[12:04] <robertj> CarlFK: do not try to bend the server for thal is impossible
[12:04] <CarlFK> robertj: ?
[12:05] <mdke> you might want to file a bug if you've found a problem
[12:06] <robertj> CarlFK: sorry, needless Matrix joke
[12:06] <CarlFK> robertj:  ah - havn't seen 2 and 3 yet
[12:06] <CarlFK> mdke: having trouble believing there is no MySql  ;)
[12:07] <mdke> CarlFK, in that case, for help, you can ask in #ubuntu
[12:07] <robertj> CarlFK: just a variation on the there is no spoon theme
[12:08] <CarlFK> or #u-bugs... which I keep forgetting about
[12:10] <Chipzz> CarlFK: do dpkg -s mysql-server first
[12:10] <Chipzz> CarlFK: what does that give?
[12:11] <CarlFK> Package `mysql-server' is not installed ...
[12:11] <CarlFK> thanks - thats a confirmation I can cut/paste 
[12:12] <Chipzz> CarlFK: there's your problem; unless you have mysql4-server or mysql5-server
[12:17] <CarlFK> no php4/5 either...
[12:18] <CarlFK> maybe they meant la Server
[03:40] <jsgotangco> good morning
[03:42] <ajmitch> hi jerome
[04:15] <zul> Kamion: question...why is their a reiserfs4 option when creating a partition in expresso when there is no kernel support in ubuntu's kernel by default?
[04:24] <bddebian> Heya folks
[05:00] <Burgundavia> mako: you around?
[05:23] <jdub> ha ha
[05:23] <jdub> "13:16 <@jrb> no.  though that would be kinda cool.  I want to send it to foo@webwnk.net and have that be resent to a bunch of other people
[05:23] <jdub> boh
[05:24] <jdub> not htat
[05:24] <jdub> Ars Technica: "Next major Ubuntu release to have worst name yet"
[05:24] <Gman> jdub, it's kinda true!
[05:25] <Gman> eft sounds more like a name for your crack
[05:25] <Gman> still, nice strategy to have fun for the next release
[05:26] <jdub> when it was first decided, i started making lame puns with words that included 'eft'
[05:26] <jdub> like
[05:26] <jdub> dapper has eaten into the edgy release cycle, we'd better be dEFT
[05:27] <jdub> gotta get the cool stuff, don't want to be lEFT behind
[05:27] <Gman> you need to get out more
[05:27] <jdub> yo' jus' jealous, solaris boy :)
[05:30] <Gman> heh
[05:33] <womble> Looking long-term, what's going to be the upgrade plan from Dapper to the next long-support release?  Will it be possible to go direct from one to the other, or will you need to step through all of the intermediates, like you do at the moment?
[05:34] <jdub> womble: i believe we're going to try to provide a direct upgrade path
[05:34] <jdub> (thoughts may have changed on that without me knowing though)
[05:36] <womble> I doubt you're going to get too many admins happy with doing 5 upgrades to go to the next long-stable release, and requiring a reinstall just seems so Slackware.  It'll need a fair bit of fiddling to make upgrades work well across that much technology change, though.
[05:36] <womble> On the other hand, Etch will probably just have been released, so you can base your upgrade fu on the Sarge->Etch upgrade.  <grin>
[05:37] <wasabi__> Buh. I just made this nice gtk.Window, and now I need to turn it into a gtk.Dialog
[05:37] <wasabi__> (or magically add a run() method that works how I want it to)
[05:40] <Amaranth> gtk.Dialog's have children already
[05:40] <wasabi__> yeah i know. i like the work flow though.
[05:41] <Amaranth> you like having it block the main app, you mean?
[05:41] <wasabi__> Naw, ya can just call run() and program flow continues right after.
[05:45] <Gman> Amaranth, i just tried to install alacarte - noticed i had to comment out the #gtk.glade.bindtextdomain('alacarte') in DialogHandler.py to get it to run :/
[05:45] <Amaranth> err
[05:45] <Amaranth> what happens if you don't?
[05:46] <Gman> Gman Traceback (most recent call last):
[05:46] <Gman> Gman   File "/usr/bin/alacarte", line 22, in ?
[05:46] <Gman> Gman     from Alacarte.MainWindow import MainWindow
[05:46] <Gman> Gman   File "/usr/lib/python2.4/vendor-packages/Alacarte/MainWindow.py", line 28, in ?
[05:46] <Gman> Gman     from Alacarte.DialogHandler import DialogHandler
[05:46] <Gman> Gman   File "/usr/lib/python2.4/vendor-packages/Alacarte/DialogHandler.py", line 24, in ?
[05:46] <Gman> Gman     gtk.glade.bindtextdomain('alacarte')
[05:46] <Gman> Gman MemoryError: Not enough memory available
[05:46] <Gman> :/
[05:46] <Gman> also get a bunch of -
[05:46] <Gman> Gman /usr/lib/python2.4/vendor-packages/Alacarte/MainWindow.py: inconsistent use of tabs and spaces in indentation
[05:46] <Amaranth> *boggle*
[05:46] <Gman> Gman /usr/lib/python2.4/vendor-packages/gtk-2.0/gtk/__init__.py: inconsistent use of tabs and spaces in indentation
[05:46] <Gman> Amaranth, yeah, that's what i thought ;)
[05:47] <Gman> Amaranth, this is on solaris fwiw
[05:47] <Amaranth> yeah, i noticed :P
[05:47] <Gman> which may be different on the whole gettext foo stuff
[05:48] <Amaranth> dunno
[05:48] <Amaranth> the tabs and spaces thing is crack
[05:48] <Amaranth> because it's lying
[05:48] <Gman> heh
[05:48] <Amaranth> (i just checked)
[05:49] <Amaranth> as for the MemoryError, i've honestly never even seen that exception :)
[05:49] <Mez> heya Travis, how's things ?
[05:49] <Gman> Amaranth, ok, no worries - seems to work removing the line, so ..
[05:49] <Amaranth> good
[05:50] <Amaranth> Gman: i think that'll break translations
[05:50] <Gman> my only half thought was doing nm on /usr/lib/python2.4/.../gtk/glade.so
[05:50] <Gman> which gives -
[05:50] <Gman> [85]     |         0|       0|FUNC |GLOB |0    |UNDEF  |bindtextdomain
[05:50] <Gman> which doesn't look too healthy :/
[05:51] <Amaranth> dunno what that means but it doesn't look good
[05:51] <Amaranth> Mez: got a new release of alacarte out today :)
[05:51] <Mez> Amaranth, ooh, shiny
[05:51] <Amaranth> already found one bug when you use Human (icons are the wrong size)
[05:52] <Amaranth> i blame Human for not having 24x24 icons for the menus :P
[06:44] <jdub> BenC: https://launchpad.net/distros/ubuntu/+source/linux-source-2.6.15/+bug/40478
[06:44] <Ubugtu> Malone bug 40478 in linux-source-2.6.15 "Promise TX4300/4310 driver update" [Normal,Unconfirmed]  
[06:54] <fabbione> morning
[07:02] <wasabi__> Crud. I need a quicky hint. Generating an apt archive to do some testing of gapti. Trying to generate/sign releases. I generate it... it only has one entry for Packages.gz, fine. Sign it, apt-get update.
[07:02] <wasabi__>  Unable to find expected entry  Packages in Meta-index file (malformed Release file?)
[07:03] <wasabi__> Using apt-ftparchive
[07:05] <wasabi__> Ahh I see.
[07:12] <rohan> is there a specific reason why the dapper live cd doesnt have gcc installed/available ?
[07:12] <seth> rohan, sudo apt-get install build-essential ?
[07:12] <rohan> seth: i know i can install it. but it was there on all versions starting from warty !
[07:13] <TheMuso> rohan: Most users do not need a compiler, and those who do generally know how to install it.
[07:14] <rohan> TheMuso: but what if those who do dont have net access ?
[07:14] <rohan> and taking in account the 650mb limit, we still have 40 mb .. enough for build-essential
[07:15] <jdub> b-e was in ship, but i don't believe we have ship on the livecd
[07:15] <rohan> oh, ok, jdub 
[07:16] <rohan> atleast putting those pkgs in /var/cache/apt/archive would help, forget installing 
[07:17] <rohan> or its still not feasible ?
[07:26] <Chipzz> ugh
[07:27] <Chipzz> does anyone have any comment on: https://launchpad.net/distros/ubuntu/+source/xfree86/+bug/38624 ?
[07:27] <Ubugtu> Malone bug 38624 in xfree86 xlibmesa-gl-dev "no packages contain libGLw.so and dev headers" [Normal,Unconfirmed]  
[07:28] <Chipzz> the headers can be found in libgl1-mesa-swrast-dev though...
[07:35] <wasabi__> Awesome. My gapti program works successfully. Anybody want to try it? :)
[07:35] <rohan> wasabi__: whats gapti ?
[07:35] <rohan> like gdebi ?
[07:35] <wasabi__> Yup.
[07:35] <wasabi__> http://wiki.ubuntu.com/ThirdPartyApt
[07:35] <rohan> ok
[07:36] <wasabi__> package is in my bazaar, builds to a .deb, installs, and pretty much works. ;)
[07:37] <rohan> yes, i read all that :D 
[07:37] <rohan> https://wiki.ubuntu.com/GAptI
[07:39] <wasabi__> You'll want to test with http://akita.larvalstage.net/~wasabi/gapti-repos/photoshop.apt
[07:41] <rohan> wasabi__: what does that install ? photoshop ? o.O
[07:42] <rohan> wasabi__: and, i am on the live cd now, so i cant test, will install soon, and then do it :)
[07:42] <wasabi__> Naw, it's just a bunch of BS empty packages.
[07:42] <wasabi__> Labeled "photoshop" because it's a good example.
[07:43] <Chipzz> wasabi__: no offence meant, but how do you intend to make ISV actually *implement* your proposal?
[07:43] <wasabi__> I don't.
[07:43] <wasabi__> Obviously it needs mindshare. ;0
[07:43] <Chipzz> I mean
[07:43] <Chipzz> allmost all of them ship tar.gz's or self-extracting shellscripts
[07:43] <wasabi__> it beats distributing .tar.gz or shell scripts.
[07:43] <wasabi__> I know.
[07:43] <wasabi__> That does their own users a disservice.
[07:43] <Chipzz> they have no smegging clue at all
[07:44] <Chipzz> I agree
[07:44] <wasabi__> Sure. Hey, don't get me wrong. I don't think I'll wake up oe morning and see everybody using this.
[07:44] <Chipzz> but... what cluebat are you/we going to use? ;)
[07:44] <wasabi__> But having the tools out there, and having ubuntu sanctify it as the "best official way", actually helps.
[07:45] <Chipzz> wasabi__: for example, nvidia doesn't acknowledge the existance of debian or ubuntu packages in their readme
[07:45] <Chipzz> (supposed they are aware of them)
[07:45] <wasabi__> That's unfortunate.
[07:45] <rohan> Chipzz: do thay acknowledge existence of _any_ packages ?
[07:45] <Chipzz> rohan: yes, suse packages
[07:46] <Chipzz> which I think they build theirselves
[07:46] <wasabi__> Chipzz: Honestly, Vmware is my prefered starting company to yell at. ;0
[07:46] <wasabi__> Because I download workstation updates every few weeks. ;)
[07:46] <wasabi__> And ya'll are "nice."
[07:46] <Chipzz> wasabi__: at least they have enough clue to make it a gtk+ 2.0 app ;)
[07:46] <rohan> Chipzz: ewwwwwwww.. rpm !
[07:46] <rohan> Chipzz: not even fedora ?
[07:46] <wasabi__> Either way, this .apt crap actually would work for RPMs
[07:46] <wasabi__> You'd just use apt-rpm.
[07:46] <Chipzz> rohan: not sure; read the readme ;)
[07:47] <rohan> Chipzz: i dont use nvidia drivers :D
[07:47] <wasabi__> One link, you click on it, yes, it has to have seperate compiled versions for your distro, but I suspect that's not That Bad.
[07:47] <wasabi__> Either way it's a seperate problem.
[07:47] <Chipzz> wasabi__: but you know... I think there's another (bigger) issue alltogether
[07:48] <Chipzz> wasabi__: ISV's don't *want* to *support* our packages
[07:48] <Chipzz> support being the imperative word here
[07:48] <wasabi__> Let them support their own.
[07:48] <wasabi__> Like they do on every other OS.
[07:48] <wasabi__> Obviously there is some software where that makes sense, and some where it doesn't.
[07:48] <Chipzz> wasabi__: they do support their tar.gz's or self-extracting shellscripts
[07:48] <Chipzz> 07:43 < wasabi__> it beats distributing .tar.gz or shell scripts.
[07:48] <Chipzz> ;P
[07:48] <Chipzz> errr
[07:48] <Chipzz> that last one was an unfortunate paste ;)
[07:49] <wasabi__> If VMware were to publish .debs, installed this way, they'd be doing their customers a service. They'd need to test them on an Ubuntu box.
[07:49] <wasabi__> I suspect they test their shell scripts there anyways.
[07:49] <Chipzz> I suspect the issue here is "more work"
[07:49] <Chipzz> wasabi__: don't get me wrong, I think gapti is an *excellent* idea :)
[07:50] <wasabi__> Maybe. Maybe not. I don't think it's that big of a deal. Ubuntu is a large installed base. It makes sense for somebody to spend 24 hours to polish the install for them.
[07:50] <wasabi__> I mean, what, third largest distro? If you can't spend 24 hours to make a few .debs for the third largest distro, we've got bigger problems.
[07:50] <Chipzz> wasabi__: btw, how does this integrate with gnome-software-manager?
[07:50] <wasabi__> Doesn't yet. Will talk to mvo.
[07:51] <wasabi__> I suspect all the software needs to do is install a .desktop file, and then they'd be able to uninstall from there
[07:51] <HrdwrBoB> wasabi__: third?
[07:51] <HrdwrBoB> http://distrowatch.com/stats.php?section=popularity
[07:52] <wasabi__> Hah nice.
[07:52] <wasabi__> anyways, POINT MADE
[07:52] <wasabi__> I don't know if it'll h elp. Having the servicve out there to be consumed by those smart enough to use it sounds like a good thing to me.
[07:53] <wasabi__> It can be promoted as the proper way to deliver software to, at least Ubuntu. There's some pushing we can do with vendors there, certainly.  heck windows makes everybody sign drivers now.
[07:53] <wasabi__> This is light compared to that. ;0
[07:53] <wasabi__> Also they gain a free auto update service (apt).
[07:54] <wasabi__> Which saves them time making their own. Maybe that equalizes? Got me.
[07:55] <Chipzz> wasabi__: have you taken a look at /usr/share/update-manager/channels/*info ?
[07:55] <wasabi__> yeah. I took a look.
[07:55] <wasabi__> Looked a bit half baked.
