[12:52] <ScottK> lamont: Fix Released with an open task for Dapper if it's SRU worthy.
[12:59] <lamont> ScottK: uh... "open task for dapper".. I assume there's a "create task" label somewhere on the page, eh?
[12:59] <Fujitsu> lool: `Nominate for release', IIRC.
[12:59] <Fujitsu> Bah, lamont ^^
[01:01] <lamont> ok
[01:03] <ScottK> WHat Fujitsu said.
[01:22] <pochu> lamont: you can't do it without nominating: change the url from lp.net/ubuntu/+source... to lp.net/ubuntu/dapper/+source..., and then click on "This bug also needs fixing here" (or something like that)
[01:22] <pochu> s/can't/can/
[01:22] <pochu> That way you don't need a driver to approve it... (dunno if you're a driver though hehe)
[01:23] <Fujitsu> pochu: It's not drivers - that piece of text is entirely wrong.
[01:23] <Fujitsu> It's uploaders to the component.
[01:24] <pochu> I thought it was drivers who had to approve them...
[01:24] <pochu> Oh, it was some time ago, wasn't it?
[01:24] <LaserJock> it used to be that core-devs where drivers
[01:24] <pochu> But there was a spec... motu-bugs-permissions, or something like that
[01:24] <LaserJock> but they've changed it to be uploaders to the component
[01:24] <LaserJock> right
[01:25] <Fujitsu> pochu: motu-bug-persmissions [sic] 
[01:26] <pochu> Anyway, I can also 'approve' a nomination for a bug, with that workaround ;)
[01:26] <Fujitsu> pochu: Yep, their permissions are very well enforced.
[01:38] <yipe> Thank you, thank you, thank you a thousand times thank you devs for gutsy!
[01:38] <yipe> I love it already!
[02:18] <cjwatson> davmor2: ctrl-alt-f1, 'dmesg | grep BUG' please
[02:18] <cjwatson> slangasek: I agree with you, "upload and have archive admins approve when ready" is a better process
[02:34] <gnomefreak> cjwatson: i had daily ISO i think tuesdays freeze at 15% during install if it happens again i should beable to save the files to usb stick right? so i can add to bug report?
[02:35] <cjwatson> gnomefreak: we know Tuesday's was hopelessly busted
[02:35] <cjwatson> unionfs was broken
[02:35] <jdong> usb works for me today
[02:35] <cjwatson> anything before this afternoon's build is not interesting
[02:35] <gnomefreak> cjwatson: ah ok cool
[02:35] <jdong> cjwatson: as I commented earlier today unionfs is still a bit busted on .1 today...
[02:35] <cjwatson> jdong: yes but it's not clear how bad yet
[02:35] <cjwatson> (it's working fine for me so it's not a universal problem)
[02:35] <gnomefreak> i will grab mondays/tuesdays than and test
[02:35] <jdong> it was bad enough that I couldn't get ubiquity to load :(
[02:36] <jdong> I posted a dmesg on the bug report
[02:36] <cjwatson> ok, I can't do anything right now and don't intend to, I have a zillion other things to fix before bed
[02:36] <cjwatson> your concern is noted for Monday :-)
[02:36] <jdong> no sweat, no hurry
[02:36] <jdong> just wanted to make sure you guys were aware of it :)
[02:36] <cjwatson> there is lots of hurry, beta is five days
[02:36] <jdong> but sleep first :)
[02:37] <cjwatson> no, fix wubi first
[02:37] <cjwatson> then sleep
[02:37] <jdong> you're a trooper
[03:43] <mthaddon> LP going down for maintenance in 15 minutes for 1 hour
[03:50] <calc> looks like i can't get rid of the java deps for ooo, but was able to disable the system lp-solve which saved some space
[03:50] <cjwatson> mthaddon: bazaar.lp.net isn't down already, is it?
[03:50] <cjwatson> I get ECONNREFUSED from it
[03:51] <calc> so we only ended up saving ~ 4MB so far with tweaking ooo
[03:51] <mthaddon> cjwatson, it is, yes - timing was for the web front end
[03:51] <cjwatson> argh
[03:51] <cjwatson> mthaddon: what's this downtime for? I thought this week was meant to be clear
[03:51] <calc> cjwatson: wow you are up early
[03:51] <cjwatson> calc: late
[03:51] <cjwatson> I was hoping to get this code committed before the downtime
[03:51] <mthaddon> cjwatson, bug fix post 1.1.9 release - I think next week is meant to be the clear week
[03:52] <cjwatson> mthaddon: beta is Thursday
[03:52] <cjwatson> we need a stable and available Launchpad
[03:53] <kiko> hi there
[03:53] <kiko> I know
[03:53] <cjwatson> this downtime will significantly reduce the chances of the Windows installer work making it into gutsy, since I was trying to land the last few fixes so that we could test them tomorrow morning
[03:53] <cjwatson> (the last few fixes that we know about)
[03:53] <cjwatson> can it at least be delayed an hour or two?
[03:54] <kiko> cjwatson, I guess that's up to mthaddon. I have already been awake for some 18 hours, I can wait another two. :)
[03:54] <kiko> let me chat with him.
[03:54] <cjwatson> I'm up late too
[03:54] <kiko> I know
[03:55] <cjwatson> kiko: I realise bug fixes are often exceptions, which is fair enough, but could you clarify whether this week or next is the clear week?
[03:55] <cjwatson> (aside from this particular case)
[03:55] <kiko> cjwatson, it's next week.
[03:55] <kiko> next week is week zero.
[03:55] <cjwatson> so the week after beta?
[03:56] <kiko> cjwatson, it is likely the rollout will not take a full hour
[03:56] <cjwatson> oh, by "week" you mean Mon-Fri?
[03:56] <kiko> the week that starts on the 24th
[03:56] <kiko> yes
[03:56] <cjwatson> sorry, I'm so used to thinking in Ubuntu weeks which start and end on Thursday :-)
[03:56] <kiko> heh
[03:56] <cjwatson> ok, that's cool
[03:56] <cjwatson> kiko: I can't realistically stay awake until 4am
[03:57] <kiko> cjwatson, how much more time do you need?
[03:57] <calc> cjwatson: inject some caffeine ;)
[03:57] <cjwatson> kiko: about 20-30 minutes I think
[03:57] <cjwatson> it's just a matter of landing commits and uploading; we can turn off the publisher at this point if the uploader stays up
[03:58] <cjwatson> (so that Tom isn't blocked on waiting for the publisher to stop)
[03:59] <kiko> cjwatson, let's try and do the rollout quick so you're not blocked. it shouldn't take more than 15 minutes tbh, and if it does we can abort. deal?
[04:00] <cjwatson> kiko: ok, I can cope with that
[04:00] <kiko> thanks
[04:00] <cjwatson> I'll go and get more coffee
[04:00] <kiko> seems like I owe many beers tonight
[04:03] <Riddell> davmor2: pong
[04:03] <Riddell> davmor2: evand does ubiquity
[04:05] <cjwatson> Riddell: I answered him
[04:12] <kiko> cjwatson, it's going smoothly, should be soon now.
[04:13] <kiko> the database updated fine.
[04:13] <cjwatson> great
[04:17] <mthaddon> cjwatson, should be back up
[04:17] <cjwatson> great
[04:18] <mthaddon> please check and let me know
[04:18] <cjwatson> yep, works fine now
[04:18] <mthaddon> cool
[04:18] <cjwatson> thanks for your speed, I appreciate it
[04:18] <kiko> cjwatson, thanks so much.
