T2 IRC Log: 2009-11-02

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 Nov 02 00:00:17 2009
00:21 -!- gw_ [n=gw@ppp118-208-73-230.lns20.bne4.internode.on.net] has quit [Read error: 113 (No route to host)]
01:06 -!- cb88 [n=tv@240.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
01:18 -!- gw [n=gw@na-224-239.wireless.griffith.edu.au] has joined #t2
01:30 < cb88> fyi there are guys from ArchLinux wanting to start an Arch SPARC port ... I would appreciate any help.... and as I seem to have problems getting on the mailing list for T2 please see here http://bbs.archlinux.org/viewtopic.php?id=55687
01:32 < cb88> I'll check there later on and here as well... I am however constrained to dialup at home
01:40 < cb88> ah! I tured off sparcv8 optimisations and it seems to be building the cross compiler....
01:41 < cb88> I handn't thought of that untill reading on the aroralinux faq that they built for sparcv7 and add v8 packages for things that need the speedup ie openssh
01:48 < cb88> yes! it built...
04:21 -!- cb88 [n=tv@240.charlotte-04rh15-16rt.nc.dial-access.att.net] has quit [Read error: 113 (No route to host)]
07:21 -!- gw [n=gw@na-224-239.wireless.griffith.edu.au] has quit [Read error: 60 (Operation timed out)]
07:48 -!- capey [n=mp@telakka.wapice.com] has joined #t2
09:42 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has joined #t2
09:42 < Baldzius> moin
10:06 < rxr> moin moin
10:06 < rxr> heyho Baldzius
10:06 < Baldzius> hey rxr
10:06 < Baldzius> my VM is down? :)
10:06 < rxr> Baldzius: yeah sorry
10:07 < Baldzius> no problem
10:07 < rxr> the f*ck old 2.6.18 + openvz patch oopsed like 7 times last week
10:07 < Baldzius> spent whole weekend to implement crypt support
10:07 < rxr> and I desperately neded to backup my new nokia Booklet, a case where the 2.6.18 virtnet driver oopsed reliable after some 1GB of traffic ...
10:07 < rxr> Baldzius: is that on the VM or your local machine ?
10:07 < Baldzius> local
10:08 < Baldzius> I will send you patch for review
10:08 < rxr> my plan is to move your VM from openvz (and rm -rf the stuid openvz crap for good and for all) and move it to a kvm instance ...
10:08 < Baldzius> but it works fine, tested everything I could
10:08 < Baldzius> just for you to see as changes goes to system init script
10:08 < Baldzius> super
10:09 < Baldzius> is it possible to get it as fresh one?
10:09 < Baldzius> with latest kernel
10:09 < rxr> sure, while you could just update yourself anyway
10:09 < rxr> but I can also tar xf ... the last t2 reference build :-)
10:10 < Baldzius> didn't want to break anything
10:10 < Baldzius> no need
10:10 < rxr> yeah - with the openvz stuff that might have been indeed good ...
10:10 < Baldzius> that VM is pointless for me now
10:10 < Baldzius> i need to do all new builds
10:10 < Baldzius> and have no changes left there
10:10 < Baldzius> only couple of ordinary packages
10:10 < rxr> I can still move your t2 files over just in case
10:11 < rxr> or use the current state to seed the new one
10:11 < Baldzius> well
10:11 < Baldzius> maybe what i need is /t2/trunk/conf dir
10:11 < Baldzius> that's all
10:12 < rxr> moving GB today is sheap, I can rather move more then less
10:12 < Baldzius> I can take it myself if you bring it up
10:12 < Baldzius> ant then you can rm -rf it
10:13 < rxr> I know run a opvenvz-less kernel
10:13 < rxr> I needed to backup some 20GB* andthe 2.6.18-52-openvz (or so) reliable oopsed after 1GB+ ...