[07:55] <wasabi__> I named my dialog "Channel" though. :0
[07:56] <Chipzz> yea maybe :)
[07:56] <Chipzz> but it would be nice to have one format, instead of 2
[07:56] <wasabi__> Yeah, I agree. The channel thing however doesn't really have a lot to do with a specific package being installed.
[07:57] <wasabi__> gapti is just proof of concept anyways. I want to get some people interested.
[07:57] <Chipzz> well, who says only one package would be in a gapti repository^Wchannel? ;)
[07:57] <wasabi__> me.
[07:57] <wasabi__> Doesn't really make much sense other wise.
[07:57] <Chipzz> vmware for example could have multiple packages in there
[07:58] <Chipzz> vmware-workstation, vmware-server, etc...
[07:58] <wasabi__> I'd have them make a .apt file for each "logical product" they have.
[07:58] <Chipzz> wouldn't that be a bit silly? ;)
[07:58] <wasabi__> Run gapti, you'll see where im going with the UI
[07:58] <wasabi__> It checks out, intalls, and works pretty easily.
[07:59] <Chipzz> debs available? ;)
[07:59] <wasabi__> Naw. bzr checkout, debuild -us -uc ;)
[08:00] <wasabi__> I want the whole .apt file part to be invisible to the user. What they are doing is clicking on a link to install a product they just bought.
[08:00] <Chipzz> shouldn't the first .apt file be gapti.apt? ;)
[08:00] <wasabi__> heh no
[08:00] <wasabi__> So, After purchasing vmware workstation, you'd have a Click Here to Installl.
[08:00] <Chipzz> except that it's free nowadays ;)
[08:00] <wasabi__> You'd have to go buy vmware-server to have a link for it.
[08:00] <wasabi__> Yeah, except that haha
[08:02] <Chipzz> wasabi__: btw, check pvt ;)
[08:04] <wasabi__> I'm not really even thinking about stuff like nvidia.
[08:04] <wasabi__> We package their stuff, and I like it that way for now.
[08:04] <wasabi__> Since they don't even support their own software, actually.
[08:04] <wasabi__> ... you ever seen a nvidia tech support line for their drivers for any platform?
[08:05] <Chipzz> internal bug-tracking system, "it will get fixed 'in a future release'", stuff like that?
[08:06] <wasabi__> This may suck, but I'm really more concerned with commercial ISVs.
[08:06] <wasabi__> ie non-open source.
[08:06] <Chipzz> uhu
[08:06] <Chipzz> but, I'm wondering
[08:06] <TheMuso> vmware free? I thought you still had to buy it.
[08:07] <wasabi__> open source stuff gets packaged in Ubuntu and other distros just fine.
[08:07] <wasabi__> And that's cool, because it can be supported centrally.
[08:07] <Chipzz> why are you concentrating on single packages? the format on http://wiki.ubuntu.com/ThirdPartyApt mentions channels, and a single package may have a chain of dependencies anyway
[08:08] <wasabi__> Because the point is to satisfy a use case... a user buys software and clicks on a link to install it.
[08:08] <wasabi__> Underneath sure, that may have deps.
[08:08] <wasabi__> And may be many .deb files, but the user is really only aware of a single logical product.
[08:09] <wasabi__> behold the photoshop example.
[08:09] <wasabi__> I'm not just trying to give a user a more convient way of browsing software offered by a repository. We already have 20 of those.
[08:12] <Chipzz> I was thinking about the following approach: have a superset of the channel .info format, in which you specify one or more packages (not the dependencies); user clicks on that, file gets download and the channel .info subset gets written in /usr/share/update-manager/channels/, you get the standard gnome-software-properties 'add channel' dialog, user selects components, clicks ok, you run apt-get update, user gets a second dialog with 
[08:13] <wasabi__> Oh, yeah, don't get me wrong, I want to reuse the gnome-software-properties UI pieces... and all of that. They just aren't done yet.
[08:13] <wasabi__> .info files too.
[08:13] <wasabi__> They also aren't currently reusable. ;)
[08:13] <wasabi__> So I whipped up my own.
[08:14] <Chipzz> uhu :)
[08:14] <wasabi__> gapti right now has a dialog that asks you if you are sure you want to subscribe to the channel.
[08:14] <Chipzz> so, like I said in private, the .info format has its deficiencies...
[08:14] <wasabi__> That should totally be the same UI in g-s-p.
[08:15] <wasabi__> The .apt files right now are pretty simple too. Just a simple multi-stanza dpkg format file with basically filters.
[08:15] <wasabi__> In fact they look quite a bit like the .info files.
[08:16] <wasabi__> i need one file for all distros though.
[08:16] <wasabi__> .info doesn't quite fit right with that.
[08:16] <Chipzz> why?
[08:16] <wasabi__> Web page can't usually properly detect the user's OS/distro. I mean, it could.
[08:16] <wasabi__> But it would suck.
[08:17] <wasabi__> So the .apt file can have a stanza for redhat, one for ubuntu, and one for debian.
[08:17] <wasabi__> And the proper stanza is chosenl
[08:17] <Chipzz> uhu
[08:17] <wasabi__> what hte heck does 'uhu' sound like?
[08:17] <Chipzz> like: "you're right" :)
[08:17] <Chipzz> (yeah)
[08:18] <wasabi__> Also, different apt lines for different releases of distro.
[08:18] <wasabi__> dapper, breezy, etc.
[08:18] <Chipzz> would they have to be different?
[08:18] <wasabi__> Just like companies now maintain win9x and winnt and win2k and winxp versions of some of their software.
[08:18] <wasabi__> They'd have to do the same with our point releases.
[08:18] <Chipzz> except for the dapper/breezy part that is
[08:18] <wasabi__> yeah of course, since it doesn't exist yet heh
[08:19] <Chipzz> couldn't you just list the releases elsewhere?
[08:19] <Chipzz> ie on a different line in your file?
[08:19] <wasabi__> That's basically what I do.
[08:19] <Chipzz> I should probably read the wiki ;)
[08:19] <wasabi__> yeah. ;0
[08:20] <wasabi__> Run it yet?
[08:21] <wasabi__> My favorite part is how I'm (ab)using Recommends/Suggests.
[08:21] <wasabi__> probably exactly how they were intended to be (ab)used. ;0
[08:22] <Chipzz> not yet ;)
[08:22] <wasabi__> ahh
[08:22] <Chipzz> but I'll definately take a look later
[08:23] <wasabi__> oh also .apt files deliver gpg keys
[08:23] <wasabi__> for apt-key
[08:23] <Chipzz> now *that* is nice :)
[08:23] <wasabi__> and the entire file is signed by the same key
[08:24] <wasabi__> so you can at least guarentee that the repository you are adding has the same key as the file you downloaded
[08:28] <Chipzz> aha
[08:28] <Chipzz> mesa build allmost done :)
[08:30] <wasabi__> me->bed
[08:30] <wasabi__> night
[08:31] <Chipzz> night :)
[08:49] <pitti> Good morning
[08:59] <Treenaks> Any X gurus in the house^Wchannel? :)
[08:59] <Treenaks> (if there are, could you please look at bug 20283? :))
[08:59] <Ubugtu> Malone bug 20283 in xserver-xorg-driver-ati "[fgl v5000]  really bad sync" [Normal,Needs info]  http://launchpad.net/bugs/20283
[09:00] <fabbione> Treenaks: hell.. please stop pushing bugs here.. :)
[09:01] <Treenaks> fabbione: ok :)
[09:01] <fabbione> Treenaks: i upload a nerw version of the driver 20 minutes ago..
[09:01] <fabbione> does it fix?
[09:01] <Treenaks> fabbione: No
[09:01] <Treenaks> unless.. *hmm*
[09:01] <fabbione> how could you have tested it, if it hasn't been published on archive yet?
[09:02] <Treenaks> didn't get that new one yet :) are there more changes than are noteed in the changelog, or is that the only one?
[09:02] <fabbione> there are 2 changes
[09:02] <Treenaks> fabbione: OK.. the GART thing might be related..
[09:18] <carlos> pitti: ping
[09:20] <dholbach> HAPPY HUG DAY
[09:23] <mdke> morning dholbach 
[09:23] <mdke> happy hug day to you too
[09:23] <dholbach> hey mdke :)
[09:34] <pitti> hey carlos, moin seb128
[09:34] <carlos> pitti: I'm still fixing translation domains exports from Rosetta
[09:35] <seb128> hi carlos pitti
[09:35] <carlos> pitti: and I detected that your report says that the buildd output includes some man pages using .po files to translate
[09:35] <carlos> seb128: hi
[09:35] <carlos> pitti: and also, you are including translation domains from universe, like beagle
[09:35] <pitti> carlos: oh, these are probably superfluous in the final langpacks
[09:35] <pitti> carlos: beagle is in main (the source)
[09:36] <carlos> oh
[09:36] <carlos> hmmm
[09:36] <pitti> carlos: we only filter by source, not by binary
[09:36] <seb128> carlos: we build nautilus by example with libbeagle
[09:36] <carlos> is there an easy way to see if a sourcepackage is in main?
[09:36] <carlos> I was using apt-get search
[09:36] <pitti> carlos: apt-cache madison srcpkg
[09:36] <carlos> but that's for binaries...
[09:36] <carlos> pitti: ok, thanks
[09:37] <carlos> seb128: I see
[09:37] <pitti> it works as well with binary packages, it figures out the corresponding source package
[09:37] <seb128> carlos: apt-cache showsrc beagle | grep Directory
[09:37] <carlos> pitti: also, I'm preparing a list of translation domains that I don't know where they come from and thus, I cannot import them into Rosetta... I will send you the list when I'm done with it
[09:38] <pitti> carlos: yes, then I'll check their origin
[09:38] <carlos> pitti: thanks
[09:38] <seb128> pitti: is there a plan for weekly translations update or something from now?
[09:38] <pitti> carlos: it's very likely that the heuristics in my script picked up some wrong ones
[09:38] <seb128> pitti: that would have be nice to get updates before beta by example :/
[09:38] <pitti> seb128: no problem from my side for that
[09:38] <carlos> pitti: I will provide you also the list of invalid .pot files, don't worry
[09:39] <carlos> seb128: I think that's doable
[09:39] <mh21> pitti: morning
[09:39] <mh21> pitti: What's broken with cups auth?
[09:39] <pitti> mh21: hey Mischa
[09:39] <carlos> the language pack exports from Rosetta are working daily
[09:39] <pitti> mh21: I saw your upstream bug report :)
[09:39] <carlos> and most .pot files are already imported
[09:40] <carlos> still need to import some .po files
[09:40] <pitti> mh21: local authentication works fine (through gnome and the command line tools), it's just the web frontend auth that's busted
[09:40] <carlos> but that will be appearing over time until Dapper release
[09:40] <pitti> mh21: it works fine if cupsd runs as root, but instead of not doing admin jobs at all if it runs as cupsys (like in breezy), it just does them without ever asking for auth
[09:41] <pitti> mh21: the idea is to disable admin stuff in the web interface by default, and reenable it with sudo adduser cupsys shadow, so that cupsd can read the passwords
[09:41] <pitti> mh21: now that beta is out, this one is very high on my list of stuff to fix :)
[09:41] <pitti> carlos: what's the latest word wrt. KDE import?
[09:41] <seb128> pitti: https://launchpad.net/distros/ubuntu/+source/hal/+bug/40437 
[09:41] <Ubugtu> Malone bug 40437 in hal "backport CVS code for is_mounted_read_only property" [Normal,Confirmed]  
[09:41] <mh21> mh21: so running cups as root should make it at least authenticate?
[09:41] <seb128> pitti: if you have some minutes for that today ... :)
[09:42] <mh21> pitti: so running cups as root should make it at least authenticate?
[09:42] <carlos> pitti: all .pot files are approved, I'm setting the flag to get them exported as part of the language packs now
[09:42] <pitti> seb128: oh, sorry for that
[09:42] <pitti> mh21: yes
[09:42] <seb128> pitti: sorry for what?
[09:42] <pitti> seb128: for delaying it that long
[09:42] <pitti> seb128: is that the same issue you pointed me to maybe two weeks ago?
[09:42] <seb128> pitti: hum, I send the bug before going to bed, I would not call that delayed ... :)
[09:42] <carlos> pitti: there are some .po files that are not yet imported, but I already wrote code to fix that and it should land on production anytime soon (Tuesday as the worst case)
[09:42] <seb128> pitti: probably yep
[09:43] <seb128> pitti: I've the patched package ready for upload, I just want your opinion before uploading, should be fine imho
[09:43] <carlos> pitti: would you do the same kind of report you send for dapper but for hoary and breezy?
[09:43] <pitti> carlos: ok, then let's schedule new packs next Wednesday; seb128, is that fine for you?
[09:43] <seb128> pitti: yeah, works for me
[09:43] <carlos> pitti: I would fix those too 
[09:44] <pitti> carlos: I can't; my script doesn't support buildd extraction for stables, just for the very latest tarballs
[09:44] <carlos> pitti: also, we should try to release a monthly update for those releases, now that language packs start to be stable enough...
[09:44] <carlos> pitti: hmm
[09:44] <carlos> ok
[09:44] <carlos> Is there anyway to fix that?
[09:44] <pitti> seb128: http://mail.gnome.org/archives/nautilus-list/2006-March/msg00046.html <= that's still on my list, that was the mail you pointed me to two weeks ago
[09:44] <pitti> seb128: so that's the same issue
[09:45] <pitti> carlos: right
[09:45] <seb128> pitti: correct
[09:45] <seb128> pitti: http://mail.gnome.org/archives/nautilus-list/2006-March/msg00045.html has the patch and is the mail to read about it probably
[09:46] <seb128> (I've pointed it on the bug)
[09:46] <pitti> seb128: looks straightforward enough, fire away :)
[09:46] <seb128> pitti: what I though, thank you ;)
[09:47] <dhonn> hey how hard would it be to put the Text-Mode Install CD on the Desktop Live CD?
[09:48] <fabbione> not too hard assuming a 1.2GB CD
[09:48] <fabbione> oh hold on..
[09:48] <fabbione> that's like the DVD :)
[09:48] <dhonn> it should have both on one cd under 650 mb
[09:49] <fabbione> impossible
[09:49] <fabbione> there is no space
[09:49] <TheMuso> The live CD can install now anyway.
[09:49] <TheMuso> With espresso.
[09:51] <dhonn> it should be easy to put all this on one cd... i mean it should be able to load a text and graphical interface 
[09:52] <HrdwrBoB> it should be
[09:52] <HrdwrBoB> but it's not.
[09:53] <Mithrandir> dhonn: not if you want it to boot in a sensible amount of time.
[09:54] <jdub> Mithrandir: a text-based espresso frontend would be interesting (and amusing)
[09:54] <jdub> (not that i really care, but it would be interesting)
[09:54] <jdub> (livecd install uber alles, etc)
[09:54] <Mithrandir> jdub: apart from the a11y benefits, I don't really see a point?
