[00:19] <superm1> slangasek, evand was going to make a new DVD image to include ubiquity 1.8.1 earlier but having internet troubles connecting to cdimage.  would you be able queue one up?
[00:25] <slangasek> superm1: just queuing an ubuntu DVD build? sure
[00:25] <superm1> yeah.  thanks :)
[00:25] <slangasek> scheduled, expect results in > 3h
[00:31] <keescook> blueprint: add -Werror to default compiler flags.
[00:31] <keescook> step 2: lose all my friends
[00:31] <StevenK> step 3: PROFIT!
[00:31] <StevenK> Ahem
[00:31] <cody-somerville> slangasek, I think Xubuntu alternative should be okay to release now
[00:32] <ScottK2> He'd have lots of free time for Step 3
[00:32] <slangasek> cody-somerville: I only see one test result for amd64?
[00:33] <cody-somerville> I think Jim is going to do another amd64 test anyhow
[00:33] <slangasek> shall I wait for results on that test before publishing?
[00:34] <cody-somerville> How about we publish them tomorrow afternoon?
[00:35] <cody-somerville> As long as there is no bad reports in by that time
[00:35] <slangasek> afternoon in whose time zone?
[00:36] <slangasek> (afternoon is fine, just want to know which noon it's to be after :)
[00:39] <cody-somerville> hehe
[00:39] <cody-somerville> How about 1600 UTC?
[00:39] <ScottK2> How about Guam?
[00:40] <ScottK2> ;-)
[00:40] <cody-somerville> :)
[00:43] <pochu> cody-somerville: congrats :)
[00:43] <cody-somerville> Thanks pochu :)
[00:56]  * calc is doing first build of OOo 2.4.0-1ubuntu1
[00:57] <slangasek> cody-somerville: ok
[00:58] <calc> anyone happen to know how to get changed properties to show up in bzr diff/stat/etc something?
[00:58] <calc> it shows me something changed but not what
[01:11] <alex-weej> if brasero is in the default installation, nautilus-cd-burn should be removed
[01:12] <_MMA_> alex-weej: Wouldnt that also remove the ability to right-click on a image and burn? I wouldnt want to lose that.
[01:12] <stgraber> calc: bug 46636
[01:12] <ubotu> Launchpad bug 46636 in bzr "meaning of '*' in bzr status is undocumented and unverbose" [Low,Confirmed] https://launchpad.net/bugs/46636
[01:12] <alex-weej> personally, i think brasero is not ready
[01:13] <alex-weej> and i prefer nautilus's integration
[01:13] <alex-weej> BUT, we currently have massive duplication
[01:13] <alex-weej> you can right click an ISO and go to "Open with Brasero Disc Burner"
[01:13] <_MMA_> Didnt answer my question and it looks like you're just trying to start something so nm.
[01:14] <alex-weej> _MMA_: NO
[01:14] <calc> stgraber: fun
[01:14] <calc> stgraber: wow thats an old bug too
[01:15] <calc> oh gah
[01:16] <calc> doko: ping
[01:16] <calc> doko: i'm getting crt1.o errors on amd64
[01:26] <calc> http://pastebin.ubuntu.com/6124/
[01:26] <calc> anyone know what the problem is there?
[01:26] <calc> if i am reading it correctly its a bug in libc6 but i think i must be misreading it
[01:27] <slangasek> (.text+0x20): undefined reference to `main'
[01:27] <slangasek> so where's your main()?
[01:27] <slangasek> that's not a glibc problem
[01:27] <slangasek> -o ../unxlngx6.pro/bin/cpp
[01:27] <slangasek> er... also, why does OOo need to build its own C preprocessor...?
[01:28] <calc> because it is crack
[01:28] <slangasek> sigh
[01:28]  * calc is looking for the missing main
[01:28] <mneptok> no, crack is what you're smoking if you think you can build OO.org without a blood sacrifice to mmeeks. :)
[01:29] <slangasek> well anyway, those undef symbols aren't anything that glibc is responsible for providing
[01:29] <calc> slangasek: ah ok
[01:34]  * calc thinks he sees where the main is, but doesn't see where it is built
[01:40] <calc> ah i see
[01:40] <calc> its in _cpp.c
[01:40] <calc> int main(int argc, char **argv)
[01:40] <calc> thats a valid main isn't it?
[01:41]  * calc will have to objdump _cpp.o and see whats in it
[01:43] <calc> slangasek: what am i looking for in objdump output to be certain it is properly compiled?
[01:43] <calc> i see this:
[01:43] <calc> 0000000000000b20 g     F .text  00000000000000c7 main
[01:44] <calc> i did a disassemble and see the assembly for the main function as well
[01:51] <calc> do any of those weird PIE things apply to debuild stuff on a local machine, or just buildds? or can it even cause weird issues like this?
[01:51]  * calc haven't looked into how PIE works, just that it does similar thing to what is done for fPIC libs
[01:52]  * calc guesses he can test by attempting to recompile current hardy OOo to see if it still works
[02:28] <calc> what does a lone -L do on a line for g++ ?
[02:28] <calc> eg if the line contains various -L/foo/path then a -L
[02:29] <calc> with nothing after it
[02:40] <calc> I think I found the bug, now how to fix it
[02:45] <emgent> calc: what bug?
[02:48] <calc> emgent: bug in OOo 2.4.0 build system
[02:49] <tritium> You gotta love $42 shipping fees + $53 customs fees for a 6 lb. order of shirts from shop.canonical.com...
[02:49] <emgent> argh
[02:51] <calc> tritium: wow
[02:51] <tritium> calc: yeah, probably our last order, even though we would have wanted some of the laptop bags that were out of stock
[02:53] <tritium> (as were the white polos)
[02:56] <calc> hmm i wonder if i can buy them directly when in london, to skip on the nice shipping/custom fees
[02:57] <tritium> calc: I certainly hope you can
[02:57] <calc> i would probably have to ask to have them have it at the office, if i wanted it
[02:57] <calc> since they sell through another company i think
[02:59] <tritium> That would be a nice arrangement.
[03:04] <emgent> calc: about bug #174112 have you saw debian patch?
[03:04] <ubotu> Launchpad bug 174112 in openoffice.org "[openoffice.org] [CVE-2007-4575] Potential arbitrary code execution vulnerability in 3rd party module (HSQLDB)" [Critical,In progress] https://launchpad.net/bugs/174112
[03:05] <emgent> ops, sorry i read now your last comment.
[03:07] <calc> emgent: yea, i am going to have to do something about it soon, the upstream guy vanished before he gave me the fix
[03:07] <calc> :(
[03:08] <calc> i'm probably going to have to find out exactly how the patch for debian works and write different one for each of ubuntu's releases :-\
[03:08] <calc> since upstream who was going to do it and send me the patches just stopped responding to emails entirely :(
[03:10] <emgent> argh
[03:21] <calc> emgent: as soon as i get OOo 2.4 resolved i will be working on that
[03:21] <calc> emgent: i have some other related work to do soon anyway
[03:21] <emgent> calc: if you like i can work with you about it.
[03:22] <calc> emgent: ok, are you emgent@ubuntu.com ?
[03:23] <calc> emgent: i think for the other work i have what is needed, just not for the old security bug
[03:24] <emgent> not now, i'm motu-swat member and ubuntu-whitehat teamleader, but not official members now.
[03:24] <emgent> calc: https://edge.launchpad.net/~emgent
[03:24] <calc> ok
[03:38] <ScottK2> emgent: Why is 174112 not marked as affecting hsqldb in Ubuntu and why is motu-swat subscribed when all the affected packages are in Main?
[03:39] <emgent> one moment
[03:40] <emgent> bug #174112
[03:40] <ubotu> Launchpad bug 174112 in openoffice.org "[openoffice.org] [CVE-2007-4575] Potential arbitrary code execution vulnerability in 3rd party module (HSQLDB)" [Critical,In progress] https://launchpad.net/bugs/174112
[03:40] <emgent> motu-swat is an error. removed.
[03:41] <emgent> about hsqldb i'm working to it and i will add it.
[03:42] <ScottK2> OK.  Sounds good.
[03:47] <ds> why is swfdec-mozilla so old in ubuntu compared to debian?
[03:49] <Fujitsu> !info swfdec-mozilla hardy
[03:49] <Fujitsu> !info swfdec-mozilla sid
[03:50] <Fujitsu> swfdec-mozilla |    0.5.4-1 | hardy/universe | source, amd64, i386, powerpc
[03:50] <Fujitsu> swfdec-mozilla |    0.5.5-1 |      unstable | arm, armel, hppa, m68k, mips, mipsel
[03:50] <Fujitsu> swfdec-mozilla |    0.6.0-2 |      unstable | source, alpha, amd64, i386, ia64, powerpc, s390, sparc
[03:50] <Fujitsu> We're 0.0.1 behind lenny. I don't think that's too bad.
[03:50] <Fujitsu> And we're not that far behind sid, considering we won't have synced in 3 months.
[03:56] <ds> it seemed like an oversight, as it fixes the outstanding issues in launchpad
[03:57] <Fujitsu> We unfortunately don't have enough developers to carefully watch every single package in Debian.
[03:57] <ds> yeah
[04:00] <Fujitsu> We would have noticed if the Debian uploads had closed RC bugs, but they didn't close any bugs at all.
[04:08] <slangasek> calc: -fPIE would apply to locally-built executables, but isn't part of the standard flags right now; -Wl,-Bsymbolic-functions also applies, and could have something to do with it, but I'm not sure what given that other apps aren't having this problem when building.  Oh, I just now noticed that you're using ccache...
[04:08] <slangasek> calc: oh, yes, -L by itself translates to "please screw up the interpretation of the following argument for me"
[04:13] <bd_> Hi, there's a bug which breaks gitk and git-gui's dependencies, and has had a trivial patch in launchpad to fix it for about two weeks... Any chance of getting an upload so it's not broken for hardy? https://bugs.launchpad.net/ubuntu/+source/git-core/+bug/196846
[04:13] <ubotu> Launchpad bug 196846 in git-core "gitk requires wish8.5 but depends on tk8.4" [High,Confirmed]
[04:32] <enrico> Riddell: I just opened debian bug #472911
[04:32] <ubotu> Debian bug 472911 in debtags "Data files are embarassingly created in the wrong place" [Serious,Open] http://bugs.debian.org/472911
[04:32] <enrico> Riddell: it affects hardy
[04:33] <enrico> Riddell: I should upload a fix within a day or two
[04:44] <PurpleBlu> I really enjoy http://brainstorm.ubuntu.com/.   With this whole suggest/vote concept, how much of it would realistically transfers over?
[04:46] <PurpleBlu> I know you can't quantitatively answer that.  I am just curious if my votes actually have better chances then me hitting megabillion lotto.
