[00:11] <lamont> cjwatson: well, the part of me wearing my buildd-guy hat would be tempted to just stub it out completely, but that would be, well, wrong.
[00:15] <cjwatson> lamont: it's a genuine request, I've spent a fair bit of time making that go faster (principally by optimising the hell out of process execution and trying to do as much in-process as possible) but perhaps other eyes could see something I missed
[00:17] <lamont> cjwatson: yeah - I'll add it to my list of things to do
[00:17] <cjwatson> I note idly that man-db can do full-text search of all manual pages on my laptop over seven times faster than Fedora's man package can
[00:18] <cjwatson> (and that's without either of them having a proper FTS engine in there, which would probably be a good idea)
[00:18] <cjwatson> (http://www.chiark.greenend.org.uk/ucgi/~cjwatson/blosxom/2009-07-14-man-db-K.html)
[00:22] <ScottK> ajmitch: :-)
[00:23] <ScottK> I only got involved with it because no one else was sorting it and one of the major users in Main is KDE.
[00:23] <ajmitch> ScottK: I've dropped the majority of changes that we had, but there's some black magic going on with dependencies from what I've seen
[00:23] <ajmitch> I only touched it because of boost-python being broken & it being a dependency of other packages
[00:25] <ScottK> ajmitch: If you want a package that uses the MPI stuff, elmerfem needs merging.
[00:25] <ajmitch> I may take a look
[06:26] <Joit> hi. where can i make suggestions for the install cd
[06:51] <dholbach> good morning
[07:28] <chilicuil> hhh
[07:33] <pitti> Good morning
[08:39] <geser> could an archive admin please kill http://people.canonical.com/~ubuntu-archive/NBS/rdoc1.8. ruby1.8 provides/conflicts/replaces rdoc1.8 and there are only unversioned dependencies left (and one versioned but ruby1.8 is the preferred alternative to rdoc1.8)
[08:40] <geser> and the same for http://people.canonical.com/~ubuntu-archive/NBS/rdoc1.9.1. ruby1.9.1 provides/conflicts/replaces rdoc1.9.1 and only unversioned dependencies left.
[08:43] <sabgenton> ubuntu servers usb install instructions http://www.ubuntu.com/server/get-ubuntu/download show how to install desktop
[08:44] <sabgenton> this is disconcerting as with say   the old  karmic this never worked for server editon
[08:44] <sabgenton> does this information even work?
[08:44] <sabgenton> (for server)
[08:46] <sabgenton> considering there was never in the  past support  for server on usb   http://www.ubuntu.com/server/get-ubuntu/download should be made a lot clearer if there really is support now.
[09:01] <bigon> robert_ancell: are you around?
[09:01] <robert_ancell> bigon, hi
[09:05] <robert_ancell> bigon, ?
[09:06] <bigon> robert_ancell: hi, about http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=550478 is this really needed? josselin was not happy with putting things for testing purpose only in the "production" package
[09:07] <pitti> is it just me, or has edge.lp been ridiculously slow since yesterday?
[09:08] <robert_ancell> bigon, it's proposed to be in the -dev package.  It is used by python-gi in their make check.  If we don't package it somewhere then that will always fail.
[09:09] <robert_ancell> bigon, since it is built and installed with make install, I'm guessing the gobject-introspection guys consider it useful in other projects.
[09:11] <bigon> robert_ancell: it's require the adding of an extra shared lib too
[09:12] <robert_ancell> bigon, does that matter?
[09:13] <cjwatson> sabgenton: please file a bug on https://bugs.launchpad.net/ubuntu-website about that
[09:14] <sabgenton> doing right now :)
[09:14] <cjwatson> sabgenton: (the webmaster isn't likely to be awake at this time and isn't in this channel anyway)
[09:14] <sabgenton> what chanel ?
[09:15] <sabgenton> cjwatson: where does he live on irc?
[09:17] <cjwatson> sabgenton: please use the bug tracking system, it's more efficient
[09:17] <sabgenton> k doing
[09:32] <ogra> pitti, hey, looking at udisks i see we dont have any ubuntu patches in it, i need to add one thats really specific to the image builds, so very ubuntuish, do you think upstream would accept that or should i just go ahead and add a quilt patch ?
