.0-alpha-sparc64-v9vis-desktop:
11:11 < sepp> :)
11:11 < sepp> 658 builds total, 656 completed fine, 2 with errors
11:12 < rxr> yeah - I gues not sparc ,-)
11:12 < sepp> yes, boring x86
11:16 < rxr> but building ,-)
12:23 < CIA-15> rene * r12948 /trunk/package/ (3 files in 3 dirs): * migrated x11/xorg/startwm.patch to xorg/xinit and xorg/xdm
12:24 < CIA-15> rene * r12949 /trunk/package/xorg/ (twm/twm.conf xdm/xdm.conf): * added xdm and twm registration to the modular xorg/ packages
12:27 < CIA-15> rene * r12950 /trunk/package/ (18 files in 2 dirs): (log message trimmed)
12:27 < CIA-15> * migrated most monolythic xorg configuration and T2 integration to
12:27 < CIA-15> modular xorg/, including:
12:27 < CIA-15> * xcfgt2 and xcfgt2.init
12:27 < CIA-15> * fontconfig configuration (/etc/fonts/local.conf)
12:27 < CIA-15> * xvfb-run
12:27 < CIA-15> * postinstall to create font metrics and fontconfig cache
12:42 < rxr> -> Compiler Cache Hits while pkg build: 99.96% (2333 hits, 1 misses)
12:43 < rxr> ^- and why compiling with T2 is soo much fun ,-)
13:10 < CIA-15> rene * r12955 /trunk/package/x11/mesa/mesa.desc: * updated mesa (6.4-rc4 -> 6.4)
13:23 < misl> hi all
13:24 < misl> mnemoc: 2.1 minimal finished
13:24 < misl> minimal: 166 builds total, 166 completed fine, 0 with errors.
13:38 < rxr> hi misl
14:17 -!- karasz[Away] [n=karasz@80.97.102.202] has joined #t2
14:19 -!- karasz[Away] is now known as karasz
14:20 < karasz> moin all
14:23 < CIA-15> rene * r12956 /trunk/package/x11/mesa/ (mesa.conf mesa.desc):
14:23 < CIA-15> * fixed mesa's driver path to match X11R7 reality and made
14:23 < CIA-15> it build and install glxinfo and glxgears
14:25 -!- sepp_ [n=sepp@p213.54.218.51.tisdip.tiscali.de] has joined #t2
14:28 < CIA-15> rene * r12957 /trunk/package/x11/mesa/mesa.desc:
14:28 < CIA-15> * downdated mesa (6.4 -> 6.4-rc4), since 6.4 breaks the unichrome driver
14:28 < CIA-15> due to last minute changes :-(
14:28 < karasz> hmmm 2.6.14 today....hmmm wondering
14:29 < CIA-15> rene * r12958 /trunk/package/xorg/xorg-server/xorg-server.conf:
14:29 < CIA-15> * reindented xorg-server with s and fixed the installation of
14:29 < CIA-15> /etc/fonts/local.conf
14:30 < CIA-15> rene * r12959 /trunk/package/xorg/xorg-server/xorg-server.conf:
14:30 < CIA-15> * fixed Mesa source file matching now that the mesa package has two
14:30 < CIA-15> download tags
14:34 < karasz> rxr can i try a bootdisk target?
14:35 < rxr> on your open power?
14:35 < karasz> yep
14:35 < rxr> well - not on trunk, since the bootdisk is to be deleted soon
14:35 < rxr> karasz: we mostly only need the system build and then could boot it
14:35 < karasz> ahm ok i get it
14:35 < rxr> however there is this strange error with glibc/gcc not beeing able to build new executables
14:35 < karasz> so bootdisk target out of the window...
14:36 < karasz> that is bad
14:36 < karasz> the gcc stuff i mean
14:36 < karasz> by the way could you do a checkin for me?
14:36 < karasz> can you still connect?
14:37 < karasz> i had an ugly stuff goin around here yesterday...so i was wondering
14:37 < rxr> yep
14:38 < karasz> power breack down and after that internet down for hours
14:38 < karasz> some stuff even got melted after that... :'(
14:39 < karasz> lucky that power survived...
14:40 -!- sepp [n=sepp@p213.54.208.191.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)]
14:43 < rxr> I try with less optimization - /me curious if that makes a difference
14:44 < karasz> what exactly is the error not that i could do anything else than googlin around...
14:48 < rxr> /usr/lib64/gcc/powerpc64-t2-linux-gnu/4.0.2/crtbegin.o:(.fini+0x0): undefined re
14:48 < rxr> ference to `.__do_global_dtors_aux'
14:48 < rxr> /usr/lib64/gcc/powerpc64-t2-linux-gnu/4.0.2/crtbegin.o:(.init+0x0): undefined re
14:48 < rxr> ference to `.frame_dummy'
14:48 < rxr> /usr/lib64/gcc/powerpc64-t2-linux-gnu/4.0.2/crtend.o:(.init+0x0): undefined refe
14:48 < rxr> rence to `.__do_global_ctors_aux'
14:50 < karasz> ok i found the file
14:50 < karasz> rxr in the mean time did you noticed http://sourceforge.net/projects/iprdd ?
14:58 < rxr> nope
15:02 < karasz> http://gcc.gnu.org/ml/gcc/2003-07/msg00008.html
15:02 < karasz> it is even abiut the same subject?
15:06 < karasz> from my understanding the problems of gcc are due to the fact that on my power you have gcc 3.3.3 on 3.4 there are no probs of this kind
15:12 < karasz> mnemoc are you around?
15:18 < misl> rxr: Is it possible to use a local directory as download mirror? Something like
15:18 < misl> ./scripts/Download -mirror /mnt/space/t2/download/mirror/ -cfg desktop -required
15:20 < rxr> there is a -alt-dir that could work - or -mirror file:/// should do as well
15:27 < misl> :) that speeds up things considerably :)
15:43 < rxr> ,-)
15:46 * karasz fingers crossed ...
15:47 * rxr has a KDE konsole with a lot of tabs full of T2 builds ...
