T2 IRC Log: 2009-09-29
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.
--- Log opened Tue Sep 29 00:00:43 2009
01:17 -!- synchris [n=synchris@pdpc/supporter/professional/synchris] has quit [Remote closed the connection]
01:17 -!- synchris [n=synchris@adsl195-81.kln.forthnet.gr] has joined #t2
08:51 -!- mqueiros [n=mqueiros@217.70.75.125] has quit ["Ex-Chat"]
09:39 < rxr> http://patchwork.kernel.org/patch/49462/
09:41 -!- mpp [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has joined #t2
09:41 < mpp> moinsen
09:45 < rxr> moin moin mpp
10:59 < CIA-41> rene * r34333 /trunk/package/xorg/xf86-video-intel/xf86-video-intel.desc: * updated xf86-video-intel (2.8.1 -> 2.9.0)
10:59 < CIA-41> rene * r34334 /trunk/package/emulators/wine/wine.desc: * updated wine (1.1.29 -> 1.1.30)
11:00 < CIA-41> rene * r34335 /trunk/package/x11/mesa/mesa.desc: * updated mesa (7.5 -> 7.6)
11:00 < CIA-41> rene * r34336 /trunk/package/develop/cln/cln.desc: * updated cln (1.2.2 -> 1.3.1)
12:21 -!- xor_ [n=oscar@80.38.62.11] has quit ["Saliendo"]
13:41 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: T-One, CIA-41
13:42 -!- Netsplit over, joins: T-One
13:47 -!- CIA-38 [n=CIA@208.69.182.149] has joined #t2
14:24 -!- kic_ [n=rfoote@99.146.149.79] has joined #t2
14:36 -!- kic [n=rfoote@99.146.151.81] has quit [Read error: 110 (Connection timed out)]
15:41 -!- the_rainbow0 [n=aaa@41.104.64.179] has joined #t2
15:44 -!- the_rainbow0 [n=aaa@41.104.64.179] has left #t2 []
16:04 -!- Baldzius [n=aldas@87.198.133.94] has joined #t2
16:05 < Baldzius> moin
16:05 < rxr> moin moin Baldzius, hi!
16:05 < Baldzius> hey rxr
16:06 < Baldzius> do you have any udev update locally?
16:06 < Baldzius> we need 146 ver. :/
16:06 < Baldzius> I can't add new package until then which would fix one of gnome packages
16:07 < Baldzius> gnome-session if I recall correctly
16:07 < rxr> well, newer udev versions are mangled more and more
16:07 < rxr> we might have to patch it and dietlibc for our size optimzed initrd ..
16:10 < rxr> what to we need the newer version for? to motivate me hacking thru all the details
16:12 < Baldzius> for devicekit-power
16:12 < rxr> urgs
16:12 < Baldzius> gnome-session fails without it
16:12 < rxr> oh my god
16:12 < rxr> more and more useles glue code all over the mostly GNOME world ?
16:12 < rxr> eventually nautilus will come with it's own OS kernel ?
16:13 < Baldzius> I can try to see if it's possible to disable it
16:13 < rxr> yeah - we can also work on getting udev updatede ...
16:14 < Baldzius> this devicekit-power might be used with more packages in the future
16:14 < Baldzius> I think fedora is going away from hal for devicekit*
16:14 < rxr> aha
16:14 < koan> indeed
16:14 < koan> fedora 11 was already a partial move from hal to devicekit
16:14 * rxr going away from this insane crap ... have to start my own micro kernel or use dragonfly bsd or haiku ...
16:15 < koan> :)
16:15 < Baldzius> so I've added devicekit yesterday and I can't add devicekit-power
16:15 < Baldzius> I now it requires udev 146
16:15 < Baldzius> as I have a dirty update here which partialy fails
16:15 < rxr> I'll keep looking at the new udev versions
16:15 < Baldzius> but that's enough for devicekit-power
16:15 < rxr> I looked at two earlier versions
16:16 < rxr> but their use of very latest glibc features made me always revert it ...
16:16 < Baldzius> to compile
16:16 < rxr> dietlibc just did not hat those very latest-gen (GNU only) features
16:16 < Baldzius> btw I have one big nasty thing
16:16 < rxr> you could do a build with udev not linking against dietlibc, that would certainly just work
16:16 < Baldzius> kdelibs freezes on compile
16:16 < rxr> oh
16:17 < rxr> on your box or my VM ?