[09:32] <pitti> ogra: depends on the patch
[09:33] <ogra> pitti, its adding another partition label to the hidden partitions in the udev rule
[09:33] <pitti> ogra: if it's sensible for Debian, I can commit it there; if it's sensible for upstream, I'd rather commit it there and then cherrypick to Debian and sync
[09:33] <ogra> its not sensible for either i think, since it fully depends on the way we build images
[09:34] <pitti> ogra: and we can't use one of the standard partition types for rescue partitions?
[09:34] <pitti>   ENV{UDISKS_PARTITION_TYPE}=="0x00|0x11|0x12|0x14|0x16|0x17|0x1b|0x1c|0x1e|0x27|0x3d|0x84|0x8d|0x90|0x91|0x92|0x93|0x97|0x98|0x9a|0x9b|0xbb|0xc2|0xc3|0xdd|0xef", \
[09:34] <ogra> omap images will have a vfat partition that fakes a NAND and is the first partition, i want to hide that on desktops
[09:34] <pitti> there's plenty to choose from, after all?
[09:34] <ogra> no, i cant change the type
[09:34] <ogra> it has to be a label
[09:34] <ogra> else the bootloader wont work
[09:35] <pitti> does the type/label come predefined by some vendor?
[09:35] <ogra> no, i can define the label, the type needs to be vfat bootable etc etc
[09:35] <ogra> (wioth a certain CHS layout even)
[09:35] <pitti> ogra: so, what's the new label?
[09:35] <pitti> ogra: we can't just use "Recovery Partition" or "RECOVERY"?
[09:35] <ogra> well, i'D call it OMAPBOOT or some such
[09:36] <pitti> ah
[09:36] <ogra> currently i use one of the SYSTEM* labels
[09:36] <ogra> works fine but its an abuse
[09:37] <ogra> if you think that abuse is fine i can also keep it but calling it OMAPBOOT or some such would be cleaner
[09:37] <pitti> ogra: I don't mind the "abuse" :)
[09:37] <ogra> ok, i'll keep the current name then
[09:37] <pitti> ogra: OTOH, any package can ship a rule to ignore it
[09:37] <pitti> ogra: i. e. if your boot loader package ships an udev rule to ignore OMAPBOOT, that's just fine
[09:37] <ogra> right, but that would mean a new package just for that
[09:38] <ogra> we dont install the bootloader packages :)
[09:38] <pitti> ok
[09:38] <ogra> flash-kernel could probably do it though ... hmm
[09:38]  * ogra will think about that, thanks pitti 
[09:38] <pitti> ogra: so, I don't mind which label it has; I'd just like to avoid a delta to Debian just for this
[09:38] <ogra> yeah, understood
[09:39] <pitti> so if it's necessary, I'd rather commit it to the Debian tree
[09:39] <pitti> (need to ask mbiebl for that, since it's debatable)
[09:39] <ogra> well, its very ubuntu specific
[09:40] <ogra> so i dont think debian wants it
[09:40] <pitti> other distros don't use that boot loader?
[09:40] <pitti> (meego, etc.)
[09:40] <ogra> its bound to the way we build the new images which i assume debian will never adopt )preinstalled images)
[09:41] <pitti> ogra: in general such things should be maintained in one place; i. e. if your image build system defines $LABEL, it could also stick in an udev rule file which ignores that very $LABEL :)
[09:41] <ogra> the new images are two partition images where we resize the second partition on the fly to the full size of the disk while the first partition has to follow a ceratin scheme
[09:41] <pitti> echo ... > /etc/udev/rules.d
[09:41] <ogra> right, butu the tool i use gets removed with oem-config
[09:41] <geser> cjwatson: could you please kill rdoc1.8 and rdoc1.9.1 from the NBS list? or should I file a bug for it?
[09:42] <ogra> and i dont really want a file on the FS thats not handled by a package
[09:42] <pitti> ogra: /etc/udev/rules.d/ already has a few
[09:42] <ogra> but rolling a package only for the rules file seems overkill
[09:42] <pitti> persistent network, etc.
[09:42] <ogra> hmm
[09:42] <cjwatson> geser: one moment
[09:42] <pitti> ogra: no, if nothign else, the image build process should do that along with creating the partitions -> define stuff in one place
[09:43] <ogra> pitti, well, then i can also do it from jasper-initramfs during fiorst boot
[09:43] <BlackZ> pitti: did you read my e-mail regarding cdbs?