[04:47] <nixternal> PurpleBlu: last time I checked, we only had mega millions here in Illinois ;p
[04:49] <PurpleBlu> nixternal, I know FOSS is a hobbyist based community driven devs.  I was just curious that someone like me who is not a 1337 coder, how I can ultimately contribute.
[04:50] <ScottK2> bd_: ubuntu-main-sponsors was not subscribed to the bug.  They are now.  Odds are no one who could upload it even saw it.
[04:51] <nixternal> PurpleBlu: talk to me in #ubuntu-chicago
[04:54] <PurpleBlu> k
[04:55] <jdong> nixternal: Mega Millions and MegaPlier are registered trademarks of Mega Ball Lottery Games. Use of trademarks is subject to the End Speaker License Agreement (ESLA). By using these trademark terms you are certifyin agreement to these terms. Please drink responsibly.
[04:59] <bd_> ScottK2: ah, okay, it wasn't clear what the procedure was for notifying the appropriate people :) or rather, the Bugs/HowToFix wikipage said to just come here...
[05:00] <ScottK2> OK.  Somewhere it should say for ubuntu-main-sponsors or ubunt-universe-sponsors to be subscribed.
[05:01] <ScottK2> I think there's a page on sponsorship that ought to be linked from that one.  If it's not, would you please add it.
[05:04] <bd_> hm, the MOTU page talks about it a bit, but in the context of universe, not main (and mostly on how to contribute new packages)
[05:04] <bd_> but I'm way behind on a school assignment atm so I think I shouldn't be futzing around with wiki pages atm, sorry :)
[05:05] <ScottK2> bd_: I understand.  Thanks for contributing and come back when you have a little time.  The documentation is best updated by those who are trying to use it.
[05:05] <calc> slangasek: ah nice :-\
[05:06] <calc> slangasek: there isn't a way to get someone to fix that issue with -L screwing up the build?
[05:06] <calc> a -L with a space following is obviously always wrong (or am I missing something?)
[05:34] <doko> calc: problem solved?
[05:38] <calc> doko: yea it was a empty -L
[05:39] <calc> doko: why does empty -L cause build failures like that, shouldn't it just ignore it?
[05:41] <doko> no, it just takes the next arg as the dir
[05:42] <calc> oh so it doesn't have to have no space like how it usually is done
[05:42] <calc> eg -L/path/to/something, i see :)
[05:43] <calc> makes debugging things annoying unless you happen to see the extra -L floating about :)
[05:43] <calc> i ended up comparing it to the working build to find it
[05:43] <calc> since there were so many -L in that gcc call
[06:46] <warp10> Good morning
[07:12] <dholbach> good morning
[07:13] <cody-somerville> lo dholbach :)
[07:13] <dholbach> hiya cody-somerville
[07:16] <pitti> Good morning
[07:16] <dholbach> hiya pitti
[07:17] <ion_> Hi all
[07:23] <orbisvicis> im going through several scripts, but hardy jack client scripts call pasuspender and that can be really annoying if pulse outputs to jackd
[07:23] <orbisvicis> *tip to packager
[07:24] <ion_> ...who probably will never notice it if you only mention it here.
[07:24] <ion_> Please file a bug report.
[07:26] <orbisvicis> not going to file a bug per package
[07:30] <orbisvicis> maybe i will if all these bash scripts are so badly written
[07:31] <dholbach> hey seb128
[07:34] <seb128> oh, dholbach is back!
[07:35] <seb128> hey hey dholbach
[07:40] <dholbach> seb128: I was in France! :)
[07:41] <seb128> dholbach: how, was it? ;-)
[07:41] <dholbach> very very nice :)
[07:43] <dholbach> seb128: Mimi and I were in Paris for a few days - we even managed to meet up with lool
[07:47] <pitti> dholbach: viva la France! *hug*
[07:47] <dholbach> wooohoo! :)
[07:47]  * dholbach hugs pitti
[07:49] <seb128> hey pitti
[07:49]  * seb128 hugs pitti
[08:36] <soren> How do I do pm-suspend-hybrid from GNOME?
[08:41] <pitti> soren: you don't, I think :/
[08:41] <soren> Alrighty. I'll stop looking :)
[09:13] <kagou> hi
[09:17] <kagou> is there a problem with gnome-system-monitor, i can't find it in administration menu ?!
[09:19] <james_w> It's moved to the System Tools menu it seems.
[09:19] <james_w> in Applications
[09:23] <kagou> james_w, diantre !
[09:24] <kagou> james_w, indeed, thanks
[09:24] <james_w> no problem
[09:31] <asac> anyone knows how to make dget also fetch the debs listed in a .changes? (i only get the source parts, though the .debs are definitly listed in changes)
[09:39] <cody-somerville> asac, According to the nice man page, you provide the package name and it grabs it from the mirror listed in your sources.list
[09:40] <cody-somerville> asac, It appears that it does not support grabbing the binary packages by providing a url :)
[09:42] <asac> thats strange. i am sure i got binary bits through dget yesterday
[09:42] <asac> anyway. thanks
[09:43] <asac> (and no, those are not in a archive i could list in sources.list)
[09:44] <seb128_> cody-somerville: what manpage do you have?
[09:45] <cody-somerville> The one that ships with ubuntu's dget?
[09:45] <seb128_> "dget downloads Debian packages.  In the first form, dget fetches the
[09:45] <seb128_>        requested URL.  If this is a .dsc or .changes file, then dget acts as a
[09:45] <seb128_>        source-package aware form of wget: it also fetches any files referenced
[09:45] <seb128_>        in the .dsc/.changes file.  When the -x option is given, the downloaded
[09:45] <seb128_>        source is also unpacked by dpkg-source."
[09:45] <cody-somerville> Ah
[09:45] <cody-somerville> Fair enough
[09:46] <cody-somerville> I was looking at the second paragraph
[10:21] <pitti> seb128_: can you please have a quick look at https://bugs.edge.launchpad.net/ubuntu/+source/jockey/+bug/202898/comments/9, the second point?
[10:21] <ubotu> Launchpad bug 202898 in jockey "gnome-appearance-properties shouldn't install restricted driver without asking" [High,Fix committed]
[10:21] <pitti> seb128_: do you know which exit code appearance-properties checks, and how it should behave?
[10:24] <seb128_> pitti: is that jockey-gtk --check-composite which does the driver installation?
[10:24] <pitti> seb128_: right
[10:24] <pitti> seb128_: we need to agree on how it should behave, and which exit code jockey shuold deliver for 'cancel', etc.
[10:24] <seb128_> pitti: it checks for == 0
[10:27] <pitti> seb128_: so it should return 0 IFF it installed the composite-aware driver correctly, and 1 on cancel/fail to install/no driver necessary?
[10:31] <seb128_> pitti: yes
[10:31] <seb128_> pitti: I'm happy to change the g-c-c code if you prefer something else though
[10:31] <pitti> hm, that's what it does
[10:31] <pitti> I don't quite understand what he means in the bug report
[10:41] <kagou> slangasek, i'm investifating the use of nautilus-share (samba usershare) as it will be installed by default. Normal (non admin) users are not in sambashare group by default. usershare is provided by samba to allow normal users to share folders without special permissions. Is it a problem to hack user creation to add all users on sambashare on creation ?
[10:45]  * kagou offer some croissants to slangasek 
[10:48] <\sh> one question to PA: the monitor input devices (showing up in padevchooser) for the soundcards are (as I see it) a passthrough device from output devices to input devices, correct?
[11:07] <pitti> YokoZar: ooh! congratulations to your MOTU badge, and welcome! *hug*
[11:08] <YokoZar> pitti: bug is now closed ;) *hug*
[11:11] <ogra_cmpc> YokoZar, hey, great to hear
[11:13] <TheMuso> Does anybody object to adding pulseaudio-module-x11 to the desktop seed, so that pulseaudio properly gets killed on user logout?
[11:14] <ogra_cmpc> TheMuso, doesnt that also bring the cool keyboard bell sounds ?
[11:14]  * ogra_cmpc is all for it
[11:14] <TheMuso> ogra_cmpc: If they are enabled, and I don't think they are by default.
[11:14] <TheMuso> I'd have to check.
[11:18] <ogra_cmpc> yup, its disabled
[11:19] <TheMuso> ogra_cmpc: One thing I intend to spec for intrepid, is to use pulse to make the bell sound, as IMO we shouldn't use the PC speaker, and more and more machines are coming without them.
[11:23] <ogra_cmpc> yeah
[11:23] <ogra_cmpc> and it just sounds a lot friendlier than BEEP !
[11:23] <TheMuso> I agree.
[11:25] <ogra_cmpc> i would actually propose it for hardy ... its two comments to remove, unlikely to break anything and there are no screenshots of it :) we should put it on next weeks platform agenda (and ask the desktop team for comments)
[11:26] <TheMuso> ogra_cmpc: Yeah but I think we need a better way for the user to set what sound to use.
[11:26] <TheMuso> Atm it can only be done in that file. It would be nice if it could be done per user session, which I need to look into.
[11:26] <ogra_cmpc> i remember we had it on in happer when we first evaluated polypaudio and people liked it (even though we dropped back to esd back then)
[11:27] <ogra_cmpc> *dapper
[11:27] <TheMuso> Right.
[11:27] <TheMuso> I still think its something that should be configured per user.
[11:27] <ogra_cmpc> thats a nice to have option imho
[11:28] <ogra_cmpc> you cant configure the beep easily either
[11:28] <TheMuso> Well, perhaps to select the sound yes, but whether the user wants a sound played, or a pc speaker beep I think should be user configurable.
[11:28] <TheMuso> True.
[11:30] <seb128> TheMuso: you forgot to revert the gucharmap patch changes otherwise lpi looks good
[11:31] <TheMuso> seb128: Oh yeah right. :) I originally changed it thinking that it was the same patch used in the actual package. I'll do that now.
[11:32] <emgent> heya people
[11:40] <jdstrand> hi evand! I saw your comment in bug #206030
[11:40] <ubotu> Launchpad bug 206030 in ubuntu "cannot activate ufw in live mode (/ is world-writable)" [Undecided,Fix committed] https://launchpad.net/bugs/206030
[11:41] <jdstrand> evand: oh, heh-- but I didn't see 'Fix committed'
[11:41] <jdstrand> nevermind... :)
[11:41] <emgent> hi jdstrand
[11:41] <jdstrand> hi emgent!
[11:43] <emgent> jdstrand: when you have time see bug #207494 && bug #207490
[11:43] <ubotu> Bug 207494 on http://launchpad.net/bugs/207494 is private
[11:43] <ubotu> Bug 207490 on http://launchpad.net/bugs/207490 is private
[11:44] <jdstrand> emgent: thanks emgent!