16:17 < Baldzius> at about 76%
16:17 < Baldzius> with NOPARALLEL at 98%
16:17 < Baldzius> VM
16:17 < Baldzius> ref build
16:17 < rxr> hm
16:17 < rxr> btw. sorry that the VM was in flux recently
16:17 < Baldzius> that's ok, enough for me
16:18 < rxr> planed power outage and then I messed up some virtual network ...
16:18 < rxr> anyway,
16:18 < rxr> strange thing with kde
16:18 < Baldzius> yeah
16:18 < rxr> how does it freeze? gcc compiler bug ?
16:18 < rxr> 7GB not eough MEM ?
16:18 < Baldzius> not sure if it's memory
16:18 < Baldzius> it just stops and there's nothing going on
16:19 < Baldzius> but doesn't crash
16:19 < rxr> maybe a script bug
16:19 < rxr> wanting to read from stdin or so ?
16:19 < Baldzius> same is for 4.3.0
16:19 < rxr> like qt license "ok" question in the past
16:20 < rxr> btw. all the latest linux developments (last 2y++) make me really desperately want to move somewhere else
16:20 < rxr> linux kernel and associated project development is just not sane anymore ...
16:20 < rxr> koan: what are your ipressions? already on haiku ?
16:22 < koan> I have not tried Haiku yet, I don't know if I will, I'm not so into BeOS
16:22 < rxr> dragonfly bsd ?
16:22 < rxr> anything?
16:22 < Baldzius> rxr: also kdelibs builds on my machine fine
16:22 < Baldzius> both 4.3.0 and 4.3.1
16:22 < koan> I have tried DragonFly, that looks interesting, but I didn't stress-test it
16:22 < Baldzius> so not sure if there's something went it to trunk lately
16:22 < Baldzius> ot it's actually a VM fault
16:22 < rxr> strange
16:23 < koan> the HAMMER filesystem does look very interesting
16:23 < rxr> Baldzius: in my VM it apparently built, modulo shared files
16:23 < koan> and I like Nexenta because it gives access to ZFS and OpenSolaris zones while giving a Linux-like (in this case Ubuntu) userland
16:24 < rxr> (that might just occured due in-place re-build with old version around)
16:24 < rxr> koan: yeah
16:24 < rxr> Baldzius: so maybe openvz's fault
16:24 < rxr> instable code anyway
16:24 < rxr> maybe I should wipe the last, yours opevz instance and give you a KVM instance
16:24 < rxr> it's just that I can not share that much ram into that
16:25 < Baldzius> I don't think I need much
16:25 < rxr> unless I go buy some 8 or 16 gig of new
16:25 < Baldzius> just to get most of the packages to compile
16:25 < rxr> well, builds, GCC eat a lot, ... 2GB is minimum today, no ?
16:25 < Baldzius> dunno
16:26 < rxr> ok - with swap 1gig may be enough
16:26 < rxr> unless some huge C++ project throws some monster files on gccc v4 ...
16:26 < koan> rxr: and I was impressed by the simple cross-building process of NetBSD, I compiled NetBSD/x86 without any problems on Linux/ppc64 and OpenSolaris/x86_64
16:26 < rxr> koan: sounds good
16:27 < koan> so if you're planning a new kernel port of T2, maybe NetBSD is an interesting start, also for embedded targets :-)
16:27 < rxr> yeah - thing is when I throw some major part of my time on it I want something modern, microkernish
16:28 < rxr> best with multi user-space servers
16:28 < koan> hm did you take a look at L4?
16:28 < koan> they have also Linux running on top of L4
16:31 < koan> the problem with all those microkernels (even Minix) is of course the limited set of device drivers, Linux has a huge advantage with this
16:31 < rxr> sorry phone
16:39 < rxr> yeah, indeed
16:39 < rxr> but it gains me nothing when every new kernel has more regressions than they fixed
16:39 < rxr> and gets slower and huger
16:39 < rxr> intel published recently that every new linux kernel gets some percentage slower ...
16:39 < rxr> depspite I would normally not want to reboot my whole machine just to fix a security hole
16:42 < rxr> btw. with all this user-space "junk" the whole linux setup becomes multi server'ish anyway
16:42 < rxr> just with all cons of a monolith, and low quality user-space hacking
16:42 < koan> wasn't minix on the T2 agenda? (at least on the website if I remember correctly)
16:43 < rxr> yeah - but minix is too protypish ...