15:47 < rxr> from sparc, powerpc, x86 live cd to remove powerpc64 ...
15:48 < karasz> rxr should I setup another machine with T2 for emergency stuff?
15:48 * karasz is wondering...
15:49 < rxr> karasz: what do you mean?
15:49 < karasz> nothing, please ingnore me from time to time :)
15:49 < rxr> ,-)
15:51 < karasz> i am redaing now some stuff abt the new CPU from IBM,Apple,Toshiba
15:51 < karasz> CELL
15:54 < CIA-15> rene * r12960 /trunk/package/graphic/sane-backends/sane-backends.desc: * updated sane-backends (1.0.15 -> 1.0.16)
15:55 < CIA-15> rene * r12961 /trunk/package/graphic/sane-frontends/sane-frontends.desc: * updated sane-frontends (1.0.13 -> 1.0.14)
16:34 < karasz> so far so good
16:36 < karasz> yuck
16:36 < karasz> i was speacking too soon :(
16:37 < karasz> same tools.chroot stuff...
16:39 < karasz> rxr shouldn't we update gcc on that SuSE?
16:41 < karasz> i must run out
16:41 < karasz> be back in 30 mins...
16:45 -!- sepp_ [n=sepp@p213.54.218.51.tisdip.tiscali.de] has quit ["Lost terminal"]
16:47 -!- sepp [n=sepp@p213.54.218.51.tisdip.tiscali.de] has joined #t2
16:49 < rxr> karasz: nope
16:49 < rxr> that gcc is only used to build the cross compiler
16:49 < rxr> T2 toolchain build is totally decoupled from the system compiler (by default)
16:50 < mnemoc> misl: thanks
16:50 < mnemoc> karasz: pong
16:50 < mnemoc> rxr: i'm having troubles on desktop target here
16:51 < mnemoc> rxr: package selection and auto-gnome detection i guess
16:52 < mnemoc> rxr: alt-dir use cp -l so it's cheaper than file:/// as mirror ;)
16:53 < rxr> hi mnemoc
16:54 < mnemoc> hi rxr
16:54 < misl> hi mnemoc :)
16:54 < mnemoc> hi missile
16:57 < mnemoc> rxr: i was about to hardcode a pkgsel for desktop based on what was enabled on 2.1.0-rc3
16:57 < misl> :)
16:57 < mnemoc> rxr: what do you think?
17:10 < CIA-15> rene * r12962 /trunk/package/filesystem/unionfs/ (compile.patch pkg_linux26_post.conf postlinux.conf): * removed unionfs/postlinux.conf stuff since it is now included in the vanilla kernel
17:10 < mnemoc> oh
17:20 < mnemoc> rxr: does noisy has access to rsync files to gsmp's ftp?
17:22 < mnemoc> t2-minimal-2.1.0-{p4,pmmx,x86_64}.iso are ready for upload .... desktop targets needs some tweaking
17:24 < misl> mnemoc: desktop target is already running
17:24 < misl> desktop: 684 builds total, 19 completed fine, 0 with errors.
17:24 < misl> currently doing linux26 at stage 1
17:25 < mnemoc> misl: yes, but it's wont finish flawless
17:27 < mnemoc> due to changes on pkgsel afacs
17:27 < misl> :( I already feared so.
17:28 < mnemoc> wel... x86_64 has other kind of troubles i will solve for 2.1.1 :D
17:29 < mnemoc> misl: do you want to make a key on noisy to be used to allow direct iso uploads?
17:30 < mnemoc> -rw-r--r-- 1 root root 161607680 Oct 29 17:18 t2-minimal-2.1.0-x86_64_cd1.iso
17:30 < mnemoc> -rw-r--r-- 1 root root 151830528 2005-10-27 17:27 t2-minimal-2.1.0-p4_cd1.iso
17:30 < misl> mnemoc: go ahead
17:31 < mnemoc> -rw-r--r-- 1 root root 152215552 2005-10-29 12:57 t2-minimal-2.1.0-pmmx_cd1.iso
17:32 < mnemoc> compare the sizes
17:33 < rxr> mnemoc: and?
17:34 < mnemoc> x86_64 is 10M bigger, with the same packages inside
17:34 < misl> mnemoc: 64 bits has some 32 bits stuff as well.
17:35 < mnemoc> ok
17:36 < mnemoc> rxr: noisy has it's pubkey... should i add it to my authotized_keys or you create an accout for minto?
17:38 < rxr> mnemoc: what do you think - is yours enough for now?
17:38 < rxr> of course I could open yet another account on gsmp ...
17:38 < mnemoc> i already added it to mine ;) but i shall ask sysadmin for permission :p
17:41 < mnemoc> i love my country, yesterday chine agreed to sign a free trade agreement with chile. same day education ministry refused to try the "One Laptop per Child" (100 bucks laptop per school student) because it's not doable
17:41 < mnemoc> china*
17:43 < karasz> i am back
17:47 < mnemoc> wb karasz
17:48 < karasz> mnemoc i was reading config files for postgresql on T2 package
17:49 < karasz> so if i want to add a compile option i should do var_append extraconfopt " " "--with my option" ?
17:51 < valentin> moin
17:51 < misl> rxr: linux24 and 26 failed
17:51 < misl> regtest-alpha: 78 builds total, 38 completed fine, 2 with errors.
17:51 < misl> hi valentin
17:51 < karasz> moin valentin
17:52 < valentin> hi misl, karasz, mnemoc, rxr
17:52 < rxr> hi valentin
17:52 < mnemoc> karasz: yes, but without spaces ;)
17:53 < karasz> valentin as i understand you are the T2 mantainer for SH cpu's, am i right?
17:53 < mnemoc> karasz: what option are you missing?
17:53 < karasz> mnemoc i didn't studied well enough i was asking teorethicaly..
17:53 < valentin> karasz: well, you could say so because i am the only one here with an sh system
17:54 < valentin> but gcc 4.0.2 does not compile sh kernel in trunk version
17:54 < valentin> you have to add some multilib options which i tried to do yesterday
17:54 < misl> mnemoc: desktop build stopped after 30 minutes compiling linux26
17:54 < karasz> valentin what do you think about the new CELL cpu?