[09:55] <jdub> Mithrandir: it'd be an amusing abuse of the espresso/d-i relationship
[09:55] <jdub> i don't think there's much point either
[09:56] <jdub> apart from theoretical interest value
[09:56] <TheMuso> d-i could be given a11y for the console, but that is not exactly an easy task.
[09:56] <TheMuso> However not possible for low vision users and those with motor difficulties.
[09:56] <TheMuso> AFAIK.
[09:57] <dholbach> hey jdub!
[09:57] <jdub> it would be easier to give a11y love to the espresso experience
[09:57] <Mithrandir> TheMuso: why would it be particularly hard?
[09:57] <jdub> yo dholbach 
[09:57] <dholbach> how's it going?
[09:58] <TheMuso> Mithrandir: FOr speech, one firstly needs to include ALSA drivers, and have some way of accessing the speech synthesizer, which would not fit into the initrd image.
[09:58] <TheMuso> And since festival/flite are rather big...
[09:59] <TheMuso> I was for a while investigating a cludgy implementation, but there were a few too many points of failure for it to be usable in all cases.
[09:59] <Mithrandir> TheMuso: there's a brltty-udeb
[09:59] <TheMuso> Mithrandir: That doesn't take care of speech however.
[10:15] <mh21> pitti: Normally, access to http://localhost:631/admin/conf/cupsd.conf fails; with usermod -aG shadow cupsys it works
[10:15] <mh21> pitti: How is it supposed to be?
[10:16] <pitti> mh21: hm, indeed
[10:16] <pitti> mh21: however, adding a printer does not require any password
[10:17] <mh21> pitti: local auth?
[10:17] <pitti> mh21: not through TCP
[10:20] <Keybuk> WTF is brltty
[10:20] <pitti> mh21: local auth with the certificate works with the gnome tools and command line tools only
[10:20] <Keybuk> and why is it using all of my CPU and MEM ?
[10:21] <seb128> Keybuk: because you didn't update to the "fixed" version dholbach uploaded yesterday?
[10:21] <Keybuk> seb128: I'm updating now ... I was wondering why APT/dpkg were so damned slow
[10:25] <mh21> pitti: I think it is more along the lines of AuthType Basic, Require user @SYSTEM in /etc/cups/cupsd.conf:<Location /admin>
[10:30] <pitti> mh21: indeed
[10:30] <pitti> mh21: that's why you can use any lpadmin user in the http auth dialog
[10:30] <pitti> mh21: the point is that it doesn't appear if you run cups as non-root
[10:31] <mh21> it does when you insert these two lines in the admin section
[10:31] <mh21> pitti: its not there normally
[10:32] <pitti> mh21: hm, it's for me, in the <Policy default> section
[10:32] <pitti> ah, I see what you mean
[10:32] <pitti> mh21: still strange that it works as root then
[10:33] <jordi> pitti: how hard is it getting to get fixes in dapper?
[10:33] <jordi> pitti: for trivial stuff like dictionaries
[10:33] <pitti> jordi: not hard at all
[10:33] <jordi> ok
[10:42] <pitti> mh21: upstream's default is to not restrict the mere access to /admin by default, but just require authentication for the 'commit' actions - i. e. when you configure a new printer, you can click through all the steps, and just the final 'do it' button asks for auth
[10:42] <pitti> mh21: that's why Location /admin is not resticted, just e. g. CUPS-Add-Printer is
[10:50] <mh21> pitti: I'll look into the policy authentication...
[10:50] <pitti> mh21: I'll do so too ASAP, I have the Debian packages for comparison here
[11:07] <pitti> Kamion: good morning! are you already awake?
[11:24] <\sh> morning
[11:25] <\sh> guys, do we have an nfs mountable ubuntu archive?
[11:30] <Kamion> pitti: hi
[11:30] <pitti> hi Kamion 
[11:30] <pitti> Kamion: can you join #ubuntu-desktop for a moment, please?
[11:36] <Mithrandir> doko: re bug 40338 ; that's an xresprobe bug and basically how X config works.  (Which is somewhat broken for LCD panels)
[11:36] <Ubugtu> Malone bug 40338 in Ubuntu "amd64-live - 1920x1200 screen detected as 1600x1200" [Normal,Unconfirmed]  http://launchpad.net/bugs/40338
[11:37] <Mithrandir> doko: and it's a duplicate, but that's ok.
[11:41] <doko> ok, thanks for looking
[11:47] <dholbach> ogra: dia 0.95.0 is in incoming
[11:48] <dholbach> ogra: and it closes loads of bugs
[11:48] <Diziet> ogra: re the firefox homepage, did you read the mails from yesterday ?
[11:48] <ogra> dholbach, yes, yes, i'll create a syc bug :)
[11:48] <ogra> Diziet, before i can read mail, evo needs to be fixed ...
[11:49] <ogra> i'll care for a replacement today, sorry, couldnt read mail the last two days :/#
[11:49] <Diziet> ogra: :-/
[11:50] <Diziet> OK, well, we have a planned answer.  Only the package dependencies need to be decided on.  I was going to write it up on the wiki page and implement it (in m-f-l-all at least) ASAP.
[11:52] <mdke> Diziet, I'm slightly worried that the ed/k/xubuntu-docs maintainers will not be keen on doing all the symbolic links, do you think it is a sane solution?
[11:52] <ogra> Diziet, my prob is, that we use the ubuntu-docs package, so just conflicting wont work
[11:52] <Diziet> mdke: Why is it so hard to do these links ?  It's about 4 lines of your favourite scripting language.
[11:53] <Diziet> ogra: You really need to read these emails.  Bear with me and I'll put them on a web page.
[11:54] <ogra> Diziet, i need my mailer, so i'll have to solve it today... 
[11:54] <mdke> Diziet, fine by me, as long as they don't mind doing it. It's right that there needs to be a link for every mfla locale? if so, we'd probably need a way to figure out when some of those are added to make sure that the link gets created for the new locale
[11:54] <ogra> (i get at east 400 mails a day, i dont want them to stack up :)
[11:54] <Diziet> http://www.chiark.greenend.org.uk/~ijackson/d/ubuntu-docs-firefox.txt
[11:55] <Diziet> mdke: There needs to be a link for every locale listed in the list on the wiki page.
[11:55] <Diziet> If the *-docs maintainers don't want to do it then I could do it myself.
[11:56] <Diziet> It might be quicker and easier for one person to do it anyway.
[11:56] <mdke> Diziet, right, so when that list changes?
[11:56] <Diziet> That way we can test it locally without having to break everyone's firefox for a day due to the path being wrong.
[11:56] <Diziet> mdke: The wiki page has a process for changing the list, which tells you what to do in what order.
[11:56] <Diziet> But we need to think about the dependencies.
[11:57] <Diziet> I'm tempted to just ditch the dependencies and if you install partial updates or something you just don't get a homepage.
[11:57] <Diziet> Since the system is otherwise quite useable.
[11:57] <heno> Kamion: xcursor-themes have been approved for main, but hasn't yet been moved in there. Can you help with that?
[11:57] <Diziet> The problem being that the dependency is really difficult to express.
[11:58] <heno> It is useful for the hight visibility theme setup
[11:58] <Kamion> heno: yeah, I'll do anastacia processing in a bit
[11:58] <mdke> Diziet, how about "depends on xubuntu-docs or ubuntu-docs or kubuntu-docs or edubuntu-artwork"?
[11:58] <Kamion> I obviously wasn't doing it during the beta freeze
[11:59] <heno> Kamion: ah, ok thanks
[12:00] <Diziet> mdke: That's not right, because it has to be specific versions and it's not `or', it's `whichever of *-docs is going to win the alternative needs to be at least xyz'.
[12:01] <pitti> Hi Diziet 
[12:01] <mdke> Diziet, gosh
[12:01] <Diziet> I think it would be expressable with Breaks and versioned Provides, but we don't have either of those.
[12:01] <Diziet> pitti: Hello.
[12:01] <pitti> Diziet: do you have some minutes today to speak about gs-gpl vs. gs-esp?
[12:01] <Diziet> Sure.
[12:01] <Diziet> In terms of ReducingDuplication, you mean ?
[12:02] <pitti> yes
[12:02] <pitti> Diziet: I just checked dependencies
[12:02] <pitti> Diziet: the only package that b-deps on gs-gpl without alternatives is quagga
[12:02] <pitti> I guess for building documentation and such
[12:02] <pitti> Diziet: do you see any reason to keep gs-gpl in main? i. e. does it have anything gs-esp hasn't?
[12:03] <Diziet> Yes.  It has a different set of bugs.
[12:03] <pitti> hpijs and scribus prefer gs-gpl over gs-esp, but have proper alternative dependnecies
[12:04] <Diziet> Unfortunately the gs area is a mess because it's not clear what upstream is.  I hope gs-esp is going to win but in the meantime I think we need to provide gs-gpl so that users who find that gs-esp breaks have a workaround.
[12:04] <Diziet> Possibly quagga's docs triggered some gs-esp bug in an earlier version.
[12:04] <Diziet> Or possibly it predates gs-esp.
[12:04] <pitti> Diziet: but it seems that most users will only have gs-esp installed anyway, since most packages alternatively depend on it
[12:05] <dholbach> Diziet: had any time to look at mozilla bug 312998? it has a patch which might fix up one of the gtkmozembed problems
[12:05] <Ubugtu> Mozilla bug 312998 in Embedding: GTK Widget "fix gtkmozembed's EmbedWindow::GetVisibility" [Major,Assigned]  https://bugzilla.mozilla.org/show_bug.cgi?id=312998
[12:05] <pitti> Diziet: under the assumption that quagga works with gs-esp (i'd change it to b-dep: gs-esp | gs), would you still keep gs-gpl in main?
[12:05] <Diziet> pitti: I think I would keep it in main, yes.
[12:05] <nomed> hi all
[12:05] <Diziet> dholbach: I was just reading that.
[12:06] <nomed> the bug #40418 has been reported in xubuntu-meta ..
[12:06] <Ubugtu> Malone bug 40418 in xubuntu-meta xubuntu-live "xubuntu live cd fails to display desktop correctly" [Normal,Unconfirmed]  http://launchpad.net/bugs/40418
[12:06] <Diziet> I'm not sure of the connection.
[12:06] <nomed> should i move it somewhere else ?
[12:06] <dholbach> Diziet: ah ok. Just wanted to make sure you knew about it. :-)
[12:06] <Diziet> Thanks.  But do we have any reason to think this is responsible for our crash ?
[12:07] <Riddell> Keybuk: your kcontrol.postinst fails with "rmdir: /etc/hotplug/usb: Directory not empty"
[12:08] <dholbach> Diziet: this problem seems to show in devhelp (not sure if it fixes other issues) - devhelp displays a really weird "start up page" due to that and I'd trust chpe with his patches, he's epiphany upstream and dug himself into quite a lot of mozilla problems.
[12:08] <Keybuk> Riddell: bah :)
[12:08] <Keybuk> forgot an || true?
[12:08] <dholbach> Diziet: So, I don't think that's a crash, it fixes.
[12:08] <Riddell> Keybuk: that would work, I'll add that
[12:09] <Diziet> dholbach: Right.
[12:09] <ogra> Diziet, ok, read the thread, thanks for uploading it
[12:09] <Diziet> ogra: NP
[12:10] <Keybuk> Riddell: already added and uploaded
[12:10] <Diziet> dholbach: I think I should put mozilla 312998 on my ff to-fix list though - the patch looks sane.  So thanks.
[12:10] <Ubugtu> Mozilla bug 312998 in Embedding: GTK Widget "fix gtkmozembed's EmbedWindow::GetVisibility" [Major,Assigned]  https://bugzilla.mozilla.org/show_bug.cgi?id=312998
[12:10] <ogra> Diziet, edubuntu rlies on the ubuntu server documentation, so both packages need to be installable alongside ... the idea with the directory sounds intresting, can FF handle that ? 
[12:10] <dholbach> Diziet: Thank YOU!
[12:11] <Diziet> ogra: FF doesn't care.  Before I sent the mail I did test that the alternatives system works with directories.
[12:11] <ogra> cool
[12:11] <ogra> then lets do it like that ! :)
[12:12] <pitti> Diziet: oh, btw, do you have updating to ffox 1.5.0.2 on your list? I'll mail you the CVE list shortly
[12:12] <ogra> the current system would require that all systems have exactly the same amount of translated pages, else you flip over to another page :)
[12:12] <Diziet> pitti: Yes, I do.  And yes, please, thanks.
[12:13] <Diziet> pitti: For form's sake, have you already requested and got a uvf exception ?  Not that we have much of an alternative ...
[12:13] <pitti> Diziet: no, I didn't
[12:13] <Diziet> pitti: OK.
[12:14] <Diziet> ogra: No, because of the symlinks, you can fall back to English for the cases where there is no translation.
[12:14] <ogra> Diziet, on edubuntu it uses the translated ubuntu page if available 
[12:15] <ogra> i only have the edubuntu page in english installs
[12:15] <Kamion> Keybuk: rmdir --ignore-fail-on-non-empty is better than || true
[12:16] <Diziet> ogra: That's because we haven't implemented yet what is described in that mail chain.
[12:16] <Keybuk> Kamion: does that work in busybox?
[12:16] <Kamion> Keybuk: no, but who cares in postinst scripts
[12:16] <ogra> Diziet, thats why i talked about "the current system" ;)
[12:17] <Diziet> Yes, we know it's broken.
[12:18] <pitti> Diziet: mailed
[12:19] <Kamion> Riddell: I'm doing the espresso => ubiquity rename at the moment; you might want to not accumulate any changes in your branch, otherwise your merge will be very complicated
[12:19] <Kamion> $ bzr diff | wc -l
[12:19] <Kamion> 35724
[12:20] <Diziet> OK, I'm not imagining it.  I don't have a mail telling me who to ask about a UVF exception.
[12:20] <Diziet> I suppose I'll just wibble here and hope some poor sod pipes up.
[12:20] <Kamion> Diziet: see DeveloperResources
[12:21] <Kamion> in the "Freezes" section
[12:21] <Diziet> dholbach: Unfortunately that gtkmozembed upstream bug seems to be languishing.
[12:22] <Diziet> (The crash one, 325884)
[12:22] <dholbach> mozilla bug 325884
[12:22] <Ubugtu> Mozilla bug 325884 in Embedding: GTK Widget "race? crash in EmbedPrivate::Realize with gtkmozembed" [Critical,New]  https://bugzilla.mozilla.org/show_bug.cgi?id=325884
[12:24] <dholbach> that's quite unfortunate :-/
[12:24] <dholbach> I wonder why they didn't ask for more debugging symbols.
[12:26] <mdz> Keybuk: should we add an external tool somewhere for creating /var/run directories in the appropriate manner?