[04:18] <cjwatson> sorry to make a fuss :-/
[04:18] <mthaddon> np
[04:54] <lamont> interesting that the unaccepted queue has -12.37.{diff.gz,dsc} and -12.38_source.changes....
[04:55] <kiko-zzz> cjwatson, everything going okay?
[05:09] <cjwatson> kiko-zzz: yep, all fine now
[05:09] <cjwatson> about to fall over
[05:09] <cjwatson> lamont: probably tomorrow after CD builds
[05:09] <kiko-zzz> cjwatson, good job. congrats :)
[05:09] <lamont> ah, right.
[05:09] <lamont> I suppose I'll live.
[05:10] <cjwatson> kiko-zzz: we'll see whether it all works ...
[05:10] <lamont> cjwatson: how hard would it be for me to find a copy of the dsc, diff.gz and source.changes files before then?
[05:10] <cjwatson> lamont: they should all be in the Launchpad librarian
[05:10] <cjwatson> or do you mean of -12.37?
[05:11] <cjwatson> lamont: http://people.ubuntu.com/~ubuntu-archive/queue/gutsy/unapproved/
[05:11] <lamont> that has -12.37 diff/dsc, and .38 source.changes...
[05:11] <cjwatson> the .changes is misnamed, that's all
[05:12] <cjwatson> if you actually look at it it's for .36
[05:12] <cjwatson> er, for .37
[05:12] <cjwatson> dunno what kyle did to that
[05:12] <lamont> there are tags for 12.37 and 12.38 in git...
[05:16] <lamont> and the next upload looks like it wants to be -12.39 atm
[05:22] <lamont> test build queued
[06:31] <manchicken> Anybody seen mvo lately?
[06:34] <desrt> pretty sure the CoC has negative things to say about skewering
[06:36] <LaserJock> manchicken: I saw him today
[06:36] <LaserJock> today my time anyway
[06:36] <manchicken> I totally need to pick his brain.
[06:36] <LaserJock> I'm actually looking for him too
[06:37] <LaserJock> gnome-app-install is acting funny on me :(
[06:37] <manchicken> I'm trying to figure out how to get changelog URL generation working under adept.  mvo *is* the documentation for libapt.
[06:50] <Hobbsee> desrt: heh :)
[06:50] <Hobbsee> desrt: you wish.
[06:53] <desrt> oh.  ow.
[07:22] <`23meg> Hobbsee: http://ubuntuforums.org/showthread.php?t=553726
[07:27] <jdong> would almost seem like the forusms are giving monkeys a bad name :D
[07:29] <IntuitiveNipple> an alternative might be a 'vote' portlet that non-commenting users can do the +1 with to indicate their interest in the bug, without creating comment spam
[07:29] <Hobbsee> jdong: *grin*
[07:29] <Hobbsee> `23meg: ah yes, i think i saw that in my email
[07:29] <jdong> :)
[07:30] <jdong> IntuitiveNipple: perhaps at UDS we can hammer out some good way of flowing between LP and the forums
[07:30] <Hobbsee> IntuitiveNipple: true - but popularity is not the only meter, nor a particularly good meter, of the priority of the bug.
[07:30] <`23meg> IntuitiveNipple, subscribing is a way of showing one's interest
[07:31] <IntuitiveNipple> Hobbsee: agreed, but getting some feedback as to the number of users actively affected is useful in deciding what to tackle and what priority to set
[07:31] <IntuitiveNipple> `23meg: Yes, but how many of those users will then send "unsubscribe" messages trying to leave?  :)
[07:32] <`23meg> unsubscribing from a bug doesn't send mail to everyone
[07:32] <`23meg> neither does subscribing
[07:32] <evand> Isn't this what the developer forum on ubuntuforums is for?  Soliciting feedback, that is.
[07:33] <evand> I fear the day when this activity spreads to most other bugs.
[07:33] <LaserJock> jdong: LP-forums stuff has been discuss for ever UDS I've been at, which is the last 3
[07:33] <LaserJock> *discussed
[07:33] <`23meg> jdong, there's also a blueprint I think
[07:33] <IntuitiveNipple> I'm working on something similar, for Hardy, automated collection of hardware profiles and association with bug, and integration with Laptop-Testing, so we have a structured view of bug impact and effects, and a 'hardware popularity' contest
[07:34] <LaserJock> I honestly think it's better for devs to go to the forums than the forums goin to devs
[07:34] <evand> +1 ;)
[07:35] <IntuitiveNipple> No reason not to have a link on a bug, to a forum thread URL, same as forums do to launchpad :)
[07:42] <Hobbsee> IntuitiveNipple: true.
[07:42] <Hobbsee> `23meg: if they do it stupidly, yes it does spam pepole.
[07:43] <Hobbsee> LaserJock: so it seems.  which is why it wont get done :P
[07:43] <`23meg> Hobbsee, people who post "+1" rarely use the mail interface
[07:43] <IntuitiveNipple> wow, first major fog of the autumn out my window... I was expecting a nice sunrise
[07:44] <LaserJock> `23meg: is there any way to anticipate when a bug report is gonna get spammed?
[07:44] <Hobbsee> LaserJock: when it's something that lots of forum users would want
[07:44] <Hobbsee> LaserJock: and set it to private, when that happens.
[07:45] <LaserJock> if you can put a link to a forum thread for all the +1 stuff that allows people to still voice themselves
[07:45] <`23meg> LaserJock, when a link is posted to a controversial forum thread, or one in which inexperienced users are involved
[07:45] <LaserJock> but doesn't get in the way of the bug reports or send bug spam
[07:45] <Hobbsee> LaserJock: i think they did - and they spammed teh report anyway
[07:45] <IntuitiveNipple> Also, if you do that, they can add a vote to the forum thread, and +1 away :)
[07:46] <IntuitiveNipple> reflect the vote results into the bug report, if there's a forum link :)
[07:46] <IntuitiveNipple> makes +1 users happy, and saves tempers fraying for devs
[07:47] <LaserJock> `23meg: I don't see a link to the forums from the bug report
[07:47] <`23meg> LaserJock, it's the opposite
[07:47] <LaserJock> or rather that was for Hobbsee
[07:48] <LaserJock> I'm thinking that the reverse direction is more helpful
[07:48] <`23meg> how exactly should that be done?
[07:48] <LaserJock> put a link in the bug report
[07:49] <`23meg> to a forum thread, and say "post the +1s here"?
[07:49] <LaserJock> yes, basically
[07:49] <IntuitiveNipple> I like the idea, because prompter text could say "to vote or discuss this bug visit this forum thread"
[07:50] <LaserJock> we've been looking at linking from forums to launchpad
[07:50] <LaserJock> but maybe going the other day around is more helpful
[07:51] <`23meg> LaserJock, the forums have a "link to LP bug" function
[07:51] <IntuitiveNipple> That's been there for a while - there's an option for a link to launchpad at least
[07:51] <IntuitiveNipple> I know I've used it a few times
[07:51] <`23meg> but I doubt it does anything beyond putting a bug URL on top of the forum post
[07:51] <LaserJock> right, I know it's there
[07:51] <`23meg> it's not possible to query for forum posts related to a certain bug, etc.
[07:52] <LaserJock> I'm just saying sending hordes of people to LP doesn't really help us much does it?