17:54 < mnemoc> :\
17:55 < valentin> never thought about it so far
17:55 < misl> mnemoc:
17:55 < mnemoc> misl: what module?
17:56 < misl> !> os-interface.c: In function `os_set_mlock_capability':
17:56 < misl> !> os-interface.c:1262: error: structure has no member name ..
17:56 < mnemoc> unionfs is that?
17:57 < misl> dunno how do I find out?
17:58 < mnemoc> look for the last tar call
18:04 < misl> Could find any tar calls in ERROR-LOG. Compiled without xtrace. However the last extract results in
18:04 < misl> Extracting linux-2.6.11.tar.bz2 (--use-compress-program=bzip2 -xf) ...
18:04 < misl> could = couldn't
18:04 < mnemoc> oh
18:05 < mnemoc> weird, it worked for minimal
18:05 < mnemoc> and for bootdisk
18:05 < misl> I chose the desktop target, not generic with some pkgsel.
18:06 < mnemoc> grep for 'Sub-building'
18:06 < CIA-15> rene * r12963 /trunk/package/network/ (10 files in 2 dirs): * removed driver source from ipw2*00, since they are not included in the kernel
18:06 < mnemoc> misl: yes, that's what you should have done
18:07 < mnemoc> misl: i mean, you did it right
18:07 < misl> last sub-build:
18:07 < misl> Sub-building 3rd party module package nvidia ...
18:08 < CIA-15> rene * r12964 /trunk/package/network/ (14 files in 2 dirs): * ipw2* -> ipw2*-fw rename to reflect those packages only contain the firmware
18:08 < mnemoc> so it is nvidia
18:08 < mnemoc> rxr: what do you think about documenting those renames to be able to make a 'migration' script between 2.1 and 2.2?
18:08 < CIA-15> rene * r12965 /trunk/package/network/ (ipw2100/ ipw2100-fw/ ipw2200/ ipw2200-fw/): * ipw2* -> ipw2*-fw rename part II
18:09 < CIA-15> rene * r12966 /trunk/package/ (4 files in 4 dirs): * moreved ipw2*-fw packages from network/ into firmware/
18:12 < valentin> going to make some fajita, cu later
18:12 < CIA-15> rene * r12967 /trunk/NEWS: * flushed NEWS file and started to document major changes since 2.1
18:12 < misl> valentin: yummi, make me one too :)
18:17 < rxr> mnemoc: can you commit the new desktop selection into 2.1 branch?
18:18 < mnemoc> rxr: i haven't done it yet
18:18 < mnemoc> rxr: but yes, every change on 2.1.0 tag will be replicated on 2.1 branch, and if necesary to trunk
18:19 < mnemoc> sent 158337782 bytes received 89471 bytes 120157.19 bytes/sec
18:19 < mnemoc> total size is 161645477 speedup is 1.02
18:20 < mnemoc> ^--- gcc-3.4.4 impact :\
18:20 < mnemoc> (2.1.0 iso over copies of 2.1.0-rc3)
18:21 < CIA-15> amery * r12968 /trunk/package/multimedia/klear/klear.desc: * updated klear (0.5.1 -> 0.5.2)
18:23 < karasz> guys a ssh question
18:23 < karasz> i use putty to login to a linux server
18:23 < karasz> after i input the username the server responds with
18:24 < karasz> Using keyboard-interactive authentication.
18:24 < karasz> how do i disable that response
18:24 < mnemoc> karasz: do you want to allow only key based auths?
18:25 < karasz> nope
18:25 < karasz> just to disable that response
18:25 < mnemoc> o.o
18:25 < karasz> it should ask for pass right after i input username
18:25 < mnemoc> ssh -q ?
18:27 < karasz> ssh -q ?
18:27 < karasz> it displays options
18:27 < mnemoc> ehm
18:31 < rxr> mnemoc: also impact of timestamps and bzip2
18:31 < karasz> rxr we have a little prob
18:32 < karasz> did you get out or it was me closing your connection?
18:32 < rxr> maybe we should default to gzip+rsyncable_patch for better rsync ISOs
18:32 < rxr> karasz: I did nothing
18:33 < karasz> then it was me restarting ssh
18:33 < karasz> sorry for that, you should be able to connect again...
18:45 < misl> rxr, mnemoc: I added 2 scripts to noisy crontab. /mnt/space/cron.reboot.t2 and /mnt/space/cron.19hour30.t2
18:45 < misl> I think it is clear what they stand for.
18:46 < misl> If you want to start a build at start up add something to the corresponding script :)
18:46 < mnemoc> no
18:46 < mnemoc> reboot means 08hour00 ?
18:46 < misl> yes
18:47 < mnemoc> :)
18:47 < misl> actually @reboot in the crontab. So if it boots some other time the script is run as well.
18:47 < mnemoc> great
18:48 < misl> I just hope the 19:30 script does bring the box down properly. and that it does not immediately reboot.
18:48 < mnemoc> rxr: on mnemosyne i use a .pause file, to stop building after the next packages finishes. do you think something like that should be merged to generic?
18:51 < karasz> rxr something starnge ocured
18:53 < rxr> karasz: ?
18:54 < rxr> mnemoc: what do you use .pause for ?
18:54 < mnemoc> rxr: touch it to make Build-Target stop after it finish current Build-Pkg
18:55 < mnemoc> instead of ctrl-c
18:55 < rxr> when do you favour that over C-c ?
18:55 < mnemoc> C-c aborts current or next, is very hard to hit the instead between
18:55 < mnemoc> instant*
18:55 < mnemoc> so you get src.* stuff
18:56 < mnemoc> and i avoid to stay watching current build
18:56 < mnemoc> if i don't want to abort current build-pkg
18:57 < mnemoc> graceful Build-Target stop ;)
19:02 * mnemoc needs to complete 'sde' and 'sam' helpers
19:02 < mnemoc> +systemwide
19:03 < rxr> feel free to commit it - though I never felt the need for that ,-)
19:03 < mnemoc> :)
19:08 < misl> rxr: there are quite some regtest failures
19:08 < misl> regtest-alpha: 78 builds total, 69 completed fine, 9 with errors.
