T2 IRC Log: 2005-01-17

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 Mon Jan 17 00:00:47 2005
00:08 < rxr> yep
00:21 < CIA-10> jsaw * r5611 /trunk/package/network/samba/ (README.xinetd samba.conf): * add swat for xinetd integration README
00:21 < rxr> hi jsaw
00:21 < jsaw> hi rxr :)
00:22 < CIA-10> jsaw * r5612 /trunk/package/mail/sylpheed/sylpheed.desc: * update sylpheed (1.0.0beta3 -> 1.0.0
00:22 < rxr> syntax error
00:22 < jsaw> ?
00:22 < jsaw> where
00:22 < rxr> @ closing bracket
00:22 < rxr> commit message ;-)
00:22 < jsaw> oh
00:22 < rxr> no problem ;-)
00:32 < rxr> jsaw: I hope you had a relaxing sunday ?
00:33 < jsaw> well mostly.
00:33 < jsaw> I'm so sleepy since a couple of days... so I sleep a lot... to much..
00:34 < jsaw> rxr: you were hacking a lot...?
00:34 < jsaw> rxr: or did you relax also?
00:34 < rxr> well - hacking a lot describes it best ...
00:35 < rxr> did you notice the parallel build stuff and x86-64 progress ?
00:35 < jsaw> yep.
00:35 < jsaw> did mnemoc progress with the uclibc thing?
00:35 < rxr> I did not saw mnemoc_ since last night
00:36 < rxr> hope he got some time to relax a bit, too
00:36 < jsaw> he, yes
00:36 < jsaw> I'm relocating myself into the lab now, cu in a few minutes.
00:36 < rxr> cu
00:37 * rxr photoshopping mouse cursors ...
00:37 < rxr> btw. photoshop is highly ugly, non intuitive, and dog slow
00:37 < rxr> compared with the real gimp ...
00:51 < rxr> photoshop is a usability nightmare ....
01:04 < CIA-10> rene * r5613 /trunk/package/ (filesystem/nfs-utils/nfs-utils.desc shells/zsh/zsh.desc): * marked zsh and nfs-utils as NOPARALLEL
01:16 < jsaw> re
01:20 < rxr> hi
01:20 < jsaw> hi rxr
01:20 * rxr switched from photoshop to inkscape ...
01:20 < rxr> good to have more than one cpu ...
01:20 < jsaw> oh, really?
01:21 < jsaw> I have to admit that I like Illustrator , actually.
01:31 < rxr> I'm not in properitary software for many years
01:31 < rxr> only for testing I had to use this ugly gimp clone recently - and I hate it ...
01:31 < rxr> I do not know illustrator - but I hate scribus ...
01:32 < jsaw> sadly there was no good vector graphics designer when I started.
01:32 < rxr> and I quite liked Auto CAD - when I has this technical drafting training ...
01:32 < rxr> Auto CAD was quite cool - just not perfect - and the LISP scripting sucked ...
01:32 < jsaw> and I have seen one recently too...
01:32 < rxr> but the rest was quite cool
01:33 < jsaw> yes, AutoCAD does an extremely well job!
01:33 < rxr> since AutoCAD was open source (AFAIK) in the beginning (pre 1 or whatever) I wonder where that source is ...
01:33 < rxr> but qcad is a quite nice 2d only replacement for it ...
01:33 < rxr> for the not too advanced jobs ...
01:34 < jsaw> Well, I have to use for 2D only...
01:34 < jsaw> s/to/no/
01:34 < Postal> Hey all
01:34 < jsaw> hi Postal
01:34 < Postal> any new fixes today?
01:35 < jsaw> xfce 4.2.0 out
01:36 < Postal> == 19:36:15 =[5]=> Building shells/zsh [4.2.1 2.1.0-beta3].
01:36 < Postal> !> The package build aborted with the following config error:
01:36 < Postal> !> zsh: Unknown package flag: UNPARALLEL
01:36 < Postal> !> Disable the 'Paranoia Check' config option to ignore suc ..
01:36 < Postal> == 01/16/05 19:36:15 =[5]=> Aborted building package zsh.
01:36 < Postal> !> Creation of binary package isn't possible, because the package was not
01:36 < Postal> !> built successfully in (at least) the current stage.
01:37 < CIA-10> jsaw * r5614 /trunk/package/www/webcdwriter/ (6 files): * add webCDwriter 2.6.7
01:38 < Postal> that seems pretty easy, just remove the -UNPARALLEL configure option...
01:39 < jsaw> rxr -^
01:42 < rxr> NOPARALLEL even ...
01:42 < rxr> sorry
01:43 < Postal> lol
01:43 < Postal> rxr, is that the fix, that will hopefully make this thing compile?
01:44 < rxr> on my x86 box yes ...
01:44 < rxr> on yours maybe not ...
01:44 < rxr> Postal: btw. I think you could prepare to setup to boot into your new t2 system ...
01:44 < rxr> or is there s.th. vital missing ?
01:44 < Postal> grub
01:45 < CIA-10> rene * r5615 /trunk/package/shells/zsh/zsh.desc: * fixed NOPARALLEL flsg in zsh
01:45 < Postal> grub is important
01:45 < rxr> well - you should use your system's grub to boot the new system anyway
01:45 < rxr> so no need for grub ...
01:45 < rxr> (yet)
01:45 < rxr> when mipe's test build finishes we know if just enabling multilib fixes it ...
01:45 < Postal> i can build a new toolchain for grub anyway
01:45 < rxr> yep
01:45 < Postal> how long will that test take..?
01:46 < rxr> don't know - most probably he will already tell use tomorrow ...
01:47 < Postal> ok, well again, i dont know if ill be around... but obiously, go and do the commits, and when i come home after exams, ill svn up..
01:47 < Postal> zsh just built
01:48 < Postal> now that leaves gcc, glibc, bdb and grub
01:48 < Postal> rxr, so how do i boot this thing?
01:49 < Postal> and dirtree
01:51 < rxr> again glibc and dirtee have a harmless shared file conflict that will not happen on the next build ...
01:51 < rxr> Postal: haven't you the log of last where I lined out the bootup procedure ?
01:51 < Postal> um, no, i dont keep logs of irc,,,
01:54 < rxr> 03:06 < rxr> chroot build/system-*/
01:54 < rxr> 03:06 < rxr> passwd
01:54 < rxr> 03:06 < rxr> exit
01:54 < rxr> 03:07 < rxr> currently all the t2 stuff is in the root ...
01:54 < rxr> 03:07 < rxr> so just do s.th. like this:
01:54 < rxr> 03:07 < rxr> mkdir t2
01:54 < rxr> 03:07 < rxr> mv * t2
01:54 < rxr> 03:07 < rxr> mv t2/build/sytem-*/* .
01:54 < rxr> 03:07 < rxr> then you have you fresh sandbox in the front of all
01:54 < rxr> 03:08 < rxr> I would then just use your current suse grub or so to load the new
01:54 < rxr> system - until you know all works fine and you want to everyday use it and you c
01:54 < rxr> an rm the suse
01:54 < Postal> ok
01:55 < Postal> mv * t2
01:55 < Postal> mv: cannot move `t2' to a subdirectory of itself, `t2/t2'
01:55 < Postal> i suppose thats easy to fix lol
01:56 < rxr> that is ok
01:56 < rxr> what does ls tell you?
01:56 < Postal> .. .svn t2
01:56 < rxr> oehm
01:56 < rxr> mv .svn t2
01:56 < rxr> ld -l t2/
01:56 < rxr> ls -l t2
01:56 < rxr> even
01:57 < Postal> its too long to paste here
01:57 < Postal> lots of src files
01:57 < rxr> ok
01:57 < rxr> then do
01:57 < rxr> mv t2/build/system-*/* .
01:58 < Postal> its thinking
01:59 < rxr> I hope you did it in the right dir ...
01:59 < Postal> well, if not we can always move it lol
01:59 < rxr> should be quite instantly finished
02:00 < Postal> well its still thinking
02:01 < Postal> and its not like i can ctrl-c it, cause then well end up with a mess
02:03 < rxr> where have you been when you executed it ?
02:03 < rxr> I hope in /mnt/t2, right ?
02:03 < Postal> no /mnt/t2/t2-trunk
02:04 < Postal> :-(
02:04 < rxr> shoudn't take that long anyway ...
02:04 < rxr> ok - just wait until it finished
02:05 < Postal> yeah well its still going... i have to wait otherwise well end up with half a distro one place and the other half god knows where
02:06 < Postal> its lagging out my box
02:07 < rxr> hm
02:07 < Postal> its making me kinda nervous lol, i dont suppose i could have sent it into a loop or somethign
02:07 < rxr> maybe some loopback filesystem was still mounted and it copis quite a lot of stuff ...
02:08 < rxr> open a new terminal and look into /mnt/t2/t2-trunk/t2
02:08 < rxr> that is where you moved it, right ?
02:08 < Postal> there is alot of stuff in there
02:09 < rxr> ls
02:09 < Postal> you want the output
02:10 < rxr> ok - you can stop
02:10 < rxr> what does "mount" output ?
02:11 < Postal> ctrl-c
02:11 < Postal> wait did you say i can ctrl-c the mv?
02:11 < Postal> /dev/hde5 on / type reiserfs (rw,acl,user_xattr)
02:11 < Postal> proc on /proc type proc (rw)
02:11 < Postal> tmpfs on /dev/shm type tmpfs (rw)
02:11 < Postal> devpts on /dev/pts type devpts (rw,mode=0620,gid=5)
02:11 < Postal> /dev/hde2 on /boot type ext2 (rw,acl,user_xattr)
02:11 < Postal> /dev/hde1 on /windows/C type ntfs (ro,noexec,nosuid,nodev,gid=100,umask=0002,nls=utf8)
02:11 < Postal> /dev/hdb on /media/cdrecorder type subfs (ro,nosuid,nodev,fs=cdfss,procuid,iocharset=utf8)
02:12 < Postal> /dev/hda on /media/dvdrecorder type subfs (ro,nosuid,nodev,fs=cdfss,procuid,iocharset=utf8)
02:12 < Postal> /dev/fd0 on /media/floppy type subfs (rw,nosuid,nodev,sync,fs=floppyfss,procuid)
02:12 < Postal> usbfs on /proc/bus/usb type usbfs (rw)
02:12 < Postal> /dev/hde6 on /mnt/t2 type ext2 (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:12 < Postal> none on proc type proc (rw)
02:13 < rxr> ohm
02:13 < rxr> in the new console
02:13 < rxr> where you are in /mnt/t2/t2-trunk
02:13 < rxr> run
02:13 < rxr> ./scripts/Cleanup
02:14 < Postal> ok its running
02:14 < Postal> can i ctrl-c the move command in the other terminal?
02:14 < rxr> not yet
02:14 < Postal> ok
02:15 < rxr> note when Cleanup finished
02:15 < Postal> the mv segfaulted
02:15 < Postal> i think the cleanup and the mv collided
02:16 < Postal> cleanup is done
02:16 < rxr> ok - how does the 1st mv behave ?
02:16 < Postal> what do you mean?
02:16 < rxr> still running?
02:16 < rxr> finished, ...
02:17 < Postal> no, it segfualted
02:17 < Postal> mv: cannot stat `t2/build/system-2.1.0-beta3-x86-athlon64-64-generic-expert/TOOLCHAIN/src.gcc.1105900946.5882.0x7f0200/gcc-3.4.3/objdir/gcc/cse.o': No such file or directory
02:17 < Postal> Segmentation fault
02:18 < rxr> segment what ?
02:18 < rxr> mv should never segfault ..
02:18 < rxr> oh my ...