10:13 -!- mjungwirth_ [n=mjungwir@213.174.234.68] has quit [Remote closed the connection]
10:13 < rxr> I sort it when I create the new kvm vm
10:14 < rxr> ma 10:15 < Baldzius> ok, no rush
10:17 -!- mjungwirth [n=mjungwir@213.174.234.68] has joined #t2
10:33 < CIA-28> rene * r34772 /trunk/package/base/dialog/dialog.desc: * changed dialog to cross build by default for mine's gasgui
10:49 < capey> puuh, soon i should have my desktop build ready for final install, for some reason xf86-video-vga doesnt build, also need to check if current nvidia version supports nv 280 chips
10:57 < Baldzius> seems like nobody from X team cares about xf86-video-vga
10:58 < rxr> I can check
10:58 < Baldzius> this isn't building for me either
10:58 < rxr> Baldzius: btw. regarding removing old packages too fast
10:58 < Baldzius> yep
10:58 < rxr> 915resolution saved my day^W week!
10:58 < Baldzius> it was removed by me?
10:59 < rxr> nope :-)
10:59 < rxr> but it is of the sort of packages you like to remove :-)
10:59 < Baldzius> almost got a heart attack
10:59 < Baldzius> :)
10:59 < rxr> http://www.t2-project.org/packages/915resolution.html
10:59 < Baldzius> I know , I am bad :)
10:59 < Baldzius> rxr: you should have an email from me about crypt
11:00 < rxr> found it useful to ineject the native resolution into my Nokia Booklet 3G with Intel GMA500, aka Poulsbo, likewise
11:00 < Baldzius> I need your ack for patch
11:01 < rxr> this does not allow encrypted rootfs ?
11:03 < Baldzius> no
11:04 < rxr> one could if that would be moved to the initrd, I think we also have some bits for lvm in the initrd
11:04 < rxr> well, I'd say commit for now
11:04 < Baldzius> yeah I know, but my goal was swap and other mp like /home
11:04 < rxr> I'm not much in this crypto business, so I'm not up-to-date with the state of the art with this things anyway
11:05 < rxr> modulizing the system stuff would be nice
11:05 < rxr> so that there would be some .d directoy and packages could install their add-in there
11:05 < Baldzius> all I wanted to know if coding style is fine for you
11:05 < rxr> so not every possible option would have to be hardcoded in the default scripts
11:06 < Baldzius> and are there any changes needed
11:06 < rxr> I see one line with imperfect indenting, but, well, just commit :-)
11:06 < Baldzius> which one?
11:06 < rxr> $crypt_setup luksOpen $crypt_src $crypt_name < /dev/console
11:06 < rxr> there is one spelling issue in a comment
11:06 < Baldzius> what you suggest?
11:07 < rxr> # be cearful , the parameter ordering differs for LUKS and non-LUKS devices
11:07 < rxr> ^- careful ?
11:07 < Baldzius> :)
11:07 < rxr> the other line would need one tab ...
11:07 < rxr> but, as I said, just commit
11:08 < Baldzius> ok
11:08 < Baldzius> thanks
11:08 < rxr> thanks to
11:08 < rxr> o
11:10 < Baldzius> I will check it once again and will commit
11:10 < rxr> btw. I'll probably stop using Firefox on Linux
11:11 < Baldzius> will make note in howto that there's no rootfs support
11:11 < rxr> i'm sick of installing half of gnome for this fat mozilla ...