16:44 < koan> protypish?
16:46 < rxr> classroom-like
16:48 < koan> well, I don't know of many microkernels that have escaped academia, maybe L4 comes the closest
16:49 < rxr> well - hybrid like dragonfly of haiku ar haiku are already a progress ...
16:49 < rxr> and less random and breaking changes another
17:08 < rxr> Baldzius: there is also a possible binutils update if you want to test it with some bigger builds * updated binutils (2.19.51.0.1 -> 2.19.51.0.10)
17:10 < Baldzius> I'll try to remember if I will start new ref build
17:10 < Baldzius> need to fix some packages first
17:17 < rxr> ok - no problem
18:06 -!- T-One [n=T-One@62.218.122.66] has quit [Read error: 110 (Connection timed out)]
18:30 -!- synchris_ [n=synchris@adsl108-49.kln.forthnet.gr] has joined #t2
18:31 -!- synchris [n=synchris@pdpc/supporter/professional/synchris] has quit [Read error: 60 (Operation timed out)]
20:11 -!- synchris_ is now known as synchris
20:30 -!- mqueiros [n=mqueiros@217.70.75.225] has joined #t2
20:43 -!- mpp [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has quit ["good night - good fight"]
20:48 < rxr> http://rene.rebe.name/2009-09-29/sony-vaio-vgn-p-nearing-refresh-or-end-of-life/
22:49 < CIA-38> aldas * r34337 /trunk/package/gnome2/rhythmbox/libmtp.patch: * removed obsolete rhythmbox/libmtp.patch
22:50 < CIA-38> aldas * r34338 /trunk/package/base/libx86/lheader.patch: * fixed libx86 to compile with new linux headers
23:02 < CIA-38> aldas * r34339 /trunk/package/network/aria2/ (. aria2.cache aria2.desc):
23:02 < CIA-38> Martin Papadopoulos:
23:02 < CIA-38> * added aria2 (1.5.2) - aria2 - The ultra fast download utility
23:34 -!- Baldzius [n=aldas@87.198.133.94] has quit ["Leaving"]
--- Log closed Wed Sep 30 00:00:43 2009
01:17 -!- synchris [n=synchris@pdpc/supporter/professional/synchris] has quit [Remote closed the connection]
01:17 -!- synchris [n=synchris@adsl195-81.kln.forthnet.gr] has joined #t2
08:51 -!- mqueiros [n=mqueiros@217.70.75.125] has quit ["Ex-Chat"]
09:39 < rxr> http://patchwork.kernel.org/patch/49462/
09:41 -!- mpp [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has joined #t2
09:41 < mpp> moinsen
09:45 < rxr> moin moin mpp
10:59 < CIA-41> rene * r34333 /trunk/package/xorg/xf86-video-intel/xf86-video-intel.desc: * updated xf86-video-intel (2.8.1 -> 2.9.0)
10:59 < CIA-41> rene * r34334 /trunk/package/emulators/wine/wine.desc: * updated wine (1.1.29 -> 1.1.30)
11:00 < CIA-41> rene * r34335 /trunk/package/x11/mesa/mesa.desc: * updated mesa (7.5 -> 7.6)
11:00 < CIA-41> rene * r34336 /trunk/package/develop/cln/cln.desc: * updated cln (1.2.2 -> 1.3.1)
12:21 -!- xor_ [n=oscar@80.38.62.11] has quit ["Saliendo"]
13:41 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: T-One, CIA-41
13:42 -!- Netsplit over, joins: T-One
13:47 -!- CIA-38 [n=CIA@208.69.182.149] has joined #t2
14:24 -!- kic_ [n=rfoote@99.146.149.79] has joined #t2
14:36 -!- kic [n=rfoote@99.146.151.81] has quit [Read error: 110 (Connection timed out)]
15:41 -!- the_rainbow0 [n=aaa@41.104.64.179] has joined #t2
15:44 -!- the_rainbow0 [n=aaa@41.104.64.179] has left #t2 []
16:04 -!- Baldzius [n=aldas@87.198.133.94] has joined #t2
16:05 < Baldzius> moin
16:05 < rxr> moin moin Baldzius, hi!
16:05 < Baldzius> hey rxr
16:06 < Baldzius> do you have any udev update locally?
