T2 IRC Log: 2007-08-25

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 Sat Aug 25 00:00:10 2007
01:37 -!- sepp [n=sepp@Z73a2.z.pppool.de] has joined #t2
01:55 -!- sepp_ [n=sepp@Z74fc.z.pppool.de] has quit [Read error: 113 (No route to host)]
02:12 -!- Ragnarin is now known as Ragnar|away
02:43 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
02:43 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
03:24 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
03:25 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
07:01 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
07:01 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
09:29 < rxr> re
09:37 < CIA-24> susan * r25754 /trunk/package/gnome2/glabels/glabels.desc: * updated glabels (2.1.3 -> 2.1.4)
09:37 < CIA-24> susan * r25755 /trunk/package/security/gsasl/gsasl.desc: * updated gsasl (0.2.20 -> 0.2.21)
09:39 < CIA-24> rene * r25756 /trunk/package/network/curl/curl.desc: * updated curl (7.15.5 -> 7.16.3)
09:39 < CIA-24> rene * r25757 /trunk/package/network/neon/neon.desc: * updated neon (0.25.5 -> 0.26.4)
09:41 < CIA-24> rene * r25758 /trunk/package/base/binutils/binutils.desc: * updated binutils (2.17.50.0.15 -> 2.17.50.0.18)
09:47 < rxr> valentin: http://www.fefe.de/c++/c++-talk.pdf
09:51 < cyda> hey rxr
09:52 < cyda> i would say "moin" but am about to go to bed in the next hour :-p
09:54 < rxr> hey cyda !
09:54 < rxr> sleep well then! and cu in the morning :-)
10:06 < rxr> valentin: http://www.fefe.de/know-your-compiler.pdf
10:18 < CIA-24> rene * r25759 /trunk/package/develop/libowfat/libowfat.desc: * updated libowfat (0.25 -> 0.26)
10:19 < CIA-24> rene * r25760 /trunk/package/network/gatling/ (gatling.desc libowfat.patch): * updated gatling (0.8 -> 0.9)
10:28 < CIA-24> rene * r25761 /trunk/package/base/dietlibc/ (4 files): * updated dietlibc (0.30 -> 0.31)
10:37 < CIA-24> rene * r25762 /trunk/package/base/embutils/ (embutils.desc mount.patch): * updated embutils (0.18 -> 0.19)
11:13 < rxr> heh - embutils mount managed to mess up my /proc/mounts
11:14 < rxr> # tail -n 1 /proc/mounts
11:14 < rxr> G` / sysfs rw 0 0
11:14 < rxr> ^[[?1;2c
11:14 < rxr> and yes, this junk is the content of the proc file, now ...
11:30 < CIA-24> rene * r25763 /trunk/package/base/embutils/mount.patch: * fixed one more memory clobber bug in embutils/mount
11:46 < DeeKey> RXR: Going to reinstall my desctop system.
11:46 < DeeKey> I remember reading your article about different file systems. Could you point the adress to me?
11:46 < CIA-24> rene * r25764 /trunk/package/base/u-boot/u-boot.conf: * removed debugging set -x/+x from u-boot.conf
11:46 < rxr> DeeKey: it's on the uni box that is down :-(
11:46 < rxr> basically it was XFS oopses for me, JFS was good in the beginning, but after trashing dog slow
11:46 < DeeKey> I see.. Put it simplier what was the best?
11:46 < rxr> I continue go use Riser3 for now ...
11:46 < DeeKey> :)
11:47 < rxr> JFS has the feature of no journal replay in kernel. keeps the kernel code tiny, but requires an userspace fsck.jfs to boot on an unclean root partition ...
11:47 < rxr> but less kernel code is nice (tm) in my opinion
11:47 < rxr> ext{2,3,4} is too slow for me (tm)
11:48 < rxr> and reiser4 did not oops unlike XFS, but o course has an uncertain future
11:48 < rxr> reiser4 is way fastest for T2 system build workload, but reiser3 comes next and has been rock solid for me over the past 5-6 years
11:48 < rxr> not a single incident since the 2.6.10 or what it was NFS glitch
11:49 < DeeKey> I am also unhappy with ext3... Too slow...