[09:43] <cjwatson> geser: done
[09:43] <geser> thanks
[09:43] <pitti> BlackZ: not yet, bit busy right now; I'll get to it
[09:43] <jmux> Hi. I'm on 8.04 LTS using upstart. I have a dhclient-exit hook, which fails, because udev starts ifplugd, which starts dhclient, which calls the hook before I have a writable local fs. How can I wait for the upstart writable fs event inside my dhclient exit hook?
[09:43] <BlackZ> pitti: thanks
[09:43] <pitti> BlackZ: I'm not in the ubuntu distro team this cycle, so I'm lagging on distro stuff, sorry
[10:02] <apw> cjwatson, was it you i was talking to about ti-omap in maverick, that we were now building those from the main package ?
[10:09] <cjwatson> apw: yeah, I hadn't got round to doing anything about the result of that conversation yet ...
[10:11] <apw> cjwatson, ahh ok then that makes sense... a quick Q on how to migrate the meta package, i _think_ i don't need anything speciial (replaces et al) to produce a replacement binary package at a higher version number from a different source package ... is that correct ?
[10:11] <sabgenton> is there any reason I should have my bug  ticked private?
[10:12] <apw> sabgenton, if apport ticked it private then it would likely mean that it thinks it added some backtrace information which may contain private information and it cannot tell
[10:12] <cjwatson> apw: correct
[10:12] <apw> cjwatson, thanks
[10:12] <sabgenton> apport?
[10:13] <sabgenton> sorry new to launchpad
[10:13] <apw> apport is the automatic bug filer, like from application crashes
[10:13] <apw> if its just your own bugs, then normally not
[10:14] <sabgenton> so if i want my bug to get traction I should make it unprivate?
[10:14] <sabgenton> I just typed the bug up my self on the website
[10:15] <apw> if you made it private, then likely  only you can see it, so noone else is even aware of it
[10:16] <cjwatson> private bugs are for (a) security bugs that need to not be disclosed to the public as yet (but in that case tick the security box so that the security team get to see it) (b) bugs with private information like crash dumps from your mail reader (but apport deals with subscribing some people so that they aren't entirely lost) (c) commercially-sensitive bugs, such as bugs from a hardware manufacturer on unreleased hardware
[10:17] <cjwatson> a private bug with no explicit subscribers other than yourself is invisible
[10:17] <cjwatson> (well, to everyone other than Launchpad admins, but they'll probably never look at it)
[10:19] <achaemenes> hi guys, im running 2.6.24-28-server on 8.04 - i want to get the kernel source, as-is currently running, disable CONFIG_DEBUG_RODATA, and build a package
[10:19] <achaemenes> i dont want the latest from git, i want the exact version, and all i want to do is rebuild it with that single change
[10:20] <apw> achaemenes, there are two ways to do that, but i would normally get the git source, the version you have is tagged in that
[10:20] <sabgenton> thanks for  the info
[10:20] <apw> achaemenes, we tag every released version
[10:20] <achaemenes> oh, i just installed linux-source-2.6.24
[10:20] <sabgenton> acording to what I understand now I'll unprivate it
[10:20] <apw> achaemenes, that is a raw source and not that useful for building
[10:21] <achaemenes> right, ok ill do a git clone then
[10:21] <apw> achaemenes, if you have trouble you can find us over on #ubuntu-kernel
[10:21] <achaemenes> excellent - thankyou apw
[10:22] <sabgenton> grr it won't  unprivate
[10:23] <sabgenton> ajax symbol spins but just  says error
[10:23] <sabgenton> probly my slow conection I guess
[11:15] <Joit> hi. where i can make suggestions for the install cd?
[11:15] <cjwatson> the bug tracking system, usually
[11:16] <Joit> ok thanks
[11:16] <Joit> and bye :)
[11:16] <cjwatson> there's #ubuntu-installer if you want to discuss things first, but IRC is in general not so good for anything you want to be remembered
[11:17] <Joit> ok i will try this channel too
[11:17] <Joit> well its actually only a option what is to add
[11:22] <tkamppeter> pitti, hi
[11:22] <pitti> hello tkamppeter
[11:24] <tkamppeter> pitti, what does the reporter of bug 597248 has to enter to get Apport making a symbolic backtrace of his segfault.