[11:44] <emgent> np :)
[11:50]  * dholbach added bitesize, upgrade, packaging, ftbfs and unmetdeps bugs to http://daniel.holba.ch/really-fix-it - hope it proves useful to you
[12:08] <james_w> I'd like some advice on debian bug 464118 (ubuntu bug 192239)
[12:08] <ubotu> Debian bug 464118 in coreutils "rm -r broken: Function not implemented" [Important,Open] http://bugs.debian.org/464118
[12:08] <james_w> the Debian maintainer says that it doesn't need to be fixed for lenny, as it requires a kernel version older than that shipped in etch.
[12:09] <james_w> can we apply the same logic for hardy/dapper?
[12:11] <james_w> I think not, as the bug was reproduced on hardy running a .15 kernel, which was the most recent in dapper.
[12:14] <mjg59> pitti: http://www.codon.org.uk/~mjg59/tmp/98-unload_network_modules.patch
[12:15] <pitti> mjg59: ah, great;
[12:15] <pitti> mjg59: I was about to ask you how the suspend script can "remember" which modules it unloaded, so that resume can modprobe them again
[12:15] <pitti> so that 'modunload' macro does that trick?
[12:22] <mjg59> Yup
[12:27] <Ng> nice, wish I'd spotted that modunload was smrt. my aborted attempt at such a network module script was a horror show of trying to remember a list of modules ;)
[12:27] <Ng> mjg59: is there any real need to do the IRDA and netconsole ones acpi-support does too?
[12:30] <ogra_cmpc_> mjg59, why dont you just use the SUSPEND_MODULES variable ?
[12:33] <pitti> seb128: hm, bug 207587, is that nautilus or gvfs?
[12:33] <ubotu> Launchpad bug 207587 in human-icon-theme "[hardy regression] Missing icon for mounted encrypted volumes" [Medium,Confirmed] https://launchpad.net/bugs/207587
[12:33] <mjg59> ogra_cmpc_: Because that's already used
[12:34] <ogra_cmpc_> ah
[12:34] <seb128> pitti: the icon theme I would say
[12:34] <ogra_cmpc_> i dont see it used anywhere yet (i set it on teh classmate to get rid of button during resume)
[12:35] <mjg59> ogra_cmpc_: I don't understand what you mean by "just use"
[12:35] <seb128> pitti: the icon used is "media-encrypted""
[12:35] <ogra_cmpc_> mjg59, instead of uloading them yourself with modunload i mean
[12:35] <ogra_cmpc_> *unloading
[12:37] <seb128> pitti: gvfs is where the icon are set if that's the question
[12:37] <ogra_cmpc_> ... vs just appending them to SUSPEND_MODULES and let pm-utils care for it
[12:37] <mjg59> ogra_cmpc_: Because that's a user-settable variable - I wasn't enthusiastic about overloading it
[12:37] <ogra_cmpc_> ah
[12:39] <pitti> seb128: ah, I see; I thought it should/would just use the same icon as for unencrypted devices
[12:40] <pitti> mjg59: I'll upload it soon, thank you!
[12:42] <mjg59> pitti: I've got another patch or two for you
[12:43] <mjg59> pitti: http://www.codon.org.uk/~mjg59/tmp/97_fix_ppc_suspend_test.patch (for 189851)
[12:43]  * \sh jumps from the next bridge...
[12:45] <mjg59> pitti: Oh, and 70-remove-pm-pmu.patch needs to be dropped
[12:45] <pitti> \sh: don't
[12:45] <mjg59> pitti: ...and pm-utils made arch: any not all
[12:46] <mjg59> Grah. Hideous debian-specific breakage.
[12:47] <\sh> pitti: I'm doomed...why is it not possible to capture audio from an output device? pulseaudio is enabling an monitor device and offers it as input .. but apps with libasound compat are not able to get this device as input device :(
[12:47] <mjg59> pitti: Hngh. This is all horribly awkward.
[12:48] <mjg59> pitti: We need (a) that patch I provided, (b) lose 70-remove-pm-pmu.patch, (c) s/all/any/, (d) make sure pm-pmu-sleep is shipped in the PPC builds
[12:48] <pitti> mjg59: hm, we can't just add some uname -m check to the script and leave it as arch:all?
[12:49] <mjg59> pitti: No, because pm-pmu is a binary
[12:49] <pitti> ugh
[12:49] <pitti> mjg59: in pm-i I remember using a perl call do do the sleep ioctl
[12:49] <sjoerd> \sh: sure they can
[12:49] <mjg59> pitti: That'd be possible, but I'm not sure there's an advantage in deviating from upstream
[12:50] <\sh> sjoerd: -device: plug:pulse? doesn't help...
[12:52] <mjg59> pitti: The alternative would be to call /usr/lib/hal/hal-system-power-pmu sleep on PPC, rather than pm-pmu
[12:53] <mjg59> pitti: That would keep us arch: all
[12:53] <pitti> ah, even easier, yes
[12:53] <mjg59> pitti: But it still needs my patch :)
[12:53] <pitti> mjg59: too bad I don't have a PPC any more to test all this
[12:54] <mjg59> pitti: Heh
[12:54] <\sh> sjoerd: what device should I use when I want to use this monitor device of pulse in an libasound app?
[12:54] <sjoerd> \sh: put pcm.monitor { \n type pulse \n device <pulse devicename> \n } in your .asoundrc
[12:54] <pitti> mjg59: so, I'll head out for lunch, then do the desktop team meeting, and then see to assembling it all to a source, unless you beat me to it with a debdiff :)
[12:54] <sjoerd> and then you can use the monitor device
[12:55] <sjoerd> I'm not sure how and if you can specify that as a device parameter instead
[12:56] <\sh> sjoerd: well, that is my problem...the device paramenter
[12:56] <\sh> parameter even
[12:57] <sjoerd> One idea we've been toying with is to create a pulse module that creates a asoundrc with all your devices listed automagically
[12:57] <sjoerd> The nasty part is ensuring that goes away when pulse dies/crashes/is shot/etc
[12:57] <\sh> sjoerd: the pulse device name is the name of the monitordevice which pulse added automatically?
[12:57] <sjoerd> yes
[12:58] <mjg59> pitti: Ok, cool
[13:02] <\sh> sjoerd: well...the problem it seems, to tell the app to use this monitor device...I wonder what the alsa name of such a device is
[13:03] <sjoerd> the device names are quite static
[13:03] <soren> Are uploads to -proposed announced anywhere?
[13:04] <seb128> soren: not on -changes
[13:05] <soren> seb128: Precisely. Hence my question :)
[13:05] <\sh> sjoerd: well, e.g. recordmydesktop is using hw:0,0 as default (it doesn#t know anything about the "default" mapping name of asound)...I could use as well plughw:0,0 .. when I set the defaults on padevchooser, this works...but whem I start to set default to the monitor device, it doesn't work :(
[13:11] <\sh> grmpf...bbl
[13:22] <\sh> sjoerd: YOU ARE DA MAN :)
[13:22] <sjoerd> heh ?
[13:22] <\sh> sjoerd: it works .)
[13:22] <sjoerd> using -D directly or through a .asoundrc thingy
[13:22] <\sh> apt-get install libasound2-plugins -> your pcm.monitor magic -> recordmydesktop -device monitor <- wokrs
[13:24] <sjoerd> cool
[13:24] <\sh> hey this costed me now one week of trying and banging my head on the table
[13:25] <sjoerd> next time just jump in to #pulseaudio :)
[13:25]  * ogra_cmpc_ thought we install that by default
[13:27] <ogra_cmpc> \sh, i thought you had pulseaudio installed ... libasound2-plugins is a dep
[13:28] <ogra_cmpc> ah, no, only a recommends
[13:29] <\sh> ogra_cmpc: it's not
[13:29] <\sh> ogra_cmpc: but fun part, without the pcm.monitor magic in .asoundrc it doesn't work
[13:30] <Hobbsee> open week...again?
[13:30] <seb128> Hobbsee: new ubuntu again, uds again, open week again ;-)
[13:30] <Hobbsee> but...the last open week was only februrary....
[13:30] <Hobbsee> -r
[13:31] <seb128> Hobbsee: that was a developer week
[13:31] <Hobbsee> oh
[13:31] <\sh> ogra_cmpc: anyways..one week of hard work...and the solution is so simple
[13:31] <Hobbsee> seb128: right, that makes sense then
[13:31] <\sh> sjoerd: I will join #pulseaudio :) if I get my project working...I think it will be a nice example what you can do with linux and sound and video :)
[13:32] <ogra_cmpc> at least if you invest a week :)
[13:33] <\sh> ogra_cmpc: can you send privmsgs now with your nick now? :)
[13:34] <ogra_cmpc> not with the cmpc one :) just set unregged to on for your client :P
[13:45] <Hobbsee> where's mvo?
[13:45] <pitti> Hobbsee: on holiday
[13:45] <Hobbsee> oh
[13:46] <Hobbsee> pitti: when's he back?
[13:46] <pitti> next week
[13:46]  * Hobbsee didn't think you guys were allowed holidays near releases.
[13:47] <seb128> Hobbsee: that's not near yet
[13:48] <seb128> Hobbsee: yeah, we have no holidays in the 3 months before and after each versions ;-)
[13:48] <Hobbsee> seb128: that's normal for software development, it appears.
[13:48] <Hobbsee> seb128: mind you, i grew up thinking that the australian working day didn't finish before 8pm, either.
[13:52] <pitti> mjg59: WDYT? http://paste.ubuntu.com/6132/
[13:53] <pitti> mjg59: that doesn't contain the e1000 fix yet, just the ppc suspend issue
[13:53] <mjg59> pitti: Looks good
[13:54] <pitti> mjg59: I'll forward the suspend fix to upstream, or is it already?
[13:55] <mjg59> pitti: Not sure. It's not /strictly/ correct, because you can have a pmu without being able to suspend
[13:55] <pitti> mjg59: ah, so you need to do that ioctl to know
[13:55] <pitti> we have that in pmi
[14:10] <emgent> i go to work, see you later people.
[14:22] <lamont> jdstrand: btw, /var/lib/misc/group.db r, for cupsd (and anything else that uses/checks groups...)
[14:23]  * jdstrand makes note of that
[14:23] <lamont> Mar 27 07:35:14 mix kernel: [391824.339457] audit(1206624914.559:79):  type=1503 operation="inode_permission" requested_mask="rw" denied_mask="rw" name="/var/lib/misc/group.db" pid=6339 profile="/usr/sbin/cupsd"
[14:24]  * lamont files that one under "WTF?"
[14:24] <jdstrand> lamont: is there a bug for that? I'll be doing a cups update soon and can roll it in
[14:24] <lamont> mid-upgrade firefox just crashed for me (gotta love transitioning libraries...)  so I didn't file one yet
[14:25] <lamont> also, if nsswitch.conf has 'ldap' in it, then global read access for /etc/ldap/ldap.conf is kinda needed too.
[14:27] <jdstrand> lamont: can you toss it all into the same bug?