11:50 < DeeKey> and unreliable..
11:50 < DeeKey> had some problems with powerofs...
11:50 < rxr> well - with noatime and some journal tweak option ext3 was not that bad ...
11:50 < rxr> btw. you probably want to run noatime
11:51 < rxr> http://kerneltrap.org/node/14148
11:51 < DeeKey> noatime is also applicable to reiserfs?
11:51 < rxr> yes for all of them
11:51 < rxr> massive speedup
11:51 < DeeKey> ok thanks...
11:52 < rxr> the above article / thread is a good read
11:52 < rxr> you need atime for nearly nthing
11:52 < rxr> nothing
14:57 < rxr> re
14:58 < rxr> China späht angeblich PCs des Bundeskanzleramtes aus
14:58 < rxr> ^-> http://www.heise.de/newsticker/meldung/94899
15:02 < CIA-24> rene * r25765 /trunk/package/emulators/wine/wine.desc: * updated wine (0.9.43 -> 0.9.44)
15:38 < sepp> do they have no better target?
15:39 < rxr> maybe the US runs SELinux isnstead of Windows ...
15:40 < sepp> they talk all the time about everything, can there still be some secret?
15:55 < rxr> re
16:10 < CIA-24> rene * r25766 /trunk/package/base/linux26/ (disable-broken.lst mactel.patch):
16:10 < CIA-24> * added minimal boot-only USB keyboard and mouse driver to
16:10 < CIA-24> linux26/disable-broken.lst, the mouse one binds to the apple
16:10 < CIA-24> touchpad and prevents it from working properly, generally
16:10 < CIA-24> this driver are way too limitted for normal use - enable
16:10 < CIA-24> exlicitly in the target if needed
16:10 < CIA-24> * updated macbook patch from upstream site
16:11 < rxr> so - finally my MacBook is more usable with trunk again
16:11 < rxr> totally messed up touchpad sucked a little when abroad ...
17:01 -!- tri [n=tri@p57AD9E43.dip0.t-ipconnect.de] has joined #t2
17:03 < CIA-24> rene * r25767 /trunk/package/network/madwifi/ (eth-naming.patch madwifi.desc no-find-previous.patch): * updated madwifi (0.9.30.13-r2351-20070519 -> r2669), mostly to build with linux-2.6.22.*
17:08 < tri> moin
17:10 < rxr> moin tri
17:12 < rxr> valentin: I have most stuff together for smooth support of our MacBook in trunk
17:13 < rxr> valentin: so if you want a major leap ahead sync let me know
17:19 < CIA-24> rene * r25768 /trunk/package/network/iwlwifi/ (iwlwifi.conf iwlwifi.desc postlinux.conf):
17:19 < CIA-24> * updated iwlwifi (0.1.1 -> 0.1.11), updated iwlwifi-3945-ucode and added iwlwifi-4965-ucode
17:19 < CIA-24> * fixed to build with it's new "auto config / re-run make" setup
17:35 < CIA-24> rene * r25769 /trunk/package/base/sdricoh_cs/sdricoh_cs.desc: * updated sdricoh_cs (2007-04-20 -> r18), mostly to build with linux-2.6.22.*
17:42 < tri> if i try to compile openoffice i get: configure: error: Library requirements (gstreamer-0.10 gstreamer-plugins-b... ? any hits what i need - gstreamer is already installed ?
17:45 < rxr> not really
17:45 < rxr> I think I got thru that part
17:46 < rxr> but it stopped for me anyway at some visual basic stuff IIRC
17:47 < tri> he ?
17:55 < tri> you use koffice ?
18:00 < rxr> we use a mix of koffice and the crappy gnomes wanna-be's, abiword and gnumeric
18:00 < rxr> but susan kicks my ass three (3) times a week to get openoffice compiling again
18:00 < rxr> because, sadly, the kde and gnome tools are not really sufficient for our daily office use and stress our nerves quite badly ...