[07:52] <`23meg> sure
[07:53] <`23meg> in this particular case, one user said something in the lines of "if you want this package included, please show your support in LP"
[07:53] <LaserJock> right
[07:53] <`23meg> which hints to lack of knowledge of bug tracker etiquette
[07:53] <`23meg> which in turn brings us to the fact that the bug tracker has no etiquette page
[07:53] <IntuitiveNipple> you know what? that K3B trumpet fanfare when it completes a burn made me spill my mug of tea!
[07:54] <LaserJock> lol
[07:54] <TheMuso> c
[07:54] <LaserJock> `23meg: would people read it?
[07:54] <TheMuso> wrong window
[07:54] <TheMuso> IntuitiveNipple: IMO the ejecting of your drive should be enough. :p
[07:54] <`23meg> LaserJock, whether people would read it is separate question
[07:55] <IntuitiveNipple> TheMuso: lol... I was stood up looking out across the fields, absorbed in thought
[07:56] <IntuitiveNipple> I think what we really want is to keep LP secret from the forum users, but have a link from LP to forums for devs to be able to look at discussions surrounding user experiences :)
[07:56] <`23meg> when it's not there, there's nothing concrete to hold people accountable against
[07:56] <StevenK> IntuitiveNipple: Ah, so Launchpad asks "Are j00 a forum user?" when signing up? :-P
[07:56] <LaserJock> true
[07:56] <IntuitiveNipple> lol
[07:56] <`23meg> in other bug trackers, when users breach the etiquette, people typically post a link to the etiquette page (see GNOME bugzilla)
[07:57] <IntuitiveNipple> You've got my drift... it's the silly time of the day
[07:57] <IntuitiveNipple> 7am and I've been programming 5 hours
[07:57] <StevenK> Or in the Gentoo bug tracker, refuse to fix the bug unless the submitter refiles it following etiquette rules
[07:58] <minghua> Hmm, I think I registered for forum before becoming a LP user.
[07:58] <IntuitiveNipple> Hmm, not sure I like that idea. I feel 'told off' if a bug report is marked 'Invalid' when it has resolved itself... I would like a "Resolved" status :)
[07:58] <minghua> Because forum predates LP IIRC.
[07:59] <IntuitiveNipple> I find forums better for documenting... code-formatting, quoting, and so forth
[07:59] <`23meg> sure, the forums are much older than LP
[08:00] <minghua> It's true that LP's web interface sucks for quoting, but email is always there.
[08:01] <IntuitiveNipple> I don't think it does HTML <> LP does it?
[08:02] <IntuitiveNipple> Grrr! I can't win. The Gutsy LiveCD fails through unionfs/i810 issues on the vaio notebooks, and now the alternate claims the CD isn't in the drive after it just booted from it!
[08:11] <kagou> good morning
[12:48] <slytherin> pitti: When you say soname should remain same for a new version of library, what exactly does it mean?
[12:50] <pitti> slytherin: i. e. the API and ABI must be backwards compatible
[12:50] <slytherin> pitti: Does it have anything to do with actual name of .so file?
[12:50] <pitti> slytherin: if you upgrade the library, all existing binary packages which linked and depend against the old version must continue to work with the newer library
[12:51] <pitti> slytherin: the 'soname' is the common means of designating the backwards compatible "ABI version"
[12:51] <slytherin> pitti: Ok
[01:09] <stgraber> pitti: The tracker worked correctly yesterday and e-mail notifications were correctly sent, so next time you'll be able to add the new ISOs yourself.
[01:10] <pitti> stgraber: cool
[01:10] <pitti> stgraber: who does it notify now?
[01:14] <cjwatson> stgraber: we might as well cancel testing of alternate images; they're screwed
[01:14] <cjwatson> I just uploaded netcfg to fix them
[01:28] <pitti> cjwatson, stgraber: alternates disabled in iso tracker
[01:29] <pitti> stgraber: ^ erm, at least I wanted to; it doesn't actually seem to work; can you please have a look?
[01:47] <stgraber> pitti: ok, I'll have a look. e-mail notifications are sent to people with e-mail notification enabled in their tracker profile (My profile)
[01:48] <pitti> cjwatson: FYI, if someone complains about ubiquity crashing with a mmap SystemError, I filed bug 144001
[01:48] <ubotu> Launchpad bug 144001 in python-apt "crashes with SystemError: E:Unable to write mmap - msync" [Critical,New]  https://launchpad.net/bugs/144001
[01:48] <stgraber> cjwatson: alternate also means Edubuntu server and all the Netboot right ?
[01:53] <stgraber> pitti: in fact the ISOs were correctly disabled (you can't post a result), it's just the ISOs list that doesn't show "rebuilding"
[01:54] <pitti> stgraber: ah, I see
[02:27] <slytherin> Do I need original maintainer field if the version I am packaging is not in debian?
[02:28] <stgraber> Anyone having problems with ekiga ? (ekiga: symbol lookup error: /usr/lib/libopal.so.2.2: undefined symbol: _ZN11PSafeObjectC2Ev)
[02:28] <Kmos> slytherin: for ubuntu you need.. XSBC-Original-Maintaner...
[02:29] <Kmos> and Maintainer:
[02:29] <stgraber> I'll try a package rebuild and see if that's just some kind of library change that's causing it
[02:29] <slytherin> Kmos: ok
[02:29] <Kmos> slytherin: better to ask on #ubuntu-motu
[02:34] <Hobbsee> stgraber: i think there may have been a bug on that filed earlier, actually
[02:37] <stgraber> Looks like bug 131569
[02:37] <ubotu> Launchpad bug 131569 in ekiga "Ekiga does not start (Gutsy) - undefined symbol" [Medium,Triaged]  https://launchpad.net/bugs/131569
[02:42] <Hobbsee> stgraber: is that amd64 weirdness, or?
[02:46] <ion_> benc:
[02:46] <ion_> Whoops
[02:46] <ion_> benc: The nvidia_supported patch doesnt seem to be included in the newest l-r-m.
[02:51] <ion_> benc: Additionally, the added echo "$start $stop $symname" line messes with its output.
[03:01] <_MMA_> ion_: Do you mean you cant use the nVidia drivers anymore? Today, with the new upload was the 1st time I could use 'em with -rt.
[03:03] <ion_> mma: No, only that the modalias pattern list is incorrect.
[03:04] <ion_> benc: As a reminder, the diff to the previous version is at <http://heh.fi/tmp/nvidia_supported.diff> and heres the version with the diff applied, <http://heh.fi/tmp/nvidia_supported>.
[03:04] <slytherin> if anyone is interested, I have created libtheora packages for 1.0beta1, should appear in my ppa in some time. http://ppa.launchpad.net/onkarshinde/ubuntu I will file bug on Monday.
[03:05] <_MMA_> ion_: Ahh... Ok.
[03:22] <yacoob> Hi there. Can anyone confirm that this http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=402332 also applies to feisty?
[03:22] <ubotu> Debian bug 402332 in coreutils "cp/mv do not respect default ACL set on the parent directory" [Normal,Open] 
[03:24] <cjwatson> stgraber: anything using d-i where the installation profile includes the network-manager package
[03:31] <asac> Hobbsee: was there a knetwormanager upgrade recently as well? or why can your .knetworkmanagerc become corrupted
[03:31] <Hobbsee> asac: not to my knowledge.
[03:31] <Hobbsee> asac: i have no idea.
[03:32] <asac> Hobbsee: do you do backups of your HOME?
[03:32] <asac> Hobbsee: maybe you can recover the broken file?