19:08 < misl> regtest-arm: 78 builds total, 2 completed fine, 1 with errors.
19:08 < misl> regtest-cris: 78 builds total, 2 completed fine, 1 with errors.
19:08 < misl> regtest-hppa: 78 builds total, 2 completed fine, 1 with errors.
19:08 < misl> regtest-hppa64: 79 builds total, 1 completed fine, 1 with errors.
19:08 < misl> regtest-ia64: 78 builds total, 18 completed fine, 6 with errors.
19:08 < misl> regtest-m68k: 78 builds total, 0 completed fine, 0 with errors.
19:08 < misl> ... ( rest not yet build )
19:16 < mnemoc> outch
19:17 * mnemoc just added his screen to cron.reboot.t2
19:22 < mnemoc> rxr: do you have a rework for postinstall in mind?
19:24 < rxr> yes
19:24 < rxr> the new install will get fine grained install support
19:24 < rxr> preinstall postinstall preremove postremove
19:24 < mnemoc> per package or also 'global'?
19:24 < rxr> I have started a next-gen-install.txt but all the other work delayed it
19:25 < rxr> what do you mean with global?
19:25 < mnemoc> scripts owned by package foo, which are run on every install/remove
19:26 < rxr> SDECFG_ABORT_ON_ERROR_AFTER is named wrong, no?
19:26 < rxr> it is NOTABORT
19:27 < mnemoc> ABORT_AFTER
19:27 < mnemoc> NOTABORT_BEFOREOREQUAL
19:27 < rxr> mnemoc: ?
19:27 < rxr> no
19:27 < rxr> when it is 0 it does abort in stage 0 - not in higher?
19:28 < mnemoc> oh
19:28 < mnemoc> 0 means never abort
19:28 < mnemoc> 9 means always abort
19:28 < mnemoc> :p
19:28 < mnemoc> NOTABORT :p
19:28 < rxr> 0 'Continue if package-build fails after toolchain (native)'
19:29 < rxr> 9 'Always abort if package-build fails'
19:29 < rxr> mnemoc: so you agree the name is wrong ?
19:29 < mnemoc> >n is allowed to continue
19:29 < mnemoc> yes
19:29 < rxr> ok - I already thought I got mad ...
19:30 < mnemoc> i have big troubles thinking these days
19:30 < rxr> mnemoc: oh - any particular reason?
19:31 < mnemoc> 'natural' degradation
19:31 < rxr> in your age ?
19:32 < mnemoc> :\
19:32 < mnemoc> maybe i'm a dog undercover and i have 7x years
19:36 < misl> :( cron 19:30 script did not work
19:36 < mnemoc> did it run?
19:36 < misl> if it did noisy should have been shut down.
19:37 < mnemoc> failed != ignored
19:37 < misl> how can I find out if it did run?
19:37 < mnemoc> syslog
19:38 < misl> where can I find it?
19:38 < rxr> misl: seems noisy get's more used every day ,-)
19:38 < misl> :)
19:40 < mnemoc> :)
19:41 * misl slaps himself
19:41 < misl> Oct 29 19:30:01 noisy /USR/SBIN/CRON[29826]: (root) CMD (text -x /mnt/space/cron.19hour30.t2 && /mnt/space/cron.19hour30.t2)
19:42 < misl> text should have been test
19:42 < misl> I will do a shutdown manually
19:43 < mnemoc> chmod +x ?
19:43 < mnemoc> text!
19:43 < mnemoc> LOL
19:45 < rxr> misl: my regtest builds !!!!!!!
19:55 < misl> rxr: ;-) that's what happens after 19:30
19:55 < misl> but you can add something to /mnt/space/cron.reboot.t2 to have a start automatically when it restarts the next day.
19:57 < misl> rxr: you will have to wait another week and I can leave noise on for a couple of days :)
19:58 < misl> november 3rd I get the keys of my summerhouse. We plan to go there on saturday for a full week. I can leave noisy on that week.
19:59 < mnemoc> misl: summerhouse = daily for minto, weekends for minto's ?
19:59 < misl> mnemoc: A daily for all of us as long as I have that project very near.
20:00 < misl> But some weekend we will be back. Those weekends noisy has to sleep when we want to sleep :)
20:00 < mnemoc> :)
20:01 < misl> Unless... My neighbors start to complain about the humming sound from our house :D
20:01 < rxr> hehehe ,-)
20:08 < mnemoc> oh, "the word of the day" claims rene = Reen = Hyppelaphus :) the animal saint claus use
20:15 * karasz off for a couple of hours, will be back
20:15 < owl> bye karasz
20:15 < karasz> rxr did you saw my query?
20:15 < karasz> by owl :)
20:20 < rxr> yes - but I could not make sense of it and thus replied with a ? ... ,-)
20:25 -!- karasz is now known as karasz[Away]
20:46 < rxr> here are too many computers powered up - I do not want to see this years electricity bill ...
20:47 < mnemoc> can't you make special high-consume contracts there?
20:47 < valentin> still expensive...
20:47 < valentin> re
20:47 < valentin> we need some method to generate power
20:47 < mnemoc> we need some method to generate more money :)
20:48 < valentin> we could invent a special power device and sell that power and get money
20:48 < sepp> no we need computers that do not consume power like a drilling machine :)
20:49 < valentin> yeah
20:49 < valentin> hi sepp
20:49 < sepp> hi valentin
20:49 < mnemoc> and capable to build t2-based targets on less than a week
20:49 < valentin> wait two years and you will need a cryogenic device per cpu for cooling
20:51 < valentin> liquid nitrogen at -196 C
20:52 < sepp> and one more for a 500W consuming graphic card
20:53 < valentin> and a plutonium battery for power
20:54 < rxr> sepp: one graphic card? I heared having two becomes normal ...