02:18 < rxr> ok
02:18 < Postal> so thats bad then
02:19 < rxr> I would say just rerun the move, so you get a consistent state of the fils in their new location ...
02:19 < Postal> ok
02:19 < Postal> v: cannot overwrite directory `/windows/system-2.1.0-beta3-x86-athlon64-64-generic-expert'
02:19 < Postal> mv: cannot move `/boot' to `/windows/boot': Device or resource busy
02:19 < Postal> mv: cannot move `/proc' to `/windows/proc': Device or resource busy
02:19 < Postal> mv: cannot move `/sys' to `/windows/sys': Device or resource busy
02:19 < Postal> thats odd
02:20 < Postal> um, im not sure what that mv command did, but it has crahsed my system...
02:20 < rxr> .oO
02:21 < Postal> i cannot open anything at all
02:21 < rxr> that is quite bad ...
02:21 < Postal> Service '/opt/kde3/share/applications/kde/konsole.desktop' is malformatted.
02:21 < Postal> everything says that
02:22 < Postal> and gaim cannot find its sound files...
02:22 < rxr> well - better try to reboto soon
02:22 < Postal> i fear if i reboot, i will not get a system...
02:22 < rxr> I hope you did not yet had s.th. imporant on the box since you reinstalled suse recently ...
02:22 < Postal> nothing but t2
02:22 < Postal> this was purley to install t2
02:22 < rxr> I have no idea ... I never has a system that crashed during file i/o ...
02:23 < rxr> good - puh
02:23 < Postal> i must have typed the mv command wrong
02:23 < rxr> now with all the fixes you can rebuild t2 in a matter of 4 hours or so, right ...
02:23 < jsaw> Postal: did you have any mount --bind's?
02:23 < Postal> right
02:23 < rxr> no - either the suse was not yet that stable
02:24 < rxr> or moving files in the possible mounted /proc dirs in the src.* package diles did s.th. very odd ...
02:24 < Postal> so what would you recomend...
02:25 < rxr> try to reboot - when it does not come you - well - thanks to god you have not left s.th. important ...
02:25 < rxr> I would hit reset for the roboo
02:25 < rxr> when it is just system memory corruption you might write back random junk to the FS during shutdown ..
02:25 < rxr> a reset would just leave the fs's in an unchecked state ...
02:25 < Postal> ok if im not back in 2min, i am reinstalling and will be back in 25 min
02:25 < Postal> ttyl
02:26 < rxr> ;-)
02:28 < rxr> I yet again found a graphical glitch in osx ...
02:35 -!- Postal [~Postal_@ool-43562a7f.dyn.optonline.net] has quit [Read error: 60 (Operation timed out)]
02:50 < rxr> hey!
02:50 < rxr> why do I mess around to get my linux toolchain generate tiff files - when the OSX framework loads the png's just fine
02:51 < rxr> again 40 mintues wasted for useless research ...
02:52 < rxr> at least I now know that my .svg -> bitmap converted cursors look quite ok - and that osx can load png's as cursor files ,-)
02:52 < rxr> in contrast to the imagemagick generated .tiff files ...
02:58 -!- rxr [~rene@p213.54.195.112.tisdip.tiscali.de] has joined #t2
02:58 -!- Topic for #t2: T2 | 2.1.0-beta2 RELEASED | The next generation of System Development Enviroments (SDE) | http://www.t2-project.org/ | CIA, don't ever leave us again!
02:58 -!- Topic set by mnemoc [] [Wed Jan 12 03:04:43 2005]
02:58 [Users #t2]
02:58 [ _Ragnar_] [ daja77] [ jsaw ] [ nzg ] [ praenti] [ sparc-kly]
02:58 [ CIA-10 ] [ ideal ] [ mnemoc_] [ Postal] [ rxr ] [ valentin ]
02:58 -!- Irssi: #t2: Total of 12 nicks [0 ops, 0 halfops, 0 voices, 12 normal]
02:58 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
02:58 -!- [freenode-info] why register and identify? your IRC nick is how people know you. http://freenode.net/faq.shtml#nicksetup
02:58 < rxr> re
02:58 < Postal> hey rxr
02:58 -!- Irssi: Join to #t2 was synced in 6 secs
02:59 < rxr> hi Postal
02:59 < rxr> how is your box ?
02:59 < rxr> reinstalled ?
02:59 < Postal> yup
02:59 < Postal> im about to pull down the new svn now
03:00 < rxr> ok ;-)
03:00 < Postal> ok svn is doing its thing
03:01 < Postal> should i build on a sperate partition, or does it not matter..?
03:02 < rxr> /mnt/t2 was a seperate partition, right ?
03:02 < Postal> yeah
03:02 < rxr> ok - then do it so again ...
03:02 < Postal> ok
03:03 < rxr> we keep care what we move around the next time ...
03:03 < Postal> lol yeah
03:04 < Postal> if you have any last minute fixes you want to put in before i build this, now would be the time, the svn is almost donw
03:05 < Postal> just so you know "Calling ioctl() to re-read partition table.
03:05 < Postal> WARNING: Re-reading the partition table failed with error 16: Device or resource busy.
03:05 < Postal> The kernel still uses the old table.
03:05 < Postal> The new table will be used at the next reboot.
03:05 < Postal> Syncing disks.
03:05 < Postal> " you do have to reboot
03:05 < Postal> brb
03:05 -!- Postal [~Postal@ool-43562a7f.dyn.optonline.net] has quit [Remote closed the connection]
03:12 < rxr> how I love my T2 ...
03:12 -!- Postal [~Postal@ool-43562a7f.dyn.optonline.net] has joined #t2
03:12 < rxr> Makefile to convert the .svg cursors using inksape ;-)
03:12 < rxr> not this murkse Photoshop crap ...
03:13 < Postal> ok, im going to start the build in just a moment
03:15 < Postal> I also changed ownership of the mnt/t2 dir so this time i can build as a regular user
03:16 < rxr> no
03:16 < rxr> this is not possible
03:16 < rxr> T2 must be build as root
03:16 < Postal> oh ok
03:16 < rxr> a) file permission preserving
03:16 < rxr> b) chroot
03:17 < rxr> we want to get rid of this limitation for T2 3.0 or so ...
03:17 < Postal> well, edward owns all the files, as opposed to root owning them
03:20 < rxr> oops
03:20 < Postal> its not a big deal
03:20 < rxr> inkscape in no-gui mode segfaults if the exporting file exists ...
03:20 < Postal> oh
03:20 < rxr> Postal: better change the permissions back to root before the build
03:20 < rxr> some files are copied - and might have the wrong ownership then ...
03:21 < Postal> rxr, pain in the a** but ok, after the dl i will
03:21 < rxr> chmod -R root:root *
03:21 < rxr> chmod -R root:root .svn
03:21 < Postal> rxr, if i change ownership of a folder all files contained within are also changed correct
03:21 < rxr> should do
03:22 < rxr> only if you specify -R ... ;-)
03:22 < Postal> i used chown
03:22 < rxr> chown even
03:22 < rxr> not my night
03:22 < Postal> lol
03:23 < rxr> the -R thing still applies ...
03:23 < Postal> i figured
03:23 < Postal> so i should just be able to build strait through correct..?
03:23 < rxr> this is my hope ... ;-)
03:24 < Postal> well we will soon see
03:24 < rxr> only 5-glibc 5-bdb and 5-zsh will fail IIRC
03:24 < Postal> and 5-grub and also 5-dirtree
03:24 < Postal> 5-zsh is fixed
03:25 < Postal> the NOPARELLEL option fixed it
03:26 < rxr> hm - if this is healthy for my Xcode project file:
03:26 < rxr> 2005-01-17 03:25:37.113 Xcode[339] PBXProject at /Users/rene/develop/avision-osx/AvisionDS.xcode still has 18 registered items after during deallocation. These items are of the following classes:
03:28 < Postal> rxr, what are you talking about?
03:30 < rxr> about not that perfect Apple products I'm currently programming in ..
03:30 < rxr> and the bugs and oddities on the way ...
03:30 < Postal> Kdevelop
03:30 < rxr> and how much I would love to boot this iBook back into T2 ...
03:30 < Postal> lol
03:30 < rxr> yeah - *lol*
03:30 < Postal> or jedit
03:30 < Postal> brb cig time
03:30 -!- Postal is now known as Postal|Cig
03:30 < rxr> not if you need to write OS X drivers and Cocoa based GUIs in ObjC ...
03:30 < rxr> cu
03:34 -!- Postal|Cig is now known as Postal
03:35 < Postal> rxr, i dont suppose that you know C++?
03:36 < rxr> yes - far better than ObjC ...
03:36 < Postal> im learning C++ now
03:37 < Postal> its hard lol
03:37 < rxr> in school or so ?
03:37 < Postal> no from a book
03:40 < rxr> ah - which one ?
03:41 < Postal> rxr, which file is makeinfo in again
03:41 < rxr> I find Accelerated C++ from A. Koenig and Babara E. Moo (--variations) a quite good beginners choice
03:41 < rxr> texinfo
03:41 < Postal> ok
03:42 < Postal> I have too look up the name of the book cause its at school gimmie a sec, trying to do 10 things at once
03:43 < Postal> C++ How to Program, Fourth Edition
03:43 < Postal> Harvey M. M. Deitel, Paul J. Deitel
03:44 < rxr> I think I do not know that one ...
03:44 < rxr> I hope it is good ,-)
03:44 < Postal> it is
03:44 < Postal> it moves very fast, but if you do all of the exercises you catchon
03:44 < Postal> i like it cause i can read the whole chapter and not have to touch the keyboard until the end
03:45 < rxr> ;-)
03:45 < rxr> I whish the backery would already be open ...
03:46 < Postal> backery=?
03:46 < Postal> btw i am using a 4 parallel jobs build
03:50 < rxr> backery should already be english ..
03:51 < rxr> a - without 'c'
03:51 < Postal> ah ok
03:51 < rxr> bakery ;-)
03:51 < Postal> thats what i thought..
03:51 < Postal> yeah, fresh doughnuts
03:52 < Postal> mmmm, but its 4am in the morning, i doubt that anythign will be open
03:53 < rxr> .oO
03:53 < rxr> damn just messed up one of my svn repositories
03:53 < Postal> how?
03:53 < Postal> build stage 0 complete without errors
03:53 < rxr> tagging a ag twice
03:54 < rxr> s/ag/tag/
03:54 < Postal> we should have timed this build
03:55 < rxr> ah - no - did not tag twice - puh
03:55 < Postal> is mplayer in the repository?
03:56 < rxr> multimeda
03:56 < rxr> media even
03:56 < Postal> so thats a yes then
03:57 < rxr> damn - I messed some tags a bit up already ...
03:57 < rxr> ah - yes - in the repository multimedia ...
03:57 < Postal> If a package is not in the repository, how does it get put in?
03:58 < rxr> Postal: oh - maybe I confused you - I was talking about another SVN code repository - not t2 that I messed up ...
03:58 < rxr> Postal: someone adds it ..
03:58 < rxr> or used some script
03:58 < rxr> e.g. there is some helper that converts a freshmeat.net entry to a t2 package and so on ..
03:58 < Postal> when we get this build up, i want you to show me..
03:59 < rxr> sure
04:00 < Postal> you are much better at linux than i am you see, so, i am trying to learn.. i hope you dont mind me asking all these questions that probably seem pretty simple to you..
04:00 < Postal> ;-P
04:00 < rxr> no - of course it is not simple or obvious how things are done or handled by some "yet another project"
04:01 < rxr> and in addition there are no dump questions - just dump answers.
04:01 < rxr> damn - what do I do with my messed up tags ...