[14:27] <lamont> yeah - I'll file it in about an hour, once the upgrade completes
[14:28] <jdstrand> lamont: cool, thanks
[14:51] <dholbach> pitti, seb128, lool: you like  http://daniel.holba.ch/really-fix-it  now? :)
[14:52] <seb128> dholbach: I'm wondering why C is before B ;-)
[14:52] <dholbach> seb128: I just wondered the same :)
[14:52] <seb128> and E before D
[14:52] <dholbach> I'll fix that
[14:52] <dholbach> no problem
[14:52] <seb128> etc
[14:52] <seb128> ;-)
[14:52] <emgent> heya
[14:52] <seb128> otherwise looks cool
[14:53] <seb128> but the fact that launchpad has no patch status does make it easy to build a pertinant list
[14:59] <mathiaz> I uploaded mysql-dfsg-5.0 yesterday. It built successfully on sparc, but I've received an email stating that it failed to upload for this arch. What should I do to fix this error ?
[15:06] <pitti> dholbach: nice set sort ordering :)
[15:06] <ScottK> dholbach: Do you have a way to exclude patches/bugs that have a fix, but the fix isn't a valid one?
[15:06] <pitti> dholbach: looks nice! just waaay too big
[15:07] <pitti> dholbach: is there a way to filter by component?
[15:07] <dholbach> pitti: that's why I split it up :)
[15:07] <dholbach> no, no filtering
[15:07] <ScottK> dholbach: As an example, Bug #165184 solves a to much mail from spamassassin problem by piping mail to /dev/null.
[15:07] <ubotu> Launchpad bug 165184 in amavisd-new "amavisd-new + spamassassin: cronjob spams root user" [Medium,Triaged] https://launchpad.net/bugs/165184
[15:07] <pitti> dholbach: big> not the fault of the messenger
[15:08] <pitti> ScottK: heh
[15:08] <dholbach> ScottK: the only thing I know to work around it is to "unflag the patch thingie of the attachment" :/
[15:08] <asac> pitti: you mean nm-editor menu entry, right?
[15:08] <pitti> ScottK: debian/patches/dry_your_email_swamp.patch? :)
[15:08] <asac> (not nm-tool)
[15:08] <pitti> asac: right; slangasek did it this morning, so all is well
[15:09] <asac> pitti: yep. I gave him green light yesterday. thanks for confirming
[15:10] <ScottK> dholbach: How often does it refresh data from LP?
[15:11] <dholbach> ScottK: every 4 hours because it takes ~40m to run - I don't want to hammer LP that bad
[15:11] <ScottK> dholbach: Understand.  I unchecked that one and I want to know how soon to check back and see if that cleared it.
[15:12] <dholbach> ScottK: I'm re-running it right now - I'll let you know
[15:13] <rexy_> xb6FFFFFF xb7FFFFFF isnt that where ussually libc and such live?
[15:25] <lamont> Error org.freedesktop.DBus.Error.Failed: Element <standard_system_servicedirs> not allowed inside <busconfig> in configuration file
[15:25] <lamont> is that me, or hardy, I wonder?
[15:26] <pitti> mjg59: https://bugs.freedesktop.org/show_bug.cgi?id=15226 FYI, if you want to subscribe
[15:26] <ubotu> Freedesktop bug 15226 in General "network modules need to be unloaded during suspend" [Normal,New]
[15:27] <pitti> mjg59: heh, cute; I just saw that I have an rmmod/modprobe in /etc/pm/sleep.d on my laptop :)
[15:27] <collusion> any thoughts on how to debug a broken suspend on a x40 with hardy?
[15:29] <pitti> mjg59: erk, s/done/fi/ in the script; fixing
[15:30] <james_w> collusion: https://wiki.ubuntu.com/DebuggingKernelSuspend
[15:31] <collusion> james_w: sweet, reading now, thx.
[15:33] <rexy_> are there any issues wrt segfaulting of the pulseaudio daemon? forcing in a bluetooth source and switching streams will segfault it
[15:33] <james_w> that sounds like an issue to me
[15:34] <rexy_> heu i ment if anyone was aware on br's filed of that nature
[15:35] <lamont> "Sorry, the package "updata-manager" failed to install or upgrade."
[15:35] <lamont> there's something funny about that
[15:37] <rexy_> there are several issues actually, depending on your options it will randomly decide to not start, it does not accept an virtual alsa bluetooth device on startup, but it will load it just fine through the pactl module loader, but crashes easily
[15:37] <rexy_> how could i start that br_reporting thing from the cli?
[15:38] <james_w> apport-cli
[15:47] <ScottK> dholbach: What would you think about removing partner from the packages you list.  They really aren't an Ubuntu distro issue.
[15:47] <rexy_> ok filed, #207796, anything i should add?
[15:47] <dholbach> ScottK: um... which package and bugs do you mean in this case?
[15:48] <ScottK> dholbach: Bug #146269 in openssl097 is the one I noticed, but I think partner should just be left out to reduce the noise.
[15:48] <ubotu> Launchpad bug 146269 in openssl097 "[openssl security] OpenSSL SSL_get_shared_ciphers() off-by-one buffer overflow" [High,Confirmed] https://launchpad.net/bugs/146269
[15:49] <dholbach> ScottK: added to my todo list - I'll try to figure something out
[15:50] <ScottK> Thanks.
[15:52] <ScottK> dholbach: One more comment: Bugs that affect multiple packages appear to be listed for all affected packages, not just the ones that have fixes.  Bug #204895 is an example.
[15:52] <ubotu> Launchpad bug 204895 in harvestman "Packages failed archive rebuild test possibly due to python-central transition" [Medium,In progress] https://launchpad.net/bugs/204895
[15:53] <ScottK> Finally, http://qa.ubuntuwire.com/bugs/rcbugs/ would be another source of information that might be worth integrating.
[15:53] <dholbach> ScottK: it's impossible to figure out which of the tasks a patch is for :/
[15:53] <dholbach> of course I could parse the filename of bug attachments
[15:54] <ScottK> That or file a Launchpad bug ...
[15:54] <dholbach> ok, adding to the list too
[15:54] <dholbach> rcbugs looks great
[15:58] <siretart> can someone from the release team please have a look at bug #204557? Is there something missing about that report?
[15:58] <ubotu> Launchpad bug 204557 in xine-lib "Freeze exception for xine-lib 1.1.11" [Undecided,New] https://launchpad.net/bugs/204557
[15:59] <dholbach> ScottK: is the bug you untagged still on the list?
[16:00]  * ScottK looks
[16:00] <smoser> sorry for offtopic question. does anyone know the difference between "Mobile Developer" and "Mobile Engineer" (http://www.ubuntu.com/employment) ?
[16:01] <ScottK> dholbach: It is not!
[16:01] <dholbach> ScottK: ROCK
[16:01] <dholbach> smoser: you could try asking in #ubuntu-mobile
[16:01] <wasabi> so at some point in the future we're going to stop asking our users to view diffs of smb.conf
[16:02] <wasabi> right? :0
[16:02] <siretart> dholbach: do you spot something missing on the xine-lib bug?
[16:02] <ScottK2> wasabi: What would you propose we do instead?
[16:02] <siretart> dholbach: do you think I should milestone it?
[16:02] <wasabi> Automatically merge the settings in some fashion. Like every other OS, I guess.
[16:03] <wasabi> I don't have a good answer. I'm wondering if anybody has thought about it.
[16:03] <smoser> dholbach, i figured it was somewhat a "generic canonical job title" question. i'll ask there though
[16:03] <dholbach> siretart: slangasek might know better than I do - it looks OK to me though
[16:04] <ScottK2> wasabi: I think the people who have thought about it mostly get headaches from corner cases.
[16:05] <wasabi> Wondering if maybe I was off base a bit. If we end up with a UI to manipulate Windows settings (join a domain, workgroup, make shares), maybe it should divert samba into a seperate private config file.
[16:05] <wasabi> And revert the distro smb.conf
[16:05] <wasabi> And just deal with ugprades and stuff on it's own.
[16:09]  * ScottK2 imagines one might separate out a .local conffile and do something like that.
[16:10] <collusion> well, /sys/power/pm_trace did not reveal anything too useful (i.e., no "hash matches ...")
[16:10] <collusion> i was able to get it to suspend once while in recovery mode tho...
[16:11] <mjg59> collusion: How's it failing?
[16:13] <collusion> it's a thinkpad x40 and it seems to do most things (video off, e.g.,) but the little half-moon light keeps blinking and i can still use the keyboard to control the led typing light.
[16:14] <collusion> i think i used to have to add acpi_support=s3_bios to the kernel config line.
[16:16] <collusion> but i don't remember how to do that now; editing /boot/grub/menu.lst and running update-grub seems to nuke the manually edited menu.lst
[16:17] <mjg59> collusion: This is suspend to RAM?
[16:18] <collusion> yes, suspend to ram.
[16:18] <collusion> i just ran through the https://wiki.ubuntu.com/DebuggingKernelSuspend instructions but the magic returned did not match any hashes.
[16:20] <mjg59> collusion: Does it work with an older hardy kernel?
[16:21] <collusion> uh. dunno.  i'm running 2.6.24-12-generic which is what came with this when i installed it.
[16:21] <lamont> hooray for gnome-keyring and ssh-agent
[16:21] <lamont> at least only one can grab the keyboard at a time...
[16:21] <collusion> i used to be running 2.6.15-51 or something like that (whatever 6.06.2 had.)
[16:22]  * lamont tries to remember how the hell to get his resolution back above 1280x1024
[16:22] <mjg59> collusion: Ok. Next kernel update should fix it, with luck.
[16:23] <mario_limonciell> mjg59, what's changing in the next kernel update that would be resolving this?
[16:23] <collusion> oh, that's cool.  when's that due out?  was there a launchpad bug about it?  (there are so many suspend bugs in launchpad it is hard to find something specific.)
[16:24] <pitti> mdz: with the e2fsprogs I'm just uploading, the remaining case of fsck usplash integration is fixed now as well  \o/ (feedback appreciated)
[16:25] <ogra_cmpc> pitti, sexy !
[16:25] <mjg59> mario_limonciell: Disabling the new DRM suspend code on i830 and i855
[16:26] <mario_limonciell> ah
[16:26] <tseliot> lamont: maybe you can use "gtf" to generate a modeline
[16:26] <mjg59> pitti: Did you tweak the quirk blacklisting to run the quirks on i830 and i855?
[16:26] <pitti> mjg59: yes, I did
[16:26] <mjg59> pitti: Sweet, thanks
[16:28] <pitti> mjg59: http://lists.freedesktop.org/archives/hal/attachments/20080326/db4dd274/attachment.patch is the current patch
[16:28] <mjg59> pitti: Yup, looking at it now
[16:28] <lamont> tseliot: given that ~/XFree86.0.log doesn't even mention 1680x1050....
