--- Log opened Thu Jun 14 00:00:15 2007
00:25 -!- tri [n=tri@p57ADCC4E.dip0.t-ipconnect.de] has left #t2 []
02:00 -!- mtr_ [n=Michael@pD9E12CA9.dip0.t-ipconnect.de] has joined #t2
02:16 -!- mtr [n=Michael@pD9E13244.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
02:16 -!- mtr_ is now known as mtr
02:39 -!- Demian [n=demian@201.206.168.86] has joined #t2
02:39 < Demian> hi
02:59 -!- Demian [n=demian@201.206.168.86] has quit ["leaving"]
03:23 -!- cbrake is now known as cbrake_away
05:42 -!- bfg [n=blacky@203-206-57-245.dyn.iinet.net.au] has joined #t2
05:43 < bfg> hi
05:47 < CIA-9> gweis * r24464 /trunk/package/xorg/rendercheck/rendercheck.desc:
05:47 < CIA-9> * changed rendercheck build priority (112.400 -> 112.420) to build
05:47 < CIA-9> after libxrender
05:53 < CIA-9> gweis * r24465 /trunk/package/x11/tuxcursor/tuxcursor.desc: * fixed License tag. tuxcursor is now unde GPL
05:59 < CIA-9> gweis * r24466 /trunk/package/zope/zopeinterface/ (gcc41.patch zopeinterface.desc):
05:59 < CIA-9> * updated zopeinterface (3.0.1 -> 3.3.0)
05:59 < CIA-9> * removed obsolete gcc41.patch
06:02 < CIA-9> gweis * r24467 /trunk/package/base/hal/hal.desc: * updated hal (0.5.8.1 -> 0.5.9)
06:09 < CIA-9> gweis * r24468 /trunk/package/editors/auctex/auctex.desc:
06:09 < CIA-9> * changed auctex build priority (135.500 -> 142.800)
06:09 < CIA-9> because it needed ghostscript in my fresh build
06:11 < CIA-9> gweis * r24469 /trunk/package/base/syslog-ng/ (head_n.patch syslog-ng.desc):
06:11 < CIA-9> * updated syslog-ng (1.9.5+20050913 -> 2.0.4) and marked it stable
06:11 < CIA-9> * removed obsolete head_n.patch
06:13 < CIA-9> gweis * r24470 /trunk/scripts/config.in: * fixed typo book => bool
06:17 < CIA-9> gweis * r24471 /trunk/package/network/tor/tor.desc: * fixed license and state tag (BSD, stable)
06:20 < CIA-9> gweis * r24472 /trunk/package/network/madwifi/ (eth-naming.patch madwifi.desc):
06:20 < CIA-9> * updated madwifi (0.9.30.10-r2257-20070410 -> 0.9.30.13-r2351-20070519)
06:20 < CIA-9> * rediffed eth-naming.patch
06:20 < CIA-9> gweis * r24473 /trunk/package/network/asterisk/asterisk.desc: * updated asterisk (1.4.2 -> 1.4.4)
06:21 < CIA-9> gweis * r24474 /trunk/package/network/luma/luma.desc: * fixed download location for luma
06:43 < CIA-9> gweis * r24475 /trunk/package/develop/commonc++/commonc++.desc: * updated commonc++ (1.4.3 -> 1.5.6)
06:44 < CIA-9> gweis * r24476 /trunk/package/scientific/r/ (r.conf r.desc): * updated r (2.4.1 -> 2.5.0)
06:55 < CIA-9> gweis * r24477 /trunk/package/java/java-dirtree/parse-config: * added parse-config support for sun-jdk-160
07:00 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
07:06 < CIA-9> gweis * r24478 /trunk/misc/share/REGISTER: * registered dovecot user and group id 76
07:07 < CIA-9> gweis * r24479 /trunk/package/mail/dovecot/ (dovecot.conf dovecot.desc dovecot.init postsysfiles.in):
07:07 < CIA-9> * added sieve support to dovecot
07:07 < CIA-9> * added dovecot user and group creation
07:15 < CIA-9> gweis * r24480 /trunk/package/gnome2/evolution/evolution.desc: * added checksum for evolution download
07:33 < CIA-9> gweis * r24481 /trunk/package/gnome2/pidgin/pidgin.desc:
07:33 < CIA-9> * changed pidgin build priority (178.200 -> 260.000) to build after
07:33 < CIA-9> libgnomekbd
07:33 < CIA-9> * source gnome2-conf.in in pidgin.desc
07:35 < CIA-9> gweis * r24482 /trunk/package/gnome2/gnome-system-tools/gnome-system-tools.desc: * changed gnome-system-tools build priority (178.800 -> 208.910)
07:36 < CIA-9> gweis * r24483 /trunk/package/gnome2/pybliographer/pybliographer.desc: * updated pybliographer (1.2.9 -> 1.2.10)
07:36 < CIA-9> gweis * r24484 /trunk/package/gnome2/gnome-netstatus/gnome-netstatus.desc: * changed gnome-netstatus build priority (151.200 -> 152.100)
07:37 < CIA-9> gweis * r24485 /trunk/package/gnome2/gnome-games/gnome-games.desc: * changed gnome-games build priority (160.000 -> 486.000)
07:38 < CIA-9> gweis * r24486 /trunk/package/gnome2/enchant/enchant.desc: * changed enchant build priority (157.900 -> 153.900) needed by gedit
08:11 < rxr> re
08:13 < rxr> oh cool
08:13 < rxr> quite some nice commits :-)
08:21 < Ragnarin> hey =)
08:21 < Ragnarin> rene: do you happen to have any hints for my question earlier?