04:01 < Postal> i must say though from what i have seen of t2 linux so far i really like it.. i like to tinker with things you see... im not happy unless there is somthing broken to fix
04:02 < rxr> ;-)
04:03 < Postal> I have to say man the flu sucks
04:04 < Postal> why would a package build in stages 1-4 but fail in stage 5
04:08 < rxr> what flu ?
04:10 < Postal> die Grippe haben...
04:10 < rxr> Postal: you have a flu ?
04:11 < Postal> yeah
04:11 < rxr> oh - keep care!
04:11 < Postal> thank god for germen translators lol
04:11 < rxr> do you know German - or did you just used some online dictonary ?
04:11 < Postal> online dictionary lol
04:12 < Postal> i know some spaish thats about it
04:12 < Postal> i thought youd be impressed none the less
04:12 < rxr> yeah - quite nice ;-!
04:13 < Postal> well, it made it easier for you to understand me lol
04:13 < rxr> I just retagged 0.1.{3,4,5,6} of one project - what a pain
04:13 < Postal> i bet
04:13 < Postal> stage 1 is comming along well so far
04:14 < rxr> Postal: I thought you meant some package and guess you misstyped glibc or so .. ;-)
04:14 < Postal> LOL
04:14 < rxr> what a pain glibc is ...
04:14 < Postal> i know, do you have any ideas on the package?
04:14 < Postal> other than to take a big hammer and smash things
04:14 < rxr> f is near g, u is near i ...
04:15 < rxr> yeah - I wanted to do s.th. like that ...
04:15 < rxr> like downdating the glibc a bit again
04:15 < Postal> i dont like that idea
04:15 < rxr> soem weeks before my xmas glibc update it built fine ....
04:15 < rxr> well - I like a building glibc more thatn a crashing one
04:15 < Postal> better to fix the newer version..
04:15 < rxr> and there is not real official release of glibc for nearly 2 years or so anyway ...
04:15 < rxr> well - glibc is ugly like hell - the real bug could be anywhere
04:16 < Postal> hmm
04:16 < Postal> couldent you backtrace it to find the bug
04:16 < rxr> and not only that glibc is ugly - no - the annoying Makefile system make is so damn hard to get to the bug
04:16 < rxr> I once wasted over an hour to hack thru the Makefiles just to get to the bug I wanted to catch on sparc64 ...
04:16 < Postal> ic... what version would you revert to..?
04:16 < rxr> 2004-08-xy IIRC ...
04:16 < rxr> the one we had before my xmas update ... ;-)
04:17 < Postal> i mean a version number like 2.3.4...
04:17 < rxr> there is no such version yet
04:17 < rxr> and there was no 2.3.3 as well
04:17 < Postal> it just built is stage 1...
04:17 < Postal> == 21:53:59 =[1]=> Building base/glibc [2.3.4-2004-12-24 2.1.0-beta3].
04:18 < rxr> the tarball the glibc people put on the homepage (half a year after red hat shipped some 2.3.3 on CD) is not near 2.3.3 Red Hat shipped and does not even function that well either ...
04:18 < rxr> we use random CVS snapshots (the ones that work) for over 2 years now
04:18 < rxr> glibc hacking and the release management sucks majorly
04:18 < Postal> ok
04:18 < Postal> ill take you word on that
04:18 < rxr> they made some incompatible ABI change in any 2.3 minor release AFAICR
04:19 < rxr> and so on and so on ...
04:19 < Postal> ABI=?
04:19 < Postal> AFAICR=?
04:19 < rxr> when this Ullrich Drepper (some German employed by RedHat (master of glibc in chief)) is in Germany the next time I shoudl kick him badly ...
04:19 < rxr> Applicatoin Binary Interface
04:19 < rxr> As Far As I Can Remember
04:20 < Postal> ok
04:21 < rxr> ABI is s.th. like an API
04:21 < Postal> ok
04:21 < Postal> your bloody kernel takes so long to build
04:23 < rxr> ;-)
04:23 < rxr> thats a cool kernel
04:23 < Postal> why would a package build in stages 1-4 but fail in stage 5
04:23 < rxr> I do not rember the last time I missed a module ..
04:23 < rxr> when I invested into the USB adapter for my (ugly Siemens) handy I asked in the shop:
04:24 < rxr> "Well I only buy this adaper when it works flawlessly with my Unix workstations"
04:24 < rxr> "hm - well - you can not bring it back when you opened it"
04:24 < rxr> "Well - I have one portable with me - could I quickly plug it in"
04:24 < rxr> "hm - grummel - well - ok"
04:24 * rxr placing the iBook on the desk"
04:24 < rxr> "Hey - thats an iBook - it does not suport OS X"
04:25 < Postal> lol
04:25 < rxr> "Well - who needs OS X ... - There is my own Linux distribution on it ..."
04:25 < Postal> haha
04:25 * rxr plugging in the tiny USB dongle
04:25 < rxr> /dev/usb/ttyUSB0 (or so) showed up
04:25 < rxr> "Ok - I take it ..."
04:25 < rxr> "What - that was the test?"
04:25 < Postal> ROTFL
04:26 < rxr> "Sure - it is recognized - standard USB serial chip"
04:26 < rxr> "ok - ehrm - well - ..."
04:26 < Postal> I think you prolly confused the poor guy
04:27 < rxr> yeah - I hope he remebers Linux guy as companionable people ...
04:28 < rxr> oh - osx ships rsync - nice ...
04:30 < Postal> lol
04:31 < Postal> im telling you dude, you a linux genus
04:31 < Postal> your*
04:31 < rxr> http://www.exactcode.de/about.html
04:32 < rxr> but it does not include all yet ...
04:32 < Postal> Do you want me to check the grammer lol?
04:32 < rxr> is it that bad ?
04:33 < rxr> (I think I also did not yet run a spell checker over it ...)
04:33 < Postal> There are some tense dissagrements
04:33 < rxr> ouhm ...
04:33 < rxr> if you have free time you could send a mail with corrections or so ;-)
04:33 < rxr> but only if you have nothin to do and so ...
04:34 < Postal> and there are some things that whilst they are correct, you just would not say...
04:34 < Postal> yeah, ill do it tomorrow when i have time..
04:34 < rxr> yeah - delay it - not vital ...
04:35 < Postal> you play guitar?
04:35 < Postal> I also play guitar
04:36 < rxr> yes - but these days I have so less really free time that I only manage to practice it quite seldom
04:36 < Postal> I just bought a new Fender Accoustic Electric
04:36 < Postal> Candy Apple Red
04:37 < Postal> ;-)
04:37 < rxr> I have a accustic one - and some very strange Ibanez one (somem custom made thing I got very cheap - which also already got repainted)
04:38 < rxr> s/some/a/ - just one ;-)
04:38 < rxr> and a Jackson 5-string bass
04:43 < Postal> ;-)
04:55 < Postal> i think stage 1 is almost done
04:57 < Postal> YEAY stage 2
05:05 < Postal> what is GSMP?
05:08 < rxr> General Sound Manipulation Program/Platform
05:08 < rxr> a nice pet project I lack time for ...
05:09 < Postal> ic
05:09 < Postal> Do you work on computers for a living, or just for fun
05:10 < rxr> living ... of course
05:10 < rxr> in theory I study it (just to get the paper)
05:10 < Postal> were in stage 3
05:10 < rxr> but in practice I'm more working on all the projects I'm involved in - including founding a company (ExactCode)
05:11 < rxr> if I would not do it for a living I would not be in OS X right now and work on some to be licensed Scanner driver ...
05:11 < Postal> you and Valentin founded the company correct?
05:11 < Postal> lol
05:11 < rxr> (it is more like we are in the progress of doing so), but yes
05:11 < Postal> that sounds exiting
05:12 < Postal> exciting*
05:13 < rxr> well - I do computer stuff since I'm - well - 10 or so ..
05:13 < rxr> it is not that exciting anymore ...
05:13 < Postal> bin/bash: error while loading shared libraries: libncurses.so.5: cannot open shared object file: No such file or directory
05:13 < Postal> !> installing ../lib/libncurses.so.5.4 as //lib/libncurses. ..
05:13 < Postal> !> /TOOLCHAIN/build/system-2.1.0-beta3-x86-athlon64-64-gene ..
05:13 < Postal> !> /bin/bash: error while loading shared libraries: libncur ..
05:13 < Postal> !> make[1]: *** [install] Error 127
05:13 < rxr> more quite hard to get into the marked these days ...
05:13 < Postal> !> make[1]: Leaving directory `/TOOLCHAIN/src.ncurses.11059 ..
05:13 < Postal> !> make: *** [install] Error 2
05:13 < Postal> !> No 3-ncurses.log file.
05:13 < Postal> -> $root/var/adm/logs/3-ncurses.out -> 3-ncurses.err
05:13 < Postal> /bin/bash: error while loading shared libraries: libncurses.so.5: cannot open shared object file: No such file or directory
05:13 < rxr> well - this damn bug ...
05:13 < Postal> == 01/16/05 23:12:27 =[3]=> Aborted building package ncurses.
05:13 < rxr> how I hate it ..
05:13 < rxr> just do:
05:14 < Postal> it was fixed yesterday
05:14 < rxr> ./scripts/Build-Target -cfg system -job 1-bash
05:14 < rxr> and it should work again ...
05:14 < rxr> yes - it seems only halfly fixed
05:14 < rxr> I need to review it in more detail today (after sleeep)
05:14 < Postal> remember we have to symlink the libs before it will chroot into stage 5
05:14 < rxr> the lib64 ?
05:14 < Postal> yeah
05:15 < rxr> that is already done - otherwise you could not have entered stage2 already ...
05:15 < rxr> long committed ;-)
05:15 < Postal> oh ok
05:15 < Postal> i know we did it yesterday
05:16 < Postal> ok bash is built, now i will continue with the system build?
05:17 < rxr> yep
05:17 < rxr> I wonder when my bakery opens
05:17 < rxr> and if I could get fresh rolls before I go into bed ... ;-)
05:17 * rxr hungry ...
05:17 < Postal> lol
05:18 < rxr> thats not funny ... ;-)
05:18 < Postal> i think once this finishes tonight, i will leave it until i come home thursday and prehaps you will have fixed all the errors by then
05:18 < rxr> yeah - maybe ;-)
05:19 < Postal> prehaps, i should have done a generic build, so that you could have a complete list of the broken packages is x86_64
05:21 < rxr> well - you want some base system ..
05:21 < rxr> and mipe is now already building a generic one I think ...
05:21 < rxr> you do not need to do all the dirty work *g*
05:21 < Postal> lol
05:21 < Postal> im not, you are lol
05:21 < Postal> your the one who actually fixes the problem
05:22 < Postal> i wouldent know how to fix it if it kicked me in the a**
05:23 < rxr> when amd128 comes out, you do the work, ;-)
05:23 < Postal> ok
05:24 < Postal> honestly, the fact that you can look at those compiler messages and just know how to fix the problem is awsome
05:24 < rxr> well - I do T2 like work since years ...
05:25 < rxr> that is just "normal" then ...
05:26 < Postal> what time does the bakery open?
05:26 < rxr> no idea
05:26 < Postal> lol
05:26 < rxr> it is 5:26 now ...
05:26 < Postal> i know
05:26 < rxr> I never was that early up to get to the bakery ...
05:26 < rxr> normally I sleep at this time ...
05:26 < Postal> prolly not till at least 6:00am
05:26 < rxr> I think so too
05:27 < Postal> this build is almost done
05:28 < Postal> do you think that mipe will commit fixes for packages like glibc and whatnot that he finds along the way?
05:28 < rxr> well - IIRC mipe has not yet write access ..
05:29 < rxr> I guess he does some stuff - but sends them to the mailing list or so and we apply them
05:29 < rxr> the normal open source procedure ..