[16:29] <pitti> mjg59: I dropped the pm-utils patch again, it was too evil IMHO, and it fits much better in the hal script
[16:29] <pitti> mjg59: evil in the sense of ignoring stuff you explicitly pass on the command line
[16:30] <mjg59> pitti: Yup, that works
[16:31] <tseliot> lamont: XFree86.0.log??? Oh, I thought you were dealing with Xorg.
[16:32] <lamont> well, that could be an ancient file...
[16:32] <lamont> heh.  aug 2004
[16:32] <ogra_cmpc> pretty ancient
[16:33] <mjg59> pitti: Hm. We probably want to hit 845 as well - that's 0x2562
[16:33] <mjg59> And 865, 0x2572
[16:36] <collusion> wow, this bug looks insanely complicated.  thanks guys.
[16:36] <wasabi> bunch of stuff just started crashing in latest updates... gtk_toolbar_insert
[16:36] <wasabi> Some critial assert that it's not really a toolbar
[16:39] <lamont> dear gnome.  what have you done with my ssh-agent session?
[16:39] <pitti> mjg59: thanks; http://bazaar.launchpad.net/~ubuntu-core-dev/hal/ubuntu/revision/223
[16:39] <mjg59> pitti: Win, thanks
[16:40] <pitti> ergh, "quirs"
[16:40] <pitti> fixed
[16:42] <stgraber> The Hardy Heron on my house :) http://www.stgraber.org/download/heron.jpg
[16:42] <ScottK> dholbach: For the really-fix-it problem about multiple affected packages, you could at least not include the bug for packages marked Invalid or Fix Released.
[16:43] <dholbach> ScottK: oh.... it should filter them out - if that'S not the case, I can fix it
[16:44] <dholbach> ScottK: made a note of that and bug 204895
[16:44] <ubotu> Launchpad bug 204895 in harvestman "Packages failed archive rebuild test possibly due to python-central transition" [Medium,In progress] https://launchpad.net/bugs/204895
[16:44] <ScottK> OK.
[16:57] <cody-somerville> slangasek, http://iso.qa.ubuntu.com/qatracker/build/xubuntu/all :)
[17:02] <lamont> Warning: Tried to connect to session manager, Could not open network socket
[17:02] <lamont> what's ^^ that mean?
[17:02] <james_w> lamont: when do you see it?
[17:04] <lamont> firing up a new uxterm
[17:05] <lamont> and then there's: Cannot mount volume. org.freedesktop.DBus.Error.AccessDenied
[17:07] <pitti> lamont: sounds like you don't have a ConsoleKit session?
[17:07] <pitti> lamont: ck-list-sessions is empty, I guess?
[17:07] <slangasek> kagou: well, I think you want to check with the desktop team about their plans for who should be added to the sambashare group by default
[17:07] <lamont> ** (ck-list-sessions:12147): WARNING **: Failed to get list of seats: Failed to execute program /usr/lib/dbus-1.0/dbus-daemon-launch-helper: Success
[17:08] <pitti> -rwsr-xr-- 1 root messagebus 259016 2008-02-29 10:25 /usr/lib/dbus-1.0/dbus-daemon-launch-helper
[17:08] <pitti> lamont: just gotta love stuff that perror()'s errno==0...
[17:10] <pitti> anyway, I need to run, cu!
[17:10] <lamont> no worries
[17:12] <slangasek> cody-somerville: great, publising
[17:30] <kagou> slangasek, ok
[17:31] <kagou> slangasek, can we talk about including "map to guest = bad user" in smb.conf ?
[17:32] <slangasek> kagou: we can, but we should get input from the server team as well on that :-)  (zul, mathiaz, or soren, most likely)
[17:32] <kagou> ok :)
[17:32] <slangasek> I can't think of any reason that it wouldn't be a reasonable default behavior, but the server team may have another opinion
[17:33] <kagou> as nautilus-share will be included bu default, and shares-admin removed, usershare will be used
[17:33] <seb128> hey slangasek
[17:33] <slangasek> seb128: hiya
[17:33] <seb128> slangasek: so there is an another issue with the shares-admin to nautilus-share switch
[17:34]  * slangasek braces himself :)
[17:34] <seb128> slangasek: what happens to users who configured shares using the old way once they upgrade?
[17:34] <seb128> slangasek: they will have no way to unshare those
[17:35] <zul> slangasek: meh...how secure is it?
[17:35] <slangasek> seb128: hmm.  I'm not sure how avoidable that is, how would you then distinguish shares that were added using shares-admin from those that were added by hand?
[17:36] <slangasek> zul: map to guest = bad user?  All that does is say that if the client provided an unknown username, the connection is automatically mapped to guest access
[17:36] <seb128> slangasek: well, worded differently "do we consider that as an issue and have a migration plan or just write something in the upgrade notes about it"?
[17:36] <slangasek> zul: not really a security issue, it's just a question of choosing between two sometimes-awkward behaviors
[17:37] <slangasek> seb128: I think "upgrade notes" is best here - I don't see that there's any safe migration path, you can't just migrate all the shares to usershares without the risk of suddenly making some shares editable by users that weren't meant to have access to them
[17:38] <seb128> slangasek: alright, what I though, I was just checking
[17:38] <zul> slangasek: I dont really have a problem with it but you might want to check with mathiaz, soren is off
[17:41] <mathiaz> slangasek: which scenarios do you refered to when you say "two sometimes-awkward behaviors" ?
[17:42] <kagou> after many tests, with xp clients and ubuntu clients i can say that "map user guest  = bad user" is a definitively great enhancement on sharing folders/informations with samba under Ubuntu
[17:43] <slangasek> mathiaz: if you map to guest on bad user, some clients won't give you any straightforward way to undo a mistyped username
[17:43] <slangasek> mathiaz: if you *don't* map to guest on bad user, some clients won't give you any straightforward way to choose guest access :)
[17:44] <kagou> hacking nautilus-share to prompt for smbpasswd will be the last step to have a fully working samba share system since ...  euh ... we never had a fully working samba share system ^^
[17:44] <kagou> sladen, if you map... wich clients are you talking about ?
[17:45] <kagou> sorry i mean slangasek
[17:45] <mdz> pitti: thanks!
[17:45] <slangasek> kagou: off the top of my head, I don't remember; some versions of Windows, though
[17:46] <kagou> not xp clients, i'v tested. wrong user/pass don't allow to enter in non guest shared folders
[17:46] <kagou> so each time xp clients try to enter in it , user+pass is asked
[17:46] <kagou> until good login is provided
[17:46] <slangasek> kagou: er, the problem is what happens when you connect this way to a share that *does* allow public access, but you want to connect as an authenticated user
[17:47] <mathiaz> slangasek: isn't the authentication process decoupled from authorization in smbd ?
[17:47] <seb128> kagou: so share have different access mode depending on whether you are guest or using an user
[17:47] <seb128> s/so/some
[17:47] <slangasek> mathiaz: in what sense do you mean?
[17:48] <kagou> slangasek, using usershare, and allowing guest cause no problem to enter in it with xp client 'loged or not)
[17:49] <slangasek> kagou: connect to a public share from XP; type in a wrong username; get mapped as a guest; now change the user you're connected as so that you can edit files.  How do you do this?
[17:49] <slangasek> how do you correct a typo'ed username when samba agrees to map your access to guest?
[17:50] <mathiaz> slangasek: I meant that authentication uses the share your want to access to check whether you can connect as anonymous ?
[17:51] <kagou> slangasek, allowing writable share is done with usershare for all the world (guest) or require login. If this is requiring login, you can't enter in it without providing user+pass inforùations
[17:53] <kagou> folders protected by password are not browsable by guest
[17:53] <mathiaz> slangasek: reading smb.conf and the map to guest option
[17:54] <mathiaz> slangasek: the paragraph after the list of options for map to guest
[17:54] <slangasek> mathiaz: yes; the authentication is done, and you're either authenticated as a user or mapped to a guest or given a "wrong password" error; and then if you're authenticated, samba then checks the share permissions and gives you a permission denied if you don't have access
[17:54] <slangasek> kagou: oh, true, I didn't think about that limitation of usershares
[17:54] <seb128> slangasek: right, but we don't provide a way to configure shares to allow password protected and anonymous login and act differently anyway
[17:55] <kagou> seb128, actualy shares-admin neither (it's buged) ;)
[17:55] <slangasek> kagou: well, it does cause problems for clients in the *general* case, when you need more advanced permission controls than what nautilus-share provides access to
[17:56] <mathiaz> slangasek: so now - if we enable Bad User by default - and someone creates a non-anonymous share - a client connects to the server and gives a wrong password, he will be mapped to anonymous rather than being asked for a new password ?
[17:56] <slangasek> seb128: yes.  So I agree this is not an obstacle for the current implementation
[17:56] <slangasek> mathiaz: "map to guest = bad user" means this only happens if they provide a wrong *username*, not just a wrong password
[17:57] <seb128> mathiaz: if anonymous is authorized
[17:57] <slangasek> mathiaz: and anyway, when they're mapped to anonymous, they still won't be able to connect if the share isn't public, so the client will handle this permission error by asking for a different username/password
[17:58] <slangasek> mathiaz: so I concur with kagou that, in the case of nautilus-share with its limited ACL interface, "map to guest = bad user" is not a UI problem
[17:59] <slangasek> cody-somerville: images are published, I leave it to you to announce where you think it's appropriate (I'm not going to send a separate mail to ubuntu-announce for these two images)
[17:59] <cody-somerville> aye
[18:04] <kagou> slangasek, mathiaz adding this line will close Bug #32067
[18:04] <ubotu> Launchpad bug 32067 in samba "public Samba SMB shares cannot be accessed anonymously from Windows XP, a password prompt appears" [High,In progress] https://launchpad.net/bugs/32067
[18:05]  * slangasek nods
[18:06] <azeem> slangasek: so, what about opensync, will it get dropped from hardy?
[18:06] <azeem> who will make a decision?
[18:07] <slangasek> azeem: what do you recommend?
[18:07] <slangasek> (I have a personal bias here, so may defer to another member of the release team for the final decision)
[18:07] <mathiaz> slangasek: hum... I don't see any problem with setting map to guest = bad user
[18:08] <azeem> I never tried kitchensync, but Mandriva seems to think it's good enough as gui
[18:08] <azeem> in any case, I don't think hardy should ship with 0.19
[18:08] <azeem> if it's not possible to update to 0.22, I'd recommend dropping it
[18:09] <slangasek> azeem: do you think updating to 0.22 is reasonable?