16:06 < Baldzius> we need 146 ver. :/
16:06 < Baldzius> I can't add new package until then which would fix one of gnome packages
16:07 < Baldzius> gnome-session if I recall correctly
16:07 < rxr> well, newer udev versions are mangled more and more
16:07 < rxr> we might have to patch it and dietlibc for our size optimzed initrd ..
16:10 < rxr> what to we need the newer version for? to motivate me hacking thru all the details
16:12 < Baldzius> for devicekit-power
16:12 < rxr> urgs
16:12 < Baldzius> gnome-session fails without it
16:12 < rxr> oh my god
16:12 < rxr> more and more useles glue code all over the mostly GNOME world ?
16:12 < rxr> eventually nautilus will come with it's own OS kernel ?
16:13 < Baldzius> I can try to see if it's possible to disable it
16:13 < rxr> yeah - we can also work on getting udev updatede ...
16:14 < Baldzius> this devicekit-power might be used with more packages in the future
16:14 < Baldzius> I think fedora is going away from hal for devicekit*
16:14 < rxr> aha
16:14 < koan> indeed
16:14 < koan> fedora 11 was already a partial move from hal to devicekit
16:14 * rxr going away from this insane crap ... have to start my own micro kernel or use dragonfly bsd or haiku ...
16:15 < koan> :)
16:15 < Baldzius> so I've added devicekit yesterday and I can't add devicekit-power
16:15 < Baldzius> I now it requires udev 146
16:15 < Baldzius> as I have a dirty update here which partialy fails
16:15 < rxr> I'll keep looking at the new udev versions
16:15 < Baldzius> but that's enough for devicekit-power
16:15 < rxr> I looked at two earlier versions
16:16 < rxr> but their use of very latest glibc features made me always revert it ...
16:16 < Baldzius> to compile
16:16 < rxr> dietlibc just did not hat those very latest-gen (GNU only) features
16:16 < Baldzius> btw I have one big nasty thing
16:16 < rxr> you could do a build with udev not linking against dietlibc, that would certainly just work
16:16 < Baldzius> kdelibs freezes on compile
16:16 < rxr> oh
16:17 < rxr> on your box or my VM ?
16:17 < Baldzius> at about 76%
16:17 < Baldzius> with NOPARALLEL at 98%
16:17 < Baldzius> VM
16:17 < Baldzius> ref build
16:17 < rxr> hm
16:17 < rxr> btw. sorry that the VM was in flux recently
16:17 < Baldzius> that's ok, enough for me
16:18 < rxr> planed power outage and then I messed up some virtual network ...
16:18 < rxr> anyway,
16:18 < rxr> strange thing with kde
16:18 < Baldzius> yeah
16:18 < rxr> how does it freeze? gcc compiler bug ?
16:18 < rxr> 7GB not eough MEM ?
16:18 < Baldzius> not sure if it's memory
16:18 < Baldzius> it just stops and there's nothing going on
16:19 < Baldzius> but doesn't crash
16:19 < rxr> maybe a script bug
16:19 < rxr> wanting to read from stdin or so ?
16:19 < Baldzius> same is for 4.3.0
16:19 < rxr> like qt license "ok" question in the past
16:20 < rxr> btw. all the latest linux developments (last 2y++) make me really desperately want to move somewhere else
16:20 < rxr> linux kernel and associated project development is just not sane anymore ...
16:20 < rxr> koan: what are your ipressions? already on haiku ?
16:22 < koan> I have not tried Haiku yet, I don't know if I will, I'm not so into BeOS
16:22 < rxr> dragonfly bsd ?
16:22 < rxr> anything?
16:22 < Baldzius> rxr: also kdelibs builds on my machine fine
16:22 < Baldzius> both 4.3.0 and 4.3.1
16:22 < koan> I have tried DragonFly, that looks interesting, but I didn't stress-test it
16:22 < Baldzius> so not sure if there's something went it to trunk lately
16:22 < Baldzius> ot it's actually a VM fault
16:22 < rxr> strange
16:23 < koan> the HAMMER filesystem does look very interesting
16:23 < rxr> Baldzius: in my VM it apparently built, modulo shared files
16:23 < koan> and I like Nexenta because it gives access to ZFS and OpenSolaris zones while giving a Linux-like (in this case Ubuntu) userland
16:24 < rxr> (that might just occured due in-place re-build with old version around)
16:24 < rxr> koan: yeah
16:24 < rxr> Baldzius: so maybe openvz's fault
16:24 < rxr> instable code anyway
16:24 < rxr> maybe I should wipe the last, yours opevz instance and give you a KVM instance
16:24 < rxr> it's just that I can not share that much ram into that
16:25 < Baldzius> I don't think I need much
16:25 < rxr> unless I go buy some 8 or 16 gig of new
16:25 < Baldzius> just to get most of the packages to compile
16:25 < rxr> well, builds, GCC eat a lot, ... 2GB is minimum today, no ?