[03:32] <yacoob> Sigh. Well, bug filed... :(
[03:33] <Hobbsee> asac: er, i should do one of them.  i rm'd it, sorry.
[03:33] <asac> ok
[03:33] <asac> ;)
[03:34] <yacoob> there's nothing more annoying when you look for something, you find the "perfect fix" and then it turns out it doesn't work as expected :/
[03:37] <pkern> If I use boot.img.gz from d-i, does it only retrieve the installer components from the net and on install use the iso provided on the stick or is there some option I could turn on so that it automatically loads everything from the provided cd image?
[03:55] <cjwatson> pkern: boot.img.gz is just a format, not an installation type; the thing you want to look at is whether it's cdrom, hd-media, or netboot
[03:57] <cjwatson> pkern: ISTR that hd-media will grab everything from the ISO on the stick if it can
[04:12] <soren> "update-manager -d" on feisty should offer to upgrade to gutsy, shouldn't it?
[04:12] <soren> update-manager version is 0.59.23
[04:14] <soren> Ah, I need to run it as root? That doesn't seem right.
[04:46] <Kopfgeldjaeger> er... ubiquity does start anymore for me on the live-cd (well, that was the first time i tried it). can somebody confirm this?
[04:57] <bigon> hi, could some buildd admin have a look at https://launchpad.net/~telepathy/+archive/+build/397159 ?
[04:59] <Hobbsee> !weekend
[04:59] <ubotu> It's a weekend.  Often on weekends, the paid developers, and a lot of the community, may not be around to answer your question.  Please be patient, wait longer than you normally would, or try again during the working week.
[04:59] <pkern> cjwatson: Oh, I guess I accidentally grabbed netboot then, thanks!
[04:59] <Hobbsee> especially for non-beta critical stuff
[05:00] <pkern> So almost time to trash the laptop with gutsy encrypted LVM install... ;)
[05:00] <Hobbsee> bigon: oh, that's a ppa bug.
[05:01] <bigon> Hobbsee: yep
[05:01] <Hobbsee> bigon: in which the correct place to ask is probably #launchpad, but again, the stuff about weekends applies
[05:03] <bddebian> Heya
[05:16] <pkern> No fun, hd-media reports kernel mismatch... Hm.
[05:18] <pkern> Published in gutsy-release 5 hours ago
[05:18] <pkern> Gah... !@#@!#
[05:18] <\sh> and on my desktop nautilus is getting mad and hitting  the harddrive...with something I can't identify...killing nauilus helps, but no gnome comes up..
[05:18] <pkern> cjwatson: Is there a possibility to get a hd-media boot.img.gz built against the current kernel?
[05:18] <\sh> strange thing, it just happens on this desktop machine...and not on my laptop
[05:19] <jdong> \sh: does a logout help? When I applied the updates yesterday gnome went crazy infinitely regenerating thumbnails
[05:19] <pkern> cjwatson: Hm, other way round, the cd image is the problem of course. \:
[05:19] <\sh> jdong, nope...reboot doesn't help nothing...
[05:19] <\sh> jdong, you say .thumbnails?
[05:19] <\sh> hmm
[05:20] <jdong> well that was my problem yesterday, yes
[05:20] <\sh> jdong, I see after dm login, two or three times my panels coming up and destroying
[05:20] <cjwatson> pkern: should already be one
[05:21] <pkern> cjwatson: Yep it is, it's just incompatible to the current cd build.
[05:21] <jdong> \sh: that's odd.... it could be the panel's crashing like crazy?
[05:21] <cjwatson> current CD build is against -12 too
[05:21] <pkern> There was a new revision 5h ago...
[05:21] <\sh> jdong, no...when I kill nautilus form console...it stops hitting the harddrive
[05:21] <cjwatson> you're just out of date with one or the other I expect
[05:21] <jdong> hmm...
[05:22] <pkern> cjwatson: Doesn't it also depend on the right kernel revision?
[05:23] <stgraber> Hobbsee: yes it's amd64
[05:23] <pkern> And how often does d-i get a rebuild pushed to the mirrors? (If it isn't that often my assumption is wrong anyway.) But I do get a mismatch between the hd-media image currently on archive and the current cd image (that's for sure, just jigdo'ed).
[05:26] <cjwatson> pkern: they both have the right kernel version as far as I can see. d-i gets rebuilt manually when I feel like it
[05:26] <Hobbsee> good morning pitti!
[05:27] <pkern> cjwatson: Ok, thanks. I will investigate this one. ;)
[05:27] <jdong> good morning, world :)
[05:27] <jdong> it's a bright and shiny day....
[05:27] <jdong> until someone brings up unionfs :D
[05:27] <pitti> hey Hobbsee
[05:28] <lamont> if apache was removed from gutsy, why were libapache-mod-* left?
[05:29] <lamont> libapache-mod-{cgi-debug,filter,index-rss,ldap,mp3,random,relocate,text2html,trigger} all build-depend apache-dev
[05:31] <\sh> lamont, check the depends...they are transitional packages
[05:32] <\sh> sry...those ones
[05:32] <bddebian> Oohh, pitti.  I was trying to use your syncpackage script for pochu to sync gnome-build but the dang thing keeps failing on signing the package.  I have DEBMAIL and DEBSIGN_KEYID set and I have no problems signing my packages or manually signing stuff.  Could you help?
[05:32] <bddebian> pitti: Oh Hi, bye the way :-)
[05:33] <lamont> \sh: ok.. I just know they're ftbfs on gutsy/hppa, and I expect that they'll be ftbfs on the rebuild testing for every other architecture...
[05:33] <pitti> bddebian: no idea; just debsign the package manually after syncing it
[05:33] <bddebian> lamont: That's a dang fine question.  Though libapache-mod-auth-radius became libapache2-mod-auth-radius.  Maybe we should do the same where possible for the oothers?
[05:33] <pitti> bddebian: it's the last step anyway
[05:34] <bddebian> And then just dput it?
[05:34] <pitti> bddebian: well, check the source.changes for sanity first
[05:34] <lamont> \sh Build-Depends: debhelper (>= 4.1.16), apache-dev (>= 1.3.34)
[05:35] <bddebian> pitti: Hrm, this is the same reason signing is failing.  I get: Changed-By: Barry deFreese <> in the .changes file
[05:36] <pitti> bddebian: ah, so it put in the wrong email address, it seems
[05:36] <bddebian> Aye it put no email address :)
[05:36] <pitti> bddebian: hm, but $DEBEMAIL is correct?
[05:36] <\sh> guys
[05:37] <bddebian> DEBMAIL=bddebian@comcast.net
[05:37] <pitti> bddebian: DEB*E*MAIL
[05:37] <bddebian> gah, wtf. I knew it was going to be something stupid I did
[05:37] <\sh> I have in ~/nautilus-debug-log.txt messages like "0x8177888 2997/09/22 17:36:38.1745 (USER): debug log dumped due to signal 11
[05:37] <\sh> and this is hitting my hd so hard..
[05:37] <\sh> what is it? any clue?
[05:38] <Vegar> how come the restricted drivers manager tells me I don't need any restricted drivers when I have an nvidia card?
[05:39] <bddebian> pitti: Excellent thanks.  Now, it shouldn't add a build1 version or anything? I didn't think we were supposed to do that?
[05:39] <pitti> Vegar: you have the nv driver ATM?