[18:09] <slangasek> I don't have a bias wrt dropping 0.19 since it's not useful to me, but I do want 0.22 in :-)
[18:11] <azeem> I didn't get any more feedback about the PPA packages, so it's really hard to say.  Windows Mobile devices won't work unless a miracle happens, that's for sure
[18:11] <calc> grr i can't even get to the OOo bug tracker :\
[18:12] <kagou> mathiaz, Glad to hear it ^^
[18:13] <calc> i need someone that knows how to use bzr
[18:13] <calc> i am getting weird stuff
[18:14] <james_w> calc: can I help?
[18:14] <calc> james_w: hopefully :)
[18:14] <azeem> slangasek: AFAIK, the kitchensync currently in hardy is not the one used by Mandriva, but I'm not sure
[18:14] <calc> james_w: i'm trying to use bzr in my build tree for OOo and it is claiming that i have removed files, but i didn't use bzr rm to remove them
[18:14] <calc> james_w: it was due to uuencoding for diff.gz
[18:15] <azeem> slangasek: ah no, it apparently is
[18:15] <calc> james_w: when i went back to uudecode them to commit for bzr it now claims i have removed them and that they are also in unknown status as well
[18:15] <calc> james_w: so is bzr just too smart for its own good?
[18:15] <slangasek> azeem: the GUI doesn't really interest me personally... obviously it's interesting to a wider end user audience if there's a GUI, but there's no way kitchensync is going to be part of the desktop task or anything
[18:15] <azeem> slangasek: of course
[18:16] <calc> james_w: from what i recall with subversion you have to explicitly tell it to remove stuff from your tree, eg svn rm foo, but it seems bzr wants to do it anytime it sees a file disappeared
[18:16] <james_w> calc: yes, bzr does autoremove files, there's some debate about whether it should
[18:17] <james_w> calc: this isn't an svn checkout is it?
[18:17] <kagou> slangasek, seb128 will contact author of nautilus-share to see if he can add a smbpasswd gui to his tool.
[18:17] <james_w> calc: but as to why you get removed and unknown, I'm not sure.
[18:17] <kagou> can we hope to have map to guest in smb.conf soon ?
[18:18] <slangasek> azeem: so I think the real questions are: 1) does 0.22 meet the use cases that 0.19 is currently satisfying? 2) what does 0.22 give us over 0.19? 3) how much work is it going to be to get the polish on the integration done to the point where 0.22 packages are a satisfactory replacement for hardy?
[18:18] <slangasek> zul: maybe you'd like to do the upload for the 'map to guest' change?  so you can see how slick the new ucf handling really is ;)
[18:18] <calc> james_w: no its a bzr checkout
[18:18] <zul> slangasek: sure :)
[18:18] <slangasek> kagou: meh, I don't think nautilus-share should have a smbpasswd gui, I think we need a fix so that smbpasswd is integrated with PAM by default
[18:19] <calc> james_w: it autoremoved them automatically :( then i uudecoded the removed files to have it check them back in since some changed and it now thinks they are removed and unknown
[18:19] <calc> james_w: so you autoremoving files is definitely a really annoying feature for packaging use
[18:19] <james_w> calc: does bzr revert file and then make it have the one you want fix it?
[18:19] <calc> s/you/yea/
[18:20] <azeem> slangasek: part of the problem is that my harddisk space is pretty low so I don't have a hardy installation currently to properly test this.  I'll try to remedy this on the weekdn
[18:20] <calc> james_w: that probably will work, i would imagine, but is really annoying since i will have to do that after every rebuild
[18:20] <james_w> calc: i.e. revert file will tell you there is a conflict I think, and then you can pick the one you want and delete the other, it should then work it out
[18:20] <calc> james_w: since the binary files are removed on debian/rules clean
[18:20] <calc> iow i would spending more time on reverting files than actually packaging
[18:20] <james_w> calc: it's definitely a bug that it doesn't see them once you recreate them. Could you file a bug with the relevant parts of ~/.bzr.log please?
[18:21] <kagou> sladen, you mean that samba passord will be the same as user password ?
[18:21] <calc> so i think i will have to go back to only using bzr for final check in of changes for an upload
[18:21] <kagou> arggl i mean slangasek
[18:21] <azeem> slangasek: 1) 0.19 packages do not have a reasonable GUI either, and I am not even sure kitchensync works with it; besides 0.22 is considered very much superior by most people including upstream it seems
[18:21] <slangasek> azeem: it would probably be helpful if you were to file a FFe using the process at https://wiki.ubuntu.com/FreezeExceptionProcess
[18:22] <slangasek> kagou: yes, I believe that should be the goal
[18:22] <calc> james_w: i rm'd the .bzr.log uudecoded the files then ran bzr stat and it doesn't really give anything useful in the log file
[18:23] <calc> james_w: do i need to do something to make it more verbose?
[18:23] <kagou> slangasek, oh, ok but since that (i don't think for hardy) if nautilus-share test for existence of samba password and prompt for one if it don't exist, this will be a nice work around before integration in PAM
[18:24] <james_w> calc: "bzr -Dhashcache stat" is the only debug flag that seems vaguely useful.
[18:24] <calc> james_w: returns nothing useful in .bzr.log
[18:24] <slangasek> kagou: but how is the nautilus-share UI supposed to know *which* user needs an smbpasswd entry?
[18:25] <calc> james_w: its probably easily reproducible though so i can just document how to cause the problem
[18:25] <slangasek> kagou: if you share it as "allow other users", any user on the system could be the one that you need the password for
[18:25] <james_w> calc: please do, a simple add rm touch test here doesn't show it.
[18:25] <alex-weej> is Fx3 going to be in the final Hardy?
[18:30] <kagou> slangasek, a shared folder, by usershare, requiring password, is only accessible with the user+pass from the user who have shared the folder. Only one user+pass is allowed to enter in it.
[18:30] <slangasek> kagou: er, no, that's not what the "allow other people to write in this folder" flag is supposed to mean
[18:31] <kagou> it's not allow other, it's allow guest or not, if not its user+samba pass
[18:32] <slangasek> kagou: you're wrong, I've just tested a usershare here connecting as a different user
[18:32] <zul> kagou: which lp bug # is that for the bad user stuff?
[18:32] <slangasek> bug #32067
[18:32] <ubotu> Launchpad bug 32067 in samba "public Samba SMB shares cannot be accessed anonymously from Windows XP, a password prompt appears" [High,In progress] https://launchpad.net/bugs/32067
[18:33]  * alex-weej has given away his login to many Windows XP machines because of that :P
[18:35] <kagou> slangasek, from another user, with it's login+pass, different than the one who'v created the share ?
[18:35] <slangasek> kagou: yes
[18:35] <kagou> wow...
[18:35] <kagou> bug or feature of samba
[18:37] <slangasek> feature
[18:38] <slangasek> to the extent that there's a bug, it's a bug of nautilus-share
[18:38] <slangasek> because nautilus-share is what populates the ACL setting within the config, and is setting it to "Everyone:R" or "Everyone:F"
[18:39] <calc> bug 207891
[18:39] <ubotu> Launchpad bug 207891 in bzr "bzr thinks files are removed and unknown at the same time" [Undecided,New] https://launchpad.net/bugs/207891
[18:40] <calc> there's my bzr eats OOo bug ;-)
[18:40] <slangasek> talk about indigestion
[18:43] <james_w> calc: do the uuencoded files have the same names as the binary ones?
[18:44] <calc> james_w: they have .uu appended to the end
[18:44] <calc> james_w: once uudecoded they have the same name as original
[18:44] <calc> so i think it probably will need to have the smarts to realize if an autoremoved file reappears to take it out of auto removed status automatically
[18:45] <james_w> that's what it should do
[18:45] <calc> preferably it would be possible to turn autoremoval off as well
[18:45] <james_w> and what it does in my testing, so I'm trying to work out what is different in your setup
[18:45] <calc> but that doesn't affect me on OOo as long as this other part is fixed
[18:45] <calc> ok i'll manually revert the removals and then uudecode and see it can make it happen again
[18:46] <zul> kagou: uploaded
[18:47] <james_w> calc: thanks, the .bzr.log from that may still be useful, even if it doesn't look so to you, so please attach it
[18:48] <calc> james_w: ok
[18:49] <calc> interesting this time it worked
[18:49]  * calc wonders what is going on
[18:50]  * calc hates non-repeatable bugs
[18:50] <james_w> yup
[18:52] <calc> i closed out my bug until i can make it happen again
[19:06] <kagou> zul, thanks :)
[19:19] <\sh> infinity: what do you need for enabling wine for building on lpia? dollars, beer, buildlogs? wine runs on lpia :)
[19:19] <\sh> at least our package ;)
[19:27] <asac> james_w: there?
[19:29] <infinity> \sh_away: Can you mail me a reminder about it?
[19:39] <lamont> jdstrand: 207912
[19:43] <kagou> slangasek, mathiaz  and zul , thanks for closing bug #32067
[19:43] <ubotu> Launchpad bug 32067 in samba "public Samba SMB shares cannot be accessed anonymously from Windows XP, a password prompt appears" [High,Fix released] https://launchpad.net/bugs/32067
[19:43] <zul> kagou: np
[19:43] <kagou> Good night
[19:43] <DktrKranz> Mithrandir, bluez-gnome has a couple of irritating bugs which are fixed by new upstream version. Do you plan to request it in Hardy?
[19:44] <jdstrand> lamont: thanks
[19:47] <james_w> asac: yep
[19:48] <asac> james_w: i wonder if there is a simple fix to make bzr bd automatically fetch an orig and put it to the tarballs dir :)
[19:49] <james_w> asac: there is, write a watchfile :-)
[19:49] <asac> no... i mean from the archive
[19:50] <james_w> it should do that as well
[19:50] <calc> doko: i'm uploading the preliminary stuff for you to take a look at
[19:50] <asac> james_w: what do you mean?
[19:51] <james_w> asac: you are using a recent version?
[19:51] <asac> here i am using the one from the archive ... but let me check
[19:51] <asac> looks like 0.92
[19:53] <james_w> asac: the changelog for 0.92 says "* Also look for upstream tarballs in the archives. Do this in preference to the watch file, for the case where the upstream was repacked."
[19:54] <james_w> so it should do it. you're not using export-upstream mode are you?
[19:54] <asac> james_w: no ... i am just running bzr bd --merge
[19:55] <alex-weej> are extended attributes enabled on ubuntu systems by default these days?
[19:56] <james_w> asac: will 'apt-get source -y --tar-only package=upstream-version' fetch the tarball without an error code in your case?
[19:56] <asac> trying
[19:59] <asac> james_w: no ... neither apt-get source -y --tar-only  firefox-2=2.0.0.13+1nobinonly .... nor apt-get source -y --tar-only  firefox=2.0.0.13+1nobinonly work
[19:59] <asac> firefox is the source package name
[19:59] <asac> firefox-2 the binary package name
[19:59] <asac> only firefox-2 works for plain apt-get source here
[19:59] <asac> (well ... firefox will give you firefox-3.0 source)
[20:00] <asac> the upstream version looks correct
[20:00] <james_w> so it tries to use the source package name
[20:01] <asac> thats wrong in the first place i guess ... and using upstream-version doesn't work as well :/
[20:01] <james_w> is this a problem with apt where it will prefer getting the source for a binary package of that name over the source package of that name?