[12:26] <Riddell> Kamion: ok, let me know when I'm good to merge
[12:27] <Diziet> Kamion: Ah, thanks.  I've added some cross-references to the wiki.
[12:28] <Keybuk> mdz: I've been thinking about that
[12:28] <Diziet> dholbach: I'm tempted to try to set up helgrind.
[12:29] <dholbach> Diziet: I suppose 'helgrind' is not a typo. :-)
[12:29] <Diziet> Apparently not ...
[12:30] <Diziet> Although the thought of running mozilla under an obscure but strict validator isn't giving me a good feeling.
[12:32] <Kamion> Diziet: thanks
[12:32] <mdz> Keybuk: all the dependencies would be sort of a pain though
[12:35] <Keybuk> Launchpad has suddenly lost its monospace-ness for the text entry box
[12:35] <Keybuk> Diziet: is that a firefox bug?
[12:36] <Diziet> I don't think it should be.
[12:36] <Diziet> But I wouldn't rule it out.
[12:37] <Keybuk> seems they're using
[12:37] <Keybuk>   font-family: sans-serif;
[12:37] <Keybuk>   font-family: caption;
[12:37] <Diziet> Well, that's not my fault then, at least not entirely :-).
[12:38] <Keybuk> yup
[12:38] <Keybuk> sticking font-family: monospace !important; in my "make LP not suck" stylesheet fixed it for me
[12:41] <Keybuk> Diziet: apparently it's your fault for not picking the right "caption" font
[12:41] <mdz> pitti: is cups basically maintained by one person upstream?
[12:42] <mdz> pitti: I think every commit in that changelog was by the same person
[12:42] <pitti> mdz: yes, Michael Sweet
[12:42] <pitti> mdz: there are some contributors, mainly from Apple, but mikes commits everything
[12:42] <Diziet> I have no idea what `caption' is.
[12:43] <pitti> mdz: i. e. patch proposal go through bug reports (STRs)
[12:43] <Keybuk> Diziet: whatever font the GUI is using
[12:43] <Keybuk> apparently
[12:44] <Keybuk> so for me that would be "Sans 8"
[12:45] <Keybuk> the font it's picked looks terrible; (a) it's a Serif font, and (b) it looks more like one of those silly printing fonts than one designed for the screen
[12:45] <Fjodor> Hi. Though this isn't a support channel, I'd like to report a problem. Only thing I did to cause it was dist-upgrade to dapper. Now I can't enter Danish chars in emacs, and most graphical apps report "Warning: locale not supported by Xlib, locale set to C"
[12:47] <Diziet> Though it isn't a support channel, you would like support ?  Sorry, but it's still not a support channel even for you.
[12:47] <Keybuk> http://people.ubuntu.com/~scott/firefox-font-bug.png
[12:48] <pitti> Keybuk: hm, what is this /var/run/cups chgrp'ing to 'lpadmin' fixing?
[12:48] <Keybuk> pitti: the fact that /var/run was in the wrong group?
[12:48] <mvo> Fjodor: does it help if you run the language-selector again?
[12:48] <Fjodor> Keybuk: No, I don't seek an answer right now, but would like to report it as a problem
[12:48] <Diziet> Keybuk: That's pretty cruddy, yes.
[12:49] <pitti> Keybuk: ah, so really just cosmetics? or had it some actual impact?
[12:49] <Diziet> Hrm.  Are you sure about that css ?
[12:49] <Keybuk> pitti: wasn't aware of a bug, but I figured there could be one, so fixed it anyway
[12:49] <Keybuk> Diziet: yes, the CSS looks ok
[12:50] <pitti> Keybuk: ok, thanks
[12:50] <pitti> since it's 0755, it shouldn't matter
[12:50] <Fjodor> Keybuk: Oh, sorry. Read the lines wrong...
[12:50] <Fjodor> Diziet: : No, I don't seek an answer right now, but would like to report it as a problem
[12:51] <Diziet> Keybuk: No, what I mean, is are you sure that that's what the CSS says the font should be.
[12:54] <Keybuk> Diziet: yeah, used dominspector on it
[12:54] <Keybuk> firefox thinks the font should be "caption"
[12:54] <Diziet> I say `fc-match caption' and it wants to use DejaVu Sans.
[12:55] <mdz> Keybuk,Diziet: speaking of which font the GUI is using, my firefox has been doing this all week: http://people.ubuntu.com/~mdz/temp/Screenshot.png
[12:55] <Keybuk> quest scott% fc-match caption
[12:55] <Keybuk> DejaVuSans.ttf: "DejaVu Sans" "Book"
[12:55] <mdz> I have no idea how it chooses that font
[12:55] <mdz> but it's ginormous
[12:56] <Diziet> Keybuk: You want `caption:anymetrics=1' at the very least.
[12:56] <mdz> it's only my profile which is busticated
[12:56] <Keybuk> Diziet: huh?
[12:56] <Diziet> What do you have set in the fonts preferneces ?
[12:56] <Keybuk> assume I know nothing about font selection
[12:57] <Diziet> (And damn, the font prefs box is still too damn small!)
[12:57] <Keybuk> Diziet: http://people.ubuntu.com/~scott/firefox-fonts-prefs.png
[12:57] <mdz> absolutely everything with 'font' in the name under about:config is set to default
[12:58] <Diziet> Keybuk: Not you, I can reproduce your LP problem.
[12:58] <Diziet> mdz: That's nice.  Ummm.
[12:58] <Keybuk> Diziet: by the magical process of going one-by-one through the font list, I can tell you that it's picked FreeSerif as my font-of-choice for textarea
[12:58] <Diziet> mdz: What about the dpi setting ?
[12:59] <mdz> Diziet: 72
[12:59] <Diziet> Is that correct ?
[12:59] <mdz> which is pretty close to correct for my laptop display
[12:59] <mdz> which is 14" diagonal, 1024x768
[01:00] <mdz> aha, setting that to 96 (default) fixes it
[01:00] <mdz> how ridiculous that a lower dpi setting causes larger fonts to be used; that's backwards
[01:00] <mdz> Diziet: thanks
[01:00] <Diziet> The whole dpi thing is a complete nightmare.
[01:00] <Keybuk> mdz: what's your font-render
[01:00] <mdz> it doesn't take effect until firefox is restarted
[01:01] <Keybuk> ignore that
[01:02] <Diziet> We've made rendering randomly break for the vast majority of people who want fonts to be a particular size in pixels, to make life easier for the few people who can afford absurdly high res displays just to make it pretty.
[01:02] <mdz> how have we done that exactly?
[01:03] <Diziet> By making things depend (in a flakey way) on the (often inaccurate) dpi setting.
[01:03] <lifeless> Diziet: most screens sold these days are > 100 dpi
[01:03] <Diziet> lifeless: Surely not.  Most of them are LCDs nowadays ?
[01:03] <lifeless> also note that gnomes handling of DPI is rather broken.
[01:03] <lifeless> even though X often gets it right
[01:04] <lifeless> Diziet: yes, and the resolutions they are selling at are high dpi - my last three laptops have all been > 100 dpi, without me trying for that specifically.
[01:04] <lifeless> last one was 144, current one is 120
[01:04] <dholbach> pitti: you have time to join us at the HUG DAY in #ubuntu-bugs?
[01:04] <lifeless> I dont recall the last time I saw a < 100 dpi new machine of any sort.
[01:04] <mdz> lifeless: I'm typing at one right now
[01:04] <Diziet> I don't think most people are buying such fancy laptops :-).
[01:04] <pitti> dholbach: I'll join and try to read a bit
[01:04] <Keybuk> Diziet: >100 dpi screens are standard these days
[01:04] <lifeless> mdz: ok, I've seen one now :)
[01:04] <Keybuk> even on the cheap laptops
[01:05] <Diziet> Pretty much all desktop flatpanels are more like 72.
[01:05] <lifeless> Diziet: Seriously, not fancy to get this.
[01:05] <mdz> and in fact I just had a friend return a laptop because the display resolution was too high
[01:05] <lifeless> mdz: wow.
[01:05] <Keybuk> Diziet: my cheap desktop panel is 120 dpi or something
[01:05] <mdz> he's a web designer, and needs to see things layed out similarly to most casual users
[01:05] <Diziet> 1280x1024 and really tiny ?
[01:05] <mdz> laid
[01:06] <lifeless> mdz: yah. dpi matters tonnes. I find things are usually ok if I go into every possible place and set it to match the hardware
[01:06] <lifeless> but thats asking a lot of users
[01:06] <mdz> lifeless: that'll still never give you a good correspondence between images and text on web pages
[01:07] <lifeless> images have a defined dpi
[01:07] <lifeless> let me grab the reference
[01:07] <mdz> you need to use larger fonts to make things readable, and that'll change the ratio
[01:07] <mdz> images for the web are usually ~72dpi
[01:07] <lifeless> larger font? I usually run at 7pt these days
[01:08] <Fjodor> mvo: Thanks for suggesting, but no. And Danish chars actually work fine in other apps, just not emacs
[01:08] <Keybuk> the whole dpi/font thing is a mess
[01:08] <Keybuk> especially not helped by gnome hard-coding it at 96dpi for everything because "most monitors lie"
[01:09] <lifeless> mdz: the /defined/ dpi for gif etc images is 72 dpi, and browsers are expected to scale the images.
[01:09] <jdub> are malone+reportbug doing useful things for us in dapper?
[01:09] <lifeless> Keybuk: 100% ack on that
[01:09] <mdz> lifeless: they don't, and if they did, it would look like crap anyway
[01:09] <dholbach> jdub: no.
[01:10] <Keybuk> jdub: rewriting reportbug from scratch would be a large task
[01:10] <Keybuk> despite some people's claims to the contrary, it isn't generic and "pluggable"
[01:10] <Kamion> jdub: Malone's mail interface needs a GPG signature; if that weren't the case it would be fairly trivial
[01:10] <jdub> Keybuk: it can't mail to some useful malone entry point?
[01:10] <jdub> Kamion: aha
[01:10] <lifeless> sorry, my bad - 96 dpi is the reference.
[01:10] <jdub> is that fixable?
[01:10] <lifeless> http://www.w3.org/TR/CSS21/syndata.html#x39
[01:10] <Kinnison> Kamion: I thought malone no longer required gpg sigs
[01:10] <Kamion> it's policy ...
[01:10] <Keybuk> jdub: reportbug does a huge amount of wget and grep on the web interface to offer lists of packages, currently open bugs, etc.
[01:10] <Kinnison> Kamion: but I might be wrong
[01:11] <Kamion> Kinnison: on new bugs, AFAIK, it does?
[01:11] <Kamion> maybe that's changed
[01:11] <Keybuk> Kamion: if you think it's trivial, *you* can rewrite reportbug
[01:11] <mdz> Kamion: removing that requirement would be trivial as well, if not for the fact that we don't accept anonymous or unauthenticated bug reports
[01:12] <Kamion> Keybuk: I've hacked reportbug in small ways before, it's not hard to make it talk to a slightly different mail interface. you're right that the web interface munging is hard to change
[01:12] <Kamion> but disabling that would be better than the current situation
[01:12] <Keybuk> Kamion: whoever's been maintaining reportbug recently has been making it very hard to bypass the web interface stuff
[01:12] <mdz> lifeless: everyone I know who publishes web content for a broad audience would disagree; I think it's safe to say that your hardware purchases (and most of the rest of ours) are atypical
[01:12] <jdub> Kamion: (i didn't expect to hear complaints about it as often as i have)
[01:13] <Kamion> AFAIK the person who's been maintaining it recently is the same person who always has; have you talked to him about it?
[01:13] <Keybuk> personally I think we should just drop reportbug
[01:14] <lifeless> mdz: I can agree that our purchases may be atypical. But I'm not sure that the atypicality we display implies that the 'typical' purchase is still 96 dpi.
[01:14] <lifeless> anyway, I've injected the info about the standards that I wanted to
[01:14] <Keybuk> the only useful purpose reportbug serves is for allowing Debian developers to file bugs into the Debian bts
[01:14] <lifeless> so it over to you guys 
[01:14] <Keybuk> likewise "bts"
[01:14] <Keybuk> I would be surprised if there's anyone here who hasn't got "bts debian" in their .reportbugrc
[01:16] <Mithrandir> .reportbugrc? :-P
[01:16] <mdz> Keybuk: I agree that shipping reportbug by default has turned out to be not such a good idea
[01:19] <Keybuk> reportbug isn't that useful for us
[01:19] <Keybuk> it's handy in Debian where the BTS has requirements about package names, versions, tags, etc.
[01:19] <Keybuk> Malone doesn't have any of that; you just give the name of the package and the bug description
[01:20] <mdz> it's handy in Debian where it can talk to the bug tracking system
[01:21] <Keybuk> dholbach: ping
[01:22] <dholbach> Keybuk: pong
[01:23] <Keybuk> dholbach: is it easy to turn Tango icons into Tangerine icons?
[01:24] <Keybuk> and how would I get icons added to both themes?
[01:24] <dholbach> Keybuk: you might want to ask andreasn or lapo - they really know their stuff.
[01:24] <Keybuk> where would I find them?  they're not on here
[01:24] <dholbach> Keybuk: for tango upstream mailing list or bugs.freedesktop.org
[01:25] <Keybuk> gnargh
[01:25] <Keybuk> that's not really the kind of answer I was hoping for
[01:25] <Keybuk> this is Ubuntu-specific really
[01:25] <Keybuk> nm ships GNOME-style icons
[01:25] <Keybuk> someone's done Tango-style icons for it
[01:25] <Keybuk> and obviously we'd want Tangerine-style
[01:25] <dholbach> Keybuk: for tangerine branch from http://daniel.holba.ch/bzr/tangerine/ and tell Andreas Nilsson to merge from it or file a bug in launchpad for tangerine-icon-theme
[01:25] <Keybuk> how do I get it to have all three?
[01:26] <mdke> how does one file a bug about words which aren't in spell checkers? Go straight upstream to openoffice, or to launchpad?
[01:26] <dholbach> tangerine-icon-theme bug report
[01:26] <Kinnison> mdke: people have been filing bugs in launchpad for aspell dictionaries
[01:26] <Kamion> BenC: linux-source-2.6.15 FTBFS on i386, if you didn't see; I'd like to avoid NEWing the others until i386 is fixed, if you don't mind
[01:26] <Keybuk> what about tango?  tango-icon-theme bug report?
[01:26] <Mithrandir> mdke: depends on the language, I'd guess.  Most aren't built by ooo upstream.
[01:26] <Kinnison> mdke: but I have yet to work out how to add them sanely
[01:26] <mdke> Kinnison, openoffice uses a different dictionary, doesn't it?