08:21 < rxr> initrd? right now this is not possible
08:22 < rxr> you can edit the /sbin/initrdinit and inject some hardcoded modprobe early
08:22 < rxr> and then regen the initrd
08:22 < rxr> we definetly need mechanisms to control stuff like this
08:22 < rxr> (aside that you could use udev to control persistent naming)
08:22 < rxr> one thing on our TODO is to use some script directories in the initrd to run custom code
08:23 < Ragnarin> yea
08:23 < rxr> like RAID, LVM, LUKEs whatever with some script directories iterated thru
08:23 < Ragnarin> well in this case it would even be enough to have a reproducible sequence of loading modules. but the sequence seems to be pretty random
08:25 < rxr> it should be PCI bus number based, but maybe the udevtrigger is doing strange stuf
08:25 < Ragnarin> maybe
08:26 < Ragnarin> inho there needs to be some flexible list so you can for example prevent an E-sata device from becoming sda
08:28 < rxr> maybe rather use labels and do not care about the actual sdX names
08:29 < rxr> or configure presistent devices names based on the serial or manufactor
08:29 < Ragnarin> well I need to put something into the fstab
08:29 < rxr> yeah - mount by label
08:29 < Ragnarin> hmm udev can do that?
08:29 < rxr> yes, just google a bit for presistent naming with udev
08:30 < rxr> and mounting by label works since ages in Linux, no matter if static nodes, devfs or udev
08:30 < Ragnarin> ah ok thanks
08:30 < Ragnarin> will do =)
08:36 < CIA-9> susan * r24487 /trunk/package/gnome2/libgksu/libgksu.desc: * updated libgksu (2.0.4 -> 2.0.5)
08:46 < CIA-9> susan * r24488 /trunk/package/gnome2/loudmouth/loudmouth.desc: * updated loudmouth (1.2.2 -> 1.2.3)
09:07 < CIA-9> rene * r24489 /trunk/package/base/linux24/linux24.desc: * updated linux24 (2.4.34.3 -> 2.4.34.5)
09:21 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
09:23 < rxr> hi write_erase
09:24 < write_erase> hi
09:44 < bfg> hi
09:44 < rxr> hi bfg
09:45 < bfg> a short question. since a while it is possible to override desc files in the taget dir. is it also possibl to override init scripts in the target dir?
09:46 < bfg> rxr: thanks for applying the new hast :)
09:46 < bfg> hsat => hash
09:46 < rxr> I think the auto-init generate code does not look into the target directory, yet
09:46 < rxr> but that should be a 2 line change if that is desired
09:46 < bfg> it is just because i need ldap to start before dbus and hal...
09:47 < bfg> because i have some unix users and groups in ldap and dbus angs quite a long time if ldap is not up
09:47 < rxr> we could change the priority in general as well
09:47 < bfg> would be ok for me :)
09:48 < bfg> if it is desired, i commit the patches
09:48 < rxr> but init re-define would be probably nice to have as well
09:48 < rxr> yes, just commit it
09:48 < bfg> ok
09:48 < rxr> should not regress the current users AFAICS
09:48 < bfg> jep init redefine could be nice (maybe some other files too ? :) )
09:49 < rxr> yes
09:50 < rxr> of course even right now one could manually create / copy differing files via hook_add in the overloaded .conf or so
09:50 < bfg> there will be only on issue on redefining the init priorites.... if you do a rebuild you have then both, the old and the new init-script in rc.d/rcx
09:50 < rxr> but we should add the usual automatic files to just work automatically
09:50 < bfg> yeah... would be much less confusing
09:51 < rxr> the same duplicate files would happen when the prority if just changed in the package/... file as well
09:51 < rxr> should be add a solution for this?
09:51 < rxr> I mean is that a quite common case ?
09:51 < bfg> not really... just for them, they build their system from trunk i think
09:52 < rxr> ah - you mean when we update it for openldap, now
09:52 < rxr> indeed
09:52 < rxr> maybe we should then add code to remove all init files of a package before generting the new ones
09:52 < rxr> btw. when we are at this topic
09:53 < bfg> yes, but the normal use case, i think, is, that one does a fresh build from one of the releases and so this file duplication won't happen
09:53 < rxr> we found out that our current "meta init" format sucks majorly, because it closely resembles sysv
09:53 < bfg> yes that's true
09:53 < rxr> we already have changing this to be able to generate upstart, minit and runit et al. formats from a single description
09:53 < rxr> just in case this attracts you and you might want to work on that as well
09:54 < bfg> i had a look on some sysv init alternatives (especially the parallel ones), but none of the solutions convinced me
10:05 -!- bfg_ [n=blacky@203-206-57-245.dyn.iinet.net.au] has joined #t2
10:05 < rxr> bfg: can you register your gweis T2 ID at: http://www.t2-project.org/startpage/ ?