05:29 < Postal> right
05:29 < rxr> I wanted to add s.th. to the wiki - but forgot what - damn ...
05:29 < Postal> you have to be very careful about who you give commits to
05:30 < rxr> http://www.t2-project.org/svn-access.html
05:30 < Postal> what does wiki stand for
05:30 < rxr> oehm - no idea
05:30 < rxr> it some commonly used term - not from use
05:30 < Postal> i read that eairler
05:30 < rxr> maybe due to some apache wiki
05:30 < rxr> to come together or so
05:31 < rxr> http://en.wikipedia.org/wiki/Wiki
05:32 < Postal> cool
05:33 < rxr> you now our wiki @ http://wiki.t2-project.org/ ?
05:34 < Postal> i do now
05:35 < Postal> what are Bread Crumbs?
05:35 < rxr> ;-)
05:37 < Postal> I just registerd
05:38 < Postal> stage 5 now
05:39 -!- mnemoc [~amery@200.75.27.34] has joined #t2
05:39 < rxr> ,-)
05:39 < rxr> hi mnemoc !
05:39 < rxr> mnemoc: is that you ?
05:39 < Postal> hi mnemoc!!
05:43 < Postal> the wiki site is very nice
05:43 < Postal> i think i shall visit more often
05:43 < rxr> our wiki is quite new - feel free to add stuff ;-)
05:43 < Postal> like what lol
05:44 < rxr> I add a happy users section *g*
05:44 < Postal> lol
05:45 < Postal> there you are i added somting
05:46 < rxr> URL ?
05:46 < Postal> http://wiki.t2-project.org/index.cgi?WhyT2Rules
05:47 < rxr> ah ;-)
05:47 < rxr> ;-)
05:47 < rxr> hey - a new PR slogan:
05:48 < rxr> "T2 where bug fix cycles are minutes - and not years as in Redmond"
05:48 < Postal> if you had not answered my question when i first came in here, i prolly would have just written it off as another underdeveloped distro
05:49 < Postal> How many users would you say you have?
05:50 < rxr> well - we just craeted T2 some months ago - previous most of us have been working on ROCK Linux
05:50 < rxr> ROCK Linux exists since 1998 and has quite some users
05:50 < Postal> why did you fork?
05:50 < rxr> however many of us where not pelased with the development, release schedule, and so on ...
05:50 < Postal> if i may ask
05:51 < rxr> well - that is documented at so many places ...
05:51 -!- mnemoc_ [~amery@200.75.27.47] has quit [Read error: 110 (Connection timed out)]
05:52 < Postal> i really like the fact that with T2, I actually got to watch it evolve right in front of me
05:52 < rxr> well - I do not find good texts why we forked
05:52 < rxr> http://svn.exactcode.de/t2/trunk/NEWS
05:52 < rxr> http://svn.exactcode.de/t2/trunk/NEWS
05:52 < rxr> err
05:52 < rxr> http://svn.exactcode.de/t2/trunk/README
05:53 < rxr> but that is far not all ..
05:53 < rxr> the biggest ones are:
05:53 < rxr> ROCK is pretty closed - nearly only one may add code - clifford wold himself ...
05:53 < rxr> this is slow, error prone - and he pretty much dictates what happens ...
05:54 < rxr> another is instability
05:54 < rxr> then there is lack of professionalism (e.g. despite of my fair and magazine work it never got to an momentum like Gentoo or so ...
05:55 < rxr> and then there are too many technical issues - I (and most of us) ust do not like how ROCK Linux 2.1 develops ...
05:55 < Postal> what names are you thinking off..?
05:56 < rxr> Postal: _and_ I was the ROCK Linux 2.0 stable release maintainer - the onely one who managed to ever release a release that could have been called /stable/ and the only one ever releasing more than one release in such a release series ...
05:56 < rxr> well - I think t2 is already too much in use - we most probably keep it now
05:56 < Postal> lol
05:57 < rxr> http://www.rocklinux.net/lurker/message/20040730.162344.36d7b512.html
05:57 < rxr> ^- and when I revise it - and look at the last months it was one of my best decisions ever ;-)
05:58 < Postal> cool
05:58 < rxr> this was ~ 6 months ago - and t2 exist shorter
05:59 < Postal> so im really getting in on the ground floor as they say
06:00 < rxr> yes
06:00 -!- praenti [~praenti@82.139.198.99] has quit [Read error: 104 (Connection reset by peer)]
06:00 -!- praenti [~praenti@82.139.198.99] has joined #t2
06:02 < Postal> http://www.rocklinux.org/98+M5837bab91b5.html
06:03 < rxr> yep
06:04 < Postal> T2 should go
06:05 < rxr> hm?
06:05 < Postal> to the next confrence
06:05 < rxr> well sure
06:05 < rxr> we already have been at the Linux World Expo
06:05 < rxr> and hopefully CeBit this year
06:06 < rxr> btw - this camp you linked to was no conference - but some chaor hacker meeing ;-)
06:06 < rxr> meeting even
06:06 < _Ragnar_> hi
06:06 < Postal> chaor..?
06:06 < _Ragnar_> anyone know how to build a 64bit system on a 32bit host?
06:07 < Postal> i do in theory...
06:07 < rxr> rxr: well - cross compile it ;-)
06:07 < rxr> _Ragnar_: like using T2 for this ...
06:07 < rxr> just hit the "this is a cross build button"
06:07 < Postal> rxr, its really picky about the version numbers
06:07 < _Ragnar_> this seems to not work
06:07 < rxr> we just made sure x86-64 does buildd using T2 now ...
06:08 < rxr> _Ragnar_: I just did this 2 days ago or so
06:08 < _Ragnar_> k
06:08 < Postal> its still in stage 5
06:08 < _Ragnar_> that's why I asked, I want to exclude user error
06:09 < Postal> dirtree built by the way
06:09 < rxr> Postal: http://gsmp.tfh-berlin.de/rene/photos/lxexpo-2004/
06:09 < Postal> Building. Writing output to $root/var/adm/logs/5-glibc.out
06:09 < Postal> !> CPP='gcc -E -x c-header' /TOOLCHAIN/src.glibc.110593651 ..
06:09 < Postal> !> make[2]: *** [/TOOLCHAIN/src.glibc.1105936514.14686.0x7f ..
06:09 < Postal> !> make[2]: Leaving directory `/TOOLCHAIN/src.glibc.1105936 ..
06:09 < Postal> !> make[1]: *** [sunrpc/others] Error 2
06:09 < Postal> !> make[1]: Leaving directory `/TOOLCHAIN/src.glibc.1105936 ..
06:09 < Postal> !> make: *** [all] Error 2
06:09 < rxr> Postal: valentin and me ...
06:10 < rxr> http://gsmp.tfh-berlin.de/rene/photos/cebit-2004/gallery.html
06:10 < rxr> ^- doig CeBit that time presenting ROCK
06:10 < rxr> Postal: yep - glibc is on my list ....
06:11 < Postal> i know
06:11 < Postal> in the wiki
06:12 < Postal> i belive it says "hämmern glibc"
06:12 < Postal> ;-)
06:16 < Postal> does that even make sense?
06:17 < rxr> to some degree ...
06:17 < Postal> I was trying to say.. beat glibc with a hammer
06:19 < Postal> como se dice
06:20 < rxr> yes I understood ;-)
06:20 < rxr> how long do you paln to stay awake ?
06:21 < Postal> about as long as u
06:21 < Postal> i diddnt know you spoke spanish
06:21 < rxr> that is? Be in bed since 6 hours ?
06:21 < Postal> ..?
06:22 < rxr> it is 6:22 over here ...
06:22 < Postal> i know
06:24 < rxr> so - do I try the bakery - or the straight way into bed ....
06:24 < Postal> you could call the bakery and see if its open
06:25 < rxr> getting that number might be harder / take longer than just getting there ... - it's berlin here - so that's just to the next crossing
06:26 < Postal> well then i suppose that it depends on how hungry you are
06:28 < rxr> NVU 0.70
06:28 < rxr> ^- package anyone ?
06:28 < Postal> ???
06:29 < rxr> mnemoc: Xfce 4.2.0 Released
06:29 < Postal> well are you going to the bakery?
06:29 < rxr> no - to bed ...
06:30 < Postal> ok
06:30 < Postal> ok im going to bed, cause im sick lol
06:31 < Postal> ill see you prehaps in the morning, if not ill see you on thursday after my exams
06:32 -!- Postal is now known as Postal|Sleeping
06:32 < rxr> much success!
06:32 < rxr> cu Postal|Sleeping
06:33 < rxr> http://www.xiaprojects.com/www/prodotti/kxdocker/main.php
06:37 < CIA-10> rene * r5616 /trunk/package/textproc/libxml/libxml.desc: * updated libxml (2.6.15 -> 2.6.17)
06:38 < rxr> http://svn.exactcode.de/big.png
06:38 < rxr> ^- how I like that graph ...
06:39 < rxr> ok - I'm off into bed ...
06:39 < rxr> (at least I think so ...)
06:39 < rxr> or drive I first the bank and bakery ..
06:40 < rxr> oh - life is so difficult ...
07:06 < rxr> re
07:06 < rxr> damn - just cycled to the bank for nothing
07:07 < rxr> the money transfer (paper scanner) box was stuck ...
07:07 < rxr> damn
07:07 < rxr> at least the bakery was open *g*
07:19 < _Ragnar_> hmm for me even building dietlibc fails when trying to crossbuild for 64
07:21 < rxr> you are at t2/trunk:HEAD ?
07:22 < rxr> e.g. dietlibc was recently fixed
07:22 < _Ragnar_> I assume so
07:22 < _Ragnar_> hmm is that something bad: svn: Working copy 'source/utility/tests' not locked
07:23 < rxr> well
07:23 < rxr> cd source
07:23 < rxr> svn cleanup
07:23 < rxr> cd ..
07:23 < rxr> svn up
07:24 < _Ragnar_> okay:)
07:24 < _Ragnar_> still the same error
07:24 < rxr> cd srouce
07:24 < rxr> oops
07:24 < rxr> cd source
07:24 < rxr> rm -rf utility
07:24 < rxr> cd ..
07:24 < rxr> svn up
07:24 < rxr> ;-)
07:24 < _Ragnar_> k;)
07:25 < _Ragnar_> revision 5616. <--
07:26 < rxr> ok
07:26 < rxr> I have another idea why dietlibc fails
07:26 < rxr> what is your host system ?
07:27 < rxr> is dietlibc failing linking due to some __guard and so on stack smashing protector symbols ?
07:27 < _Ragnar_> yes exactly
07:27 < _Ragnar_> (even tho I have stack protector disabled in the config)
07:28 < rxr> ok - I'm curious - what is the host? ROCK or T2 ?
07:28 < _Ragnar_> rock
07:28 < rxr> hehe ,-)
07:28 < _Ragnar_> *g*
07:28 < rxr> ok - I wanted to modify the dietlibc build anyway - so the defect compiler should not matter ...
07:28 < rxr> but I wanted to go into bed ...
07:29 < _Ragnar_> no problem, I've got time
07:29 < rxr> could you just touch build/your-config-*/var/adm/log/1-dietlibc.log
07:29 < rxr> (or the one that is failing - you got the idea)
07:29 < rxr> and continue the build?
07:29 < rxr> e.g. dietlibc is not really vital for normal (e.g. non install-disk or dietlibc target) builds ...
07:30 < _Ragnar_> oh ... so should I build normal instead of rescue?
07:30 < rxr> you mean bootdisk?