11:11 < rxr> I'll probably switch to the tiniest webkit based thing I can find
11:11 < Baldzius> I think there is one
11:11 < Baldzius> with qt4 interface
11:11 < rxr> yeah - there are some webkit things
11:11 < rxr> (beside the example thing that ships inside the webkit package :-)
11:11 < Baldzius> which looks lovely
11:12 < rxr> midori and friends
11:12 < Baldzius> don't recall how it's named
11:12 < rxr> midori is one :-)
11:12 < rxr> there are others
11:12 < rxr> installing half of a sub-system which I'll *NEVER* use is just plain sick
11:13 < capey> chrome? dont know what depencies it has, but should be bit fewer than firefox
11:14 < capey> ah, that's not event officially released for linux, seems to be work in progress
11:15 < rxr> yeah - I thought about that option
11:15 < rxr> but their build page lists 1.5G disk sapce needed
11:15 < rxr> this kind of scares me
11:15 < capey> :)
11:15 < capey> heh
11:15 < rxr> sounds like an open office monster
11:15 < rxr> I could try though
11:15 < rxr> however, I fear the javascript and other stuff might be x86 and arm only
11:16 < rxr> and I need powerpc, too :-)
11:16 < capey> that might be
11:16 < rxr> but we should try packaging it in t2 in any case, of course
11:16 < rxr> btw. I already committed to the v8 javacript engine
11:16 < rxr> when chrome came out a patch of mine was the first external contribution to make it into the codebase :-)
11:16 < capey> :)
11:17 -!- mjungwirth [n=mjungwir@213.174.234.68] has quit [Remote closed the connection]
11:17 < rxr> btw. there are currently 2 packages we should really try (hard) to get into t2: virtualbox and google-chrome
11:24 < CIA-28> rene * r34773 /trunk/package/lua/luajit/ (luajit.conf luajit.desc): * updated luajit (1.1.5 -> 2.0.0-beta1)
11:29 -!- gw [n=gw@ppp118-208-73-230.lns20.bne4.internode.on.net] has joined #t2
11:30 < Baldzius> rxr: I don't use initrd and I have my custom built kernel, those recent kernels spawn very fast to init so I end up in having no device nodes, those logical like sda5 and so on ..
11:31 < Baldzius> do you know how could I slow it little in professional way?
11:31 < Baldzius> without sleep
11:31 < Baldzius> now I have sleep 1 in /etc/initscript
11:32 < Baldzius> but I need sleep 2 for sda8 to come up :/
11:32 < Baldzius> and my kernel boots up in 3 sec
11:32 < rxr> hm?
11:33 < rxr> what do you need the sleep for? why no initrd ?
11:33 < Baldzius> I compile kernel manually
11:33 < rxr> the initrd has a 15s loop probing for the initrd
11:34 < rxr> well, when I comopile (a test kernel, otherwise I use t2 proper) I still use an initrd, just mkinitrd 2.6.32-git984350983hpwiugn
11:34 < rxr> but instead of a fixed sleep, you could have a simillar "wait for rootfs loop" instead the hardcoded sleep
11:34 < rxr> ah - you talk about your crypto setup?
11:35 < Baldzius> no, it's in general
11:35 < Baldzius> it affects those with higher device numbers
11:35 < Baldzius> I have 8 partitions created so far
11:35 < Baldzius> might have more in the future :)
11:35 < CIA-28> rene * r34774 /trunk/package/lua/luajit/ (luaL_narg.patch maxn.patch): * imported luajit2-beta hotfixes
11:36 < Baldzius> I will test with initrd
11:41 < CIA-28> rene * r34775 /branches/7.0/package/lua/luajit/: * merged the stable luajit to the 7.0 branch
11:41 < Baldzius> rxr: are you going to do a final 7.0 release?
11:42 < rxr> well, we could tag one for the sake of it
11:42 < rxr> but I think it's wasted time to relesae binaries
11:43 < rxr> the commit above is just for documenting history, have another fallback option, have the stable file on our mirrors, and prominently improve our web view:_)
11:43 < rxr> http://www.t2-project.org/packages/luajit.html
11:46 < rxr> mostly because I just posted the t2 luajit package update to the lua list :-)
11:58 < koan> rxr: if you want a browser without bloat, try uzbl ;-) http://lwn.net/Articles/341245/
12:04 < rxr> sounds good :-)
12:08 -!- gw [n=gw@ppp118-208-73-230.lns20.bne4.internode.on.net] has quit [Read error: 60 (Operation timed out)]
12:10 < rxr> how my t2 and own blog pr looks like: http://rene.rebe.name/2009-11-02/native-res-nokia-booklet-3g-and-linux/
12:11 -!- gw [n=gw@ppp118-208-83-27.lns20.bne4.internode.on.net] has joined #t2
12:12 < rxr> koan: ok the "no key bindings" by default sounds a little drastic
12:13 < rxr> I would expect the usual arrows and mouse point and click to to a form just-work
12:17 < CIA-28> rene * r34776 /trunk/package/security/opensc/opensc.desc: * updated opensc (0.11.10 -> 0.11.11)
12:18 < koan> yes it's drastic, but the example config file has some usable vim-like key bindings
12:18 < CIA-28> rene * r34777 /trunk/package/network/bluez/bluez.desc: * updated bluez (4.56 -> 4.57)
12:18 < CIA-28> rene * r34778 /trunk/package/base/libelf/libelf.desc: * updated libelf (0.8.12 -> 0.8.13)
12:18 < CIA-28> rene * r34779 /trunk/package/textproc/iso-codes/iso-codes.desc: * updated iso-codes (3.11 -> 3.11.1)
12:18 < CIA-28> rene * r34780 /trunk/package/graphic/libsdl_image/ (CVE-2008-0544.patch libsdl_image.desc): * updated libsdl_image (1.2.6 -> 1.2.8)
12:18 < CIA-28> rene * r34781 /trunk/package/xorg/xf86-input-joystick/xf86-input-joystick.desc: * updated xf86-input-joystick (1.4.99.2 -> 1.5.0)
12:31 < rxr> koan: dependency-wise it's also a bad joke that Gimp even requries Python and cairo, gtk+ bindings, ...