10:06 < bfg_> :)
10:09 < rxr> lol: http://www.localthoughts.com/node/47?page=9
10:11 -!- CIA-9 [n=CIA@208.69.182.149] has quit []
10:11 -!- CIA-2 [n=CIA@208.69.182.149] has quit []
10:12 < rxr> bfg: you are in brisbane ?
10:18 -!- bfg [n=blacky@203-206-57-245.dyn.iinet.net.au] has quit [Read error: 113 (No route to host)]
10:24 -!- pdinoto [n=pdinoto@200.117.246.103] has quit [Read error: 60 (Operation timed out)]
10:48 < bfg_> rxr: yes i am
10:49 < rxr> heh, from your .at address I though you where in austria
10:49 < rxr> Susan's best friend and his husband will migrate to brisban in some weeks ...
10:49 < bfg_> i moved over here last year
10:49 < bfg_> it's a great area here :)
10:59 < rxr> :-)
10:59 < rxr> guess we'll visit them sooner than later when they moved there
10:59 -!- pdinoto [n=pdinoto@200.117.246.103] has joined #t2
10:59 < rxr> they planned the migration for quite some time now already, and they where half a year in .au to check it out
11:15 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has quit [Remote closed the connection]
11:27 < bfg_> if i'm still here then, maybe we have time to meet
11:27 < bfg_> ( sorry for the delay i had to brin my children to bed)
11:28 < rxr> do you plan to go somewhere else ?
11:28 < bfg_> not planned
11:28 < bfg_> but who knows what happens in the next year.
11:29 < bfg_> i'm sure i stay here till august or november next year, after tat we will see
11:29 -!- bfg_ is now known as bfg
11:45 < bfg> as you might know there is currently quite a drought here, i would advise to susan's friends, that they should try to move to redland. (it's located south east next to brisbane). redland has its own water dams which are quite full compared to the surrounding area and so there are quite less water restrictions there
11:46 < rxr> I think their choose where to go is mostly coupled with where their new employers will be
11:49 < bfg> yeah i think so... but distances are not counted that important here as we are used to in europe :)
11:51 < bfg> lately somone told me his relatives live just down the street... in fact that meant going half an hour by car
11:51 < rxr> lol
11:56 < sepp_> moin :)
11:56 < rxr> moin sepp_
11:58 < sepp_> can i /query you a new htpasswd ?
11:58 < sepp_> my old one was so secure that i can not remember what it was
11:58 < sepp_> lol
11:59 < rxr> sepp_: sure
11:59 -!- sepp_ is now known as sepp
12:08 < rxr> sepp: there are e17 releases: http://enlightenment.freedesktop.org/files/
12:08 < rxr> is it worth basing on those tarballs ?
12:10 < sepp> they are not reularly updated :/
12:11 < rxr> heh - I feared
12:11 < sepp> most are >1 year old
12:19 -!- tri [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has joined #t2
12:19 < tri> re
12:19 < tri> does the xf86-video-avivo driver also support the x1400 chipset ?
12:20 < bfg> i would suggest two package renamings. cyrus-sasl => cyrus-sasl1 and cyrus-sasl2 => cyrus-sasl... and move the sasl2 headers to sasl/ and sasl1 headers to sasl1...
12:20 < bfg> because most progs which depend on cyrus-sasl2 have problems to find the headers
12:20 < rxr> tri: tri not yet I think, that is the bare minimal driver anyway
12:20 < rxr> tri: no 3d yet
12:20 < bfg> and i don't know any pogramm which depends on cyrus-sasl1
12:21 < tri> but 2d ?
12:21 < tri> the only driver which is running here for my x1400 chipset is the vesa driver
12:22 < tri> but with that one i have problems: when i start x and switch back to console the fonts etc. are much to big...
12:23 < tri> hm t2 also supports git checkout - nice :-)
12:23 < rxr> ah x1400 should work according to their release mail
12:23 < rxr> we still lack proper initialisation
12:23 < rxr> for r520 & r580 (X1800 and above, some X1600) because of lack of time
12:23 < rxr> and hardware.
12:23 < rxr> tri: yes, but git checkout is preleminary
12:23 < rxr> we still need to add support checking out a specific point of time from git
12:23 < rxr> currently always the master is archived
12:24 < rxr> but we only have a few git checkouts right now so this was mostly ignored
12:25 < rxr> bfg: I think the rename is ok
12:25 < rxr> bfg: we mostly prefer to have the current version named just without version additions and the old legacy support junk to have like $pkgname123
12:25 < tri> from the status: 4 VT switching: 5 - Works.