20:55 < sepp> lol, yes in 2 years it will be std to have 2 nvidia monsters
20:56 < valentin> gamers will have a small box as workstation and a big fridge attached containing the rendering hardware
21:01 < valentin> anyone seen this before: http://www.symphonyos.com/ their site is down since they tagged beta 1
21:01 < sepp> its not down, they are just hiding it
21:01 < sepp> http://symphonyos.com/screenshots.html
21:01 < sepp> ;)
21:02 < valentin> thanks - that is what i wanted to see
21:03 < rxr> I bet it is dog slow since run on Mozilla's Gecko
21:05 < valentin> oh
21:07 < valentin> so, now i have to find out about this SDECFGSET stuff
21:07 < valentin> mnemoc: what was it you told me ?
21:07 < misl> seems like they have the screens that require 2 nvidia monsters http://symphonyos.com/screenshots/a3/3.jpg
21:08 < mnemoc> valentin: i told you that config/$config/config is noe generated as a dump of SDECFG_ vars
21:08 < misl> hehe, maybe we should add a bike to our workplace and generate our own electricity. Then we'd get some excersize too :)
21:09 < valentin> misl: looks like a normal knoppix startup
21:09 < misl> valentin: Yeah, with lots of colors :D
21:09 < valentin> mnemoc: so every variable has to be dumped explicitly ?
21:09 < valentin> misl: i have a new bike
21:10 < mnemoc> valentin: yes, but helpers do it for you
21:10 < misl> valentin: does it give you any power your computer run on?
21:10 < valentin> but the dynamo is inside the front wheel center
21:11 < valentin> mnemoc: how ?
21:12 < mnemoc> for x in VSERVER=1 GRSECURITY=0 KERNEL='linux' LVM=1 \ SSP=1; do
21:12 < mnemoc> eval SDECFGSET_TRG_MNEMOSYNE_$x const SDECFG_TRG_MNEMOSYNE_${x/=/ }
21:12 < mnemoc> done
21:12 < mnemoc> aarg... damn paste
21:13 < mnemoc> valentin: http://svn.exactcode.de/mnemosyne/sheba/trunk/config.in
21:15 < valentin> ok, so you set the SDECFGSET_... variable like i do. but what is the relevant thing here to make it be exported ?
21:15 < mnemoc> const
21:16 < valentin> just a 'const' ? /me trying
21:17 < mnemoc> const VAR VAL
21:17 < mnemoc> which does the dump
21:17 < mnemoc> and export the var
21:19 < valentin> did not work - just resetet all my config settings ?!
21:19 < valentin> or mom
21:20 < valentin> i restore my config - maybe the var is there again
21:22 < valentin> i hate scripts/config
21:22 < valentin> my computer is to slow for it
21:22 < rxr> every CPU is too slow for Config
21:22 < rxr> we badly need to redo it ...
21:22 < mnemoc> Config design stinks badly
21:22 < mnemoc> the sh* is recursive
21:23 < valentin> it is not only recursive but also cursed
21:23 < mnemoc> self calling including awk replacements on the fly
21:23 < valentin> yes, now i have the variable in my list but with wrong value
21:24 < valentin> it takes the const value
21:24 < valentin> but what do i need the sdecfg set then
21:24 < valentin> \
21:24 < valentin> ?
21:24 < mnemoc> SDECFGSET_ alters UI
21:24 < mnemoc> to not let you change the value by UI
21:24 < valentin> ok
21:25 < valentin> and why do i need to give it a value ?
21:25 < mnemoc> used by UI helpers to know what's value give when doing 'const'
21:26 < valentin> ok, it seems to work now. in theory i could commit it, but i fear it will kill your config's, too ...
21:26 < mnemoc> test it deeeeeeeply
21:27 < valentin> mnemoc: i just placed theese to lines in architecture/sh/config.in :
21:27 < valentin> const SDECFG_MULTILIBLIST "m4,m4-nofpu"
21:27 < valentin> SDECFGSET_MULTILIBLIST="m4,m4-nofpu"
21:27 < valentin> and as i started ./scripts/Config -cfg dream my old config was reset
21:27 < mnemoc> shouldn't affect us
21:27 < valentin> why ?
21:28 < mnemoc> previous abort i guess
21:28 < valentin> ok
21:28 < mnemoc> Config is self calling and self destructing
21:28 < rxr> especially the self destruction is annoying me like h*ll
21:28 < valentin> you are right, i hit ^C twice before
21:28 < mnemoc> i have moved parts of it to src/$config/
21:28 < valentin> maybe this killed it
21:29 < mnemoc> but 'config' generation is too deeply hardcoded
21:29 < mnemoc> what 'this'?
21:29 < mnemoc> ^C... yes
21:29 < mnemoc> ^C will always kill you config
21:31 < rxr> valentin: never hit C-c on Config
21:31 < rxr> you most probably cripple or loose your config
21:31 < rxr> I thought everyone here knows taht
21:32 < valentin> i did sometimes before
21:32 < valentin> and i never had .... mom now that i think about it
21:32 < valentin> i sometimes had problems like this
21:32 < mnemoc> Config is learnt by blood
21:33 < valentin> i thought by reasonable presumptions that config files are written on exit only
21:33 < mnemoc> most of the time that file is empty
21:34 < mnemoc> and current design _needs_ it that way
21:35 < valentin> ??
21:36 < valentin> Removing unsupported '--with-multilib-list=m4,m4-nofpu' from configure option string.
21:36 < valentin> why this - i placed it there in gcc.conf
21:36 < valentin> how about canceling a BuildTarget -job ?
21:36 < valentin> is that harmfull too ?