[05:39] <Vegar> no
[05:39] <Vegar> pitti: vesa
[05:39] <pitti> bddebian: nope
[05:39] <Vegar> pitti: it used to give me the choice to install the nvidia driver
[05:39] <bddebian> pitti: OK, thanks again and sorry to bother you
[05:39] <pitti> Vegar: it doesn't even appear in the list?
[05:39] <Vegar> no
[05:40] <pitti> Vegar: lspci | grep 0300
[05:40] <pitti> ?
[05:40] <Vegar> it tells me my hardware doesn't need any restricted drivers
[05:40] <Vegar> sec
[05:40] <Vegar> I'll log on on the laptop in question
[05:41] <Vegar> no matches on that, pitti
[05:41] <pitti> Vegar: erm... no graphics card at all??
[05:43] <Vegar> 01:00.0 VGA compatible controller: nVidia Corporation Quadro NVS 140M (rev a1)
[05:43] <Vegar> oh, this is the -devel channel
[05:43] <Vegar> I thought I was in ubuntu+1
[05:43] <pitti> Vegar: what's the corresponding "lspci -n" line for that?
[05:43] <pitti> lamont: that doesn't give me the vendor/product ID :/
[05:44] <lamont> pitti: ah
[05:44] <Vegar> pitti: 01:00.0 0300: 10de:0429 (rev a1)
[05:44] <pitti> Vegar: ah, so 0300 *does* match
[05:44] <Vegar> I might have forgotten the -n flag
[05:45] <Vegar> meh
[05:45] <pitti> Vegar: right, this model isn't in /usr/share/linux-restricted-modules/2.6.22-12-generic/modules.alias.override/nvidia
[05:45] <Vegar> ah
[05:45] <pitti> Vegar: please file a bug against linux-restricted-modules-2.6.22 and include above information (vendor/product ID and model name)
[05:45] <Vegar> so that's a bug then?
[05:46] <pitti> Vegar: you can manually edit that file to include your device ID, if you prefer
[05:46] <Vegar> will do
[05:46] <Vegar> yeah, I'll do that
[05:46] <Vegar> I was looking forward to trying the new nvidia drivers
[05:46] <pitti> Vegar: depends; if the card is actually supported by nvidia-glx, it's a bug of that restricted driver
[05:46] <Vegar> it's nvidia-glx-new
[05:47] <Vegar> isn't nvidia-glx the legacy driver?
[05:47] <pitti> Vegar: you can also configure it manually, of course (install nvidia-glx and use displayconfig-gtk to switch the driver)
[05:47] <pitti> Vegar: right, -new
[05:47] <lamont> pitti: if you get bored feel free to look at the util-linux 2.13-{6,7} diff and tell me if you want that in pre-beta, post-beta, or in hardy.  (hwclockfirst.sh is debian only, btw)
[05:48] <Vegar> what's the format of the nvidia alias override file?
[05:50] <pitti> Vegar: just copy&paste a line and insert your product ID
[05:50] <pitti> alias pci:v000010DEd00000429sv*sd*bc03sc*i* nvidia_new
[05:50] <pitti> ^ Vegar
[05:51] <sponix> anyone in here seen the /dev/null permission issue, every boot it goes to 600 instead of 666, _might_ be udev related (but udev rules are set for it at 666) ?
[05:51] <Vegar> pitti: ah, wonderful
[05:51] <Vegar> pitti: thanks a lot
[05:52] <lamont> sponix: but 600 is far more secure than 666. :-D
[05:52] <cjwatson> sponix: I'd bet more on some broken init script personally
[05:52] <pitti> Vegar: please file a bug, though; in the end it's a bug in that stupid nvidia driver, but we can work around it with that list
[05:52] <cjwatson> and we would love to know which it is ...
[05:52] <cjwatson> for example something might remove /dev/null which would lead to it being recreated with bogus permissions
[05:52] <Vegar> pitti: I'm working on that
[05:52] <cjwatson> (doesn't happen here, though)
[05:53] <lamont> sponix: is it still a character device file?
[05:53] <sponix> lamont: perms are crw-r-r is what they keep getting reset to instead of crw-rw-rw
[05:53] <Vegar> pitti: how come it suddenly stopped recognising my card? could it be because it was removed from that list in the latest restricted-modules upgrade?
[05:53] <cjwatson> interesting
[05:53] <cjwatson> so it's not just being removed
[05:54] <pochu> pitti: gnome-build uploaded. since it's in universe, will it need to wait for beta release, or can you accept it?
[05:54] <pitti> Vegar: yes, we updated the driver to a new version; the new nvidia driver might have decided to not support your card any more, or so
[05:54] <pitti> pochu: I can accept it
[05:54] <cjwatson> sponix: I think you need Keybuk to investigate, when he's around
[05:54] <Vegar> ah, that's evil
[05:54] <sponix> lamont: see xD card reader on these boxen doesn't wort without udev active in runlevel 2,3,4,5 so I turned that on... after that the problem seems to appear, but even after turning it off except S (startup), the issue still seems to persist
[05:55] <lamont> interesting
[05:55] <sponix> lamont: found a quick hack on a forum that makes it _work_ but its a bit dirty, and would like to know what causes the issue so I can fix it properly
[05:55] <pochu> pitti: cool, thanks. that way I'll be able to request the anjuta UVFe.
[05:55] <lamont> that is the challenge with quick hacks
[05:55] <sponix> I don't mind a simple hack myself, but its on my wifes boxen also, and I have to leave for a year, would like to make sure she is setup correclty
[05:56] <sponix> correctly
[05:56] <lamont> understandable
[05:57] <lamont> remote puppet-master support technology is still in its infancy
[05:57] <sponix> lamont: quick h4x0rz was just to drop chmod 666 /dev/null at the end of /etc/init.d/rc.local to reset it to proper state at the _end_ of each boot
[05:58] <sponix> but if anything updates bootscripts it could overwrite my change, leaving it so she cannot login
[05:58] <cjwatson> sponix: err, you mean you added S links for udev in /etc/rc[2345] .d/ ?
[05:59] <cjwatson> sponix: you really shouldn't do that, it's already started in rcS
[05:59] <lamont> and stays running forever
[05:59] <cjwatson> sponix: it would probably be easier to find out why you need to start it again there
[06:00] <cjwatson> sponix: forget about runlevel editors, look in /etc/rc2.d with a shell to see if there's anything matching *udev* in there
[06:00] <cjwatson> I bet your runlevel editor threw in a K link when you disabled it ...
[06:00] <sponix> cjwatson: see I uses sysv-rc-conf and it was default set to just S, I had to add 2,3,4,5 before the card reader would work
[06:01] <sponix> cjwatson: had to start udev by hand with /etc/init.d/udev start the first go... prior, her card reader found the card, but didn't make the device for the partition
[06:01] <cjwatson> sponix: 'sudo rm -f '/etc/rc[2345] .d/[SK] ??udev'
[06:01] <cjwatson> sponix: seriously, that's bad news, I understand it fixed one problem (by luck) but it clearly introduced more
[06:01] <cjwatson> lose it
[06:01] <sponix> example, it showed /dev/sdg1 or /dev/sdd1 and would have /dev/sdg and /dev/sdd but no the /dev/sdd1 and /dev/sdg1 devices
[06:02] <cjwatson> sorry, amend the above
[06:02] <sponix> cjwatson: so you want me to try removing udev runlevels _completely_ and then give it a boot ?
[06:02] <cjwatson> sudo rm -f /etc/rc[2345] .d/[SK] ??udev
[06:02] <sponix> or you want just S and K ?