12:26 < bfg> xr: yeah.. but it will involve quite some work checking all dependan packages and fixing the patches or removing them :)
12:26 < rxr> yeah
12:28 < bfg> the jabbe project should be renamed too.... the new name should be jabberd2... i is actually not a newer version of jabbed1, but it is a rewrite and now both are developed independent of each other....
12:29 < bfg> is there anythin i have to care about renaming a package? or just svn move....
12:30 < rxr> there was a svn mv order that worked better than the other
12:30 < rxr> first you should svn up the fils, otherwsie you move old revisions
12:30 < rxr> the revision you currently have
12:30 < rxr> I do not remeber off-hand if it worked better to rename the files first and the the directory or the other way round
12:31 < rxr> you will notice - in one order svn complains about renaming an revision not checked in
12:31 < rxr> probably we should document that in some developer faq page
12:31 < bfg> i just remember that sometimes i confused svn totally on renaming and modifying files.... but i'm not sure how i did this
12:31 < rxr> if other packages depend on the package I prefer when the .cache files are sed'ed to reflect the rename
12:31 < bfg> ah yes good idea
12:38 < rxr> heh- our company homepage finally renders correctly in IE: http://www.exactcode.de/ - juche
12:39 < rxr> so - lunch - cu then
12:41 [Users #t2]
12:41 [@ChanServ ] [ emte] [ mtr ] [ rxr ] [ TobiX ]
12:41 [ bfg ] [ LMJ ] [ pdinoto ] [ sepp ] [ tri ]
12:41 [ cbrake_away] [ misl] [ Ragnarin] [ Stelz] [ valentin]
12:41 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
12:43 -!- write_erase [n=olivier@pleiades4.data-mobiles.com] has joined #t2
12:48 -!- tri [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)]
13:14 -!- tr1 [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has joined #t2
13:21 -!- bfg [n=blacky@203-206-57-245.dyn.iinet.net.au] has quit [Remote closed the connection]
13:22 -!- _Robert_ [n=merob@lab-ce-26.informatik.tu-chemnitz.de] has joined #t2
13:22 < _Robert_> hello
13:39 < rxr> hi _Robert_
13:42 < _Robert_> ah, ok, hello :)
13:42 [Users #t2]
13:42 [@ChanServ ] [ emte] [ mtr ] [ rxr ] [ TobiX ] [ write_erase]
13:42 [ _Robert_ ] [ LMJ ] [ pdinoto ] [ sepp ] [ tr1 ]
13:42 [ cbrake_away] [ misl] [ Ragnarin] [ Stelz] [ valentin]
13:42 -!- Irssi: #t2: Total of 16 nicks [1 ops, 0 halfops, 0 voices, 15 normal]
13:42 < _Robert_> my first question is: can I write in german?
13:43 < rxr> only if necessary, as there are internation people as well
13:43 < _Robert_> ok, I'll try it in english...
13:43 < rxr> but if your english is too bad or whatever some lines of German are ok as well :-)
13:43 < _Robert_> I've a problem with T2
13:44 < _Robert_> I'm a student at the university of Chemnitz and I'm trying to port T2 to a Xilinx Board with a PowerPc processor
13:44 < rxr> ah, ok
13:45 < _Robert_> I created my own target, my own kernel package and kernel header (because we need the 2.4 kernel) and I built my own architecture
13:45 < _Robert_> and now I've a problem with the architecture
13:46 < _Robert_> the point is: the variable $cpu which is used for example in the file dietlibc.conf
13:46 < rxr> why don't you just use the existing powerpc?
13:46 < rxr> you want to use T2 for the PowerPC core, or for something else ?
13:46 < _Robert_> :) good question
13:46 < _Robert_> yes, for the powerpc core
13:47 < rxr> then the existing powerpc architecture is what you need to use :-)
13:48 < _Robert_> of course, I know :) but I need to update the archtest.out file (the section arch_target=..
13:48 < _Robert_> and I just wanted to modify as less as possible
13:48 < rxr> what value do you want to use ?
13:48 < _Robert_> powerpc-405-linux-gnu, because this is our toolchain with the crosscompiler
13:49 < rxr> but T2 is about to build the actual cross compiler for you
13:49 < rxr> --
13:49 < rxr> in any way you need to use the powerpc define
13:49 < _Robert_> oh..
13:49 < _Robert_> ok
13:50 < rxr> if you want some special arch_target name (which I strongly suggest not to do) you can modify the architecture/powerpc/ files in your working copy
13:50 < _Robert_> so, T2 will build the correct cross compiler for me?
13:50 < rxr> but creating a second architecture that is "sort of powerpc but with some custom modifications" will make your life very hard
13:50 < _Robert_> :) thats my problem
13:50 < rxr> yes, in the beginning T2 will bootstrap a recent cross compiler for you
13:50 < _Robert_> cool
13:51 < _Robert_> ok, I'll use the powerpc target
13:51 < _Robert_> I'm just running a build with the powerpc target and anything is ok at this time
13:51 < _Robert_> I think this is my solution
13:51 < _Robert_> thanks for your help!