21:37 < mnemoc> gcc's configure script doesn't contain explict declaration for with-multilib-list
21:37 < mnemoc> no problem with that ;)
21:37 < valentin> that is not the configure script
21:37 < valentin> our build system seems to to that
21:37 < mnemoc> we could even add checkpoint/restore features on Build-Pkg
21:38 < mnemoc> valentin: our build system grep pkg's configure script
21:38 < valentin> hm
21:38 < mnemoc> valentin: and remove 'not supported' arguments
21:38 < valentin> i am quite sure that option exists
21:38 < valentin> maybe i have a copy and paste typo
21:41 < valentin> GCC_EXTRA_CONFIG="--with-multilib-list=m4,m4-nofpu"
21:41 < valentin> that is from crosstool for gcc 4.0.1 or so
21:41 < valentin> var_append confopt ' ' "--with-multilib-list=$SDECFG_MULTILIBLIST"
21:41 < valentin> that is what i did
21:42 < rxr> valentin: could you grep all gcc'c configure scripts ?
21:42 < rxr> /we dinner
21:42 < valentin> have a good lunch
21:43 < rxr> maybe extraconfopt was not sanity grepped
21:45 < rxr> hm - no - the check already greps all subconfigs ...
21:45 * rxr dinner - finally
21:48 < mnemoc> uff... finally doing rsync over dropbear by key
21:48 < valentin> wtf ... i grepped whole src dir and the only occurance is in changelog
21:48 < valentin> * config.gcc: Also set cpu_type / need_64bit_hwint for sh[be]*-*-*.
21:48 < valentin> (sh*linux configurations): Merge into:
21:48 < valentin> (sh*elf / sh*kaos configurations). Support --with-endian, --with-cpu,
21:48 < valentin> --with-multilib-list options. Support sh-superh-elf configuration.
21:49 < mnemoc> grep for --with-multilib
21:49 < mnemoc> it can be 'hidden'
21:52 < valentin> now i get another Changelog entry for --with-multilib-top, but that one is neither in the configs
21:52 < mnemoc> -multilib ?
21:53 < valentin> gcc/config.gcc: sh_multilibs=${with_multilib_list}
21:54 < mnemoc> :)
21:56 -!- owl [n=owl@62.4.80.218] has quit ["brb."]
21:56 < valentin> mnemoc: look at this: http://sources.redhat.com/ml/crossgcc/2005-10/msg00035.html
21:56 < valentin> what does it tell you ?
21:56 -!- owl [n=owl@62.4.80.218] has joined #t2
22:00 < mnemoc> uhm
22:01 < karasz[Away]> wb karasz
22:01 -!- karasz[Away] is now known as karasz
22:01 < karasz> oops i shouldn't say that should I?
22:02 < karasz> anyway i am back... :)
22:02 < valentin> wb means 'wellcome back'
22:02 < karasz> i know valentin :)
22:03 < valentin> feel free to wellcome yourself
22:03 < karasz> it gives me a warm fuzzy feeling, :)
22:03 < mnemoc> valentin: can you give that argument by export?
22:03 < valentin> mnemoc: maybe i should try to find out what GCC_EXTRA_CONFIG="--with-multilib-list=m4,m4-nofpu" actually does in crosstool
22:04 < mnemoc> karasz: a 'wb' from another person is warmer
22:04 < valentin> wb karasz
22:04 < mnemoc> valentin: looks exactly as extraconfopt, but without the sanity check
22:06 < valentin> you talking in riddles
22:06 < mnemoc> :)
22:06 < valentin> mom
22:07 < valentin> i have an idea
22:09 < valentin> nope
22:09 < valentin> i have no idea :(
22:09 < valentin> strange
22:09 < mnemoc> valentin: patch in a line to simulate --with-multilib-list
22:10 < mnemoc> valentin: and of course trace ${with_multilib_list}
22:11 < valentin> you can google for "--with-multilib-list" and get hundreds of hits, so i guess it is no legend
22:11 < mnemoc> valentin: you could also add a boolean var to enable/disable sanity checks on extraconfopt
22:13 < mnemoc> ftp://ftp.fsl.cs.sunysb.edu/pub/unionfs/unionfs-1.1.0.tar.gz
22:19 < valentin> mnemoc now multilib-list is not rejected ... but i doubt this will result in anything different
22:21 < mnemoc> how?
22:23 < valentin> just a dirty trick: echo --with-multilib-list > /dev/null
22:24 < mnemoc> # allow --with-multilib-list isn't cleaner ?
22:28 < valentin> mnemoc: i thought the sanity checker might care about comments
22:28 < valentin> maybe i should have used extra-confopt variable
22:34 < rxr> re
22:40 < karasz> wb rxr
22:48 < mnemoc> :)
22:48 < mnemoc> wb rxr
22:49 < valentin> mnemoc: it worked ?!!?!!??!
22:49 < valentin> now linux26 compiles
22:49 < mnemoc> :D
22:50 < valentin> but i find it embarrassing to commit my shitty echo > /dev/null patch
22:50 < mnemoc> how your patch looks now?
22:51 < valentin> +# fake following option:
22:51 < valentin> +echo --with-multilib-list > /dev/null
22:51 < mnemoc> # allow --with-multilib-list is ignored?
22:52 < valentin> dunno
22:52 < mnemoc> true --with-multilib-list would avoid the >
22:52 < valentin> that is even better
22:53 < mnemoc> i think # is the best, if it's accepted
22:53 < valentin> i would like to have a better comment than the one i have, too
22:54 < mnemoc> :)
22:56 < valentin> +# declare following option so we can pass it from gcc.conf
22:56 < valentin> +true --with-multilib-list
23:01 < mnemoc> rxr: jeffpc changed the topic of #unionfs to: UnionFS 1.1.1 available
23:02 < valentin> mnemoc: true does not work, the option is rejected again
23:03 < rxr> mnemoc: ,-)
23:04 < mnemoc> valentin: :(
23:05 < valentin> what is this super smart filter doing ?
23:05 < mnemoc> no idea, i didn't write it
23:06 < mnemoc> or reviewed it
23:06 < mnemoc> actualy i have never seen it :p
23:06 < valentin> do you at least know where it is ?