07:31 < _Ragnar_> yeah
07:31 < _Ragnar_> btw: for gzip: checking for C compiler default output... configure: error: C compiler cannot create executables
07:31 < rxr> yeah - better boot some generic or so (optionally with package selection)
07:32 < rxr> just dietlibc cross-builds will fail fur you with your compiler ...
07:32 < rxr> s/boot/build/ ;-)
07:32 < _Ragnar_> okay
07:32 < rxr> if you want to get past dietlibc without tricks, you can:
07:32 < rxr> a) disable the stop on build error in the Config
07:33 < _Ragnar_> well for generic, I should be able to just disable dietlibc, correct?
07:33 < rxr> b) rebuild your system compiler - when you already have 3.4.c you can just Build-Pkg gcc into your running system using the T2 one ...
07:33 < rxr> _Ragnar_: ah - yes - or just disable dietlibc ;-)
07:33 < _Ragnar_> =)
07:34 < _Ragnar_> what did clifford do wrong with the rock gcc?:)
07:35 < rxr> someone injectd a defect SPP patch ;-)
07:35 < rxr> we had it in T2, too - but only for a few days ...
07:36 < _Ragnar_> *lol*
07:36 < rxr> but I plan to improve dietlibc to not stress the host compiler that much ...
07:36 < rxr> (e.g. not build a complere host dietlibc for cross builds ...)
07:37 < _Ragnar_> .oO(?) ... only half of the lib?:)
07:37 < rxr> no - not half
07:37 < rxr> only one
07:38 < rxr> currently dietlibc is build twice in this case
07:38 < _Ragnar_> oh
07:38 < _Ragnar_> okay that's kinda pointless
07:38 < rxr> just for the tiny diet wrapper, that of course is only a normal tiny C program that could also be linked with the system libc ...
07:39 < _Ragnar_> ah okay
07:42 < _Ragnar_> well thanks for the help :) I'll let you rest now :)
07:44 < rxr> thanks ;-)
07:44 < rxr> cu later
07:44 < rxr> have a nice day you all ;-)
07:50 < _Ragnar_> thanks :) goodnight :)
08:07 -!- mipe [~mipe@dsl10040.japo.fi] has joined #t2
12:36 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: sparc-kly
12:40 -!- Netsplit over, joins: sparc-kly
13:38 -!- sparc-kly [~sparc-kly@64.237.130.154] has quit ["Leaving"]
13:40 -!- sparc-kly [~sparc-kly@65-23-204-29.prtc.net] has joined #t2
13:43 -!- sparc-kly [~sparc-kly@65-23-204-29.prtc.net] has quit [Client Quit]
13:55 < rxr> re
14:09 < rxr> ok - /me quickly off to the bank again - and then continueing the work ...
14:43 < mipe> hi rxr
14:43 < mipe> rxr:gcc fails with multilib enabled on stage 2
14:43 * mipe food
14:45 < rxr> hi mipe
14:59 < mnemoc> moin
15:00 < rxr> hi mnemoc !
15:00 < rxr> how are you?
15:01 < mnemoc> with an amazing headache, but ok, and you?
15:01 < mnemoc> glibc-manual?
15:04 < rxr> yeah - the usual "does not build in stage 5 thing"
15:04 < rxr> why headache ?
15:04 < mnemoc> stress mainly
15:04 < rxr> .oO
15:04 < rxr> poor mnemoc
15:07 < mnemoc> did you read what Postal wrote on wiki?
15:07 < rxr> if you mean his one entry, then yes
15:07 < rxr> or has he dome more?
15:07 < mnemoc> no, only one
15:08 -!- ideal [~idealm@211.100.227.107] has quit ["Leaving"]
15:09 < rxr> do you like it ?
15:09 < mnemoc> it's cool to see someone else writing
15:10 < rxr> I also injected more TODO ...
15:10 < mnemoc> i'm diffing it :\
15:11 < rxr> why diffing? to track the changes ?
15:11 < mnemoc> yep
15:11 < rxr> manually y?
15:11 < mnemoc> no, there is an option
15:11 < mnemoc> not very usable, but it is there
15:11 < mnemoc> click on revisions (arrow) and then difference
15:12 < mnemoc> click the arrow twice and difference again
15:12 < mnemoc> click the arrow twice and difference again
15:12 < mnemoc> ...
15:14 < rxr> ah
15:14 < rxr> btw. is it already running using svn?
15:14 < mnemoc> yep
15:14 < mnemoc> but without message
15:14 < mnemoc> .oO( i should have used svn diff )o
15:15 < valentin> mnemoc: does the wiki support commands to insert "plain ascii"-like verbose text ?
15:15 < mnemoc> it has a modulable parser
15:15 < valentin> so one can insert fragments of package files into it - as some kind of tutorial
15:15 < mnemoc> it has .shell blah blah .shell for esxample
15:16 < valentin> i wrote a ascii to html convertor in sed
15:16 < mnemoc> kwiki is designed to use perl plugins for such stuff
15:16 < valentin> perl is ok, too
15:17 < mnemoc> that's why i choose kwiki :)
15:18 < mnemoc> it has .irc .... .irc too ;)
15:18 < valentin> ok
15:18 < mnemoc> a .package may be nice
15:19 < mnemoc> look at cpan for Kwiki::*, there are lot of nice samples
15:19 < mnemoc> bye, indented lines are block-ed and not parsed
15:19 < mnemoc> (two spaces)
15:20 < valentin> there already is some shell code mode in it
15:20 < mnemoc> .shell
15:20 < valentin> (at least we have some in the wiki content)
15:20 < valentin> we could use that one
15:20 < mnemoc> it's very very basic but works
15:20 < valentin> yes
15:22 < mnemoc> fill free to improve it (Kwiki::ShellBlocks) i can apply patches on my local copy
15:38 < mipe> /usr/x86_64-unknown-linux-gnu/bin/ld: skipping incompatible /usr/x86_64-unknown-
15:38 < mipe> linux-gnu/lib/libc.so when searching for -lc
15:38 < mipe> huoh
15:49 < rxr> mipe: can you try to rebuild using this patch:
15:50 * rxr creating patch
15:50 < rxr> |
15:50 < rxr> /
15:51 < rxr> _
15:51 < rxr> |
15:51 < mnemoc> progress looping char?
15:51 < rxr> \
15:52 < rxr> mnemoc: yep
15:52 < rxr> iBook took some time under OS X to fetch t2 HEAD ... ;-)
15:53 < rxr> --- 00-dirtree.conf (revision 5616)
15:53 < rxr> +++ 00-dirtree.conf (working copy)
15:53 < rxr> @@ -50,8 +50,8 @@
15:53 < rxr> # TODO: research this and really sophisticated multilib
15:53 < rxr> # support -ReneR
15:53 < rxr> if [ $arch_sizeof_char_p = 8 ] ; then
15:53 < rxr> - echo "l lib64 lib"
15:53 < rxr> - echo "l usr/lib64 lib"
15:53 < rxr> + echo "m lib64"
15:53 < rxr> + echo "m usr/lib64"
15:53 < rxr> fi
15:53 < rxr> )
15:53 < rxr> }
15:53 < rxr> I could send a plain text diff over ...
15:55 < mnemoc> rxr: what will we do with genrpc
15:55 < rxr> hm - on a amd64 x86-64 box there also no lib64 somewhere ...
15:55 < rxr> mnemoc: downdate glibc I think ...
15:55 < rxr> to the last version we used
15:55 < rxr> I somehow think this is flist wrapper related
15:55 < mnemoc> and new-year-edition?
15:55 < rxr> mipe diffed it - no useful changes ...
15:56 < rxr> I could try though ...
15:56 < rxr> mnemoc: this rpcgen is the first program run with the freshly built glibc
15:56 < rxr> so it is not "the rpbgen (or so) per se" but just some bug in the wohle libc ...
15:56 < mnemoc> i'll run a downdated build to see if it's _only_ that
15:57 < rxr> and I think it is not eevn safe to assume that we are allowed to use the old wrapper for the new libc ...
15:57 < rxr> mnemoc: I already have a patch where I do LD_PRELOAD="" make all
15:57 < mnemoc> and that works??
15:57 < rxr> strangely it does break a bit later - not at rpcgen
15:57 < rxr> but calling some iconv or so converted during installation ...
15:57 < mnemoc> but it breask
15:58 < rxr> maybe also wrapper related ...
15:58 < rxr> I could disable for a install just for a test, too ..
15:58 < rxr> just that that would yield an empty flist ...
15:59 < mipe> hmm,whats going on? you want me to test with that patch?
16:00 < rxr> I guess glibc installs a 32bit and 64bit version of some libs itself
16:00 < mnemoc> rxr: find and strace flist detection are broken
16:01 < rxr> which are overwriting each other when you symlink the lib64 dirs to lib for mutlilib bulds
16:01 < rxr> mipe: can you send me the glibc build log?
16:01 < mipe> sure
16:01 < rxr> and the wrapper is not only fragile but unportable ....
16:01 < rxr> we need to do some research in this area ....
16:02 < mnemoc> i haven't been able to make the wrapper work for uclibc
16:03 < mipe> glibc log? failing part is gcc
16:03 < mnemoc> find is fragile because some packages build faster that what find can detect
16:04 < rxr> find is even not possible when you do parallel builds
16:04 < rxr> and I do this often
16:04 < rxr> no only for in-system builds - but also for accelerated development doing Build-Target -job ?-pkg for testing fixes while the main one is building ...
16:04 < mnemoc> strace?
16:05 < rxr> might be quite slow IIRC
16:05 < rxr> needs a benchmark ..
16:05 < rxr> but in case of partability strace style solutions fit best
16:06 < rxr> but it should not be that much slower - e.g. the wrapper is quite fast ... ;-)
16:07 < mnemoc> the only problem i found on strace is that detects creation of every level of $xroot
16:07 < mnemoc> but that should be easy to hack
16:10 < mipe> rxr:email addr to send
16:10 < mipe> ?
16:10 < rxr> reen@exactcode.de
16:10 < mipe> do i send gcc and glibc or just gcc?
16:10 < rxr> rene even
16:10 < rxr> glibc - but gcc could be interesting, too ;-)
16:11 < mnemoc> ren.und.stimpy@exactcde.de :)
16:11 < mipe> :)
16:12 < rxr> valentin: it seems to be a 340MB SCSI 2.5" HDD on the PPC board ...
16:12 < rxr> some part replace ment store where I found the part no. wants Cost: $306.73
16:13 < rxr> for it - *lol*
16:14 < valentin> haha
16:14 < rxr> the gpu is:
16:15 < rxr> CL-GD5434
16:15 < rxr>  
16:15 < rxr> 32bit
16:15 < rxr> 64bit
16:15 < rxr> 2M
16:15 < rxr> DRAM
16:15 < rxr>  
16:15 < rxr>  
16:15 < rxr>  
16:15 < valentin> i have an old 10Gb IDE drive - anyone wants to buy it for $1760 ?
16:15 < rxr> 1994
16:15 < rxr> valentin: I have some old RLL discs and such in Lauenau ... *g*
16:15 < rxr> including many 5 1/4" floppies ... ;-)
16:16 < rxr> some 160kB (or what it was) 5 1/4" floopy anyone?
16:16 < rxr> 1000 EUR and up ...
16:16 < valentin> rxr: did you boot the board ?
16:16 < rxr> no - not yet
16:16 < rxr> I just pasted some part no. into google ...
16:17 < rxr> I have no time to really play with it
16:17 < rxr> one of the boards already has some debian installed - but I do not know which one
16:17 < rxr> maybe yours - since mine does nto feature the scsi cable between the board and the 2.5" SCSI HD ...
16:17 < valentin> i can try tonight, if i have the time
16:18 < rxr> only if you need to relax or so - this scanner driver is quite important to get out ..