13:52 < rxr> you are welcome
13:53 < _Robert_> but, just for knowing and understanding the build process: where is $cpu set?
13:53 < rxr> there is no global $cpu variable
13:54 < rxr> some packages like dietlibc or the linux2* packages have a local variable they genereate out of the $arch_... variables to assign the internal code to use
13:54 < rxr> like the arch passed to the linux kernel build system
13:54 < _Robert_> ok
13:54 < rxr> most variables are suppost to be documented in the handbook
13:54 < rxr> and doc/PKG-BUILD-VARS
13:54 < rxr> http://www.t2-project.org/documentation/
13:54 < _Robert_> I'll have a look :)
13:54 < _Robert_> thanks
14:25 [Users #t2]
14:25 [@ChanServ ] [ emte] [ mtr ] [ rxr ] [ TobiX ] [ write_erase]
14:25 [ _Robert_ ] [ LMJ ] [ pdinoto ] [ sepp ] [ tr1 ]
14:25 [ cbrake_away] [ misl] [ Ragnarin] [ Stelz] [ valentin]
14:25 -!- Irssi: #t2: Total of 16 nicks [1 ops, 0 halfops, 0 voices, 15 normal]
14:26 -!- cbrake_away is now known as cbrake
14:28 [Users #t2]
14:28 [@ChanServ] [ emte] [ mtr ] [ rxr ] [ TobiX ] [ write_erase]
14:28 [ _Robert_] [ LMJ ] [ pdinoto ] [ sepp ] [ tr1 ]
14:28 [ cbrake ] [ misl] [ Ragnarin] [ Stelz] [ valentin]
14:28 -!- Irssi: #t2: Total of 16 nicks [1 ops, 0 halfops, 0 voices, 15 normal]
14:32 -!- pdinoto [n=pdinoto@200.117.246.103] has quit [Remote closed the connection]
14:46 -!- _Robert_ [n=merob@lab-ce-26.informatik.tu-chemnitz.de] has quit [Remote closed the connection]
14:47 -!- pdinoto [n=pdinoto@200.117.246.103] has joined #t2
14:52 < tr1> puppy linux is based on t2 ?
14:52 < tr1> is there a target anywhere ?
14:53 < rxr> no, they did not summit it
14:53 < rxr> google for puppy from scratch
14:53 < rxr> there is a modified 6.0 or so tree (or was it 2.1?) they used for building
14:53 < rxr> we are in contact with them to hopefully integrate a fully functional target in t2 in the future
14:54 < rxr> (I also have a pending request from a company that eventually would pay re-doing Puppy in T2)
14:54 < tr1> http://www.puppyos.net/test/
14:54 < tr1> http://www.puppyos.net/pfs/
15:00 < tr1> what are they using for package managment
15:00 < tr1> mine ?
15:00 < tr1> are there any other distri based on t2 ?
15:04 < rxr> tr1: I think they have no package management
15:04 < rxr> they have some add-on overlay mechanism using squashfs images placed near the main image
15:04 < rxr> puppy is about a tiny compact piece that works
15:06 < tr1> ah ok
15:07 < tr1> do you know slax-linux - i used it some times ago for a work project and it also have such a overlay mechanism
15:07 < rxr> our livecd has as well
15:08 < rxr> http://www.t2-project.org/targets/live/overlays.html
15:12 < tr1> hm slax use aufs
15:12 < rxr> in the meantime we also have aufs as alternative to unionfs
15:12 < rxr> the livecd code works with both
15:16 < tr1> is there a linux-src package in t2 ?
15:16 < rxr> nope it was removed like 3 years ago for good reasons
15:16 < rxr> why do you ask ?
15:18 < tr1> hm i work on a fglrx package...
15:18 < rxr> heh, ok
15:18 < rxr> the concept is that all third party modules are build when the kernel is build
15:18 < rxr> otherwise you always have lingering modules around that do not match the kernel
15:18 < rxr> this is what the postlinux.conf is about
15:19 < tr1> hm right
15:19 < tr1> thats ok
15:19 < rxr> the linux* kernel packages build all packages that have this postlinux.conf after the kernel was built
15:19 < tr1> what are the reasons to drop the linux-src package - i don
15:19 < tr1> 't need it
15:19 < rxr> this is also why you get all the -> Building xyz withing linux26 (using postlinux.conf) or so for each kernel build
15:19 < tr1> just to know it
15:20 < tr1> yes i saw this for the ipw3945
15:20 < rxr> well - as most people only build the kernel and not always rebuild their linux...-src package the sources are outdated and unmatching to the available kernels anyway
15:21 < rxr> also at that time all the third party modules had to be build manually after each kernel and linux-src package build
15:21 < rxr> of course cumbersome and error prone
15:21 < rxr> most people ended up with moduels that oopsed the kernel because some data structured changed
15:21 < rxr> or, if they where lucky. only tained the running kernel due to version mismatch
15:21 < rxr> plus, last but not least, the linux-src just wasted a whole lot of space on each system and cd
15:21 < tr1> hm right the conecpt right now is clear - i just need to use the postlinux.conf thing
15:22 < rxr> as most end-users did not need it and those who need usually download the latest greatest kernel from kernel.org themselves
15:22 < tr1> what about providing binary packages ?