23:06 < mnemoc> functions _or_ Build-Pkg
23:07 < CIA-15> rene * r12969 /trunk/package/network/acx100/ (5 files):
23:07 < CIA-15> * updated acx100 to another latest snapshot to get it building
23:07 < CIA-15> with linux-2.6.14 - major hackery involved
23:07 < rxr> valentin: your at76c503a does not build anymore :-(
23:08 < valentin> arrr
23:08 < valentin> since when ?
23:08 < mnemoc> 2.6.14
23:08 < rxr> ack
23:08 < rxr> 2.6.14
23:09 < rxr> you can track on haudraufinix
23:09 < rxr> I temp. disabled it for the livecd now
23:09 < rxr> I rebuilt linux26 for the 8th time now getting acx100 built
23:09 < rxr> no more willing to hack another rotting w-lan driver to build ...
23:10 < mnemoc> you don't want a acx100 card
23:10 < rxr> ?
23:10 < rxr> you had one ?
23:10 < mnemoc> i have roasted two
23:10 < rxr> hehe
23:12 < valentin> it is ./scripts/functions.in in line 267 ff
23:12 < valentin> grep -q "[[ ]${x%%=*}[]= ):]"
23:13 < valentin> i love stuff like that
23:13 < mnemoc> :)
23:14 < valentin> it actually transforms options with optional values like --with-bla=blub to [[ ]--with-bla[]= ):]
23:15 < valentin> and now i wonder why that does not match the "true" line
23:15 < mnemoc> or # foo
23:16 < rxr> I do not find occurances of this multilib thing in gcc
23:16 < valentin> i do not understand that expression
23:16 < rxr> have you grepped manually for it ?
23:16 < valentin> rxr: but it works
23:16 < rxr> do you know where it is parsed ?
23:16 < valentin> rxr: i think somewhere the variable with_multilib_list is stored in config after build
23:17 < rxr> config.gcc: sh_multilibs=${with_multilib_list}
23:17 < rxr> config.gcc: echo "with_multilib_list=${sh_multilib} not supported."
23:17 < valentin> after config even
23:17 < valentin> yes
23:17 < rxr> ah - there it was hiding
23:17 < valentin> yep
23:18 < rxr> -with-* | --with-*)
23:18 < rxr> elif [[ $x = --*able-* ]] && egrep -q "\-\-(en|dis)able-\*" $configscript ; then
23:18 < rxr> echo "Autodetection for option impossible: " \
23:18 < rxr> "$x passed thru."
23:18 < rxr> config_command="$config_command $x"
23:18 < rxr> ^- try duplicating this for with-* ...
23:18 < rxr> and it will work
23:18 < rxr> or should I ?
23:19 < valentin> if you want
23:19 < mnemoc> so a bug
23:19 < valentin> why do we have that stuff at all ?
23:19 < mnemoc> some configure scripts abort for unknown options
23:19 < karasz> glibc32 failed
23:19 < valentin> but i think --with- options are always translated in shell variables no matter if you use them in configure or not
23:20 < mnemoc> and many parse-config of our packages massivly add --options
23:20 < valentin> ok
23:20 < mnemoc> "just in case" is the policy :)
23:22 < valentin> rxr: do you or shall i do ?
23:22 < rxr> - elif [[ $x = --*able-* ]] && egrep -q "\-\-(en|dis)able-\*" $configscript ; then
23:22 < rxr> + elif [[ $x = --*able-* ]] && egrep -q "\-\-(en|dis)able-\*" $configscript ||
23:22 < rxr> + [[ $x = --with* ]] && egrep -q "\-\-(with|without)-\*" $configscript; then
23:22 < rxr> should I commit or do you wanna cut'n paste before I commit ?
23:22 < valentin> commit it when you have it now
23:22 < rxr> ok
23:22 < rxr> one second I test one thing
23:23 * mnemoc will wait for the regressions :)
23:23 < valentin> then i do not need this patch anymore
23:23 < valentin> mnemoc: you are wellcome to find them all :)
23:23 < mnemoc> Baldzius_ will find them :)
23:24 < rxr> try it
23:24 < CIA-15> rene * r12970 /trunk/scripts/functions.in:
23:24 < CIA-15> * imrproved configure option sanity filter to handle --with-* and
23:24 < CIA-15> --without-*
23:24 < mnemoc> i don't have any build with heimdal currently
23:25 < rxr> karasz: thanks for the report ;-)
23:25 < CIA-15> valentin * r12971 /trunk/package/base/gcc/ (gcc.conf with-multilib-list_declaration.patch): * set multilib list via SDECFG_MULTILIBLIST variable
23:26 < mnemoc> does it work for an empty list?
23:26 < CIA-15> valentin * r12972 /trunk/package/base/gcc/with-multilib-list_declaration.patch: * this patch is not needed anymore due to r12970
23:26 < mnemoc> *G*
23:26 < valentin> mnemoc: if that variable contains "", this option will not be set
23:27 < mnemoc> every varialbe _contains_ ''
23:27 < valentin> sure
23:27 < CIA-15> valentin * r12973 /trunk/architecture/sh/config.in: * set SDECFG_MULTILIBLIST variable for gcc
23:27 < CIA-15> rene * r12974 /branches/2.1/package/base/linux-header/linux-header.conf: * merged r12916 from trunk: fixed linux-header to generate seperated 32bit and 64bit header for powerpc64
23:28 < valentin> i mean, if '' contains the value of the variable, then....
23:28 < valentin> :)
23:28 < mnemoc> :)
23:29 < mnemoc> rxr: only for ppc64?
23:29 < CIA-15> valentin * r12975 /trunk/package/base/uclibc/patches/ (2 files): * disabled patch for gcc bug 22167 (neither working not needed anymore)
23:29 < mnemoc> rxr: are you building 2.1.0 tag right?
23:29 < rxr> mnemoc: nope - we had this for other 64 bit arches fot a long time
23:29 < rxr> for sparc64 since ROCK 1.7 when I introduced this
23:29 < mnemoc> :)
23:30 < rxr> in fact this only adds powerpc64 into the case
23:30 < rxr> mnemoc: no need to merge this into 2.1.0 tag
23:30 < mnemoc> [[ $libdir = *64 ]] ?