[06:02] <cjwatson> sponix: udev should not be mentioned AT ALL in runlevels 2, 3, 4, 5
[06:02] <cjwatson> sponix: it should only be in runlevel S
[06:03] <lamont>  ls /etc/rc*/*udev*
[06:03] <lamont> /etc/rcS.d/S10udev
[06:03] <cjwatson> sponix: your runlevel editor is dumb and did evil things to your system when you did that
[06:03] <sponix> not an problem.. I can mod that wht sysv-rc-conf
[06:03] <cjwatson> no, do it at a shell
[06:03] <sponix> ha
[06:03] <cjwatson> I don't trust that program
[06:03] <cjwatson> it broke your system
[06:03] <cjwatson> (albeit at your request, but still)
[06:03] <sponix> think it was it ?
[06:03] <sponix> no, it was just acting on my dumb wishes ! ;)
[06:04] <cjwatson> I bet you it left K links in runlevels 2, 3, 4, 5 when you asked it to reverse your previous action
[06:04] <cjwatson> which shuts down udev which is really bad news
[06:04] <cjwatson> run the command I gave above
[06:04] <sponix> cjwatson: I'll double check on my broken laptop real fast
[06:04] <cjwatson> then make sure that /etc/rcS.d/S10udev exists, as lamont showed above
[06:05] <cjwatson> it's not worth trying to debug anything else until you've straightened that out
[06:07] <sponix> yep, that command nuked K90udev from 2345
[06:07] <sponix> going to temp remove my h4x0r and reboot, see if it still fsked
[06:08] <cjwatson> now, the card reader may be broken again of course, but I bet there are better ways to fix that
[06:09] <sponix> card reader is more for this desktop I'm on now.. but I implimented the "hose" on my laptop also just to make sure I was consistant at fscking things up :P
[06:12] <stgraber> Hobbsee: Looks like rebuilding libopal (opal) helped, my ekiga no longer crash
[06:12] <stgraber> Hobbsee: I'll use it for a while (trying to make it crash) and if it's stable ask a core-dev to do a zero-change upload for opal
[06:13] <sponix> cjwatson: yep, that unhosed the boot, and udev is still running when I "ps ax|grep udev"
[06:13] <cjwatson> great
[06:13] <cjwatson> might be worth filing a bug on sysv-rc-conf - it shouldn't have let you do that
[06:14] <sponix> cjwatson: I'll have the wife get me the digital camera card next time I catch her
[06:14] <cjwatson> bug 77885 may be related
[06:14] <ubotu> Launchpad bug 77885 in sysv-rc-conf "Minor: sysv-rc-conf inconsistency displaying rcS.d or not" [Undecided,New]  https://launchpad.net/bugs/77885
[06:14] <cjwatson> certainly, one ought not to be able to set something up to have an entry (whether S or K) in both S and 2/3/4/5
[06:16] <sponix> K=Kill correct ?
[06:17] <sponix> I've been using sysv-rc-conf for a while, first time I've had it hose something up, I'll have to remember to double check its work, or quit being lazy and do it by hand
[06:18] <sponix> Vista is So Sloooowwwwww, I thought it was locked up on boot... Been running ubuntu-7.04 on the same box for about a week, almost forgot how pathetic vista makes the same hardware look
[06:19] <sponix> cjwatson: ever used Linux software raid5 ?
[06:19] <Hobbsee> stgraber: i can throw it thru regardless, but it's easier after the main freeze, yes.
[06:20] <Hobbsee> stgraber: of course, it's not exactly much I/O to rebuild it anyway, in the grand scheme of things, so then more people can test it.
[06:20] <cjwatson> sponix: no
[06:21] <cjwatson> sponix: yes, "S" means "run this script with 'start' on entry to this runlevel", "K" means "run this script with 'stop' on entry to this runlevel"
[06:23] <lamont> sponix: root on my feisty machine is sw raid5
[06:24] <sponix> lamont: what do you think of it? I've got 4-6 IDE 500G drives looking to backup my dvd collection, and then either nfs share them out to the other boxen, or make a web interface and vlc stream them
[06:24] <lamont> cjwatson:  ls /etc/rc*/*hplip*
[06:24] <lamont> /etc/rc0.d/K19hplip  /etc/rc3.d/S19hplip  /etc/rc6.d/K19hplip
[06:24] <lamont> /etc/rc1.d/K19hplip  /etc/rc4.d/S19hplip
[06:24] <lamont> /etc/rc2.d/S19hplip  /etc/rc5.d/S19hplip
[06:25] <cjwatson> lamont: yes ...?
[06:25] <lamont> sponix: I'd rather have hardware raid.  but then, I'm cheap
[06:25] <lamont> that's a S in multiple run levels...
[06:25] <cjwatson> surely software RAID is the cheap option
[06:25] <stgraber> Hobbsee: at least I can now call without having ekiga to crash, and doing the same with another ekiga (without the rebuilt libopal) make it crash instantly
[06:26] <Hobbsee> stgraber: OK.  got the bug #?
[06:26] <lamont> cjwatson: and that's why I have sw raid instead of hwraid like I'd prefer...
[06:26] <stgraber> bug 131569 IIRC
[06:26] <ubotu> Launchpad bug 131569 in opal "Ekiga does not start (Gutsy) - undefined symbol (dup-of: 117732)" [Medium,Invalid]  https://launchpad.net/bugs/131569
[06:26] <ubotu> Launchpad bug 117732 in opal "ABI changed without a corresponding soname change" [High,Triaged]  https://launchpad.net/bugs/117732
[06:26] <sponix> doesn't have to be super fast, just be able to recover if a drive goes down, hate to lose 1.5+ TB of media over a single drive failing
[06:26] <sponix> that is a _very_ long time of ripping my dvd's again :P
[06:26] <cjwatson> lamont: I didn't say that was a bad thing
[06:26] <cjwatson> lamont: I said putting an entry in S as well as in 2/3/4/5 is probably silly
[06:26] <cjwatson> you have nothing in S (by which I meant rcS) there
[06:26] <cjwatson> oh, right
[06:27] <lamont> ah, ok
[06:27] <sponix> cups-pdf is installed by default now, isn't it ... went to put that on the wifes box the other day, told me was already there
[06:27] <Hobbsee> stgraber: do i need to mess with sonames and such?
[06:28] <stgraber> Hobbsee: only thing I did was to rebuild it using pbuilder
[06:33] <stgraber> Hobbsee: hmm, in fact it looks like I replaced an "undefined symbol" error by a segfault :)
[06:33] <Hobbsee> stgraber: woot.
[06:33] <Hobbsee> stgraber: looks like debian have packaged it slightly differently, anyway.
[06:34] <Hobbsee> you know, for something that's apparently the same upstream version, but named differently...surely it should have the same upstream changelog?
[06:34] <stgraber> receiving a call make it segfaults now :) I should try with : rebuild libopal + rebuild ekiga maybe that helps
[06:35] <stgraber> but IIRC ekiga doesn't build with pbuilder (configure fails on a library version)
[06:36] <Hobbsee> uh......wha?
[06:36] <Hobbsee> since when does g++ need to be a build depend?
[06:36] <jdong> Hobbsee: just to be absolutely sure we have it ;-)
[06:37] <Hobbsee> stgraber: ask dholbach about that.