[01:26] <Kamion> oh, but it's a weird segfault on the buildd, hmm
[01:26] <Kinnison> mdke: not sure
[01:27] <mdke> bah, /me just adds the word
[01:27] <Kamion> hmm, maybe not, looks like a file that the doc processing system wants got moved
[01:27] <Mithrandir> mdke: ooo uses myspell
[01:28] <mdke> Mithrandir, ok. Is there a good reason there is more than one type of dict?
[01:29] <mdke> gah, openoffice itself seems to have three potential dicts to add words to: standard.dic, Ubuntu.dic and soffice.dic. Confusing...
[01:30] <Mithrandir> mdke: unsure, I just maintain the Norwegian dictionary and build them all from the same source.
[01:30] <Kamion> heno: please arrange for xcursor-themes to be added to some appropriate seed; we don't generally promote things until the seeds tell us to
[01:32] <heno> Kamion: so that would be the desktop seed then, since it should appear when choosing F5-option 1
[01:33] <Keybuk> hmm
[01:33] <Keybuk> I wonder if I ship the icons as /usr/share/icons/Tango (as well as /icons/hicolor) does that work?
[01:35] <Kamion> heno: that sounds reasonable
[01:35] <Lathiat> a
[01:35] <mdke> pitti, my computer appears not to have remember that I activated cups detection yesterday. I had to activate it again today. Known?
[01:35] <Kamion> heno: although check with the desktop team
[01:35] <Keybuk> hmm
[01:35] <Keybuk> how does hicolor work?
[01:35] <heno> Kamion: ok, I've pinged dholbach about it
[01:35] <Keybuk> I have my icon theme set to "GNOME", so how do the hicolor icons get picked?
[01:37] <Diziet> Well, the use of `caption' as a `font-family' is wrong according to CSS 2.1.
[01:38] <pitti> mdke: no, actually not
[01:39] <mdz> BenC: i386 kernel build failed after 4 hours, https://launchpad.net/+builds/+build/186668
[01:39] <pitti> mdke: before and after you enable it, what's the contents of /etc/cups/cups.d/browsing?
[01:39] <pitti> mdke: /etc/cups/cups.d/browse.conf even
[01:39] <mdke> pitti, I'll do a bug with that info. Anything else you want?
[01:40] <pitti> mdke: your main /etc/cups/cupsd.conf could be interesting, if you modified it
[01:40] <mdz> BenC: looks like the kernel itself built fine but docproc segfaulted
[01:40] <mdke> pitti, no, haven't touched anything
[01:40] <pitti> mdke: I still have to test browsing and such with the latest version, but thanks for the bug
[01:43] <jdong> are there any updates on us and the new fglrx?
[01:45] <holy_cow> mornin guys
[01:45] <holy_cow> i need to add some icons to the top gnome panel via command line ... where is that config stored?
[01:46] <Treenaks> holy_cow: in gconf
[01:46] <Treenaks> so you should/could use gconf-editor and gconftool-2
[01:46] <Treenaks> (theoretically)
[01:46] <holy_cow> Treenaks, really?ah!
[01:47] <holy_cow> well i'm doing this via ssh ... would be neat if i could have a gconf-editor manager remote boxen :)
[01:47] <holy_cow> k, lemme lookup where this info might be
[01:47] <holy_cow> thank Treenaks 
[01:47] <mdke> pitti, bug #40539
[01:47] <Ubugtu> Malone bug 40539 in cupsys "Failure to remember that Browsing was activated" [Normal,Unconfirmed]  http://launchpad.net/bugs/40539
[01:47] <pitti> mdke: thanks
[01:47] <Treenaks> holy_cow: this is more of a #ubuntu question btw, and you might want to look at sabayon too
[01:48] <holy_cow> really? i didn't think anyone there would have a clue what gconf is :)
[01:48] <Kamion> Mithrandir: could you please merge http://people.ubuntu.com/~cjwatson/bzr/casper/ubiquity/ ? We need to coordinate uploads of that and ubiquity (nee espresso), so if you want me to do both of them at once, I can
[01:48] <Treenaks> holy_cow: really, they do :)
[01:48] <holy_cow> k. sorry didn't meant to raise the noise level here :)
[01:48] <pitti> mdke: ok, browse.conf looks fine; does /etc/cups/cupsd.conf include it?
[01:49] <pitti> mdke: "Include /etc/cups/cups.d/browse.conf" at the very end
[01:50] <Mithrandir> Kamion: Do you want it done now or soon?
[01:50] <Kamion> Mithrandir: now's good
[01:50] <Kamion> Mithrandir: (if convenient)
[01:50] <Mithrandir> Kamion: I have a few bugs I want to squash in casper, but if you want it now, just upload yourself.
[01:50] <Kamion> Mithrandir: ok, I'll do it so that we can get the rename out of the way and carry on with real work
[01:51] <Mithrandir> Kamion: sure.
[01:51] <mdke> pitti, yes. All of this data is from the situation that it is working, I can't tell you what those files said when I saw the bug
[01:52] <pitti> mdke: oh, if it's working, then they are useless, right
[01:52] <pitti> mdke: do they change after a reboot?
[01:53] <mdke> pitti, no, I can't reproduce it just by rebooting. I think it must have been yesterday's upgrade that did it
[01:54] <pitti> mdke: hm, yesterdays update did nothing but change the group of /var/run/cups, but that seems quite irrelevant
[01:56] <mdke> oh
[01:56] <mdke> pitti, I'm fairly sure I didn't touch anything else
[01:57] <Kamion> Mithrandir: done. The changelog diff for release is in that branch now as well.
[01:58] <Mithrandir> Kamion: thanks; merged.
[02:19] <dholbach> Keybuk: if you want to talk to lapo, he's in #ubuntu-desktop
[02:22] <Keybuk> dholbach: thanks
[02:38] <Kamion> mvo: I'm testing a backport of that apt-ftparchive .dsc handling fix now. If it works, would you mind if I uploaded it to breezy-updates, so that we can easily install it on little and maybe drescher and stuff?
[02:39] <Kinnison> Kamion: changing apt-ftparchive will result in needing to regenerate the caches IIRC
[02:40] <Kinnison> Kamion: just as a warning
[02:40] <Kamion> ok - definitely want it on little though, even if not drescher
[02:41] <mvo> Kamion: just the patch that fixes the issue? sure, uploading it to breezy-updates should be ok
[02:42] <mvo> Kinnison: I don't think that this is required (if we only backport the fix)
[02:42] <Kamion> yeah, just that four-line patch or whatever it is
[02:42] <Kinnison> mvo: drescher currently runs hoary's apt-ftparchive I think
[02:42] <mvo> only the current (not-yet-released) apt-ftparchive needs cache regeneration
[02:42] <mvo> oh
[02:42] <mvo> ok :)
[02:42] <Kinnison> mvo: because of the bug :-)
[02:45] <Mithrandir> TheMuso: any chance you could look at the a11y-related casper bugs?  (Possibly together with heno?)
[02:46] <TheMuso> Mithrandir: I'm on it. Just waiting for a package to be promoted to main and seeded to add the final piece to one of the profiles.
[02:46] <Mithrandir> TheMuso: cool, thanks.
[02:49] <zul> heylo
[03:02] <Max_Littlemore> don't mean to interrupt, but is anyone looking at realtime audio? - (it is kind of important now that mac is intel)
[03:03] <mjg59> Max_Littlemore: Yes, but I don't understand your justification...
[03:04] <ogra> what makes it more important on mac intels than on other machines ? 
[03:07] <dholbach> mdke: bug 40492 for you and the gang.
[03:07] <Ubugtu> Malone bug 40492 in ubuntu-docs "Error in 4.1.10. Open files with administrative privileges from the file manager" [Normal,Confirmed]  http://launchpad.net/bugs/40492
[03:08] <jsgotangco> i see it
[03:08] <dholbach> jsgotangco, mdke: if you guys want me to make an update, that one should be easy to fix with it
[03:09] <dholbach> and maybe you should make another call for test reading
[03:09] <jsgotangco> its from the desktop guide nice spotting
[03:12] <mdke> dholbach, not sure about whether to fix it.
[03:13] <dholbach> hm? it mentions two different file names
[03:13] <mdke> yes, it's a bug, but will require a string freeze exception
[03:14] <dholbach> who approves these exceptions?
[03:14] <jsgotangco> its an obvious string bug, you're only going to change 1 string it won't even break translations
[03:14] <dholbach> the release team?
[03:14] <mdke> dholbach, we kinda decide ourselves
[03:15] <dholbach> ah i see
[03:15] <mdke> jsgotangco, it will break the translation, but probably it's worth it at this stage
[03:15] <jsgotangco> i dunno, your call there's a reason why we're in public beta
[03:16] <Max_Littlemore> sorry, left the room. mjg59,ogra. Mac has a good history of music types who understand nix bout pootahs making sound. - for me debian is good, but i wish i didn't have to futz around for so long
[03:16] <Max_Littlemore> we need a new jack
[03:17] <mjg59> Max_Littlemore: I still don't understand the justification, though I'll agree that it's a worthwhile goal (and is being worked on)
[03:19] <pitti> oh DEAR
[03:19] <Mithrandir> hm?
[03:19] <pitti> mh21_, Keybuk: it seems I finally found out why cups' web interface doesn't ever ask for authentication
[03:19] <mdz> dholbach: I purged dasher yesterday, but it's still in the menu.   icon cache problem?
[03:19] <pitti> due to the clever code, the cups server presents its own certifiate to itself and thus accepts everybody
[03:19] <pitti> GO CUPS
[03:20] <Mithrandir> pitti: hahahaha. :-)
[03:20] <pitti> sorry for the noise, but this just drove me up the wall...
[03:20] <fabbione> pitti: ahahaha
[03:20] <mh21_> pitti: Why not for root access?
[03:20] <dholbach> mdz: dasher-data problem
[03:21] <pitti> mh21_: I assume it tries to setuid() away from root to an user/group which can't read the local cert any more
[03:21] <pitti> mh21_: and that fails if cupsd already runs as non-root
[03:21] <mdz> dholbach: aha, thanks
[03:21] <pitti> mdz: thanks for the approval, btw, I'll upload cups as soon as I fixed this damn authentication bug
[03:21] <dholbach> mdz: we got a bug report about that - we really should get the automatic apt removal stuff in for edgy :-)
[03:21] <Keybuk> pitti: lol
[03:22] <Max_Littlemore> mjg59 if it is being worked on, let me help. (I have a few ideas to make it easier - like a new jack) The justification is that other systems do it 'out of the box'. While this is not Ubuntu's mission, the handles should be created to make it easy. Otherwise, people like me prefer OSX to do audio, and then have to log out to use gaim. WHERE DO i GO TO HELP?
[03:22] <mdke> dholbach, fixed, but no need to upload unless you want to.
[03:23] <dholbach> mdke: you have some other good stuff pending?
[03:23] <dholbach> oh, it's been a week - let's do it :)
[03:23] <mdke> dholbach, no
[03:24] <mdke> dholbach, I'm going to upload some translations tomorrow, so you can wait until then if you like
[03:24] <dholbach> mdke: ok, just drop me a mail, once you did it
[03:24] <mdke> will do, thanks
[03:24] <dholbach> thanks
[03:25] <mdz> dholbach: any reason not to move the .desktop file into the dasher package? it should accompany the executable program
[03:26] <Amaranth> you could always just hide it with alacarte :)
[03:27] <dholbach> mdz: hum... yes - it makes sense
[03:28] <Max_Littlemore> < just wait till you have children - if you have children > hello? -- do we like music>
[03:28] <tseng> Max_Littlemore: you don't have to exagerate and talk in capital letters to make your point
[03:29] <Max_Littlemore> I didn't mean to. sorry
[03:29] <tseng> Max_Littlemore: crimsun is the developer most interested in sound related bits, maybe see if he is interested without pressuring him
[03:30] <Max_Littlemore> I would like to find crimson... but how?
[03:31] <jdub> seb128, dholbach: shouldn't python2.4-avahi be in desktop seed?
[03:31] <Max_Littlemore> newbie to all this anarchy:-[
[03:31] <tseng> he's right here, probably working on something else right this minute
[03:31] <dholbach> jdub: what is using it?
[03:31] <seb128> jdub: anything needs it?
[03:31] <dholbach> jdub: or for the PythonUberAlles goal? :-p
[03:31] <jdub> avahi-discover, service-discovery-applet, general python love
[03:31] <jdub> it should at least be in main
[03:32] <seb128> avahi-discover is not to desktop
[03:32] <Max_Littlemore> thank you tseng
[03:32] <jdub> but i think it's important to have as one of our python developer platform libs
[03:32] <seb128> bah
[03:32] <jdub> seb128: python2.4-avahi is stuck in universe atm - doesn't seem right 8)
[03:32] <seb128> I would rather like have better l10n that ton of useless python if you ask me
[03:32] <jdub> seb128: well, there is a bunch of junk in there, but avahi is more useful than a lot of the junk python stuff we ship
[03:33] <seb128> jdub: no, main would be nice. But space on CD is limited ...
[03:33] <Kamion> bloody hell, the stupid xserver-xorg-input-synaptics version number managed to break source CDs
[03:33] <seb128> jdub: right, I agree with that
[03:33] <jdub> bzr is still in universe too
[03:33] <seb128> jdub: if you find some stuff to clean and put it instead go for it I would say :)
[03:34] <jdub> seb128: sounds like a challenge
[03:35] <seb128> yeah
[03:35] <jdub> hrm, can thoggen use gstreamer 0.10 yet?
[03:35] <seb128> slomo was discussion with tim__ about that yesterday I think
[03:35] <seb128> discussing
[03:35] <jdub> cool
[03:36] <seb128> avr 20 19:02:42 <sxpert-work>   __tim: how is thoggen 0.10 coming ?
[03:36] <seb128> avr 20 19:02:44 *       sxpert-work runs :D
[03:36] <seb128> avr 20 19:02:57 <__tim> sxpert-work: got it working locally, minus cropping dialog and preview
[03:36] <zul> Kamion: : ping
[03:36] <Kamion> zul: hi
[03:36] <seb128> jdub: maybe next week ;)
[03:36] <zul> why is their reiserfs4 support in the installer when there isnt in the kernel?
[03:37] <Kamion> zul: bug 32075
[03:37] <Ubugtu> Malone bug 32075 in gparted "installing to reiser4 partition doesn't seem to work" [Normal,Unconfirmed]  http://launchpad.net/bugs/32075
[03:37] <zul> ah ok...
[03:42] <jdong> zul: is it a subtle hint that someone wants a certain kernel patch for Edgy? ;)
[03:42] <jdub> seb128: also, gimp-svg should be added to main/desktop
[03:42] <Kamion> I certainly hope not
[03:43] <Kamion> (to jdong)
[03:43] <jdong> lol
[03:43] <zul> jdong: no way in hell
[03:43] <jdong> just a joke, guys
[03:43] <jdong> take it easy
[03:43] <jdong> :)
[03:44] <dholbach> Kamion: did you move xcursor-themes to main yet?