23:30 < rxr> mnemoc: no - because it needs some more defines
23:30 < mnemoc> ok
23:30 < rxr> like preprocessor macro to depend on ...
23:30 < rxr> and for x86-64 the 2nd dir is i386 and so on
23:30 < rxr> the case is fine ,-)
23:31 < mnemoc> :)
23:31 < CIA-15> rene * r12976 /branches/2.1/package/base/dietlibc/dietlibc.conf:
23:31 < CIA-15> * merged r12922 from trunk: added the same clutch for powerpc64 to dietlibc as we have for powerpc
23:31 < CIA-15> where our names differ from dietlibc internals
23:32 < rxr> also no need for 2.1.0 - 2.1.1 is fine
23:32 < mnemoc> what about 2.1.0 minimal iso for plain ppc?
23:32 < mnemoc> g3 or so
23:33 < rxr> I'm building desktop
23:33 < rxr> I think desktop is fine for PPC
23:33 < mnemoc> builds fine?
23:33 < rxr> most Apple user's can't handle a minimal anyway
23:33 < mnemoc> :)
23:33 < rxr> mnemoc: bootdisk finished desktiop is building
23:33 < rxr> oh - I wrote users where I meant in fact loosers ...
23:33 < rxr> == 23:24:07 =[5]=> Building gnome2/libgnomeui [2.10.1 2.1-stable].
23:33 < rxr> ^- powerpc build
23:34 < mnemoc> 2.1-stable != tag
23:34 < rxr> and yep I choose to use 2.1 so I can commit fixes immediatly ...
23:34 < rxr> I think I ever build release from the branch so I can fix and at the end ...
23:34 < mnemoc> you could commit to tag :\
23:34 < mnemoc> sdever is different
23:34 < rxr> mnemoc: I know
23:35 < rxr> few people care ...
23:35 < rxr> Error logs from default-2.1-stable-powerpc-powerpc-750-desktop:
23:35 < rxr> [5] base/linux24 [5] develop/librep
23:35 < rxr> 679 builds total, 391 completed fine, 2 with errors.
23:35 < rxr> librep is broken on ppc for a long time :-(
23:35 < mnemoc> 'trac' cares
23:35 < rxr> think I should finall look into the stack grow direction that is mismatched
23:35 < karasz> rxr some weird stuff about not finding files but still goes on
23:35 < karasz> iSeries?
23:36 < rxr> copying file alsa-kernel/ppc/tumbler.c
23:36 < rxr> patching file tumbler.c
23:36 < rxr> Hunk #2 succeeded at 918 with fuzz 2 (offset 71 lines).
23:36 < rxr> Hunk #3 FAILED at 971.
23:36 < rxr> Hunk #4 succeeded at 1313 with fuzz 1 (offset 148 lines).
23:36 < rxr> ^- that is the ppc linux24 failure :-( due to alsa
23:37 < mnemoc> weird... that should have failed also for other archs
23:37 < rxr> nope - I think alsa patches on the fly based on kernel version and what is built
23:37 < mnemoc> ic
23:37 < rxr> in fact this 2.4 alsa self-patching is quite ugly ...
23:37 < rxr> mnemoc: I fix this when the rest of the build finished
23:37 < mnemoc> :)
23:38 < mnemoc> i'll merge your fixes to tag so tarball is fixed
23:39 < rxr> mnemoc: ok - those ppc64 is not important for the tag
23:39 < rxr> we can save that noise and release 2.0.1 with that enhancements
23:39 < rxr> valentin: please notify if my with-* stuff does work
23:40 < mnemoc> rxr: ack, 2.1.1 will include *64 fixes, ppc64 and adm64
23:40 < valentin> rxr: i'm testing
23:40 < mnemoc> i want to do a 2.1-test branch and not release -rc of 2.1.x with x>0
23:44 < valentin> rxr: it works
23:44 < valentin> and now we have some more --with- vars in there
23:44 < valentin> Autodetection for option impossible: --without-libpam passed thru.
23:44 < valentin> Autodetection for option impossible: --without-pam passed thru.
23:44 < valentin> Autodetection for option impossible: --with-multilib-list=m4,m4-nofpu passed thru.
23:45 < valentin> the first one were rejected before
23:45 < valentin> s/one/two
23:46 < valentin> but we also have --disable-libpam --disable-pam so i do not think it matters
23:46 < mnemoc> "just in case"
23:47 < mnemoc> packages try to enable auto their support
23:47 < mnemoc> without exceptions
23:49 < rxr> mnemoc: nope - many packages have options that default to no
23:49 < rxr> mplayer has many e.g. but also the --with-xinerama often defaults to no ...
23:49 < rxr> mnemoc: you want one more branch?
23:49 < rxr> I would not do that
23:49 < rxr> to many merging around
23:49 < rxr> just commit conservatively in 2.1-branches test build and when all is fine tag
23:50 < rxr> would yet another branch be of much help ?
23:51 < mnemoc> i want to be sure 2.1-stable branch is always stable
23:51 < mnemoc> i test branch will let me test regressions on other architectures
23:51 < mnemoc> and on different test cases
23:52 < mnemoc> before merging the the _stable_ branch
23:52 < mnemoc> them
23:54 < rxr> ok - as you like
23:54 < rxr> however when I do 2.2 it will most probably less fine grained
23:54 < valentin> so you want three branches: HEAD -> testing -> stable ?
23:54 < rxr> however maybe someone else volunteers to do 2.2 ... - though I have not too much hope
23:56 < mnemoc> rxr: i need to be able to automatize updates on remote production machines
23:57 < rxr> mnemoc: another idea
23:57 < rxr> work commit into 2.1-stable and tag the release early and merge known to work stuff into that
23:57 < rxr> tag
23:57 < rxr> however - as you like
23:58 < mnemoc> .oO
23:59 < mnemoc> but i would need to do a release for every important update
23:59 < karasz> rxr does it matter that dietlibc had some unfound files?
--- Log closed Sun Oct 30 00:00:30 2005