[06:38] <Hobbsee> stgraber: (in particular, ask if we want to take their changes)
[06:39] <Amaranth> arg
[06:40] <Amaranth> session management is no fun
[06:41] <broonie> Hobbsee: Things often have a versioned build dep if they don't build with old g++
[06:45] <Hobbsee> greetings, Burgundavia
[06:47] <Burgundavia> hey Hobbsee
[06:53] <sponix> cjwatson: I'll be on the right box in a sec to see if I can make the card reader work without the mod to udev runlevels
[07:03] <sponix> cjwatson:  ok... back...
[07:05] <Vegar> is there a list of the patches applied to the ubuntu kernels?
[07:06] <sponix> cjwatson:  still with me ?
[07:13] <pkern> cjwatson: Sorry, I found the problem, I used the wrong CD image. But is it possible that cryptsetup is missing from the package pool on the CD?
[07:14] <pkern> d-i insists on overwriting my sources.lists with the mirror source added, and is thus unable to find cryptsetup.
[07:16] <pkern> The file list only lists the udeb, not the deb...
[07:20] <pkern> And `chattr +i /etc/apt/sources.list' lists it break badly, but I expected that somehow.
[07:37] <cjwatson> sponix: I'm no expert on udev rules, so at this point you need the udev maintainer (Keybuk) for further debugging
[07:38] <pkern> It's a binary, not a script that forcefully overwrites the sources.list, right? \:
[07:38] <pkern> Rough edges. *cough*
[07:38] <cjwatson> pkern: good point - I've moved cryptsetup to ship for the next build
[07:38] <sponix> cjwatson:  well, wanted to thank you for helping me unhose the base system anyway...
[07:38] <cjwatson> pkern: it's a script, but there are plenty of *supported* ways to add sources.list entries ...
[07:38] <pkern> cjwatson: Do you have a clue what I need to do at this point? (It installs the base system but breaks on cryptsetup.)
[07:38] <cjwatson> pkern: wait for next build
[07:39] <pkern> Hah.
[07:39] <pkern> Possible but bad solution. \:
[07:39] <cjwatson> or wget cryptsetup.deb and install it
[07:39] <pkern> cjwatson: Is this sufficient?
[07:39] <cjwatson> perfectly good solution from my point of view, you're working with a buggy daily build
[07:39] <cjwatson> dunno
[07:39] <pkern> cjwatson: That's my point. Or do I have to do some housekeeping?
[07:39] <cjwatson> my priority is to fix the bug for the beta
[07:39] <pkern> apt-get works fine in the chroot.
[07:39] <cjwatson> you can always temporarily change sources.list again
[07:40] <pkern> Of course, but the HDD is nuked now. It's just that I cannot successfully complete base-installer.
[07:40] <pkern> If this is the last step in the list I'm fine.
[07:40] <pkern> (i.e. I could ignore the failure and ask d-i to continue.)
[07:40] <cjwatson> it's not, but you can probably install cryptsetup and then rerun base-installer
[07:40] <cjwatson> it might be a bit confused
[07:41] <cjwatson> it won't have run post-base-installer.d hooks or installed a kernel yet
[07:41] <cjwatson> or processed the apt-install queue
[07:41] <cjwatson> so you should definitely rerun base-installer and let it keep going
[07:41] <pkern> cjwatson: From d-i that does not work, because it will try to start from the beginning again, and change the sources.lists just again before wanting to install cryptsetup.
[07:42] <cjwatson> pkern: edit /var/lib/dpkg/info/base-installer.postinst and comment out the lines starting with 'waypoint' from install_base_system up to apt_update
[07:42] <cjwatson> I think
[07:43] <cjwatson> you'll get a message asking for confirmation of install over unclean target, but ignore it
[07:43] <cjwatson> you do want to leave get_mirror_info I think because it sets several interesting variables
[07:45] <pkern> cjwatson: That one worked, thank you.
[07:45] <pkern> I hope there are no missing base packages or missing housekeeping *for cryptsetup* left now, but well...
[07:46] <cjwatson> shouldn't think so
[07:47] <LaserJock> cjwatson: you happen to know if mvo will be around this weekend?
[07:47] <pkern> Ok, it's installing the system now, thanks. Next fun will be the reboot.
[07:47] <cjwatson> LaserJock: don't know, sorry
[07:48] <Amaranth> LaserJock: What's up?
[07:49] <LaserJock> I think I've got a bug in g-a-i that is messing up the Edubuntu Addon CD
[07:49] <Amaranth> Ah
[07:49] <LaserJock> I've tried looking in the code a little bit but was hoping I could get ahold of mvo
[07:49] <Amaranth> Dunno about that stuff, except for the pyxdg part
[07:51] <LaserJock> well, thats kinda the stuff I'm trying to figure out
[07:51] <LaserJock> for some reason it's not using my icons
[07:52] <LaserJock> and I can get it to work doing very strange, at least to me, things
[08:01] <Amaranth> LaserJock: what, exactly, is the problem?
[08:01] <Amaranth> LaserJock: what is your Icon line?
[08:01] <LaserJock> it's just got the name of the icon
[08:02] <Amaranth> which is?
[08:02] <LaserJock> various things
[08:02] <LaserJock> Icon=xfce4 for instance
[08:02] <Amaranth> are these icons installed in /usr/share/hicolor/<size>/<category>/<iconname>.png?
[08:02] <LaserJock> no
[08:03] <LaserJock> they are on a cd
[08:03] <Amaranth> LaserJock: there is your problem
[08:03] <LaserJock> I'm trying to supply custom icons
[08:03] <Amaranth> pyxdg only reads from the icon theme unless you give it an absolute path
[08:03] <LaserJock> right, which I would think it would be getting
[08:03] <LaserJock> or hmm
[08:03] <LaserJock> it should be given the app-install/ dir
[08:04] <Amaranth> no no
[08:04] <Amaranth> the Icon line needs to be an absolute path
[08:04] <Amaranth> otherwise it uses the icon theme
[08:04] <LaserJock> hmm
[08:04] <LaserJock> then how come sometimes it works?
[08:04] <LaserJock> although I guess that's when I set --desktopdir and --cachedir
[08:05] <LaserJock> in g-a-i
[08:05] <Amaranth> gnome-app-install might have a fallback to check the current dir or something
[08:05] <Amaranth> or --desktopdir
[08:05] <LaserJock> it's just weird
[08:05] <Amaranth> not really
[08:06] <LaserJock> becuase it only works when I cp my ro app-install/ from the CD to the hard drive
[08:06] <LaserJock> and then I can point it to either directory, CD or Hard driver and it'll work
[08:06] <LaserJock> I wondered if it was because the CD was ro
[08:06] <Amaranth> gnome-app-install must have a fallback if pyxdg can't find the icon
[08:07] <LaserJock> Amaranth: gotta run to do some shopping. I'll be back later if you're interested
[08:07] <LaserJock> I'm gonna have to dig into the code and figure out how it does the icons
[08:07] <Amaranth> err, alright
[08:37] <ScottK> I'd appreciate it if someone from the archive/release team would accept wine 0.9.45-0ubuntu1.
[08:48] <cjwatson> ScottK: done
[09:23] <pkern> There is more broken. After reboot it doesn't get the root fs, but lvm seems missing, too. And I have a hard time getting rescue access to the disk.
[09:25] <pkern> vgscan/vgchange did not list s.th. The partition type (using guided) was set to Linux, maybe that has caused it, but changing it did not fix it in the same session. *cough* Anyway, I'll close this for today, this is giving me headaches somehow (or they originate elsewhere, whatever).