16:18 < valentin> ack
16:19 * mnemoc starting a build with glibc 2.3.4-2004-08-01
16:23 < mipe> rxr:did you get the logs?
16:24 < rxr> looks so .-)
16:27 < rxr> can you run file on the build lib/libc.so ?
16:27 < rxr> .6
16:27 < rxr> in build/bla.../
16:27 < mipe> ?
16:28 < rxr> file build/xyz-*/lib/libc.so.6
16:29 < mnemoc> plain text :)
16:29 < rxr> ouhm?
16:29 < mnemoc> jk
16:30 < rxr> in fact I just noticed file will tell symbolic link ...
16:30 < mipe> symbolic link to libc-2.3.4.so
16:30 < rxr> thus you need to use file on libc-2.3.4.so
16:30 < rxr> ;-)
16:30 < mipe> ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), not stripped
16:30 < mnemoc> elfls is nice too :)
16:30 < rxr> hm - then ld might just judge that it is incompatible due to the filename ... ?
16:31 < rxr> strange
16:31 * rxr needs some x86-64 box here ...
16:32 < rxr> I think when you use the above patch (really creating lib64) you also need to patch Build-Pkg to pass lib64 to all normal package builds ...
16:32 < rxr> :-(
16:33 < rxr> mipe: you can not actually grant root access to that box ? ;-)
16:33 * rxr booting into linux
16:33 * mnemoc just noticed beta3 was scheduled for yesterday :\
16:33 < mipe> yeah,i can
16:35 < mnemoc> rxr: did you read today's thread between cliff and blindy?
16:35 -!- Postal|Sleeping is now known as Postal
16:35 < Postal> Hey rxr gcc complied along with dirtree
16:36 < mnemoc> with lib64 as a real dir instead of a symlink?
16:37 < rxr> re
16:37 < rxr> no
16:37 < rxr> I only read the rock list in the web interface every 3-4 weeks when I need s.th. to laught
16:38 < mnemoc> that thread is funny :)
16:38 < Postal> make[3]: Leaving directory `/TOOLCHAIN/src.zsh.110597624 ..
16:38 < Postal> !> make[2]: *** [modules] Error 1
16:38 < Postal> !> make[2]: Leaving directory `/TOOLCHAIN/src.zsh.110597624 ..
16:38 < Postal> !> make[1]: *** [modules] Error 2
16:38 < Postal> !> make[1]: Leaving directory `/TOOLCHAIN/src.zsh.110597624 ..
16:38 < Postal> !> make: *** [all] Error 1
16:39 < mipe> rxr:need that access to box?
16:39 < rxr> ah - my beloved konqueror ;-)
16:39 < mnemoc> Postal: from src.zsh.*/ERROR_LOG, what there is before the lines you pasted?
16:39 < rxr> mipe: I could investigate such build details ...
16:39 < Postal> mnemoc, how many lines you want ill tail it
16:40 < mipe> rxr:i'll email you the password and stuff
16:41 < mnemoc> Postal: can you publish it?
16:41 < Postal> mnemoc, i dont have an ftp server...
16:41 < mnemoc> http?
16:41 < Postal> nope
16:41 < Postal> i could email it..
16:41 < mnemoc> amery@geeks.cl
16:42 < mipe> rxr:mail sent
16:42 * mipe off to shower
16:45 < Postal> mnemoc, mail sent
16:45 < mnemoc> gcc -s -shared -o complist.so complist..o -ldl -lnsl -ltermcap -lm -lc
16:45 < mnemoc> /usr/lib/gcc/x86_64-unknown-linux-gnu/3.4.3/../../../../x86_64-unknown-linux-gnu/bin/ld: /usr/lib/gcc/x86_64-unknown-linux-gnu/3.4.3/../../../../lib64/libtermcap.a(lib_tputs.o): relocation R_X86_64_32S against `_nc_outch' can not be used when making a shared object; recompile with -fPIC
16:45 < mnemoc> /usr/lib/gcc/x86_64-unknown-linux-gnu/3.4.3/../../../../lib64/libtermcap.a: could not read symbols: Bad value
16:46 < rxr> mnemoc: I already got this
16:46 < mnemoc> how will we solve this massive -fPIC issue?
16:46 < rxr> it is not massive
16:47 < mnemoc> according to cliff it is
16:47 < rxr> I think injecting it generically is a bad idea
16:47 < mnemoc> me too
16:47 < rxr> it will break a lot of stuff here and there silently ...
16:47 < mnemoc> GCC_WRAPPER_APPEND on packages which need it?
16:47 < rxr> well - like that
16:48 < rxr> in the long term the packages will be fixed anyway ...
16:48 < rxr> and so far it only have been 7 patches or so ..
16:48 < rxr> and the situation is not new - is is already the case for sparc64 and PowrPC
16:48 < rxr> PowerPC even
16:49 < rxr> for example I patched nas or so last Octover due to symbol resolution issues on PowerPC
16:49 < rxr> and we patched even more for sparc64 over the years
16:49 < rxr> all of the patches are of course no longer there today since the packages got fixed
16:50 < rxr> adding -fPIC for everything is as if you only implement PIC in the compiler and doing nothing else - so the choice should be left ...
16:50 < rxr> (aside from other technical details)
16:50 < mnemoc> yes
16:52 < rxr> mipe: is that IP permanent?
16:53 < mipe> ÿeah
16:54 -!- mipe is now known as mipe|AFK
17:00 < rxr> mipe|AFK: hm - your sandbox looks good
17:00 < rxr> it can generate 64bit binaries - just no 32 bit ones - gcc output is accurate and reasonable ...
17:15 -!- mipe|AFK is now known as mipe
17:15 < rxr> ouhm
17:15 < rxr> when I see the biarch mess in gcc and friend my somarch turns over and I wish I would be back to my scanner driver ...
17:16 < mipe> did you find a reason for build fail?
17:16 < mnemoc> :)
17:16 < rxr> yes
17:17 < rxr> gcc indeed builds it's support libraries (libgcc_s and friends) for 64 and 32 bit if multilib is not disabled
17:17 < rxr> this is why it (the linker) tries to find a 32bit libc on your system ...
17:17 < mipe> theres no rush for get this working,atleast for me. i can set it to build normal pmmx or other builds to do,until you get your own work done.
17:18 < rxr> mipe: I 'll work on a fix in an hour or so ...
17:18 < rxr> first I need to get lunch and then I need to write some news article ...
17:18 < mipe> ok :)
17:18 < rxr> (my new just for fun job)
17:18 < rxr> and then I try to fix this (hopefully quickly) before I go back to scanner driver tinkering ...
17:20 < mipe> hmm,whats the best option for optimise in config for amd64 cpu?
17:20 < mipe> in 32bit mode
17:20 < mipe> should i just do pmmx build :)
17:23 < rxr> mipe: I guess athlon-xp or so shoudl fix well enough
17:24 < rxr> OT: Should we maybe host a glibc-users list?
17:24 < rxr> Ullrich Drepper has again his ranting mode (as always) and flames away normal questioning users ...
17:25 < mnemoc> do you know how to remove a hunk using sed/awk from a patch?
17:25 < rxr> where questions are like:
17:25 < rxr> - why does libc segfault here and there
17:25 < rxr> - how do I cross build it - it fails any way
17:25 < rxr> and so on ...
17:26 < rxr> mnemoc: I would do a awk loop with state like I do in the Commit.sh in our archive ...
17:27 < mnemoc> :\ i assume that's the only solution you found
17:30 < rxr> I did not search
17:30 < rxr> that is theonly solution I can propose in 10 seconds ..
17:30 * rxr getting foot ...
17:45 < rxr> I think I even patch glibc's dynmiac linker to not use lib64 ...
17:46 < rxr> re btw
17:56 < mnemoc> rxr: awk newbie question, how do i initialize an state (before detecting what i want to detect)?
17:57 < rxr> with BEGIN you can execute before main loop
17:58 < mnemoc> how do i print the whole line as it was?
17:58 < rxr> was it just print
17:59 < rxr> I'm not a awk geek ..
17:59 < rxr> I usual place the sed/awk oreilly book on my desk when I hack on it ..
17:59 < mnemoc> it worked :)
18:05 * mnemoc proud of his new pkg_linux_pre.conf
18:05 < mnemoc> vserver's
18:13 < Postal> mnemoc, did you get my email?
18:13 < rxr> ;-)
18:19 < mnemoc> Postal: yes, an we even discuss it here
18:19 < mnemoc> and*
18:25 < CIA-10> mnemoc * r5617 /trunk/package/security/vserver/ (no_extraversion.diff pkg_linux_pre.conf vserver.desc):
18:25 < CIA-10> * improved vserver to remove extraversion using awk instead of a patch
18:25 < CIA-10> * improved vserver to inject extraversion on linux-s builds
18:25 < CIA-10> * improved vserver to detect the right patch between availables considering $ver
18:27 < Postal> mnemoc, and*...
18:27 < mnemoc> good question
18:27 < mnemoc> some pacakges need -fPIC to build in your arch
18:27 < mnemoc> in your case ncurses
18:28 < Postal> what is -fPIC?
18:28 < mnemoc> a gcc argument
18:28 < Postal> ah ok
18:28 < mnemoc> try adding this to ncurses.conf
18:28 < mnemoc> var_append GCC_WRAPPER_APPEND ' ' -fPIC
18:29 < mnemoc> on last line
18:29 < Postal> where is the .conf file
18:30 < mnemoc> package/*/ncurses
18:31 < mnemoc> and rebuild ncurses of course
18:31 < Postal> what line would i add that to
18:32 < mnemoc> var_append GCC_WRAPPER_APPEND ' ' -fPIC
18:32 < Postal> var_append confopt " " "--with-normal --with-shared"
18:32 < Postal> is what it reads now
18:32 < Postal> so just change it
18:33 < mnemoc> what you changed?
18:34 < Postal> no i havent changed anything, im asking if thats the line i should change
18:34 < rxr> no
18:34 < rxr> just add the new line
18:34 < rxr> e.g. as the last line of the file
18:34 < Postal> ok thats easy
18:34 < rxr> do not touch any of the existing lines
18:36 < Postal> should i schedual the package for a rebuild and then build it then
18:36 < rxr> yes
18:36 < Postal> ok building
18:37 < mnemoc> ncurses?
18:37 < Postal> yup
18:37 < mnemoc> :)
18:37 < Postal> but the log for the file that i sent was for zsh.. so im a little confused..
18:38 < rxr> but the error was most probably caused by a "bug" in ncurses
18:38 < Postal> how do you figure that out..?
18:40 -!- ops [~ops@202.138.54.130] has joined #t2
18:40 < rxr> hi ops
18:40 < Postal> What is ncurses?
18:40 < mnemoc> a library to disable nice in console
18:41 < Postal> and nice is
18:41 < mnemoc> /usr/lib/gcc/x86_64-unknown-linux-gnu/3.4.3/../../../../lib64/libtermcap.a: could not read symbols: Bad value
18:41 < mnemoc> err
18:41 < mnemoc> a library to display nice in console
18:41 < Postal> right but what is nice lol
18:41 < mnemoc> brain and fingers not conected
18:42 < mnemoc> an adjetive?
18:42 < Postal> lol, but in the linux world what is nice..?
18:42 < mnemoc> not as a terminal
18:42 < Postal> ok your confusing me lol
18:43 < mnemoc> not as a printer i mean
18:43 < mnemoc> line by line from bottom
18:44 < mnemoc> termcap and ncurses let you make better use of your console
18:44 < Postal> ok
18:44 < mnemoc> cleaning, saving, moving, changing (X,Y), etc..