[03:44] <seb128> jdub: right
[03:44] <Kamion> dholbach: needs to be seeded
[03:44] <dholbach> Kamion: i did that :)
[03:45] <Kamion> dholbach: done, then
[03:46] <dholbach> Kamion: merci beaucoup!
[03:46] <Kamion> de rien
[03:46] <dholbach> see you in a bit
[03:46] <ogra> bah, we're turning ito a french channel here 
[03:47] <Kamion> dholbach: you don't need to say "merged" in seed commits when all you did was pull
[03:47] <Kamion> it looks a bit weird to see a log message that says "merged" but there's no actual merge ...
[03:48] <dholbach> Kamion: you're right - I'll respect that the next time.
[03:48] <doko> infinity: do the buildd's have /etc/services and /etc/protocols installed?
[03:49] <pitti> Riddell: sorry
[03:50] <Mithrandir> doko: you can't count on it.  Netbase's is priority: important.
[03:50] <Riddell> pitti: just so long as they havn't fallen off the top of your todo list :)
[03:51] <doko> Mithrandir: thanks, currently walking through testsuite results
[03:52] <Mithrandir> would anybody be very sad if casper never picked !utf-8 locales by default?
[03:53] <pitti> Mithrandir: oh, does it ATM?
[03:53] <pitti> Mithrandir: since we optimized our system to be UTF-8 and assume that in several places...
[03:53] <bddebian> Morning peoples
[03:53] <Mithrandir> pitti: it just picks the first one it finds.  Which may or may not be UTF-8.
[03:53] <Mithrandir> (first one from /usr/share/i18n/SUPPORTED)
[03:54] <pitti> Mithrandir: would it be possible to take the first UTF-8 one, and only if that doesn't exist, take the first one?
[03:54] <Mithrandir> pitti: anything's possible. :-)
[03:55] <Mithrandir> pitti: but there are no locales where we don't have an UTF-8 equivalent.
[03:55] <pitti> now, that's good! :)
[03:57] <Kamion> I think I'd agree with changing it to pick only UTF-8
[03:58] <Kamion> Riddell: could you merge the Kubuntu seeds? I tried but it seems kinda complex due to the merge of -server into dapper
[03:58] <Kamion> Riddell: you're OK to merge my ubiquity branch now, btw
[03:59] <Riddell> Kamion: two merges coming up
[03:59] <Mithrandir> Kamion: will gfxboot ever pass debian-installer/locale=.*UTF-8 on the kernel command line?
[04:00] <ogra> Kamion, thanks
[04:01] <Kamion> Mithrandir: I suppose it could if you want it to
[04:01] <Mithrandir> Kamion: nah, I'll just make it cope either way.
[04:01] <Kamion> although that would make life awkward for d-i, so probably not
[04:01] <Kamion> because the intent is only to preseed the language, not the country
[04:01] <Kamion> (unless the country matters to the language, i.e. pt and zh)
[04:01] <Mithrandir> yup
[04:05] <lifeless> gnight
[04:07] <Diziet> Is there some random Ubuntu-specific package I can make {ubuntu,kubuntu,xubuntu,edubuntu}-docs and m-f-l-all build-depend on without trouble ?
[04:07] <pitti> Keybuk: hm, cupsd itself runs as cupsys:lp, not as group lpadmin; so what I would prefer instead is to reflect that in the adduser call and make lp the primary group of cupsys (and lpadmin an auxiliary one); is that fine for you?
[04:07] <Diziet> I want a common source for a locale list.
[04:07] <pitti> Keybuk: (instead of changing the group of /var/run/cups)
[04:08] <Keybuk> pitti: I didn't _change_ the group of /var/run/cups
[04:08] <Keybuk> it had that group in the package
[04:08] <Mithrandir> Diziet: why should it be ubuntu-specific?
[04:09] <Keybuk>         chown -R cupsys /var/run/cups
[04:09] <Keybuk>         if [ -f /etc/printcap ] ; then
[04:09] <Keybuk>             chgrp lpadmin /etc/printcap
[04:09] <Keybuk>             chmod 664 /etc/printcap
[04:09] <Keybuk>         fi
[04:09] <Keybuk> ^ from cupsys postinst
[04:09] <Keybuk> uh, just the first line you wanted, obviously :p
[04:09] <Keybuk> I just fixed the init script to match the package
[04:10] <pitti> Keybuk: ah, I see; right
[04:11] <Diziet> mithrandir: Because this mad scheme for firefox start page translation is Ubuntu specific ?
[04:12] <Mithrandir> Diziet: it's therefore a problem if the list of locales exists in other distributions too?
[04:12] <Diziet> Err, I meant a package which *ubuntu owns and which is only in soyuz, as opposed to (say) debhelper :-).
[04:13] <Diziet> (Which is also in Debian and other Debian derivatives.)
[04:13] <Diziet> So I mean *ubuntu-specific, not specific to ubuntu as opposed to *?ubuntu.
[04:13] <Mithrandir> locales has a list of all supported locales in /usr/share/i18n/SUPPORTED
[04:13] <Diziet> No, I don't want that because this list has to change less often
[04:14] <Mithrandir> uh, less often than the list of locales we support?
[04:15] <Diziet> Possibly, yes.
[04:15] <Diziet> Changing this list is a faff and involves rebuilding at least 5 packages.
[04:15] <Diziet> And the resulting binaries aren't quite runtime compatible in a way that isn't expressible as a dependency.
[04:16] <Diziet> (No great disasters happen if there's a mismatch, but you can have firefox come up and not be able to find its start page.)
[04:20] <infinity> doko: Only if you build-dep on netbase, which provides those files.
[04:21] <doko> infinity: yep, already added as a build-dep
[04:22] <Lathiat> so the LTS cd is sexy
[04:23] <Lathiat> mjg59: what was your horrible alps workaround?
[04:24] <mjg59> Lathiat: If it detects that it's an alps, it sets different default values
[04:24] <mjg59> I think the Debian guys were going to push it upstream
[04:25] <Lathiat> ah ok
[04:26] <bddebian> w000t, go pitti
[04:27] <giftnudel> I also want to mention here, that the live cd out of the box installation doesn't hibernate, see bug 39237 (since you're all praising the cd)
[04:27] <Ubugtu> Malone bug 39237 in linux-source-2.6.15 "Resuming from hibernation fails" [Normal,Confirmed]  http://launchpad.net/bugs/39237
[04:27] <Diziet> I see no-one has an answer for my `I need some random package to depend on' question.
[04:28] <Diziet> I suppose I could invent one.  Or I could make {k,x,ed}ubuntu-docs and m-f-l-all build-depend on ubuntu-docs.
[04:28] <infinity> Diziet: Hrm?  What are we mangling here?
[04:29] <Diziet> It's to do with DapperFirefoxStartPageTranslation, which is a bit of a hack really.
[04:29] <Diziet> There's a master list of relevantly-translateable locales which has to be the same in *ubuntu-docs and m-f-l-all.
[04:29] <infinity> Right.  Just read backscroll.
[04:30] <Diziet> All that these packages need is the list, but there's nowhere convenient for them to get it.
[04:30] <infinity> I'd just toss the master list in ubuntu-docs, since any change made there needs to be mirrored in the other *-{docs,artwork} packages anyway.
[04:30] <infinity> It's not ideal, but it works fine for now.
[04:30] <Diziet> (The build arrangements for *?ubuntu-docs would probably appreciate a little script to help out with generating the links.)
[04:30] <Diziet> OK, so it's OK for those to build-depend on ubuntu-docs ?
[04:31] <Diziet> I'm always a bit wary of adding new build dependencies in these kind of situations.
[04:31] <infinity> Or, you could make the m-f-l-all source package spit out a binary package that provides the current list.
[04:31] <Diziet> infinity: Right, I did consider that.  But its build system is a complete nightmare.
[04:31] <Diziet> Weirdo templateified debian/control.
[04:31] <Diziet> (For very good reasons, but.)
[04:31] <infinity> It's ugly, but simple enough to add one little binary package.
[04:32] <Diziet> If I'm going to invent a new package I'll have ubuntu-docs spit it out.
[04:32] <infinity> And generating it is as simple os listing the XPIs.
[04:32] <Diziet> Oh, no, it's much more complicated than that.
[04:32] <Diziet> Because XPIs might turn up later.
[04:32] <infinity> Well, what SHOULD define the master list, really?  The list of XPIs we ship (from m-f-l-all) would seem to be the definitive list.
[04:32] <bddebian> Heya LaserJock
[04:32] <infinity> So, if pitti adds a new locale, rebuilding the -docs packages just picks it up magically.
[04:33] <Diziet> The list is everything we ship now _plus_ anything we think might plausibly turn up in the future.
[04:33] <Diziet> Because changing the list is much harder than adding an xpi.
[04:33] <infinity> Err, why that?  It doesn't hurt to rebuild the packages to match the current XPI list.
[04:33] <infinity> And if the other packages are automating their mess based on the XPI list, then yay.
[04:34] <Diziet> To change the list you have to rebuild at least five packages and the dependencies don't stop the user from producing a system which can't find its firefox start page.
[04:34] <Diziet> So really we have to settle on the list before release and not change it later, or at least try hard not to.
[04:35] <Diziet> This is why the master list has to be somewhere separate and really under human control.
[04:35] <infinity> Mmkay.  Then ubuntu-docs seems a reasonable fit for now.
[04:35] <Diziet> Right.  Fair enough.
[04:35] <Diziet> If the extra build-dep isn't a problem then that's a good answer.
[04:35] <infinity> Hopefully, the whole hack can be rethought later.
[04:36] <infinity> I don't see that anyone would complain about the build-dep.
[04:37] <infinity> It even, in an odd way, makes the fact that *-docs are all related and all need to get along be a bit more obvious.
[04:37] <infinity> Which may prevent people from repeating the messes we saw in breezy. :)
[04:38] <mjg59> So I booted my machine in SF time
[04:38] <mjg59> Now I'm back in the UK and I've changed my timezone
[04:38] <mjg59> I haven't rebooted
[04:38] <mjg59> cron.daily now runs during the middle of the afternoon
[04:39] <mjg59> DEAR LEGACY UNIX, PLS DIE, KTHXBI
[04:39] <infinity> Makes perfect sense to me, since cron's been running from init.
[04:40] <infinity>  /etc/init.d/cron restart && echo Yay
[04:40] <infinity> Of course, syslog and the rest of the world is off too.
[04:40] <mjg59> infinity: Oh, I know /why/ it happens, and it makes sense in the same way that dying in a car crash when you weren't wearing a seatbelt makes sense
[04:40] <infinity> Dear world, stop computing in timezones other than UTC.
[04:40] <sladen> have the locales crack to restart cron
[04:40] <mjg59> It's predictable, it's just not /right/
[04:41] <LaserJock> Diziet: got a second? I was wondering how the Ubuntu Developer's Reference was going
[04:41] <infinity> Well, it's a fair wishlist bug to ask cron to reinit its envionment when it does its cleanup each minute, I suppose.
[04:41] <infinity> Much as it's a decent wishlist to ask logging daemons to just log in UTC, and ignore local time altogether.
[04:42] <mjg59> infinity: Utterly agreed on the latter
[04:42] <infinity> Well, it nicely solves the headache of DST, too.
[04:43] <infinity> DST + logfiles = confused sysadmin, twice a year.
[04:43] <mjg59> Yeah, logviewers should eal with it
[04:43] <infinity> I don't think I ever really got over that confusion (except by setting all my machines to UTC)
[04:43] <Keybuk> actually, it'd be nice if the log file viewer automatically converted the log file times to whatever localtime you're currently viewing them in
[04:43] <infinity> OH LOOK, 2:30 HAPPENED TWICE LAST NIGHT, YAY!
[04:44] <Keybuk> sadly when the boffins came up with UNIX, they didn't expect anyone to be moving machines across timezones
[04:44] <infinity> The scarier one, of course, is when you go ther other direction and you immediately assume that "all my logfiles are missing an hour" means "oh shit, we were broken into, and someone did a crap job cleaning up"
[04:44] <Keybuk> after all, they were the size of large buildings
[04:45] <infinity> Then you remember the whole clock changing business and calm down.
[04:48] <LaserJock> heh, you're supposed to actually look at logs? ;-)
[04:48] <infinity> This is another one of the hilarious ironies about Windows NT, actually.
[04:49] <infinity> WinNT does everything internally in UTC, converts all values to local time for the user in displays... But STILL ISISTS ON HAVING THE HARDWARE CLOCK IN LOCAL TIME.
[04:49] <infinity> So, what's the first tihng NT has to do on boot?  Convert the local clock to UTC, so its internal clock is correct.
[04:49] <infinity> Say what?
[04:49] <LaserJock> wha? that seem a bit dumb
[04:50] <mjg59> infinity: Not strictly true. There's a registry key that gets it to assume the hardware clock is UTC
[04:50] <jsgotangco> yeah
[04:50] <infinity> It's just legacy, since PCs traditionally have local time in the BIOS, I assume.  Ease of dual-booting with DOS-based systems when they were first developing NT.
[04:50] <infinity> But still messed up.
[04:50] <mjg59> Except that breaks over suspend/resume or if you use the Visual Studio debugger
[04:50] <LarstiQ> mjg59: ooh, that would be useful
[04:50] <infinity> mjg59: Ahh, I've not found that key.  But it's clearly not set by default.
[04:51] <infinity> Breaks suspend/resume?  Cute bug.
[04:51] <mjg59> http://www.cl.cam.ac.uk/~mgk25/mswish/ut-rtc.html
[04:51] <infinity> Nice to know we're not the only ones that can't always get that right.
[04:52] <mjg59> infinity: NT used to run on machines that had the hardware clocks in UTC, so
[04:52] <mjg59> But they never suspended
[04:52] <infinity> Yeah, fair point.
[04:52] <mjg59> HKLM\SYSTEM\CurrentControlSet\Control\TimeZoneInformation\RealTimeIsUniversal - not sure if it's in XP, was there in 2000
[04:53] <infinity> If it's there in 2K, it's in XP.
[04:53] <mjg59> Oh, it's still in XP
[04:53] <infinity> Can't vouch for Vista, though.
[04:53] <infinity> But 2K->XP was a pretty minor version bump, they barely touched internals like this.
[04:54] <mjg59> "There is RealTimeIsUniversal registry entry and it almost work but during
[04:54] <mjg59> the hours around the daylight change system will use 100% of processor time."
[04:56] <infinity> Interesting read.
[04:56] <infinity> I may just flip that bit ony Zofia's machine and see what happens.
[04:56] <infinity> I've never had to attach a kernel debugger to her machine, so that concern is out.
[04:57] <infinity> And it never suspends.
[04:58] <infinity> I can see why both those operations would get messed up by it, though, since both are very early boot-time things that probably happen before walking the registry.