15:22 < rxr> just copy the nvidia package and edit the postlinux.conf to extract the ati junk and run the make with the appropriate switches
15:23 < rxr> binary packages of t2? yes sure we long plan to provide them
15:23 < tr1> there is allready a build server ?
15:23 < rxr> well, not publically accessible, but sure we have build servers
15:24 < tr1> i now the arch way - where people vote for (source) packages
15:25 < tr1> and then if there are enough votes these packages are maintained "better" and there are binary packages provided by the build server
15:26 < rxr> well all of our packages should be "perfect" .-)
15:26 < tr1> :-)
15:26 < rxr> when we provide binaries we certianly will do so for all
15:27 < tr1> ok let provide binarie packages for hm x86
15:27 < rxr> as it is just a matter of building them (they should build in any case) and providing them online
15:27 < tr1> because i use x86 :-)
15:27 < rxr> yeah :-)
15:27 < tr1> i may be able to provide some mirror
15:27 < rxr> even providing binaries for all arches and all the c library combinations ilke glibc, dietlibc and uclibc is not so much of the problem for the packages that cross build :-)
15:28 < tr1> hm damm i hate ntfs permissions...
15:28 < rxr> yes, with homepage binary packages should be the next big thing TODO
15:29 < rxr> that should read "together with a nifty new homepage, binary packages ..."
15:29 < tr1> i just have tried to build video-avivo and i get a error - are you able to build this ?
15:29 < tr1> ah ok
15:29 < rxr> you need the util-macros
15:29 < tr1> that would be a nice subject
15:29 < rxr> util-macros and libpciaccess
15:29 < rxr> if you have both and still get an configure error run:
15:30 < rxr> /etc/postinstall.d/automake
15:34 < tr1> a util-macros was the issue
15:35 < rxr> hm - damn the util-acros DEP is not in .cache
15:36 < tr1> ah !
15:36 < tr1> (15:22:58) rxr: well all of our packages should be "perfect" .-)
15:36 < tr1> ;-)
15:38 -!- tr1 [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)]
15:41 -!- tri [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has joined #t2
15:41 < tri> re
15:47 -!- tri [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)]
15:47 -!- tr1 [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has joined #t2
15:47 < tr1> hm
15:48 < tr1> postmake hook is executed before "make..."
15:48 < tr1> but after configure ?
15:48 < rxr> premake is between configure and make
15:48 < rxr> postmake is after all make runs
15:53 < tr1> a so i need premake
15:53 < tr1> pre* - post* <- hm...
15:54 < tr1> after all make runs means also make install
15:57 < tr1> ?
15:58 < tr1> what means Corrected ./ltmain.sh ?
16:04 < tr1> when are the flist created ?
16:11 < rxr> yes
16:11 < rxr> it is:
16:11 < rxr> premake hooks
16:11 < rxr> make
16:11 < rxr> inmake hook
16:11 < rxr> make install
16:11 < rxr> postmake hooks
16:12 < rxr> Corrected ./ltmain.sh means the package file was automatically corrected
16:12 < rxr> these files come with most autoconf packages from libtool and often are outdated or otherwise defect for builds with libdir != lib, like lib64 on x86-64 and thus we choose to implement automatic correction in t2
16:13 < rxr> as manually hunting and fixing them takes ages
16:16 < tr1> hm when is the flist created ?
16:19 < rxr> at the end of the package build
16:19 < rxr> after everything was built and installed
16:19 < rxr> the files created are of course tracked all the time
16:21 < tr1> hm
16:21 < tr1> still trying to build silc-toolkit
16:22 < pdinoto> rxr: by the way, what is the use of lua in T2?
16:31 < rxr> currently not much
16:32 < rxr> in the former times we had a bash plugin written in C for the Config
16:32 < rxr> we converted it to Lua so that it is easier to maintain and harder to segfault
16:32 < rxr> and accelerated some more parts now that it is easier from Lua than in C as before
16:32 < rxr> but that is currently the only thing we use Lua for
16:33 < rxr> when the Config is entirely rewritten more parts are likely rewritten in Lua as it is such a slick and fast language
16:34 -!- pdinoto [n=pdinoto@200.117.246.103] has quit [Read error: 60 (Operation timed out)]
16:35 [Users #t2]
16:35 [@ChanServ] [ LMJ ] [ Ragnarin] [ Stelz] [ valentin ]
16:35 [ cbrake ] [ misl] [ rxr ] [ TobiX] [ write_erase]
16:35 [ emte ] [ mtr ] [ sepp ] [ tr1 ]
16:35 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
16:43 < rxr> tr1: http://bugs.t2-project.org/wiki/Svn
16:44 < tr1> hm that asks somebody else ;-)
16:45 < tr1> ah now
16:45 < tr1> if i rename the stringprep.h from the silc-toolkit
16:45 < tr1> i have to change every inlcude ?