18:44 < mnemoc> that is 'nice' :)
18:44 < Postal> haha
18:45 * mnemoc moving home to lunch, bbl
18:45 < mipe> who
18:45 < mipe> ups
18:45 < mipe> nvm that
18:47 < Postal> mipe, how did the multilib build go?
18:47 < mipe> gcc is failing on stage 2,i think rene is looking on that
18:48 < mipe> unless hes doing own work
18:48 < rxr> yep - right now
18:49 < Postal> cant you only set certain packages to multilib?
18:55 < Postal> /usr/X11R6/lib64/libXaw.so: undefined reference to `XpSe ..
18:55 < Postal> !> /usr/X11R6/lib64/libXaw.so: undefined reference to `XpGe ..
18:55 < Postal> !> collect2: ld returned 1 exit status
18:55 < Postal> !> make[1]: *** [vim] Error 1
18:55 < Postal> !> make[1]: Leaving directory `/TOOLCHAIN/src.vim.110598442 ..
18:55 < Postal> !> make: *** [first] Error 2
18:59 < Postal> thats odd, i never got an error on vim the last time around
19:00 < rxr> so far you have never built vim, hm ?
19:00 < rxr> Postal: did zsh built now ?
19:00 < Postal> rxr, yes
19:00 < rxr> cool
19:00 < Postal> vim grub glibc and bdb did not build
19:00 < Postal> couldent we just pass an -ia32 option to grub
19:01 < Postal> and then it should build..?
19:02 < rxr> no - you need a differently built gcc for it
19:02 < CIA-10> rene * r5618 /trunk/package/base/ncurses/ncurses.conf: * fixed ncurses for certain architectures including x86-64
19:02 < rxr> I'm working on it right now ...
19:02 < rxr> Postal: can you send me the vim output ?
19:02 < Postal> i could just chroot in, and build a 32 bit toolchain, ive done that before on and LFS system...
19:02 < Postal> rxr, yeah i can
19:03 < rxr> Postal: we need to really fix this anyway ...
19:04 < Postal> sent
19:05 < Postal> ok, but im about to leave for exams in about 15 min so...
19:05 < rxr> no problem - we fix it in the meantime
19:05 < rxr> I wish you success for your exams
19:07 < Postal> thank you... ill see you thursday about 12:30am your time
19:07 < Postal> i may be on irc during exams but i will be on my 32bit laptop, so i cannot test anything
19:08 < Postal> gtg guys much luck fixing the bugs
19:17 < mipe> rxr:any luck with it?
19:18 < rxr> just committed s.th.
19:18 < rxr> I test it on your box
19:18 < CIA-10> rene * r5619 /trunk/package/base/gcc/biarch-wo-multilib.patch:
19:18 < CIA-10> * attempt to convice to built-in biarch support without full multilib
19:18 < CIA-10> linking
19:18 < rxr> ^- untestd
19:20 < mipe> btw. are those c++ stuff in source the base of new t2 3.x stuff?
19:21 < rxr> no - just random tests ..
19:21 < rxr> no real egeneering work ...
19:26 < CIA-10> rene * r5620 /trunk/package/base/gcc/biarch-wo-multilib.patch: * fixed the new gcc patch to be not reversed - added sparc64 likewise
19:26 < mipe> cd
19:26 < mipe> ups
19:26 < mipe> :)
19:28 < mipe> have to shutdown the server in ~30min,dont know if its up till tomorrow morning.
19:28 < rxr> ouhm
19:28 < rxr> really .. ?
19:28 < rxr> why shutdown?
19:29 < mipe> yeah. server is in my parents house atm. and my dad uses it for work. i'm bringing pc here on my own house sometime next week or so. after i have done some remodelling
19:30 < rxr> I hoe the current gcc build finishes before you shut it down ...
19:31 < mipe> yeah.i'll wait it to build. wont shutdown before it
19:34 -!- Postal [~Postal@ool-43562a7f.dyn.optonline.net] has quit [Read error: 110 (Connection timed out)]
19:38 < mnemoc> i586-unknown-linux-gnu-gcc -isystem include -pipe -nostdinc -Os -fomit-frame-pointer -falign-functions=1 -falign-jumps=1 -falign-loops=1 -mpr
19:38 < mnemoc> eferred-stack-boundary=2 -Wall -W -Wchar-subscripts -Wmissing-prototypes -Wmissing-declarations -Wno-switch -Wno-unused -Wredundant-decls -fP
19:38 < mnemoc> IC -nostdlib -o pic-i386/diet-dyn-i pic-i386/start.o pic-i386/dyn_start.o diet.c -DDIETHOME=\"/usr/dietlibc\" -D__DYN_LIB -DVERSION=\"dietlib
19:38 < mnemoc> c-0.28\" -Lpic-i386 -lc -lgcc pic-i386/dyn_stop.o -Wl,-dynamic-linker=/usr/dietlibc/lib-i386/libdl.so -DINSTALLVERSION
19:38 < mnemoc> pic-i386/start.o(.text+0x19): In function `_start':
19:38 < mnemoc> i386/start.S: undefined reference to `environ'
19:38 < mnemoc> pic-i386/start.o(.text+0x26):i386/start.S: undefined reference to `exit'
19:38 < mnemoc> pic-i386/dyn_start.o(.text+0x17): In function `_dyn_start':
19:38 < mnemoc> this is weird
19:38 < rxr> built
19:38 < rxr> gcc
19:39 < mipe> cool
19:39 < mnemoc> on 1-dietlbc native pentium
19:41 < mnemoc> brb
19:41 < CIA-10> rene * r5621 /trunk/package/base/ncurses/ncurses.conf: * fixed typo ncurses
19:48 < mipe> so its fixed, i can continue building?
20:00 < mipe> ok,rxr have to shutdown. bash thing is just to recompile it?
20:09 < rxr> yep - gcc seems to be "hacked or workarounded" however you want to name it ... ,-)
20:17 < CIA-10> rene * r5622 /trunk/package/x11/xorg/driver-enable.patch:
20:17 < CIA-10> * enabled the X.org cirrus driver for powerpc - builds but otherwise
20:17 < CIA-10> untested
20:25 < daja77> jsaw: ping
20:25 -!- mipe [~mipe@dsl10040.japo.fi] has quit [Read error: 110 (Connection timed out)]
20:41 < mnemoc> rehi
20:43 < rxr> hi mnemoc
20:45 < mnemoc> hi rxr
20:45 < mnemoc> look:
20:45 < mnemoc> make[1]: Entering directory `/TOOLCHAIN/src.ncurses.1105989110.4243.7f0100/ncurses-5.4/ncurses'
20:45 < mnemoc> installing ../lib/libncurses.so.5.4 as //lib/libncurses.so.5.4
20:45 < mnemoc> /TOOLCHAIN/build/minimal-2.1.0-beta3-x86-pentium4-32-mnemosyne-expert/TOOLCHAIN/tools.chroot/wrapper/install -c -m 755 ../lib
20:45 < mnemoc> /bin/bash: error while loading shared libraries: libncurses.so.5: cannot open shared object file: No such file or directory
20:45 < mnemoc> make[1]: *** [install] Error 127
20:45 < mnemoc> 3-ncurses
20:45 < rxr> mnemoc: I saw this too
20:46 < rxr> the first fix attempt was not yet that effective ...
20:46 < rxr> mnemoc: I think we can not migrate to some lib(32) and lib64 layout in the near future
20:46 < rxr> and it seems debian wants to do this neither
20:46 < rxr> I even think this is very unwated
20:46 < mnemoc> multilib to the toilette?
20:47 < rxr> I do not want to waste aprox. 50% additional stuff on my sparc64 systems
20:47 < rxr> or x86-64 or powerpc64 if I ever get those
20:47 < rxr> I just want pure 64bit, no?
20:47 < rxr> I mean - the only beneifit I see would be on x86-64 to run properitary binary only stuff that is only avail in 32bit ...
20:48 < rxr> debians unofficial amd64 port does not seem to deal with lib64 either
20:48 < _Ragnar_> .oO(like ndiswrapper?)
20:48 < mnemoc> can we build a minimal using pure64?
20:48 < rxr> they just patch their toolchain too
20:48 < rxr> yep
20:48 < rxr> but then grub does not build sudo the need to generate 32bit code ...
20:48 < rxr> this is why we tried enabling mutlilib
20:49 < mnemoc> we need to stabilize a 'bit' to get beta3 out
20:49 < rxr> then gcc wants to build 32bit support libaries - where we have no glibc fore ..
20:49 < mnemoc> :\
20:49 < rxr> so I patched this away from gcc for x86-64 and sparc64 some minutes ago
20:49 < rxr> and I think I patch the lib64 ld-linux.so binbehaviour away from glibc ...
20:49 < rxr> (as debian most probably does, too ...)
20:50 < rxr> so we also do not need this ugly lib64 -> lib symlinks ...
20:50 < mnemoc> what G* guys do?
20:50 < rxr> this would be 3 patches as far as I can see (gcc, glibc and x.org I think)
20:50 < rxr> Gentoo ?
20:50 < mnemoc> yep
20:50 < rxr> hey have some strange /emul/linux/x86 layout or so
20:50 < mnemoc> o_O
20:51 < rxr> and link lib32 to -> /emul/linux/x86/route-to-hell
20:51 < rxr> I never saw such a freaky construct ...
20:51 < mnemoc> i don't want to see such an insane construct
20:51 < _Ragnar_> *rofl*
20:51 < rxr> I think we get 64 bit clean first (including a mutlilib compiler to be able to build 32bit grub or 32bit kernel on sparc)
20:51 < mnemoc> ACK
20:52 < rxr> and whoever want's s.th. else can remove this 3 patches and hack up the build system to build *every* package twice - once 32bit and once 64bit
20:52 < daja77> jsaw: thx :)
20:52 < rxr> and deal with all the many packages that would need fixup then
20:52 < mnemoc> hi daja77
20:52 < rxr> as well as dealing what bianry to install
20:53 < rxr> e.g. one would just want to install all the 32bit libraries - but discard the resuting /bin applications
20:53 < rxr> unless one wants bin and bin64 or so ...
20:53 < mnemoc> is 1-dietlibc brokeness due to this too?
20:53 < rxr> _Ragnar_: nope - ndiswrapper is another storry - it is _in_ kernel and will never work on x86-64
20:53 < daja77> mnemoc: yes
20:53 < rxr> unless you hack it up to be able to load 64bit windows drivers ... ;-)
20:53 < rxr> mnemoc: no
20:53 < _Ragnar_> *g*
20:54 < rxr> mnemoc: the dietlibc was already fixed
20:54 < mnemoc> yes or no?
20:54 < rxr> mnemoc: no
20:54 < rxr> mnemoc: dietlibc was configured wrongly
20:54 < mnemoc> ic
20:54 < rxr> t2:HEAD does it right and builds for a week or so now ...
20:54 < mnemoc> # svn up <--- no changes on diet
20:54 < rxr> the only remaining problem with dietlibc is, that it does not build when you have a ROCK-a-like broken system gcc with a defect SPP patch ...
20:54 < _Ragnar_> btw generic still tries to build dietlibc even if it's disabled in the dietlibs menu
20:55 < rxr> mnemoc: long fixed ;-)
20:55 < rxr> ouhm
20:55 < daja77> it does not occur n every optimization
20:55 < _Ragnar_> had to - dietlibc in the pkgsel rules
20:55 < rxr> no idea what's in the diet menu - have you tried the normal pacakge selection menu ?
20:55 < rxr> ah - ok
20:56 < rxr> daja77: it does occur with broken gcc - if you take about the __guard and friends symbols ...
20:56 < rxr> but I will improve dietlibc to not show this error - for other reasons discussed here ...
20:56 < daja77> huh?