[09:56] <cjwatson> I think I might have a workaround
[10:31] <IntuitiveNipple> Is the 20070921.1 desktop CD supposed to have a gnome menu-bar when running LiveCD !?
[10:38] <cjwatson> IntuitiveNipple: yes
[10:39] <cjwatson> (does here)
[10:40] <IntuitiveNipple> ha, ok, so some kind of weird bug then. I'm testing it on an Acer Travelmate C104TCi - very strange, desktop's there, Examples and Install, right-click desktop brings up context menu, but no main-menu. It is all just desktop background. Ctrl+Alt+F1-F5 do't get a reaction, but Ctrl+Alt+Backspace restarts GDM correctly.
[10:41] <IntuitiveNipple> (still waiting for the CD-ROM to restart the gnome session completely)
[10:42] <cjwatson> some people have reported continuing unionfs breakage; may be what's happening to you
[10:42] <cjwatson> dmesg would be helpful if you can find a way to get at it
[10:43] <jdong> cjwatson: has anything changed since yesterday's .1 that'd be worth me testing for the unionfs bug?
[10:44] <cjwatson> doubt it, kernel team likes having weekends
[10:44] <jdong> what are those?
[10:44] <cjwatson> (though there were kernel uploads recently, admittedly)
[10:44] <cjwatson> (but they aren't in today's image)
[10:45] <IntuitiveNipple> cjwatson: possibly, that's been stopping me testing on the Sony Vaio notebooks
[10:45] <jdong> ok, I'll sit tight and fool around with power consumption then.
[10:45] <jdong> I'm a bit curious why OS X can pull off almost 1:30 extra runtime on macbooks compared to gutsy
[10:46] <cjwatson> www.lesswatts.org
[10:46] <IntuitiveNipple> argh, I can get a virtual terminal but it looks like the fbcon isn't loaded so no way to use it, blank video
[10:46] <cjwatson> IntuitiveNipple: you aren't selecting a resolution at the boot menu are you?
[10:46] <jdong> yep, working on those tips currently
[10:46] <jdong> though it should be *fewer*watts.org ;-)
[10:46] <cjwatson> jdong: I did think that, but that's not correct - watts are continuous not discrete
[10:47] <IntuitiveNipple> cjwatson: no, just standard CD startup testing... next test is to try Safe Graphics mode
[10:47] <cjwatson> at least I think "less" is legitimate there though I think "fewer" would also be correct
[10:47] <jdong> cjwatson: good point. Somehow less watts just doesn't sound correct to me
[10:47] <cjwatson> yeah
[10:47] <cjwatson> "less power" would be better :)
[10:48] <jdong> indeed
[11:11] <pkern> cjwatson: I in-place fixed d-i to generate the correct sources.list (thanks for the pointer) but after reboot it failed to mount the root device. luks actions on cryptsetup within initramfs segfaulted.
[11:12] <pkern> (And on an unrelated sidenote with the cli install: my network interfaces were named eth1 and eth2 in persistent-rules but eth0 was set up in network/interfaces.)
[11:21] <pkern> Or is `cryptsetup luksOpen crypt0 /dev/sda5' wrong syntax?
[11:33] <slangasek> jdong: fewerwatts.org is an alias anyway ;)
[11:33] <slangasek> pkern: wrong syntax - it's realdev decrypteddev, you have the last two args backwards
[11:35] <pkern> cryptsetup segv in LUKS_dealloc_masterkey at keymanage.c:57
[11:35] <pkern> slangasek: Hm, thanks.
[11:37] <pkern> slangasek: The other way 'round it asks for a passphrase, yeah. (create has its arguments reversed.) So, thanks. *cough*
[11:38] <slangasek> no worries :)
[11:39] <IntuitiveNipple> pkern: Is there any provision to mount LUKS crypted volumes that *only* have a key-file (no password) at boot/startup ?
[11:39] <pkern> slangasek: `[ -x /foo/bar ]  || exit 0' is a bashism, right? `/etc/cryptsetup/cryptdisks.functions' has two of them near the beginning. (Just another thing I encountered in initramfs... *cough*
[11:39] <slangasek> no, what would be bashish about it?
[11:40] <slangasek> it's a "not-set-e-ism"
[11:40] <slangasek> but it's a "not-set-e-ism" in all dialects :)
[11:40] <slangasek> (sorry, "non-set-minus-e-ism"
[11:40] <slangasek> )
[11:40] <slangasek> anyway, initramfs sure doesn't use bash in any configuration I've seen
[11:41] <pkern> I thought that [ is only allowed after if... ok, fine. (It somehow failed in initramfs, though.)
[11:41] <pkern> But... nevermind.
[11:41] <Mithrandir> [ is just a normal binary in /usr/bin
[11:41] <slangasek> nope.  "if [" is "if the [ command returns a zero exit code"
[11:41] <davmor2> anyone here work on netboot?
[11:41] <ion_> [ is a program (which may be implemented as a built-in), and you put any program after if.
[11:42] <pkern> Right... \:
[11:43] <pkern> Ok, off to another reinstall.
[11:43] <Mithrandir> ion_: any program which is part of posix might be a built-in..
[11:44] <ion_> Well, *any* program might be a built-in. :-P
[11:45] <Mithrandir> no?  Having debconf as a built-in program would fuck up how dependencies work.
[11:46] <slangasek> Mithrandir: nevertheless, POSIX doesn't /prohibit/ shells from adding random programs as built-ins :)
[11:46] <IntuitiveNipple> Who'd be responsible for adding a patch to a hal script & gnome-mount to allow LUKS crypted volumes protected *only* by a keyfile (no password) ?
[11:47] <Mithrandir> slangasek: true.
[11:52] <Luke> asac: you on now?
[11:57] <davmor2> cjwatson: ping
[12:07] <asac> Luke: yes
[12:12] <pkern> Fun. So I have a crypted root fs now, but I need to debug the initramfs because it does not find it at boot. (I suspect the missing `/etc/crypttab' might be the culprit, but I may be wrong of course, as day has shown. *cough*)
[12:22] <cjwatson> pkern: after d-i, I have no idea how cryptsetup works, so I'm the wrong person to ask
[12:22] <cjwatson> davmor2: pong, briefly
[12:22] <davmor2> are you incharge of netboot?
[12:22] <cjwatson> davmor2: d-i in general, yes
[12:22] <cjwatson> (but trying to find the person "in charge" of something is usually the wrong plan, just ask the question)
[12:22] <davmor2> netboot is installing the wrong desktops
[12:23] <cjwatson> netboot is supposed to ask
[12:23] <davmor2> edubuntu is listed twice both install kubuntu
[12:23] <cjwatson> the task looks right ...
[12:24] <cjwatson> but it's half-eleven on a Saturday night, I'm not going to debug it now. please file a bug
[12:24] <davmor2> filed
[12:24] <davmor2> bug 144145 + 144147
[12:24] <ubotu> Launchpad bug 144145 in ubuntu "Netboot edubuntu desktop is listed twice both install kubuntu" [Undecided,New]  https://launchpad.net/bugs/144145
[12:25] <davmor2> just pointing out while I get chance to as I'm still isotesting the other installs :)
[12:33] <Luke> cjwatson: did you end up taking dell's disk remaster util under your wing?
[12:33] <Luke> or will that stay with dell?
[12:35] <Luke> asac: thanks for the replys. Sorry about the delay but I've been super busy with university this semester
[12:36] <Luke> asac: are you looking for anything in specific in my syslog?