[20:01] <asac> james_w: i am not sure. does apt-get source SOURCE_PKG_NAME work at all?
[20:02] <james_w> ah, no, sorry, I misread, it uses the binary version of the first package it finds that has the same upstream version
[20:02] <asac> yes it does
[20:02] <asac> hmm
[20:02] <asac> that should work
[20:02] <asac> but nothing happens here :(
[20:04] <james_w> (cd /tmp && apt-get source -y --tar-only firefox=2.0.0.13+1nobinonly-0ubuntu1) works for me it seems
[20:04] <james_w> Get: 1 http://archive.ubuntu.com hardy/universe firefox 2.0.0.13+1nobinonly-0ubuntu1 (tar) [34.9MB]
[20:04] <asac> yes ...  but with full-pkg-version .. not upstream-version
[20:05] <james_w> that's what it should be trying to use, unless I misunderstand
[20:06] <james_w> I didn't actually write this code
[20:07] <asac> james_w: https://code.edge.launchpad.net/~asac/firefox/ubuntu-2.0.0.x
[20:07] <asac> that branch doesn't work for me
[20:07] <james_w> asac: you might like to do (mkdir -p .bzr-builddeb && echo -e "[BUILDDEB]\nmerge = True" > .bzr-builddeb/default.conf && bzr add .bzr-builddeb && bzr ci -m"Make merge mode the default for the package")
[20:08] <james_w> thanks, I'll grab that and have a look. Off to the pub now though, so it will be tomorrow probably, sorry.
[20:08] <asac> thx
[20:09] <LaserJock> bryce: congrats, about time ;-)
[20:11] <bryce> LaserJock: thanks
[20:13]  * slangasek shakes his fist at people trying to induce pitti to implement package-specific apparmor hacks for things that belong in the apparmor abstractions
[20:14] <slangasek> jdstrand: new apparmor bug for you, then :)
[20:15] <jdstrand> slangasek: you referring to cupsys?
[20:15] <jdstrand> bug #207912 ?
[20:15] <ubotu> Launchpad bug 207912 in cupsys "apparmor profile needs additions for nsswitch.conf" [Medium,Invalid] https://launchpad.net/bugs/207912
[20:15] <slangasek> jdstrand: I'm referring to the bug about nss backends that for some reason ended up assigned to cupsys, yes
[20:16] <slangasek> this is not cupsys-specific by any means, it should be part of the nameservice abstraction
[20:17] <jdstrand> slangasek: I haven't looked at the cupsys profile yet, but figured that was all I was going to do to it
[20:17] <jdstrand> slangasek: so you saved me a couple minutes ;)
[20:17] <slangasek> heh, k :)
[20:17] <slangasek> yeah, cupsys's profile is still correct, in this regard - it's just that apparmor is missing a few files in the nameservice abstraction
[20:17] <slangasek> for the db and ldap backends
[20:17]  * jdstrand nods
[20:23] <siretart> is some uptodate d-i documentation for hardy online somewhere?
[20:23] <calc> doko: done uploading now
[20:24] <LaserJock> siretart: actually the doc team would like to know that too :-)
[20:24] <siretart> :/
[20:24] <LaserJock> all we have on help.ubuntu.com is a link for Edgy
[20:25] <siretart> that isn't really up-to-date... hmm
[20:25] <calc> anyone know how to find out what to put in /usr/lib/mime/packages/foo for a given example document?
[20:26] <asac> siretart: for bug 50214 it looks more and more like it would be a blessing to wpa_drv_set_ssid _before_ doing the scanning in ap_scan 2 mode. you see any issues with that?
[20:26] <ubotu> Launchpad bug 50214 in network-manager "can't connect to hidden network" [High,Confirmed] https://launchpad.net/bugs/50214
[20:26] <asac> siretart: (in wpasupplicant)
[20:26] <calc> also in /usr/share/applications/foo.desktop MimeType= ?
[20:26] <calc> i need to update the types and i have example files but i don't know what to run against them to find out the official name of the MimeType
[20:27] <asac> siretart: at least iwl drivers don't find any bssid when scanning a hidden network if the essid has not been set in advance
[20:28] <asac> and i get the feeling that its the case for more
[20:29] <siretart> asac: I'd be curious if that also happens with wpasupplicant 0.6.3 (which can be found in my PPA, and happens to work just fine with ath5k and WPA-PEAP, but I haven't tested hidden ssids, though)
[20:29] <asac> siretart: this is just for hidden
[20:30] <siretart> right.
[20:30] <asac> siretart: and given that i had similar "hidden" complains with my nm 0.7 + wpasupp 0.6.x combo
[20:30] <siretart> in general I recommend users playing with the ap_scan parameter
[20:30] <asac> i doubt that its fixed
[20:31] <siretart> default is 1, users have reported success with setting ap_scan to 2
[20:31] <asac> siretart: which comment do you refer to?
[20:31] <siretart> however AFAIK, nm doesn't offer a gui for setting the ap_scan parameter
[20:31] <asac> siretart: the ap_scan is not a problem
[20:31] <siretart> it isn't?
[20:32] <asac> the problem is that ap_scan 2 doesn't even succeed if to find the bssid if you don't do a iwconfig essid XXXX before
[20:32] <asac> and wpasupplicant doesn't do that (from code)
[20:32] <asac> currently i forcefully do that in NM for some chipsets, but i'd prefer to move that to wpasupplicant as it looks more and more like ap_scan 2 always needs that
[20:33] <asac> (in gutsy i did that for _every_ driver and thats how i explain that there are still users reporting regressions)
[20:33] <asac> siretart: from what i see it would be one line in wpa_supplicant.c :)
[20:35] <siretart> asac: you mean in wpa_supplicant_scan()?
[20:35] <asac> YES
[20:35] <asac> oops
[20:37] <wasabi> launchpad_integration_add_bonobo_ui(0x6f3f00, 0x419b95, 8, 11, 0**
[20:37] <wasabi> ** Gtk:ERROR:(/build/buildd/gtk+2.0-2.12.9/gtk/gtktoolbar.c:3021):IA__gtk_toolba
[20:38] <wasabi> suspecting that's a problem. =/
[20:38] <wasabi> Explains why most of my apps are crashing, anyways!
[20:42] <siretart> asac: I don't see concrete problems with that. as a naive guess, I think a short sleep to settle things down on 'strange' drivers wouldn't be a bad idea. However, I'm not too experienced with that area of wpasupplicant, TBH
[20:43] <asac> yeah ... i have to think about it as well
[20:50] <alvarezp> Hi. I've noticed that some fonts in Hardy don't show anymore under xfontsel. Bitstream Vera Sans is affected. Gutsy Gibbon worked correctly. Where is this configured, or what package takes care of this? I'd like to file a good bug report.
[21:13] <keescook> lamont: 207912> /var/lib/misc ?  that's goofy.  ldap is already handled in the nameservice abstraction -- which paths should be added?
[21:14] <slangasek> keescook: according to the bug report, /etc/ldap.conf (not just /etc/ldap/ldap.conf)
[21:14] <slangasek> er, or the other way around
[21:14] <slangasek>   # all openldap config
[21:14] <slangasek>   /etc/openldap/*         r,
[21:14] <slangasek> keescook: ^^ buggy, wrong path
[21:15] <slangasek> and /var/lib/misc isn't goofy, it's the longstanding path for the db backend
[21:15] <keescook> slangasek: well, it may be the path, but it's still goofy.  :)
[21:15] <slangasek> well, ok. :)
[21:15] <slangasek> blame BSD!
[21:16] <keescook> slangasek: what's the expected file mask?  /var/lib/misc/*.db ?
[21:16] <jwendell> slangasek, any news on bug 201440?
[21:16] <ubotu> Launchpad bug 201440 in anjuta "Please sync anjuta (universe) 2:2.4.0-2 from Debian unstable (main)" [Wishlist,Confirmed] https://launchpad.net/bugs/201440
[21:17] <slangasek> keescook: yes
[21:17] <slangasek> jwendell: none from me, but I'm not on archive duty today
[21:36] <salty-horse> shouldn't this be a priority? for some production systems, those 7 minutes are important: https://bugs.launchpad.net/ubuntu/+source/ntp/+bug/157608/
[21:36] <ubotu> Launchpad bug 157608 in ntp "Adjust Time -> Sync with Internet Time Servers never syncs." [Undecided,Confirmed]
[21:37] <crevette> salty-horse: do you use Network-manager with Wifi ?
[21:37] <salty-horse> no. I'm connected directly to the router
[21:37] <salty-horse> (cable)
[21:37] <crevette> (I'm reading the bug)
[21:38] <crevette> I have this, I thought it was sometyhing with ntpupdate ruinning too early in the stack
[21:40] <dmb> is oo going to be in hardy?
[21:40] <dmb> oo 2.4 i mean
[21:40] <seb128> crevette: ntp and ntpdate are different things
[21:40] <xivulon> slangasek, is anything moving re metalinks?
[21:43] <Fujitsu> Is Nautilus meant to bug me to install Samba every time I want a directory's properties window?
[21:46] <slangasek> xivulon: mmm, I've been waiting for the first cut of the code to populate the mirror list, let me talk to the webmaster again
[21:46] <seb128> Fujitsu: no, that would be a bug in the nautilus-share patch mvo did
[21:47] <Fujitsu> seb128: OK, thanks, I'll look for a bug.
[21:47] <crevette> seb128: i'm puzzled because bother are called in the same script (/etc/network/if-up.d/ntpdate)
[21:48] <Fujitsu> Anybody know of a bug about Nautilus failing to copy directories to/from SFTP shares after a while? Often after finishing one directory, the SFTP thing will hang, eventually giving a DBUS error.
[21:48] <seb128> crevette: what this one does it stopping ntp, calling ntpdate and starting ntp again
[21:48] <crevette> seb128: ah yeah, them can't be run at the same time, I had the issue on RH and solaris
[21:48] <seb128> Fujitsu: yes, several useless bugs with similar not clear descriptions
[21:49] <Fujitsu> seb128: Lovely.
[21:49] <seb128> Fujitsu: that's likely fixed in gvfs svn, a new tarball should be rolled soon
[21:50] <seb128> Fujitsu: but since nobody managed to describe how to trigger the issue and it doesn't break on my installations I didn't bother trying
[21:50] <xivulon> slangasek, if you want to you can use static metalinks just pointing to cdimage daily for the time being and replace them later
[21:51] <Fujitsu> seb128: I've been able to reproduce it for a couple of weeks on my laptop. I just reinstalled (I managed to destroy a couple of filesystems will stuffing around with LVM), and am able to reproduce it again.