18:01 < tri> :-)
18:01 < rxr> however, openoffice is zero % fun to beat to build and I have more important stuff on my overly long TODO :..
18:01 < rxr> but yes, we really should get this ooo monster to build, again
18:02 < tri> i think first we need a build file for http://gstreamer.freedesktop.org/src/gst-plugins-base/ to avoid the gstreamer-plugin-base error
18:03 < tri> hm i this this is already there...
18:03 < tri> i see
18:04 < tri> have to leave now - cu ;-)
18:16 < rxr> cu
18:33 [Users #t2]
18:33 [@ChanServ] [ DeeKey] [ happycube] [ mqueiros_] [ Ragnar|away] [ TobiX ]
18:33 [ CIA-24 ] [ emte ] [ kic- ] [ mtr ] [ rxr ] [ tri ]
18:33 [ cyda ] [ Fatal ] [ LMJ ] [ pdinoto_ ] [ sepp ] [ valentin]
18:33 -!- Irssi: #t2: Total of 18 nicks [1 ops, 0 halfops, 0 voices, 17 normal]
18:36 * rxr right now hacking a dynamically generated package matrix including the stable trees ...
18:52 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
18:52 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
19:14 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
19:14 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
19:17 < rxr> Generated: 2007-08-25 19:15 in 0.37s —
19:17 < rxr> ^- new CGI code on my laptop ...
19:17 < rxr> (package matrix)
19:18 < mqueiros_> rxr: Is that the code to allow to see the packages for different versions ?
19:19 < rxr> yep
19:20 < rxr> well - I could of couse also create static versions for the other t2 stable trees
19:20 < rxr> but the static generator already run for 15 (!!!) minutes on the aging athlon server
19:20 < rxr> (well - the downsite of shell code :-)
19:20 < rxr> and I wanted to have this pressing to go CGI for all trees at once
19:21 < rxr> should be ready in some minutes
19:21 < mqueiros_> ok, nice...
19:21 < rxr> probable around 20:00
19:21 < mqueiros_> ok, I will probably be around if you need any kind of testing...
19:22 < rxr> Generated: 2007-08-25 19:21 in 4.83s —
19:22 < rxr> ^- heh - cold cash read of 3000 files .-)
19:23 < rxr> (laptop HD)
19:44 < rxr> so - aboot dynamic gen: Generated: 2007-08-25 19:42 in 0.20s
19:49 -!- DeeKey [n=denis@84-50-17-207-dsl.kjj.estpak.ee] has quit [Read error: 110 (Connection timed out)]
19:53 < rxr> soooo
19:54 < mqueiros_> is it done ? :)
19:55 < mqueiros_> I'm looking at it now...
19:56 < rxr> hunting some quirks
19:56 < rxr> but yes
19:56 < rxr> the sub-trees
19:57 < rxr> need some fix
19:57 < rxr> one minuted
19:57 < rxr> I drop a note when I have that fixed
19:57 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
19:57 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
19:59 < rxr> ah /me stupid
20:01 < rxr> ok
20:03 < rxr> mqueiros_: you can go check it out
20:04 < rxr> two things that I see so far: sometimes there are stray "-->" in the page
20:04 < rxr> and the indexes are not yet as complete as before
20:04 < rxr> formerly the index included the version number
20:04 < rxr> I guess we want that back ?=
20:04 < rxr> I skipped that for now as it would take longer to render the index then ...
20:05 < rxr> but just hardcoding th grab the version number would still be ok
20:05 < rxr> and we have a frontside cache proxy on that server anyway, we could configure the proxy to cache t2-project.org as well
20:05 < mqueiros_> yes, we want version numbers back... :)
20:06 < mqueiros_> and theres should be something like "T2 Project - Trunk", "T2 Project - 7.0"... etc... at the top of the page...
20:06 < mqueiros_> anyway, looking good...
20:07 < rxr> do yo also sometimes ahve --> before the table on the index page ?
20:07 * rxr wonders where they come from ...
20:08 < mqueiros_> no... I don't.