[05:05] <carlos> Riddell: hi
[05:05] <carlos> around?
[05:06] <Riddell> carlos: hi
[05:06] <carlos> Riddell: seems like kdegraphics lacks the .pot files
[05:07] <carlos> and I have a bunch of .po files pending to be imported because that
[05:07] <carlos> are you going to upload an update soon that would create the .pot files?
[05:09] <Riddell> carlos: buildlogs do indeed lack any use of gettext, let me see why
[05:09] <Riddell> carlos: in the mean time you can import koffice 1.5 if you want http://archive.ubuntu.com/ubuntu/pool/universe/k/koffice-l10n/koffice-l10n_1.5.0.orig.tar.gz
[05:10] <carlos> samething with kdeaddons, kdesdk and I think others... I'm checking all them now...
[05:10] <carlos> Riddell: are you going to promote it to main?
[05:11] <Riddell> carlos: I thought you said importing it directly to rosetta would be the best idea
[05:11] <Riddell> hmm, possibly those are the packages that don't use cdbs
[05:12] <carlos> Riddell: oh, ok, sorry, that's one of those 'empty' packages
[05:15] <carlos> Riddell: hmmm, I really hate the way KDE stores .po files... that package needs more code on my side to support that specific case...
[05:16] <carlos> I'm going to do it locally for this firts import, but I will need to develop something for new updates :-(
[05:16] <carlos> to do it automatically
[05:17] <carlos> Riddell: hmm, this is interesting... kde-i18n-* packages have translations for binaries that we don't have in main, is that what we want? (I'm talking, for instance, about kdevelop)
[05:18] <carlos> and kdeaccessibility
[05:18] <carlos> aren't those core parts of KDE?
[05:18] <carlos> well, not core but important...
[05:21] <Riddell> carlos: kde-i18n includes all the .po files for all the KDE modules, but not all the KDE modules are in main
[05:22] <Riddell> kdevelop3 gets broken randomly mostly because it's renamed from kdevelop
[05:22] <Riddell> kdeaccessibility I'd like for dapper+1 so we can do a CD boot option like ubuntu has
[05:22] <carlos> Riddell: I guess is ok If I ignore those .po files for dapper, right? I don't think you are going to move them to main before release :-P
[05:23] <Riddell> carlos: yeah, that's fine
[05:30] <Riddell> carlos: I've fixed kdegraphics, kdeaddons and kdesdk, doing a test compile them uploading
[05:30] <Riddell> carlos: any others having that problem?
[05:30] <carlos> Riddell: cool, thanks
[05:30] <carlos> Riddell: no that I'm aware of
[05:31] <carlos> Riddell: the others are packages that are on universe
[05:31] <carlos> If I detect anyone I will ping you again
[05:31] <carlos> Riddell: thanks
[05:31] <Riddell> carlos: are you going to import k3b-i18n too?
[05:32] <carlos> btw, would be possible to fix the .pot files that use UTF-8 but that the header doesn't say it?
[05:32] <carlos> Riddell: yeah, now that I'm with KOffice, I will do it at the same time
[05:32] <Riddell> carlos: I suspect that will need some sed scripting or the like, I presume gettext-kde doesn't care about utf8 too much
[05:32] <Riddell> carlos: what's the syntax needed?
[05:33] <carlos> Riddell: as an easy fix, you could generate all .pot files as being UTF-8 encoded as KDE uses UTF-8 and the ones in ASCII are aslo UTF-8 valid
[05:33] <carlos> "Content-Type: text/plain; charset=UTF-8\n"
[05:33] <carlos> Riddell: instead of whatever gettext-kde uses
[05:33] <Riddell> is there a gettext switch for that?
[05:33] <carlos> Riddell: hmm, let me check
[05:34] <carlos> I think it has one
[05:35] <carlos> Riddell: --from-code=UTF-8
[05:35] <carlos> Riddell: but I'm not sure if KDE version has it, I think it appear one or two years ago
[05:35] <Riddell> doesn't seem like kde-xgettext knows that
[05:36] <carlos> ok
[05:36] <Riddell> so I can play around with adding it directly using a script
[05:36] <carlos> then you will need to use sed
[05:36] <carlos> Riddell: yeah
[05:36] <carlos> you should check that the .po file is using UTF-8 as a sanity check
[05:36] <ogra> who does the torrent seeds ? seems edubuntu isnt seeded 
[05:36] <carlos> konqueror is a good file to check it
[05:36] <carlos> I mean konqueror.pot
[05:37] <Riddell> yep
[05:44] <azeem> Kamion: ping
[05:56] <Lathiat> hey guys
[05:56] <Lathiat> where does espresso hide its logs?
[05:57] <jdub> checking for passwords? :)
[05:58] <Lathiat> haha
[05:58] <Lathiat> no
[05:58] <Lathiat> filing too many bugs :)
[05:58] <Lathiat> got one instance here where it keeps crashing out after partitioning
[05:58] <Lathiat> so trying to get some more data
[06:02] <Chipzz> jdub: happy belated birthday ;)
[06:02] <Lathiat> hrm.. it seems bittorrent has a tracker, and while theres no rc?d/S links
[06:02] <Lathiat> it has K links, so you see shutting down bittorrent tracker 
[06:02] <Lathiat> on default installs
[06:02] <Lathiat> thats a bit bad and ugly and theres a rather old bug about it
[06:02] <Lathiat> how do i harass someone about that? :)
[06:03] <dholbach> Lathiat: /var/log/installer/espresso
[06:03] <wasabi> Morning folks.
[06:04] <Lathiat> ah cheers dholbach 
[06:05] <nmsa> Kamion
[06:05] <nmsa> can I ask you something ? are u available !?
[06:06] <tseng> Kamion: debian policy says to ask base-passwd maintainer about using useradd, i need an unprivelaged user (beagleindex) for beagle system-wide indexer
[06:06] <tseng> Kamion: do you care about the policy in the scope of ubuntu?
[06:16] <Lathiat> so 
[06:16] <Lathiat> how do i raise awareness of a bug report?
[06:18] <LaserJock> spam #ubuntu-* ? ;-)
[06:18] <wasabi> Fix it.
[06:18] <Lathiat> hehe
[06:18] <infinity> Lathiat: The bttracker bug?  It's on my radar, it's just not top priority.
[06:19] <Lathiat> hrm good fix seems to read the START_BTTRACK vara and not disaplay a message on shutdown if its not set
[06:19] <Lathiat> i guess
[06:19] <Lathiat> or have the tracker in a different package
[06:19] <Diziet> Should I bump the version of ubuntu-docs to 6.06 ?
[06:20] <infinity> Lathiat: Splitting it to another package is pointless, the init script just needs to not be broken.
[06:20] <ogra> Diziet, i'd keep that step for final ;)
[06:20] <LaserJock> Diziet: I thought that was going to be done on release
[06:20] <LaserJock> Diziet: but I'm not positive
[06:21] <Diziet> ogra: Fair enough.
[06:21] <Diziet> I just noticed it was wrong :-).
[06:21] <Lathiat> infinity: does not broken = not showing a message on shutdown is START_BTTRACK is set to 0?
[06:21] <LaserJock> Diziet: well, it is still 6.04 ;-)
[06:21] <infinity> Diziet: No, it's right, the version is date-dependant, check the changelog.
[06:21] <lamont> Lathiat: that's getting close to asking him to pull it to the top of his todo list...
[06:21] <Lathiat> heh
[06:22] <infinity> Lathiat: "Not broken" means "working according to policy".
[06:22] <lamont> infinity: heh
[06:22] <Lathiat> guess i better read the policy then :)
[06:23] <lamont> debconf: Obsolete command TITLE Configuring mdadm called
[06:23] <lamont> neato
[06:29] <Kamion> Lathiat: /var/log/installer/espresso; use ESPRESSO_DEBUG=1 to make it more useful
[06:30] <Kamion> (or, as of tomorrow or so, UBIQUITY_DEBUG=1 and /var/log/installer/syslog)
[06:30] <Kamion> nmsa: sure; quicker to just ask rather than waiting for me to get back and then asking
[06:30] <Kamion> tseng: yes, very much do, the policy is sensible regardless and I don't want to fork base-passwd. However, can't you just call 'adduser --system beagleindex' in your postinst?
[06:30] <Kamion> tseng: you normally don't need a static id
[06:32] <Kamion> tseng: oh, policy tells people to check with me first even for dynamic ids; I see. I think you're the first person who's actually paid attention to that ;-)
[06:32] <Riddelll> carlos: can you give me the .po line for UTF8 again?
[06:32] <Kamion> I think that bit could be removed from policy; the base-passwd maintainer is not an effective registry of dynamic user and group names, only of static ids
[06:32] <carlos> sure
[06:33] <carlos> instead of "Content-Type: text/plain; charset=CHARSET\n"
[06:33] <carlos> you should have "Content-Type: text/plain; charset=UTF-8\n"
[06:33] <Riddelll> carlos: great, thanks
[06:33] <carlos> riddell and be sure that the file is using UTF-8 or ASCII as its encoding...
[06:33] <carlos> I guess that would be the default... but just in case
[06:34] <carlos> Riddelll: k3b and koffice are now in the Rosetta import queue
[06:34] <carlos> should be imported today
[06:34] <carlos> the queue is a bit busy atm with the post beta uploads... :-P
[06:41] <nmsa> Kamion: I have failed this Bug #40560, almost a duplicate on Bug #33504 with a diff S3 video card. I think is a Critical one, I wonder until a fix gets out is there a w/a ? 
[06:41] <Ubugtu> Malone bug 40560 in xserver-xorg-driver-s3 "GDM session does not start" [Normal,Unconfirmed]  http://launchpad.net/bugs/40560
[06:41] <Ubugtu> Malone bug 33504 in xserver-xorg-driver-s3 xserver-xorg "Server failure with S3 86c764/765 [Trio32/64/64V+] " [Normal,Confirmed]  http://launchpad.net/bugs/33504
[06:49] <Kamion> nmsa: I'm sorry, I have no relevant expertise; why did you pick me to ask?
[06:50] <nmsa> your name is there beside the bug in launchpad
[06:50] <nmsa> as xserver-xorg-driver-s3 Creator
[06:51] <ogra> heh
[06:51] <Kamion> nmsa: I only uploaded the most recent version; if you look at the changelog you'll see that the change I made is a trivial dependency change that doesn't actually require any knowledge of X
[06:51] <Kamion> I'll file a bug on Launchpad for that incredibly misleading terminology
[06:51] <nmsa> :)
[06:51] <Kamion> so sorry, can't help you
[06:51] <ogra> Kamion, if you need another example, ltsp shows fabbione as creator
[06:52] <ogra> :)
[06:54] <Diziet> ogra: The edubuntu-artwork package doesn't seem to use the alternatives system to manage index.html.
[06:54] <Diziet> It doesn't divert it either.
[06:54] <Diziet> Surely this is just completely broken ?
[06:55] <Diziet> Oh, I'm looking at an old version.
[06:55] <ogra> yes
[06:55] <ogra> look at dapper
[06:55] <ohoel> any ETA on fresh langpacks?
[06:55] <ogra> i'd be astonished if it would be broken, postinst is derived from infinitys code 
[06:55] <jdub> Kamion: are we labelling the old installer CDs as 'text install & upgrade CDs' or something?
[06:56] <Kamion> https://launchpad.net/products/soyuz/+bug/40602
[06:56] <Ubugtu> Malone bug 40602 in soyuz ""Creator" terminology is wrong and misleading" [Normal,Unconfirmed]  
[06:56] <Kamion> jdub: "text-mode install CD" is the current phrase
[06:56] <jdub> Kamion: might be worth mentioning that it's the one to download if you want to do an upgrade
[06:56] <jdub> Kamion: http://www.tectonic.co.za/view.php?src=rss&id=973
[06:56] <Kamion> jdub: for the three people who actually upgrade that way? :)
[06:56] <jdub> Kamion: useful notes there
[06:57] <jdub> Kamion: you're not bandwidth challenged - there are a fuckload more than three
[06:57] <ohoel> they'll be using shipit when the time comes
[06:57] <jdub> Kamion: it's a very common use case
[06:57] <Kamion> jdub: shipit won't cater for them ...
[06:57] <Kamion> ohoel: ^--
[06:57] <jdub> s/jdub/ohoel/ :)
[06:57] <Kamion> jdub: both
[06:58] <ohoel> Kamion: the bandwidth-challenged? ;)
[06:58] <ohoel> I suspect they won't be downloading a beta image
[06:58] <Kamion> ohoel: we will not be sending the text-mode install CD out in shipit
[06:58] <ohoel> oh
[06:58] <Kamion> it's a Canonical decision to save money
[06:58] <ohoel> probably wise
[06:59] <ogra> (you could order an edubuntu text-mode CD and remove the edubuntu parts post install though)
[06:59] <ohoel> I'll be installing both ubuntu and kubuntu on a load of machines later today... hope I won't get too ill when faced with a livecd install
[07:01] <jdub> Kamion: likelihood of having ship seed on the livecd?
[07:02] <Kamion> jdub: anyway - I'm not wild about the unwieldiness of "text install and upgrade CDs" as a short description, but I've added "can also be used as a package source for upgrading older installations" to the longer description on releases; does that help
[07:02] <Kamion> jdub: I've been meaning to talk to somebody about that; I'm not sure where to put it, and I'm not sure that all of ship is appropriate, although I realise that some of it is needed
[07:03] <Kamion> there are some awkward technical difficulties, although good reasons to look into solving them; it's not something I can just flick a switch and fix though
[07:03] <jdub> Kamion: yeah, tough one
[07:03] <Kamion> I think it may have to be a separate seed
[07:03] <jdub> miniship ;)
[07:04] <Lathiat> Kamion: mm ok, i'll have to re-do that bug then, whast with 'uniquity', new name for it?
[07:05] <desrt> what would be really great: some way for the user to precompile their volatile modules so that they don't have to relink at every startup and be stored in a tmpfs
[07:05] <desrt> and since the user is doing it themselves there is no GPL violation
[07:07] <Lathiat> anyone else noticed fn+escape on their laptop no longer suspending?
[07:08] <Burgwork> Kamion, why the second rename>
[07:08] <Burgwork> ;/
[07:08] <ogra> second ? 
[07:08] <Burgwork> ok, that was a typo, not a smiley
[07:09] <Burgwork> espresso --> ubiquity
[07:09] <ogra> yes
[07:09] <LaserJock> Burgwork: because everybody called it expresso? ;-)
[07:15] <ohoel> LaserJock: the baristas nightmare
[07:15] <LaserJock> heh
[07:18] <dholbach> have a nice WE!