[21:51] <Fujitsu> There's no easy way to trigger it; sometimes copies just don't start, sometimes they hang after finishing one directory (in the same place each time)...
[21:51] <seb128> Fujitsu: what are the steps to trigger the issue then?
[21:52] <slangasek> xivulon: well, tbh I haven't even learned how to write a valid metalink file because writing the generation code has been delegated, so I probably can't sanely create a static metalink file on my own either :)
[21:53] <xivulon> you can copy the ones in http://wubi-installer.org/metalinks/8.04 , they are a bit lean (no md5, 1 mirror)
[21:54] <slangasek> xivulon: thanks, I'll try to push that out later this afternoon then
[21:54] <xivulon> also there is a link in the bug for a couple of scripts and if you need any help I can contact the devs in the metalink forum (one of them replied in the comments)
[21:57] <bryce> ppa question - I'm trying to put older versions of the -intel driver into my ppa for users to test and help narrow down which upstream patches caused issues
[21:58] <bryce> however, when uploading an earlier version, I got a rejection:
[21:58] <bryce> Rejected:
[21:58] <bryce> xserver-xorg-video-intel_2.2.0+git20080318.ac763634-bwh1.dsc: Version older than that in the archive.
[21:58] <bryce> 2:2.2.0+git20080318.ac763634-bwh1 <= 2:2.2.1-1ubuntu3
[21:59] <bryce> do I need to fake ppa out by numbering it from the latest version or something?  (seems kludgy) or is there a way to force ppa to allow old packages?  Or should I not be using ppa in the first place?
[21:59] <TheMuso> bryce: Do you have an earlier version in your PPA already?
[21:59] <Fujitsu> bryce: Increment the epoch, perhaps. PPA version-ratchets like primary, though you can remove packages.
[22:00] <bryce> I'd built an earlier version a month ago, but deleted it; no difference
[22:02] <TheMuso> hrm
[22:03] <bryce> oh well, guess I'll go with the epoch bump.  that's less kludgy than futzing the version number
[22:03] <slangasek> but also less reversible :)
[22:05] <Fujitsu> slangasek: It's a PPA. They're erasable.
[22:06] <slangasek> Fujitsu: not from the systems of those who've installed it <shrug>
[22:06] <bryce> slangasek: yeah that's my worry... but I can just document around it
[22:06] <Fujitsu> slangasek: They'd be downgrading packages otherwise; I'm sure they can install the primary archive ones again.
[22:06] <mjg59> bryce: Hey - any idea if the Poulsbo kernel code is able to bring the video back up itself after S3?
[22:07] <slangasek> Fujitsu: they'll have to downgrade packages again once the official fix is released to the archive
[22:08] <bryce> mjg59: heya.  Not offhand.  (I'm also not sure how to test S3 on the test equipment)
[22:08] <Fujitsu> slangasek: I meant that they also had to downgrade to the PPA package in the former case. THey have to downgrade either way, so I don't see what's so terrible about having to downgrade.
[22:09] <bryce> slangasek: yeah I'd sort of prefer they had to downgrade to install my version, so if after testing they forget to restore to the original (or *gasp* prefer to use the old version), they won't get updates again after that, and it may not be clear why they're not
[22:09] <mjg59> bryce: Ok. Do you know who would know? :)
[22:09] <bryce> mjg59: amitk would probably know
[22:09] <slangasek> Fujitsu: one can conceivably rig a version number that won't require downgrades either way, if you ship all the git bisecting bits in a .diff.gz ;)
[22:09] <mjg59> bryce: Just chatted - I don't think he did
[22:10] <Fujitsu> slangasek: True, but that sounds bad.
[22:10] <bryce> mjg59: hrm... then next would be Calvin, via David Mandala and Don Johnson
[22:10] <mjg59> Ok
[22:11] <slangasek> Get:2 http://mirrors.cat.pdx.edu hardy/universe kde-icons-oxygen 4:4.0.2-0ubuntu2 [45.4MB]
[22:11] <slangasek> what a hateful build-dependency
[22:11] <Fujitsu> How is an icon set 50MB?
[22:11] <RAOF> How is an icon set a build dependency?
[22:11] <bryce> Calvin is the tungsten graphics rep.
[22:11] <slangasek> why is it a build-dependency? :)
[22:11] <Fujitsu> How is an icon set?
[22:14] <seb128_> re
[22:14] <seb128_> I was saying
 bryce: I would not use an epoch
[22:14] <seb128_>  bryce: use current-version.is.really.version rather
[22:14] <seb128_>  bryce: so they will get next upgrades
[22:16] <bryce> wow, this really reduces ppa's usefulness
[22:17] <seb128_> I think that's a bug
[22:17] <seb128_> ask to cprov-out about it
[22:17] <seb128_> or I would consider it a bug rather
[22:19] <TheMuso> Hrm whatever is responsible for managing brightness etc is not showing the little graphic when I adjust the screen brightness on my notebook like it used to...
[22:19] <mjg59> TheMuso: What hardware?
[22:20] <TheMuso> mjg59: Thinkpad R50.
[22:20] <TheMuso> no wait, an update has fixed it.
[22:20] <mjg59> Ha, excellent
[22:20] <LaserJock> easy fix ;-)
[22:21] <LaserJock> mine just keeps going up and down if I want to manually set the brightness
[22:22] <mjg59> LaserJock: What hardware?
[22:22] <LaserJock> it's an HP Pavilion dv6000ish with mostly all intel hardware
[22:23] <LaserJock> I assume it's the "normal" behavior
[22:23] <mjg59> Ah, there's one fix I've pushed for the next kernel that may help there
[22:24] <LaserJock> so if I manually set the brightness with special keys or applet it should stay there?
[22:24] <mjg59> Yes
[22:24] <LaserJock> that's handy, I just thought it was a missing feature
[22:24] <seb128_> what do you call stay there?
[22:25] <LaserJock> well, like if it resets the the gnome power management settings
[22:25] <LaserJock> *to the
[22:25] <mjg59> Oh, I see
[22:26] <mjg59> If you're idle, it'll eventually dim it if you have that box checked
[22:26] <seb128_> the dim is normal
[22:26] <LaserJock> I would have thought that manually changing the brightness would override other settings
[22:27] <seb128_> what is a bug is that changing the backlight value using the key doesn't update the stored value
[22:27] <slangasek> how about those of us who don't have the 'dim when idle' box checked but still get dimming? :)
[22:27] <seb128_> or rather than it doesn't back to whatever you were using when you are back
[22:27] <slangasek> oh, correction, I do have it checked
[22:27] <slangasek> teehee
[22:28] <LaserJock> so if I uncheck the "dim when idle" then it should just stay?
[22:29] <slangasek> what's frustrating to me is that I don't have control over the degree of dimming; so if it dims when I'm on battery, and I manually adjust it, then it dims for idle, then I unidle, I have to adjust again
[22:30] <LaserJock> slangasek: exactly
[22:30] <seb128_> slangasek: right, that's what I wrote
[22:30] <seb128_> you have the control of the dim value though
[22:30] <slangasek> seb128_: ah, so you did :)
[22:30] <slangasek> seb128_: where?
[22:30] <seb128_> it just doesn't go back to whatever you were using
[22:31] <slangasek> I don't see anything under g-p-m that lets me control how much it dims
[22:31] <seb128_> gconf-editor, /apps/gnome-power-manager/backlight/idle_brightness
[22:31] <slangasek> ah, heh
[22:31] <slangasek> go go GNOME HIG
[22:31] <seb128_> I though that was somewhere in the dialog
[22:32] <slangasek> :)
[22:32] <mjg59> seb128_: Nope, just AC and battery defaults
[22:32] <seb128_> mjg59: that's the value it's using on idle
[22:33] <seb128_> it's not clear to me how it's supposed to work
[22:33] <seb128_> the applet updates the gconf keys
[22:33] <seb128_> where changing the backlight using the keyboard doesn't
[22:33] <mjg59> I'd guess the idle value is some function of the set values
[22:33] <mjg59> There's nothing in the dialog to let you change the idle value
[22:33] <seb128_> indeed
[22:33] <seb128_> I consider gnome-power-manager quite buggy
[22:33] <seb128_> and upstream is not really responsive
[22:33] <LaserJock> is that the wonderfully consistent dialog "Dim display brightness by" and "Set display brightness to"?
[22:35] <tjaalton> when it dims my display and I resume working, it dims it even more :)
[22:36]  * slangasek grins
[22:37] <TheMuso> bryce: BTW congrats.
[22:37] <bryce> TheMuso: oh thanks
[23:03]  * slangasek nods at the topic change and queues up his work locally :)
[23:06] <slangasek> asac: hmm, no n-m upload yet?  (maybe I've lost track of the days...)
[23:07] <seb128> slangasek: do you have an user friendly description of sambashare for gnome-system-tools? ;-)
[23:08] <slangasek> seb128: "Share files with the local network"?
[23:09] <slangasek> or maybe "Manage network file sharing"
[23:16] <seb128_> re
[23:16] <seb128_> hate my dsl provider
[23:16] <seb128_> slangasek: dunno if you received what I wrote
 works for me
[23:16] <seb128_>  thanks
[23:16] <seb128_>  you share files "with" the network and not "on" the network?
[23:19] <slangasek> seb128_: I believe either is grammatically valid, but "with" the network has nuances that remove some ambiguities
[23:20] <slangasek> did you see my alternative suggestion? "Manage network file sharing"
[23:20] <seb128_> no I didn't
[23:20] <seb128_> hum
[23:20] <slangasek> either is ok with me :)
[23:20] <seb128_> I think I like the Share files option better
[23:20] <slangasek> ok
[23:20] <seb128_> I'll use this one
[23:23] <slangasek> seb128_: hrm, any chance you can comment on bug #203328?  nautilus, tracker, and eog are all reverse-dependencies - do you know what the implications are of this upstream API change?
[23:23] <ubotu> Launchpad bug 203328 in exempi "[Sync request] exempi source package from Debian sid" [Wishlist,New] https://launchpad.net/bugs/203328
[23:25] <seb128_> slangasek: debian has the new version for over 3 weeks and we didn't get any nautilus or eog issue due to exempi that I know, they do a really limited use of it anyway
[23:25] <slangasek> ok
[23:25] <seb128_> slangasek: I would say it should be safe to sync, we didn't get any bug about the ABI breakage anyway
[23:26] <slangasek> +#define XMP_OPEN_OPNLYXMP XMP_OPEN_ONLYXMP
[23:26] <seb128_> so the ABI changed and fixed seems to not be used or create really used in those applications
[23:26] <slangasek> heh, that's the only change
[23:26] <slangasek> so that's entirely trivial
[23:27] <seb128_> right
[23:29] <slangasek> syncbugbot++
[23:30] <seb128_> ;-)