16:45 < rxr> every include in the silc-toolkit sources, yes
16:46 < tr1> it is interesting that they use a silcstringprep.h also...
16:47 < tr1> how to find every include find ./ -exec grep "#include stringprep.h" {} \;
16:48 < tr1> doesn't work
16:53 < sepp> "#include \"stringprep.h\"" ?
16:53 < tr1> jop :-)
16:55 < sepp> no cia today
16:55 < tr1> for every changed file a *.patch file ?
16:56 < tr1> hm but there where same svn updates ?
16:56 < sepp> svn diff package/.../.../
16:57 < tr1> and how is this patch then provided ?
16:57 < tr1> silc-toolkit... .patch
16:57 < sepp> svn diff package/.../.../ > silc-toolkit.patch ?
16:57 < tr1> ah ok
16:58 < tr1> the build looks good now
16:59 < sepp> :)
17:00 < tr1> ah ok have to go now - might be able post patch to the mailinglist later...
17:00 < tr1> cu
17:00 -!- tr1 [n=Richter@pd95b2df8.dip0.t-ipconnect.de] has left #t2 []
17:15 [Users #t2]
17:15 [@ChanServ] [ LMJ ] [ Ragnarin] [ Stelz ] [ write_erase]
17:15 [ cbrake ] [ misl] [ rxr ] [ TobiX ]
17:15 [ emte ] [ mtr ] [ sepp ] [ valentin]
17:15 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
18:04 < rxr> yuche - down to just 230 mail in my t2-updates mail-box ...
18:04 < rxr> unfortunatly my reference building box needs to progress a bit, so the next updates will be lingering some time until they are tested, ...
18:31 -!- tri [n=tri@p57ADCC24.dip0.t-ipconnect.de] has joined #t2
18:31 < tri> re
18:31 < sepp> wb :)
18:36 < tri> wups nice there are new e17 packages :-)
18:36 < rxr> yeah - too bad CIA is dead today, again
18:36 [Users #t2]
18:36 [@ChanServ] [ LMJ ] [ Ragnarin] [ Stelz] [ valentin ]
18:36 [ cbrake ] [ misl] [ rxr ] [ TobiX] [ write_erase]
18:36 [ emte ] [ mtr ] [ sepp ] [ tri ]
18:36 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
18:37 < tri> hm
18:37 < tri> CIA is nice but doesn't run that stable ?
18:37 < rxr> it's a remote service
18:37 < tri> maybe they use not t2 ?
18:37 < rxr> it just stucks or disappears now and then
18:37 < rxr> yeah
18:38 < rxr> but IRC bot stuff often is quite mess busyness
18:38 < rxr> and they run tons of projects these days
18:38 < rxr> guess they get quite some in and out traffic these days
18:38 < tri> and its free ?
18:38 < rxr> yep
18:43 < tri> hm the irc logs are nice
18:43 < sepp> e17 xdg support is nice :)
18:43 < tri> yes right the efreet lib is a good one
18:44 < rxr> sepp: you run on x86-64, right?
18:44 < sepp> yes
18:44 < rxr> did you do something with the harcoded /lib/ already ?
18:45 < sepp> yes, that works ok now
18:45 < rxr> ah - cool
18:45 * rxr emerging on the MacBook, ...
18:46 < tri> bfg is living in brisbane - nice :-)
18:46 < sepp> though evidence is just a big sed expression, that is too ugly to make a patch for
18:47 < tri> so when he is only its time to go sleep :-)
18:47 < tri> hm
18:47 < tri> i get this error / warning: so when he is only its time to go sleep :-)
18:47 < tri> ups
18:47 < tri> i mean this one: -> LUA accelerator (C) 2006 by Valentin Ziegler & Rene Rebe, ExactCODE
18:47 < tri> lua bash error: scripts/config-functions.lua:37: bad argument #1 to 'match' (st
18:47 < tri> ing expected, got nil)
18:47 < rxr> on a new package?
18:47 < rxr> probably a tag missing or typo
18:48 < rxr> we should make the parsing more sensitive and output a pretty diagnostic
18:48 < tri> maybe i named a *.conf file to *.new just for testing ?
18:50 < sepp> ?
18:50 < tri> i have a *.conf file but i don't want it to be included right now - so i just renamed it to *.new
18:51 < rxr> that can not be a problem
18:51 < rxr> the problem will be in a .desc file
19:10 < tri> hm
19:12 < rxr> oh my god, the new dock is ugly: http://www.thinksecret.com/archives/wwdc2007/source/p6110104.html
19:19 -!- write_erase [n=olivier@pleiades4.data-mobiles.com] has quit ["Leaving"]
19:19 < sepp> hehe
19:20 < rxr> I hope the Open Source desktops do not clone that ...