[07:26] <peterpowell> hi
[07:45] <Kamion> Burgwork: Mark wanted a name that made the association with Ubuntu clearer
[07:47] <Kamion> with Espresso it wasn't obvious at all; with Ubiquity there's at least a suggestion. We didn't want "Ubuntu" blatantly in the name if possible, because that introduces branding difficulties
[07:48] <Kamion> I'm sorry it came late in the cycle, but we only dreamt up a suitable name shortly before the beta, and renaming was too intrusive at that point
[07:49] <Burgwork> Kamion, no worries, just wondered
[07:49] <Diziet> {x,k}ubuntu-docs have an alternative for the firefox home page which is at the same priority as the one from ubuntu-docs.
[07:49] <Diziet> edubuntu-artwork has a higher-priority one.
[07:49] <Diziet> Is this all correct ?
[07:49] <Kamion> Burgwork: (and Ubiquity is a better name, too - suggests putting Ubuntu everywhere)
[07:50] <mdz> Kamion: would it be non-trivial to have a ship seed for live?  or even to use the existing one?
[07:52] <LaserJock> mdz: sorry to bug you. I've got two packages from Debian (that aren't in Ubuntu) that I'd like to get into dapper Universe. Was a procedure for that hammered out or is it ok to just upload?
[07:52] <Kamion> mdz: I haven't looked into it in detail yet, and I have to go out just now; I have some thoughts but perhaps it would be better if I took a little while to think through the implications for germinate and cdimage before coming out with an opinion
[07:52] <ogra> Diziet, yes
[07:53] <Kamion> mdz: I think, regardless of technical constraints, it might be wise to have a separate seed so that we have more flexibility in terms of language packs and things
[07:53] <Kamion> we want to put language packs in the ship seed, but probably not in the ship-on-live seed
[07:53] <Kamion> anyway, out now ...
[07:53] <Diziet> ogra: Oh, good.  Well, the fixes for everything are uploading now.
[07:53] <ogra> yay+
[07:54] <mdz> LaserJock: you should request a sync, procedure on DeveloperResources
[07:54] <LaserJock> mdz: oh, sorry. I should have clarified. They need to be tweaked a bit in debian/control (GL and python deps)
[07:55] <LaserJock> mdz: so I have ubuntu1 packages ready
[07:56] <mdz> LaserJock: in that case, I guess just note in the changelog that it's a new package from Debian
[07:56] <LaserJock> mdz: k
[07:56] <mdz> so that the archive admins know where it came from
[07:57] <mdz> Diziet: why is this start page translation stuff landing post-featurefreeze? (and indeed, post-beta)
[08:03] <Diziet> mdz: The feature was there before feature freeze but it broke {k,ed,x}ubuntu.
[08:04] <Diziet> It was designed for Breezy in fact but not implemented there.
[08:07] <Diziet> So we had to extend the design to cover the derivatives, which also meant being more rigorous about the locale list (which it turns out seems to have been handled very loosely anyway).
[08:11] <infinity> mdz: How would a ship-on-live seed work?  live, by definition, is completely installed in the livefs, are we talking about adding a repository of non-installed debs on the livecd or something?
[08:12] <ogra> infinity, and link it to /var/cache/apt/archives ?
[08:15] <mdz> Diziet: if it needed redesign, it should have been the subject of a feature freeze exception discussion
[08:15] <mdz> infinity: yes, I am talking about a repository of non-installed debs
[08:16] <mdz> infinity: use case: third-party network driver needs linux-headers+build-essential
[08:17] <infinity> mdz: Ahh, cool.  If the seed side of that were sorted, I can easily mangle the livefs build to stuff them somewhere and even build a Packages.gz for them (if we want it as a static file:// repo, rather than having it in /var/cache/apt)
[08:18] <mdz> infinity: yeah, though tying it into the livefs directly probably isn't dapper material
[08:18] <mdz> infinity: it's a lot of stuff to install in the live environment anyway, but useful on the installed system
[08:19] <infinity> mdz: <shrug>... Tying it into the livefs isn't rocket science, really, and it just makes it that much simpler for espresso to not bother worrying about, since it can just copy it blindly with the rest of the filesystem.
[08:20] <infinity> Either way, something needs to be hacked, and if we want it on the livefs eventually, why bother hacking debian-cd to put it in some other odd place?
[08:29] <mdz> infinity: if espresso is copying sources.list, then we don't want to start changing it at this point
[08:30] <mdz> I think it might be generating a new one though
[08:30] <mdz> infinity: P.S. s/espresso/ubiquity/g ;-)
[08:31] <mdke> Diziet, around?
[08:33] <infinity> mdz: ;)
[08:34] <mdke> hey infinity 
[08:34] <mdke> infinity, is suspend to ram on your t43 working alright? it's playing up a lot here
[08:35] <infinity> mdke: Worked last time I tried, but I don't try often.  My laptop's generally up 24/7 for weeks on end.
[08:35] <mdke> ah
[08:35] <mdke> np then
[08:45] <Diziet> mdke: Not really but if it's quick ...
[08:46] <Kamion> infinity: I think ship-on-live probably wants to be handled by debian-cd anyway
[08:46] <Diziet> mdz: redesign> I'll make sure to have it discussed next time.
[08:46] <mdke> Diziet, np then, thanks for doing those uploads
[08:46] <Diziet> NP.  Tell mdz they're a good thing :-).
[08:46] <Kamion> infinity: I had a good reason for this in my head earlier but I seem to have forgo^W^W^W^W^Wthis IRC channel is too narrow to contain it
[08:46] <mdke> Diziet, will do
[08:47] <mdke> can I download older source packages other than the current ones? the man page says I can do it with an =versionnumber but it doesn't seem to be working.
[08:47] <Kamion> mdz: ubiquity doesn't really have to bother much actually - all I have to do would be to turn on the cdrom generator in apt-setup ...
[08:47] <Kamion> advantages of using d-i code++
[08:48] <tseng> Kamion: yeah i was planning on adduser --system in postinst, I'll consider it ack'd
[08:48] <tseng> Kamion: thanks.
[08:48] <infinity> Kamion: Alright, if we think the use-case for having the stuff available to the live system is marginal at best, and we just want it for the installed system, I'll keep my dirty hands off it.
[08:49] <infinity> Kamion: Just made sense to me to have it available in the live system too, "just cause".
[08:49] <Kamion> infinity: actually I think we can have our cake and eat it there
[08:49] <Kamion> infinity: if we know it's going to be on the CD, the live system can add a line to its sources.list referencing it
[08:50] <infinity> Ahh, I see what you mean.  Have a pool/dists structure on the ISO filesystem, then an apt-cd line in sources.list.
[08:50] <Kamion> infinity: one reason to do it this way is that it's more economical of disk space on the installed system; we don't have to copy stuff that not everyone needs to use
[08:50] <infinity> Clever.
[08:50] <Kamion> right
[08:50] <infinity> That does make more sense, yes, I wasn't thinking clearly. :)
[08:50] <Kamion> all the code to do that is already there - it's just a matter of switching some ifs around and figuring out the germinate end
[09:00] <Kamion> (after reboot)
[09:03] <Kamion> hmm, maybe it's not that the package isn't installed, but that the files are mysteriously broken
[09:06] <infinity> Kamion: I have some shiney new bug reports that claim that depmod hasn't been run.
[09:06] <infinity> Kamion: Which is utter BS... Unless you're not copying modules.dep from the installed system...
[09:07] <infinity> Kamion: Or something elsewise goofy.
[09:09] <infinity> Kamion: FWIW, I've confirmed here that LRM works on the livefs itself.  I haven't done an espresso install to see what might be breaking between A and B.
[09:14] <Riddellll> Kamion: overwriting problem in ubiquity-casper http://kubuntu.pastebin.com/673875
[09:31] <ozamosi> Henrik Nilsen Omma. Does that guy have an IRC-nick?
[09:32] <LaserJock> ozamosi: yes
[09:33] <ozamosi> Better question: what irc-nick has he?
[09:33] <LaserJock> ozamosi: something like hno7 , I can't remember. He isn't on all that often
[09:34] <infinity> ozamosi: heno
[09:34] <infinity> ozamosi: He's on #ubuntu-accessibility
[09:34] <ozamosi> Ok, thanks.
[09:34] <LaserJock> infinity: ah yes, thanks
[09:34] <RoeyInDisgust> HI!
[09:34] <RoeyInDisgust> I KNEW this place existed
[09:34] <RoeyInDisgust> quick quesiton:
[09:35] <RoeyInDisgust> I found this bug https://launchpad.net/distros/ubuntu/+source/ieee80211/+bug/35095  
[09:35] <Ubugtu> Malone bug 35095 in ieee80211 "Please update al last version" [Wishlist,Confirmed]  
[09:35] <RoeyInDisgust> seemed applicable
[09:35] <RoeyInDisgust> when will this be addressed?
[09:35] <LaserJock> RoeyInDisgust: usually when somebody has time to get to it.
[09:35] <infinity> RoeyInDisgust: ipw3945 is already in our latest kernel.
[09:36] <RoeyInDisgust> ah! cool.
[09:36] <RoeyInDisgust> infinity:  what about the module?
[09:36] <RoeyInDisgust> infinity:  I'm using DD
[09:36] <RoeyInDisgust> infinity:  I suppose you're referring to grump?
[09:36] <RoeyInDisgust> *grumpy?
[09:36] <infinity> There is no grumpy.  I'm referring to dapper.
[09:36] <RoeyInDisgust> infinity:  hmm.
[09:36] <infinity> The latest kernel uploaded a few hours ago has ipw3945 built in.
[09:37] <infinity> Just be patient.
[09:37] <infinity> As for the ieee80211 source, someone will look at your bug in time, wandeing in here to hilight *your* pet bug doesn't help anyone get their work done any faster.  Really.
[09:37] <infinity> IRC is not a bug tracker.
[09:37] <RoeyInDisgust> infinity:  thanks for the information.
[09:38] <RoeyInDisgust> infinity:  and please don't take it personally.
[09:38] <RoeyInDisgust> infinity:   *I* feel like a bug tracker.  That's why i cam here.
[09:38] <RoeyInDisgust> but anyhow, thanks a lot =)
[09:39] <infinity> RoeyInDisgust: It's nothing personal at all.  It's just that if every user felt the need to come in here and hilight their bugs for us, we'de never have any development discussion in here, which is what we TRY to do here.
[09:39] <RoeyInDisgust> heh
[09:39] <RoeyInDisgust> alright.
[09:40] <RoeyInDisgust> no one answered in #k/ubuntu.
[09:40] <RoeyInDisgust> my bad.
[09:40] <RoeyInDisgust> anyhow I'll get outta here
[09:40] <RoeyInDisgust> thanks again!
[09:41] <neuralis> infinity: i expect that the malone featureset will eventually expand to include sentience, after which it *will* come on irc. and probably get into comically pointless ALICE-style nonsense-babble with ubugtu every time someone mentions a bug.
[09:41] <infinity> Don't even joke about that.
[09:42] <jeroenvrp2> hi folks
[09:42] <jeroenvrp2> I had a almost locked/freezed system this evening (using dapper current revision)
[09:43] <infinity> jeroenvrp2: /topic
[09:43] <jeroenvrp2> I have the erroroutput (from /var/log/messages)
[09:43] <jeroenvrp2> infinity: seems like a bug
[09:43] <jeroenvrp2> http://kubuntu.pastebin.com/673936
[09:43] <infinity> jeroenvrp2: Then file it in Malone, please.
[09:44] <jeroenvrp2> infinity: off course, but first I want to know if its me or a bug and what kind of bug
[09:44] <neuralis> jeroenvrp2: there is no bug. you ran out of memory and swap, and the oom killer does what it does best: not work well. nothing to see here, please move on.
[09:45] <Treenaks> doko: OOo has a 'Use hardware acceleration' option in Options -> Display -- any reason for it to be disabled?
[09:45] <jeroenvrp2> neuralis: thank you for your nice answer
[09:46] <infinity> jeroenvrp2: If I had to guess, I'd say perhaps you ram out of memory because you had brltty running which had a hideous memory leak (ate 1.5GB on my system in under 2 days).  Update/upgrade, reboot, enjoy the extra RAM.
[09:46] <infinity> s/ram out of /ran out of /
[09:46] <doko> Treenaks: what does it do? ;-)
[09:46] <jeroenvrp2> !brltty
[09:47] <jeroenvrp2> infinity: oh thanks, so that was it
[09:47] <Treenaks> doko: Well, it's in the "Display" part of the preferences.. it also has a (disabled by default) OpenGL flag (with its own 'use acceleration' flag)
[09:47] <Treenaks> doko: it might be Xv ?
[09:47] <Treenaks> doko: or RENDER ?
[09:48] <Trewas> I had to upgrade in dapper to a self-compiled 2.6.17-rc2 to get suspend-to-ram working. consequently network-manager dies with message "Warning: Driver for device eth1 has been compiled with version 20 of Wireless Extension, while this program supports up to version 19.", does anyone have idea if just re-compiling network-manager will get support for new wireless-extension from kernel sources, or where are they defined?
[09:49] <doko> Treenaks: that might be the support for cairo-glitz in Impress' full screen mode
[09:49] <mjg59> Trewas: Did you file a bug about the suspend to RAM issue?
[09:49] <mjg59> We would aim to fix that in our kernel
[09:50] <Treenaks> doko: I've enabled them locally here.. no ill effects so far (but I'm not really a heavy user, and I have a proper OpenGL driver)
[09:50] <Trewas> mjg59: somebody had filed the bug already
[09:50] <mjg59> "Suspend to RAM doesn't work" isn't a bug
[09:50] <mjg59> "Suspend to RAM doesn't work on this specific machine" is a bug
[09:51] <jeroenvrp2> infinity: I dont have brltty
[09:51] <jeroenvrp2> as you proberly knew
[09:51] <jeroenvrp2> bye
[09:52] <Trewas> mjg59: well, all I have this specific machine and somebody had filed a bug concerning the specific x41 already :)
[09:53] <mjg59> Trewas: Oh, an X41? Should be fixed in the next kernel package (uploaded today)
[09:55] <Trewas> mjg59: I knew it would get fixed in ubuntu just as I'll manage to fix myself :)
[10:35] <siretart> infinity: re bug #26341 - I tested the linked patches from freedesktop bugzilla, and they work for me. I switched xserver-xorg-i810 to used patched on that occasion. may I upload my fix?
[10:35] <Ubugtu> Malone bug 26341 in xserver-xorg-driver-i810 "[i855]  dual-head configs are ill" [Normal,Unconfirmed]  http://launchpad.net/bugs/26341
[10:36] <siretart> s/to used patched/to use dpatch/
[10:43] <jcole> how does cdrom-detect determine if a cd is an ubuntu cd?
[10:44] <jcole> i've remastered an ubuntu install cd for netinstall and now i'm getting that error... something has changed
[10:46] <siretart> uploaded
[11:48] <jcole> Kamion: is the ubuntu /.disk/ dir on any of the ubuntu mirrors?