[11:25] <pitti> tkamppeter: easiest would be to enable apport for the current session and replicate the crash, and have apport pick it up; https://wiki.ubuntu.com/Apport#How%20to%20enable%20apport
[11:25] <pitti> tkamppeter: if he doesn't want that, the fallback is to install debug packages etc. locally and run under gdb, see https://wiki.ubuntu.com/DebuggingProgramCrash
[11:31] <tkamppeter> pitti, thank you.
[12:10] <Riddell> ev, cjwatson: as you may or may not have gathered we'd like to merge the kubuntu desktop and netbook images and load the correct desktop based on screen size, where is the best place to detect screensize and edit the config file on starting the live CD and installing the system?
[12:13] <cjwatson> somewhere in casper I suppose
[12:28] <thangam_arun> Hello all
[12:28] <thangam_arun> i am trying to trebuild kernel on ubuntu 8.10
[12:29] <thangam_arun> but i am getting only 2 .deb files, image and header
[12:30] <thangam_arun> is there any body can direct me to the right path ??
[12:30]  * ogra directs thangam_arun to #ubuntu-kernel :)
[12:31] <thangam_arun> Thanks
[14:37] <ccheney> doko, hello?
[14:38] <doko> ccheney: ?
[14:38] <ccheney> doko, did you happen to see my comments about the OOo MIRs that need processing?
[14:38] <ccheney> doko, they were assigned to you on jun 7
[15:37] <ccheney> pitti, hi
[15:38] <ccheney> pitti, iirc you are the person who assigns MIR bugs to people to process, is that correct? there is a server MIR I need to have processed for a2, bug 594372
[15:38] <pitti> ccheney: hello
[15:38] <pitti> ccheney: that's asac now (since I'm OEM this cycle)
[15:38] <ccheney> pitti, ah ok
[15:38] <ccheney> asac, ping, see above :)
[15:38] <pitti> :)
[15:38] <ccheney> pitti, have a great day :)
[15:39] <pitti> ccheney: thanks, you too!
[15:39] <pitti> in fact, was a pretty good "boot faster" day so far
[15:46] <asac> pitti: hey
[15:47] <pitti> hello asac, how are you?
[15:47] <asac> pitti: quite well ;)... hope you too! did you discuss with other linaro folks on how we can use your "reduce disk footprint"
[15:47] <pitti> oh, was I supposed to?
[15:47] <pitti> asac: I moved the stuff to the public wiki now, so it should be easy to grab
[15:47] <asac> pitti: no ;) ... just wanted to check so i dont ask questions already discussed
[15:48] <pitti> https://wiki.ubuntu.com/ReducingDiskFootprint
[15:48] <asac> pitti: cool. did the dpkg changes land in maverick?
[15:48] <pitti> it links to all the bzr changes, patches, etc
[15:48] <pitti> asac: not yet, just in debian's trunk so far
[15:48] <pitti> and I have a backport for lucid
[15:48] <asac> pitti: is it risky to make that available in maverick?
[15:48] <pitti> no, it's fairly well tested by now
[15:48] <pitti> we have run that in an OEM project for two weeks
[15:49] <asac> pitti: ok we have a https://blueprints.edge.launchpad.net/ubuntu/+spec/arm-m-disk-footprint spec
[15:49] <asac> pitti: i would like to give you an item to make the dpkg  available in the archive ... if you dont mind ;)
[15:49] <pitti> asac: the apt changes are interesting as well, but I'd like mvo to review the branch first
[15:49] <pitti> that didn't land upstream yet
[15:49] <pitti> asac: sounds fine
[15:49] <pitti> please do
[15:49] <asac> pitti: ok. i will read through your wiki and reuse most of that for the footprint spec which needs some love
[15:50] <asac> pitti: thanks!!
[15:50] <asac> stay tuned
[15:50] <pitti> gern geschehen
[15:54] <asac> ccheney: pushed that MIR to mterry
[15:54] <ccheney> asac, thanks!
[15:54] <mterry> asac, I have been remiss with MIRs, sorry
[15:54] <asac> mterry: heh. well, we all failed ;)
[15:54] <asac> mterry: do you have capacities at all atm?