20:57 < mnemoc> HOSTCC's libgcc.a want's SPP symbols
20:57 < mnemoc> wants*
20:57 < daja77> have no stack protector stuff enabled
20:58 < rxr> it does not need to be enabled if you inject a broken SPP patch into gcc
20:58 < daja77> ic
20:59 < _Ragnar_> if something fails in stage 1 due to autoconf/automake, this is still the host autoconf/automake, not the build environments?
21:00 < mnemoc> rxr: that's why i consider a pseudo-cross compiler always available a need.
21:00 < mnemoc> rxr: we can't trust host system
21:00 < rxr> yes
21:01 < rxr> but not for 2.1 ...
21:01 < rxr> when we dance the stages and clean all the mess up it is a good time to do this ..
21:01 < rxr> 0: pseudo native pseudo cross
21:01 < rxr> err
21:01 < mnemoc> can't we use BUILDCC=CC on native builds?
21:01 < rxr> 0: native pseudo cross
21:02 < mnemoc> that deserves a wiki page ;)
21:02 < rxr> for cross builds 1: real cross - built with the pseudo cross
21:02 < rxr> what do you mean with: 21:01 < mnemoc> can't we use BUILDCC=CC on native builds?
21:02 < mnemoc> BUILDCC=HOSTCC=cc used to build 'native' stuff
21:03 < mnemoc> CC={pseudo,}cross-compiler
21:04 < rxr> mnemoc: sorry - you re-outline this in more detail - for me to understand it ;-)
21:04 < mnemoc> the problem with SSP is when BUILDCC, defined as cc, is broken
21:04 < mnemoc> cc, in my case is a broken 3.3.3
21:05 < mnemoc> when crossbuilding we just have a cross-compiler
21:05 < mnemoc> but when we do native we have a pseudo-cross we can use instead of host's cc
21:06 < rxr> exactly
21:06 < rxr> the best fix is to just always build a fresh system compiler first
21:06 < rxr> the 0-pseudo cross compiler
21:07 < rxr> that is used for anything then ...
21:07 < rxr> and from cross builds the cross toolchain is build from it ...
21:07 < rxr> but I would not mess with this right now
21:07 < rxr> for normal builds it works - you should just not have a ROCK based gcc ;-)
21:07 < rxr> let's wait past 2.1 for the big restructuring to fix it, no ?
21:08 < mnemoc> i have been unable to build a t2 to replace it ;)
21:08 < rxr> well - native builds are not affected, right ?
21:08 < mnemoc> if you consider it long work, let's wait.
21:08 < mnemoc> rxr: bootdisk fails
21:08 < rxr> I consider it to like break stuff
21:08 < rxr> and we want to get out a stable base out
21:08 < rxr> let's not get into a 2 year 2
21:09 < rxr> 1.7 style endless rework
21:09 < mnemoc> we need 2.1.0 on february
21:09 < rxr> what are your native x86 -> x86 build problems ?
21:09 < rxr> did you grabbed the minimal from my box ?
21:09 < mnemoc> i have been trying hard to
21:10 * mnemoc retrying
21:10 < rxr> trying to grab the *tar*s ?
21:11 < rxr> btw the cross built bash and vi behave strangely ...
21:11 < rxr> at least on 64 bit platforms
21:11 < rxr> I saw this on my sparc64 - and today on the x86-64 ...
21:11 < mnemoc> rxr: yes, my link is very unstable today and i got corrupted files
21:12 < rxr> oh
21:12 < rxr> should I start rsync ?
21:13 < mnemoc> sounds cool :)
21:14 < rxr> hm - strange
21:14 < rxr> I have no rsyncd only rsync
21:14 < mnemoc> what's strage?
21:14 < rxr> rebuilding
21:14 < mnemoc> :)
21:14 < rxr> takes a minute - it's a 370Mhz PPC box
21:16 < mnemoc> you were talking about stages
21:27 < CIA-10> rene * r5623 /trunk/package/base/ (00-dirtree/00-dirtree.conf glibc/biarch-wo-multilib.patch):
21:27 < CIA-10> * removed the lib64 -> lib symlink hack
21:27 < CIA-10> * patched glibc to not try to use /lib64 accordingly (untested)
21:27 < rxr> ehrm
21:27 < rxr> mnemoc: ?
21:30 < rxr> oh - there is no rsyncd - why doesn't anyone tell me ...
21:30 < mnemoc> --daemon?
21:35 < rxr> yeah
21:39 < jsaw> re
21:39 < jsaw> daja77: ?
21:40 < daja77> your dietlibc fix for p2/3 optimizations
21:41 < CIA-10> rene * r5624 /trunk/package/network/netkit-base/inetd.conf: * fixed the exampel inetd.conf to list rsync --daemon - not rsyncd
21:41 < jsaw> <- meantime before forgetting about 2weeks
21:41 < jsaw> daja77: when have I done that?
21:41 < jsaw> daja77: anyway, you're welcome
21:41 < daja77> 12-2004 sometimes
21:42 < jsaw> :)
21:42 < jsaw> ah, the __i686 define...
21:43 < jsaw> resp. undef...
21:43 < jsaw> but fefe obviously didn't get it...
21:44 < daja77> send it :)
21:44 < jsaw> but flames seem to reach him more easily...
21:44 < mnemoc> jsaw: hi, -nostdlib didn't work to make fl_wrapper run for uclibc :\
21:45 < jsaw> daja77: __i686 defined while optimizing for pentiumII, Date: 12/22/04 21:48, To: dietlibc@fefe.de
21:45 < jsaw> mnemoc: :(
21:45 < mnemoc> jsaw: :(
21:45 < daja77> jsaw: ic
21:47 < jsaw> daja77: oh, btw. I meant "didn't get it" in the sense of "did not understand"
21:47 < daja77> hehe :)
21:48 < jsaw> mnemoc: the uclibc thing is strange. Really strange. Some programs are bound to libc.so.6, which they shoudn't be anymore in stage 2...
21:48 < mnemoc> to libc.so.6??
21:49 < jsaw> yep. Alarm bells were ringing. But I stupidly didn't write down how it happend.
21:50 < jsaw> daja77: did praenti take the new bash patch?
21:50 < daja77> yep
21:50 < jsaw> good.
21:53 < rxr> so
21:54 < rxr> mnemoc: have you started your glibc downdated build ?
21:54 < mnemoc> what is that 'so' about?
21:55 < mnemoc> yes, but one failed when ncurses killed it's .so and the other on 1-dietlibc
21:55 < rxr> well - the one with ncurses is known
21:55 < rxr> I can work on it right now
21:56 < rxr> the ncurses will currently fail on any build
21:58 < mnemoc> install-wrapper?
21:59 < rxr> no generic bug
22:03 < jsaw> is it really true that -fPIC does not affect executables?
22:04 < mnemoc> great question
22:05 < jsaw> I mean, these are ELFs... sounds like they are handled similiar to libraries?
22:13 < CIA-10> mnemoc * r5625 /trunk/package/security/vserver/pkg_linux_pre.conf:
22:13 < CIA-10> * improved vserver to detect patches with extraversion (e.g. -rc2)
22:13 < CIA-10> * fixed vserver to only search for vquota patch when $ver == 2.4.*
22:14 < CIA-10> mnemoc * r5626 /trunk/package/security/vserver/vserver.desc: * updated vserver to 1.94-rc2
22:15 < CIA-10> rene * r5627 /trunk/package/x11/xorg/biarch-wo-multilib.patch: * fixed x.org to not use lib64 on some architectures
22:19 < jsaw> rxr: am I free to apply the non-dist. bin. packages thing?
22:24 < rxr> yeah - do so ... ;-)
22:24 < mnemoc> :)
22:24 < jsaw> oki.. :)
22:27 < CIA-10> jsaw * r5628 /trunk/Documentation/Developers/LICENSE-WARNING:
22:27 < CIA-10> * add LICENSE-WARNING text to be copied on resp. ISOs
22:27 < CIA-10> PLEASE REVIEW!
22:28 < CIA-10> jsaw * r5629 /trunk/Documentation/Developers/REGISTER: * add "Restricted" to the license tags
22:29 < CIA-10> jsaw * r5630 /trunk/scripts/config.func: * add "pkgchecklicense" function to config.func
22:33 < CIA-10> mnemoc * r5631 /trunk/package/mail/qmail/ (6 files): * added qmail (1.03) package as "restricted" honoring (i hope all) author's restrictions
22:38 < CIA-10> jsaw * r5632 /trunk/scripts/ (config.hlp config.in):
22:38 < CIA-10> * add checks to config.in (check for Restriced and Commercial packages,
22:38 < CIA-10> then ask a license awareness question...)
22:40 < CIA-10> jsaw * r5633 /trunk/scripts/Create-ISO:
22:40 < CIA-10> * don't let the user create ISO if he's not willing to check
22:40 < CIA-10> the "license awareness" question in the Config
22:49 < jsaw> relocating myself into the lab, cu later
22:50 < rxr> cu
22:54 < mnemoc> cu
22:56 < rxr> The GCC Summit, 2005 Edition:
22:56 < rxr> http://www.gccsummit.org/2005/cfp.php
22:56 < rxr> My working title is:
22:56 < rxr> "How to coerce gfortran to compile 1,000,000+ lines of weather forecasting code correctly - or not".
23:04 < CIA-10> rene * r5634 /trunk/package/x11/freetype/freetype.conf:
23:04 < CIA-10> * fixed freetype for powerpc and more by adding -fno-strict-aliasing
23:04 < CIA-10> from Dan Burcaw - Yellow Dog Linux
23:13 < jsaw> re
23:13 < rxr> welcome ;-)
23:13 < jsaw> :)
23:26 < jsaw> gcc 4.0.0: The specs file is no longer installed on most platforms.
23:26 < jsaw> hehe
23:26 < jsaw> glad we got rid of editing it.
23:27 < mnemoc> poor gentoo
23:28 < rxr> jsaw: what ?
23:28 < jsaw> (well, it's still possible...: However, if you are accustomed to editing [snip] now have to use the -dumpspecs option to generate the specs file [snip]
23:28 < jsaw> )
23:28 < jsaw> http://gcc.gnu.org/gcc-4.0/changes.html
23:30 < CIA-10> mnemoc * r5635 /trunk/package/security/checkpassword/ (. checkpassword.conf checkpassword.desc errno.patch): * added checkpassword (0.90)
23:37 < rxr> http://www.macnews.de/gallery/displayimage.php?album=16&pos=18
23:40 < jsaw> nice
23:46 < valentin> for the google log:
23:46 < valentin> if anyone uses adobe photoshop under osx 10.x
23:47 < valentin> never create a link to the executable on your desktop
23:47 < valentin> otherwise PS will not load all plugins and some menu bar items are missing
23:47 < rxr> wsan't it that you move it? It was not a link, hm ?
23:47 < valentin> esp the import menu
23:48 < valentin> yes, but still bad
23:48 < rxr> yeah
23:48 < rxr> annoying to track ...
23:48 < valentin> took me 4h
23:48 < rxr> and especially annoying since the osx bundles can normally freely be dragged arround ...
23:48 < valentin> yes
23:53 < nzg> moin
23:54 < rxr> hi nzg
23:56 < nzg> rxr: just installed beta2 on my nb
23:56 < rxr> ,-)
23:56 < rxr> problems?
23:57 < nzg> as there is no kernel source, which do i take best? 2.6.10 or 2.6.9
23:57 < nzg> stone didn't manage to set the root password
23:58 < rxr> yes - the root password thing is known - as well as some keyboard layout oddity
23:59 < rxr> forgot to track those - and shold do to in the next days if noone else does ...
23:59 < nzg> *g* yep, i experienced that one too
--- Log closed Tue Jan 18 00:00:21 2005