20:09 < rxr> hm - maybe it is khtml misreading my commented version table entreies
20:09 < rxr> strange that the number of --> in the page of konquerer vary based on reload count ...
20:09 < rxr> strange
20:09 < rxr> ok - I hack the version grabbing in ...
20:09 < mqueiros_> I'm using Firefox...
20:09 < rxr> the time spent generting the html is included in the header ...
20:10 < rxr> mqueiros_: yes, in firefox I do not see them neither
20:10 < rxr> also I se no matching code that could produce them in the konqueror source view ...
20:14 < mqueiros_> You do have a "" ont he source... but... It should not cause any trouble...
20:14 < mqueiros_> but you do not close the table row tag...
20:18 < rxr> ok
20:18 < rxr> with version it takes long
20:19 < rxr> 45.41s
20:19 < rxr> on m core2duo lapto with SATA laptop HD
20:19 < rxr> ok - with hot cache:
20:19 < rxr> 6.81s —
20:21 < rxr> guess I'll have fun configuring our cache to also cover http://www.t2-project.org/ ...
20:21 < rxr> so - you shoudld see the changes
20:22 < mqueiros_> yes... I'm looking at it...
20:22 < rxr> heh 12s on the old Athlon server
20:22 < rxr> 17.57s even
20:22 < rxr> ok - that needs some sort of cache
20:23 < rxr> ok - now mostly better headlines listing the tree are missing
20:23 < rxr> (beside index.html performance ...)
20:24 < mqueiros_> what about the "T2 Project - Trunk", "T2 Project - 7.0"... ?
20:24 < rxr> yes
20:24 < rxr> and I see another error
20:24 < rxr> the source link is mangled
20:24 < rxr> and needs to be coded to honor the tree ...
20:24 < rxr> like trunk,branches/2.1, ...
20:27 < mqueiros_> yes, and the information on "package source" also needs to be corrected...
20:31 < rxr> yeah - that is what I ment at: "20:24 < rxr> the source link is mangled"
20:31 < rxr> nearly ready ...
20:33 < mqueiros_> ok... thought you where talking about the "T2 Source:" info...
20:34 < rxr> yeah - all of them
20:35 < rxr> and the package headline was not the one I wanted ...
20:36 < rxr> hehe - ok the headline even sucked in the old version ... :-)
20:38 < rxr> ok - tree is online
20:39 < rxr> now I fixed the source reference
20:41 < mqueiros_> In trunk I still get the wrong package source info...
20:42 < mqueiros_> and on 6.0... too...
20:44 < rxr> yes, I still fix that
20:45 < rxr> I change the title and text to include a $tree reference
20:45 < mqueiros_> yes, noticed that... It's good.
20:51 < rxr> so source links should finally be fine, can you confirm ? :-)
20:52 < rxr> the packages now have also the verison in the title, as well as the short package description
20:52 < rxr> (that is a new feature even the old staic pages did not had)
20:57 < mqueiros_> I'm going to take a look...
20:58 < mqueiros_> the source links are ok!
20:59 < rxr> any other thing that comes to mind ?
21:01 < mqueiros_> no, every thing seems fine...
21:01 < rxr> ok - I announced to the t2 list, if anything pops up just let me know :-)
21:02 < mqueiros_> ok!
21:03 < rxr> nice, even the dependency links are fine
21:03 < rxr> I did not even took care of them
21:03 < rxr> noticed them just now :-)
21:04 < rxr> are they are relative - nicer even
21:04 < rxr> ok - great
21:04 < rxr> I go home
21:07 < mqueiros_> I did noticed that ! :)
21:07 < mqueiros_> It's looking good!
22:30 < tri> re
22:40 -!- tri [n=tri@p57AD9E43.dip0.t-ipconnect.de] has left #t2 []
22:48 -!- tri [n=tri@p57AD9E43.dip0.t-ipconnect.de] has joined #t2
23:14 -!- tri [n=tri@p57AD9E43.dip0.t-ipconnect.de] has left #t2 []
--- Log closed Sun Aug 26 00:00:11 2007