[15:55] <mterry> asac, not really.  maybe this weekend
[16:17] <tseliot> pitti: do you think it would make sense if I made the -dev packages of the different nvidia packages conflict with each others? This would allow me to install the headers in (/usr/include) e.g. /usr/include/GL instead of what I'm currently doing /usr/include/nvidia-current/GL (so that all of them can be installed at the same time)
[16:18] <tseliot> slangasek: ^
[16:46] <pitti> asac: who is handling the firefox translation bits in langpack-o-matic these days?
[16:49] <asac> pitti: we said chrisccoulson_ would take that over through learning by doing
[16:49] <asac> e.g. whenever there is a need we would discuss and see how things go
[16:49] <pitti> asac: thanks
[16:49] <pitti> asac: I got a bug about adding a new language; I'll assign to Chris and sub you, does that sound ok?
[16:49] <asac> pitti: the po2xpi processor was improved by ArneGoetje in the past ... so if the problem is there he might be a good candidate
[16:49] <asac> pitti: yeah. tahts good
[16:49] <pitti> shoudl be easy, like adding a new file somewhere
[16:49] <asac> pitti: though ArneGoetje should be able to handle that too iirc
[16:50] <asac> but chrisccoulson_ should definitly be invovled so he get that knowledge too
[16:50] <pitti> *nod*
[17:09] <slangasek> tseliot: I think that makes sense, yes
[17:09] <tseliot> slangasek: ok, thanks
[17:11] <vish> pitti: hi, the sd/mmc card is also verified bug , previously we had a bigger numbering there , it is not visible now at normal view. which fixes the bug for most users. we will however be adding more icons in larger sizes for maverick which will solve the problem for users trying to zoom to 400% :)
[17:11] <vish> the humanity update ^
[17:12] <pitti> vish: ah, I see; please tag it as verification-done then (sorry, busy right now0
[17:12] <vish> pitti: cool , i didnt know i could do that. wanted to check with you first. Thanks
[17:12]  * vish doing it now
[17:43] <JFo> bryce_, bug 573601 looks like it is solved by some configuration in an X config. does this need to go back to an X package?
[17:44] <bryce_> *looking*
[17:45] <bryce_> JFo, yeah refile to nvidia-graphics-drivers if it looks like something a dev should look at
[17:45] <bryce_> JFo, if it's purely a configuration goof by the user, forward to answers or close as not a bug
[17:46] <JFo> hmm, not sure
[17:46] <JFo> he added some stuff and it fixed him
[17:46] <JFo> dunno where the goof was
[17:46] <JFo> I'll send it up to the dev
[20:34] <dupondje> pitti: you here ?
[20:58] <JontheEchidna> So, looks like libept broke API something awful yesterday. The textsearch class used for xapian is totally obliterated and left as two nearly-useless functions for getting the database path and the database timestamp
[20:58] <JontheEchidna> Since synaptic uses that class, somebody's gonna have to take care of it sooner or later. This is what I did for libqapt: http://pastebin.com/5rw3XwBV
[20:59] <JontheEchidna> It's not worth it to link to libept anymore to get a string and a time_t, imo
[21:01] <JontheEchidna> oops, wrong diff. http://pastebin.com/n5xKSqAg
[21:03] <JontheEchidna> Basically all one has to do is make member variables for the timestamp and the Xapian::Database, initialize them, then rip the expand() function out of ept::textsearch and slap it in the code somewheres
[21:05] <JontheEchidna> I'd contribute a patch, but I'm not in-tune with the build-systems of the G* world, given my K background. ;)
[21:10] <hdon> hello hard working ubuntu devs :) i submitted a bug report affecting "ALT+TAB" window switching -- a critical feature for developers and power users -- a long time ago but nobody seemed to notice it because, i believe, they were skeptical it could be reproduced. the bug still exists in Karmic Koala, and i have now updated the bug with a video demonstrating the problem: https://bugs.launchpad.net/ubuntu/+bug/405252
[21:11] <micahg> hdon: #ubuntu-bugs is more appropriate to have bugs triaged in
[21:12] <hdon> micahg, thanks for the advice
[23:31] <jono> is GIMP continually crashing for you all?
[23:38] <ddecator> the stable version was working for me earlier (on lucid)
[23:39] <arand> jono: 10.10? Randomly or on start?