16:25 < Baldzius> dunno
16:26 < rxr> ok - with swap 1gig may be enough
16:26 < rxr> unless some huge C++ project throws some monster files on gccc v4 ...
16:26 < koan> rxr: and I was impressed by the simple cross-building process of NetBSD, I compiled NetBSD/x86 without any problems on Linux/ppc64 and OpenSolaris/x86_64
16:26 < rxr> koan: sounds good
16:27 < koan> so if you're planning a new kernel port of T2, maybe NetBSD is an interesting start, also for embedded targets :-)
16:27 < rxr> yeah - thing is when I throw some major part of my time on it I want something modern, microkernish
16:28 < rxr> best with multi user-space servers
16:28 < koan> hm did you take a look at L4?
16:28 < koan> they have also Linux running on top of L4
16:31 < koan> the problem with all those microkernels (even Minix) is of course the limited set of device drivers, Linux has a huge advantage with this
16:31 < rxr> sorry phone
16:39 < rxr> yeah, indeed
16:39 < rxr> but it gains me nothing when every new kernel has more regressions than they fixed
16:39 < rxr> and gets slower and huger
16:39 < rxr> intel published recently that every new linux kernel gets some percentage slower ...
16:39 < rxr> depspite I would normally not want to reboot my whole machine just to fix a security hole
16:42 < rxr> btw. with all this user-space "junk" the whole linux setup becomes multi server'ish anyway
16:42 < rxr> just with all cons of a monolith, and low quality user-space hacking
16:42 < koan> wasn't minix on the T2 agenda? (at least on the website if I remember correctly)
16:43 < rxr> yeah - but minix is too protypish ...
16:44 < koan> protypish?
16:46 < rxr> classroom-like
16:48 < koan> well, I don't know of many microkernels that have escaped academia, maybe L4 comes the closest
16:49 < rxr> well - hybrid like dragonfly of haiku ar haiku are already a progress ...
16:49 < rxr> and less random and breaking changes another
17:08 < rxr> Baldzius: there is also a possible binutils update if you want to test it with some bigger builds * updated binutils (2.19.51.0.1 -> 2.19.51.0.10)
17:10 < Baldzius> I'll try to remember if I will start new ref build
17:10 < Baldzius> need to fix some packages first
17:17 < rxr> ok - no problem
18:06 -!- T-One [n=T-One@62.218.122.66] has quit [Read error: 110 (Connection timed out)]
18:30 -!- synchris_ [n=synchris@adsl108-49.kln.forthnet.gr] has joined #t2
18:31 -!- synchris [n=synchris@pdpc/supporter/professional/synchris] has quit [Read error: 60 (Operation timed out)]
20:11 -!- synchris_ is now known as synchris
20:30 -!- mqueiros [n=mqueiros@217.70.75.225] has joined #t2
20:43 -!- mpp [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has quit ["good night - good fight"]
20:48 < rxr> http://rene.rebe.name/2009-09-29/sony-vaio-vgn-p-nearing-refresh-or-end-of-life/
22:49 < CIA-38> aldas * r34337 /trunk/package/gnome2/rhythmbox/libmtp.patch: * removed obsolete rhythmbox/libmtp.patch
22:50 < CIA-38> aldas * r34338 /trunk/package/base/libx86/lheader.patch: * fixed libx86 to compile with new linux headers
23:02 < CIA-38> aldas * r34339 /trunk/package/network/aria2/ (. aria2.cache aria2.desc):
23:02 < CIA-38> Martin Papadopoulos
23:02 < CIA-38> * added aria2 (1.5.2) - aria2 - The ultra fast download utility
23:34 -!- Baldzius [n=aldas@87.198.133.94] has quit ["Leaving"]
--- Log closed Wed Sep 30 00:00:43 2009