12:41 -!- cmair [i=d4de3552@gateway/web/freenode/x-def8de69097bc4e2] has joined #t2
12:42 < CIA-28> rene * r34782 /trunk/package/lua/luajit/luajit.desc: * updated luajit.desc
12:44 -!- mjungwirth [n=mjungwir@213.174.234.68] has joined #t2
12:50 < rxr> http://www.t2-project.org/hardware/portable/Nokia/Booklet-3G/
12:50 < rxr> ^ - updated photo
12:55 < cmair> Hi, can someone help me getting started with t2? I just got the current svn trunk and started a build, but it fails at mpfr.
12:55 < cmair> The logfile says "make: *** No rule to make target `@MAINTAINER_MODE_TRUE@', needed by `Makefile.in'. Stop."
12:58 < rxr> hm, never seen yet
13:01 < rxr> which target, cpu config ?
13:02 < cmair> arm cpu, embedded minimal
13:02 < rxr> hm
13:02 < rxr> maybe you can check a litlte yourself what might be going wrong?
13:03 < cmair> i will try.
13:05 < cmair> Is there a way to get into a development shell which has the same environment as the build process?
13:09 < rxr> yep: cd src.$pkg$...
13:09 < rxr> and then ./debug.sh
13:10 < cmair> Thanks
13:31 < CIA-28> rene * r34783 /trunk/package/network/cherokee/cherokee.desc: * updated cherokee (0.99.20 -> 0.99.26)
13:31 < CIA-28> rene * r34784 /trunk/package/multimedia/transcode/transcode.desc: * updated transcode (1.1.4 -> 1.1.5)
13:32 < CIA-28> rene * r34785 /trunk/package/multimedia/vlc/vlc.desc: * updated vlc (1.0.2 -> 1.0.3)
13:32 < CIA-28> rene * r34786 /trunk/package/network/wireshark/wireshark.desc: * updated wireshark (1.2.3 -> 1.3.1)
13:32 < CIA-28> rene * r34787 /trunk/package/contrib/pyxdg/pyxdg.desc: * updated pyxdg (0.17 -> 0.18)
13:32 < CIA-28> rene * r34788 /trunk/package/perl/perl-date-calc/perl-date-calc.desc: * updated perl-date-calc (6.2 -> 6.3)
13:32 < CIA-28> rene * r34789 /trunk/package/gnome2/leafpad/leafpad.desc: * updated leafpad (0.8.16 -> 0.8.17)
13:36 < capey> damm its time consuming the get WPA2 to work
13:39 < capey> wicd is nice tool for quick setup (wicd.sourceforge.net), i should package it to T2
13:43 < capey> it just has small "bug", if system has wired with no configuration on it, wicd consumes a lot of cpu time
14:15 -!- capey [n=mp@telakka.wapice.com] has quit ["work"]
14:17 < cmair> I somehow managed it to compile mpfr within the devshell. The problem was a strange autotools version mismatch.
