[12:49] <BenC> ogra: checking on 73336
[03:59] <Hobbsee> hey mpt 
[04:05] <mpt> hello Hobbsee 
[04:06] <alex-weej> guys
[04:06] <alex-weej> is the plan to put beryl and xgl in feisty?
[04:06] <alex-weej> or rather stick with metacity and aiglx?
[04:06] <alex-weej> i'm becoming a little suspicious that it's the former! :S
[04:09] <LaserJock> well, it wouldn't be xgl
[04:09] <LaserJock> perhaps beryl and aiglx with metacity also
[04:41] <mpt> BenC, what would be the feasibility of merging "Suspend" and "Hibernate"? Does code for such a merged function exist for Linux like it does for OS X and Vista?
[04:42] <mpt> (I wouldn't even know where to look, and Google's not helping)
[04:42] <BenC> mpt: I'm not sure what you mean by merging it
[04:43] <mpt> Suspending to disk, then to ram
[04:43] <mpt> So it's like suspend-to-RAM, but if you lose power, you don't lose your session
[04:43] <mpt> aka "Sleep" in Vista and "Safe Sleep" in OS X
[04:44] <HrdwrBoB> surely that would take significantly longer
[04:44] <HrdwrBoB> especially on laptops with slow hard drives?
[04:45] <mpt> yes, it does
[04:46] <Lathiat> mpt: suspend2 does that
[04:46] <Lathiat> if you dont lose power it wakes up pretty quick, else resumes from disk if it dies/gets turned off
[04:46] <Lathiat> works quite nice
[04:47] <Lathiat> (but we're not using suspend2)
[04:53] <mpt> but, HrdwrBoB, it removes one of the places where people have to understand the difference between "disk" and "memory", which they rarely do
[04:53] <HrdwrBoB> true
[04:54] <mpt> and it helps reduce the "dozen different ways to log out" embarrassment :-)
[04:56] <Burgundavia> mjg59 has been looking more into this. He is probably the best person to talk to 
[04:59] <mpt> ta
[05:05] <BenC> mpt: main thing is that suspend2 is way too much of a code delta for us to maintain in our kernel
[05:05] <BenC> mpt: The onus is on the suspend2 maintainers to work out the problems with their code base and get it merged into upstream
[05:05] <mpt> ah
[05:10] <fabbione> yeah these patches are just too intrusive
[05:10] <fabbione> mpt: we have been looking at suspend2 since warty or something :)
[05:41] <Burgundavia> mpt: do you get the UWN?
[05:43] <mpt> no
[05:43] <mpt> Burgundavia, why do you ask?
[05:43] <Burgundavia> mpt: I need your advice on something
[05:44] <Burgundavia> mpt: thoughts on the "changes in Feisty" section? https://wiki.ubuntu.com/UbuntuWeeklyNewsletter/Issue22
[05:47] <mpt> "an added helping in the form of the autosyncing of all the completely new packages" could be "an added helping of auto-synced completely new packages"
[05:48] <Burgundavia> mpt: not looking for specifics, just your comments on the general readability and interestingness to non-developers
[05:48] <mpt> It's Dr?ge, not Droege :-)
[05:49] <Burgundavia> stop analysing the small stuff
[05:49] <mpt> and ... ok ok
[05:51] <mpt> Well, I don't know the target audience for UWN
[05:51] <mpt> But for me, the lists of new package versions make my eyes glaze over
[05:51] <Burgundavia> right, thanks
[05:51] <mpt> That's my only non-misspelling-of-people's-names complaint
[05:51] <Burgundavia> what wouldn't make your eyes glaze over?
[05:52] <mpt> Hmm
[05:52] <mpt> Perhaps give a total number of packages, and say "including new versions of A, B, and C"
[05:52] <elkbuntu> mpt, so the numbers are frying your brain as well?
[05:52] <mpt> where A, B, and C are some of the more interesting ones
[05:53] <Burgundavia> elkbuntu: I read that section last week and cringed
[05:53] <elkbuntu> Burgundavia, i didnt read it. i glanced and gave up
[05:54] <Burgundavia> indeed
[05:57] <Mithrandir> Burgundavia: for the list, I would consider actually using a bullet list (or drop the version numbers).
[05:57] <Burgundavia> mpt: can you come to -marketing?
[05:58] <Burgundavia> I dropping version numbers in most cases is a good idea
[05:58] <Mithrandir> hiya Hobbsee :-)
[05:58] <Hobbsee> :)
[05:59] <Mithrandir> how did/are exams?
[05:59] <Mithrandir> | grammarcorrect
[06:00] <ajmitch> hi Mithrandir 
[06:00] <Hobbsee> Mithrandir: they're over :)
[06:00] <Mithrandir> hi Andrew
[06:00] <Mithrandir> Hobbsee: that's good.