19:21 < rxr> heh - http://www.apple.com got a facelift as well
19:55 < misl> showing the first pdf page is kind of neat
19:56 < misl> that's probably the reason they want these gigantic icons
19:57 -!- CIA-8 [i=cia@208.69.182.149] has joined #gsmp
19:58 -!- CIA-9 [i=cia@208.69.182.149] has joined #t2
20:02 -!- pdinoto [n=pdinoto@200.117.246.103] has joined #t2
20:05 < tri> hm
20:06 < tri> how to use find and leave out some dirs ?
20:06 -!- CIA-8 [i=cia@208.69.182.149] has quit []
20:06 -!- CIA-9 [i=cia@208.69.182.149] has quit []
20:07 < tri> i think i have done this but can't remember how ?
20:10 -!- CIA-9 [i=cia@208.69.182.149] has joined #t2
20:10 -!- CIA-8 [n=CIA@208.69.182.149] has joined #gsmp
20:23 -!- Stelz [i=stelz@unaffiliated/stelz] has quit [Read error: 113 (No route to host)]
20:35 < rxr> find ... -prune or so
20:56 < pdinoto> rxr: some hours ago I made a question about lua, and then lost power
20:56 < pdinoto> rxr: (yes, the machine was in the non-UPS outlet: ouch!)
20:56 < pdinoto> sorry if I was "missing" :)
21:08 < sepp> 16:36 < rxr> currently not much
21:08 < sepp> 16:36 < rxr> in the former times we had a bash plugin written in C for the Config
21:08 < sepp> 16:36 < rxr> we converted it to Lua so that it is easier to maintain and harder to segfault
21:08 < sepp> 16:36 < rxr> and accelerated some more parts now that it is easier from Lua than in C as before
21:08 < sepp> 16:37 < rxr> but that is currently the only thing we use Lua for
21:08 < sepp> 16:37 < rxr> when the Config is entirely rewritten more parts are likely rewritten in Lua as it is such a slick and fast language
21:09 < pdinoto> thanks.
21:09 < sepp> np
21:23 -!- sepp_ [n=sepp@Z6eda.z.pppool.de] has joined #t2
21:25 -!- tri [n=tri@p57ADCC24.dip0.t-ipconnect.de] has left #t2 []
21:27 < sepp_> does qemu work on x86-64 built with gcc-4.2?
21:38 -!- sepp [n=sepp@Z496b.z.pppool.de] has quit [Read error: 110 (Connection timed out)]
21:44 < pdinoto> i've heard that qemu only builds with gcc 3.x, IIRC
21:51 < sepp_> t2 had patches to support it with 4.x , just dont know if it works with 4.2 ...
21:56 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
22:05 < pdinoto> I have a kernel and a squashfs rootfs: Do I need any special kernel
22:05 < pdinoto> parameters to run them into qemu?
22:06 < pdinoto> I get a kernel panic because no root filesystem.
22:10 < sepp_> i fear qemu is currently broken
22:12 < pdinoto> Do you mean that running a T2 build under qemu is not possible,
22:13 < pdinoto> or that building qemu with T2 is impossible?
22:13 < pdinoto> (I am trying to to do the former: run my recently built embedded target
22:14 < sepp_> i think it needs some patch for gcc-4.2
22:14 < pdinoto> into my working qemu install on the host system)
22:14 < pdinoto> oh, we are talking two different things then :)
22:15 < sepp_> yup
22:15 < sepp_> :)
22:17 < write_erase> I build T2 under debian/qemu ... :-)
22:17 -!- d-marc [n=d-marc@HSI-KBW-082-212-032-020.hsi.kabelbw.de] has joined #t2
22:18 < write_erase> I can't see why qemu would not boot a T2 filesystem
22:19 < pdinoto> well, that is why i am almost sure I am missing something very simple:
22:19 < pdinoto> after a successfull T2 embedded target build, I get a kernel
22:20 < pdinoto> and a squashfs root. The kernel boots under qemu, and
22:20 < pdinoto> loads the squashfs root in the initrd (using qemu's -initrd option)
22:21 < pdinoto> perhaps I should -append root=
22:22 < write_erase> If you suspect Qemu to fail.... Install all your stuff on some file disk with a bootloader, and try to boot it in a traditionnal way
22:26 -!- d-marc [n=d-marc@HSI-KBW-082-212-032-020.hsi.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
22:39 < pdinoto> will try that next
22:55 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has quit ["Leaving"]
22:55 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
22:59 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has quit [Remote closed the connection]
22:59 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
22:59 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has quit [Client Quit]
23:02 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
23:38 -!- pdinoto [n=pdinoto@200.117.246.103] has quit [Remote closed the connection]
23:39 -!- cbrake is now known as cbrake_away
23:57 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
23:57 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has quit ["Leaving"]
--- Log closed Fri Jun 15 00:00:16 2007