14:17 < cmair> Can I resume the build process from there? It should use the existing, compiled source tree.
14:18 < rxr> t2 will try to rebuild it until the t2 scripts itself managed to really build it
14:19 < rxr> ideally you sould figure out the true root cause of the problem
14:21 < cmair> Unfortunately I'm not very familiar with autotools.. I did the following to get it working:
14:22 < cmair> aclocal; autoconf; automake; ./configure ... ; rm aclocal.m4; make (make runs configure again..)
14:22 < rxr> hm
14:25 < Baldzius> rxr: I think you again updated wireshark to unstable version :)
14:25 < rxr> oh
14:57 -!- csiadmin [n=csiadmin@81.144.152.52] has joined #t2
15:01 < csiadmin> hi everyone
15:01 < rxr> hi csiadmin
15:02 < csiadmin> hey rxr - hope you are well :)
15:02 < rxr> thanks and you?
15:03 < csiadmin> I'm good thanks - I've a couple of t2 questions if that's alright?
15:03 < rxr> sure
15:03 < rxr> little busy though :-)
15:03 < rxr> just start to ask
15:04 < csiadmin> no problem, thank you
15:06 < csiadmin> During build there are a lot of file put in the /var build directory (file lists, logs etc). When I make my t2 build do can I do without these? Or is there a utility to clean these up?
15:06 < rxr> well most of them are for package management
15:06 < rxr> despite the logs
15:07 < rxr> when the t2 build system packages the install / live cd then the logs are ommitted
15:07 < rxr> but the flist et al. should normally be kept
15:07 < csiadmin> Excellent thank you.
15:10 < csiadmin> I'd like to create a custom build, but a 3rd party app has some dependancies (don't know what exactly yet but it's likely to be old libraries etc) I'm likely to have newer versions of these in my build. Is it possible for me to edit the package desc files and just include the versions I need? Or does this go against what I should be using T2 for?
15:11 < rxr> you can just edit it, of course
15:12 < rxr> but you might loose that changes somehow, if you version control your stuff then it's better to create the overlay package in target/.../package/...
15:12 < rxr> but that might be overkill, depending on your work behaviour, just editing the files in-place might be a good start to get it going
15:13 < csiadmin> Excellent, I thought perhaps these could break infinite numbers of other deps and cause misery
15:13 < rxr> well, if you edit gtk+ from 2.x to 1.x then you might get tons of builds errors indeed :-)
15:13 < rxr> I thought you change would be "lightweight"
15:15 < csiadmin> I'm hoping to keep it as "lightweight" as possible. I'm also considered setting up some sort of chroot for the app and perhaps giving it what it needs in it's own chroot filesystem
15:18 < csiadmin> Appreciate your help rxr - afk
15:19 < rxr> welcoem
15:19 < rxr> welcome even :-)
15:57 -!- mqueiros [n=mqueiros@217.70.75.212] has quit [Read error: 104 (Connection reset by peer)]
15:57 -!- mqueiros [n=mqueiros@217.70.75.212] has joined #t2
16:00 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has quit ["Leaving"]
16:00 -!- csiadmin [n=csiadmin@81.144.152.52] has left #t2 []
17:28 < rxr> the standard webkit for linux buld also includes more deps than I would like, enchant, libsoup, not to mention gtk+ itself, sigh
18:49 -!- mqueiros [n=mqueiros@217.70.75.212] has quit [Read error: 60 (Operation timed out)]
21:20 -!- Stealth [i=stealth@sourcemage/guru/stealth] has quit [Read error: 131 (Connection reset by peer)]
21:22 -!- Stealth [i=stealth@sourcemage/guru/stealth] has joined #t2
23:05 -!- gw [n=gw@ppp118-208-83-27.lns20.bne4.internode.on.net] has quit [Read error: 113 (No route to host)]
23:05 -!- gw [n=gw@ppp118-208-83-27.lns20.bne4.internode.on.net] has joined #t2
23:21 -!- gw [n=gw@ppp118-208-83-27.lns20.bne4.internode.on.net] has left #t2 ["Leaving"]
--- Log closed Tue Nov 03 00:00:17 2009