[06:01] <Hobbsee> yup
[06:01] <Hobbsee> Mithrandir: i'm told "go to spain, we're going on a cruise"
[06:01] <Hobbsee> by the parents :P
[06:01] <Hobbsee> or wherever UDS next is
[06:02] <ajmitch> lucky you
[06:02] <ajmitch> get them to pay for it as well :)
[06:02] <Hobbsee> hah
[06:03] <Mithrandir> Hobbsee: yay. :-)
[06:03] <ajmitch> alright, vmware fixed, I have active directory back to play with :)
[06:04] <sfllaw> You make that sound like a pleasant thing.
[06:04] <sfllaw> I remember AD.
[06:04] <sfllaw> I had nightmares.
[06:04] <ajmitch> sfllaw: it's wonderful, isn't it?
[06:04] <Mithrandir> sfllaw: UDS is wonderful!
[06:04] <Mithrandir> it's insanity distilled into pure form and spread out over a whole week.
[06:04] <sfllaw> Mithrandir: Insanity?  That's my real life.
[06:05] <sfllaw> Only in a different city.
[06:05] <wasabi> hi
[06:05] <ajmitch> hello wasabi 
[06:05] <wasabi> AD rocks.
[06:05] <wasabi> just had to drop by to give my two cents.
[06:05] <ajmitch> wasabi must have his irc client highlighting on AD
[06:06] <wasabi> "directory" actually.
[06:06] <wasabi> Speaking of which, I'm reconstructing cyrus imap right now.
[06:06] <wasabi> And I'm pretty pissed and wish I was using Exchange instead. ;)
[06:06] <wasabi> take that!
[06:06] <ajmitch> it's that bad?
[06:06] <wasabi> it's just a lot of silly manual work.
[06:06] <wasabi> I upgraded to 2.2, which seems to require me to manually upgrade each BDB file.
[06:07] <wasabi> Which of course in fact corrupted each bdb file.
[06:07] <Mithrandir> I'm very, very wary of products using BDB
[06:07] <wasabi> Generally I am as well.
[06:07] <wasabi> But, ya know. It works. Barely.
[06:07] <wasabi> I suspect it's less BDB's fault these days and more the consumers.
[06:08] <wasabi> I have discovered yet another core lacking factor which will have to be overcome to "do what AD does".
[06:08] <Mithrandir> my current policy for bdb-using software these days is "if it just uses bdb for caching, I'm fine.  If it uses it for permanent storage, I'm not happy"
[06:08] <wasabi> Sasl has no concept of a unique immutable ID.
[06:09] <wasabi> Which basically means cyrus names the mailbox based on the user that logs in, not some unique property of said user.
[06:09] <Mithrandir> you seem to want to support user name renames, which I think is just plain silly.
[06:09] <wasabi> Whatever. AD does it.
[06:09] <wasabi> Novell does it.
[06:09] <wasabi> You try to tell a customer they can't do it, you get shut out.
[06:10] <wasabi> There's also nothing fundamentally wrong with it.
[06:10] <wasabi> If built right.
[06:11] <Mithrandir> *shrug*; it's adding yet another abstraction layer.  Add a user name => uuid mapping and carry the uuid around everywhere you use user names today.
[06:12] <wasabi> Yes, it is an abstaction. It can be argued it is useful though. Users type in their username. Typing in jhaltom is easier to remember than 23494843
[06:12] <lifeless> there are ways and ways of supporting user renames
[06:12] <wasabi> Just need to build some sort of authid->uniqueid mapping into all these various libs, and get people using them.
[06:13] <wasabi> Doesn't matter if a particular installation chooses a posix id or a uuid or a sid then.
[06:13] <Mithrandir> wasabi: yes, and if you want to support user renames, you then have to change the user name everywhere it's used, not really hard.
[06:13] <Mithrandir> that's an easier way to do it.
[06:13] <wasabi> Not exactly.
[06:13] <wasabi> Consider, oh, for instance, Google, my prime use case.
[06:13] <wasabi> Mostly because they came up to me at UDS and said it sucked.
[06:14] <wasabi> They have 44 kerberos KDCs. Same number of lDAP servers. For any given user, his name could be strewn across a hundred different servers.
[06:14] <wasabi> Apache authz files, crontab files, whatever.
[06:14] <wasabi> Consider AD then. You right click on the user and hit "rename". Done.
[06:14] <wasabi> Nothing breaks.
[06:15] <wasabi> Since everything was keyed on the SID to begin with.
[06:16] <Mithrandir> you're trying to change fairly fundemantal bit of how unix apps work here; I see a long and uphill battle in your future.
[06:16] <maswan> Well, for the directory (kdc,ldap) you just need to build a click-rename thingie just like AD has.
[06:16] <lifeless> wasabi: acrually, thats a vast handwave
[06:16] <maswan> For the rest, well, don't use usernames.
[06:16] <maswan> Use uids instead
[06:16] <wasabi> lifeless: Not really.
[06:16] <lifeless> wasabi: renames of the account name in AD do not rename everything 
[06:16] <wasabi> lifeless: They don't rename files named "bob" on the HD. But roaming profiles keep working.
[06:16] <lifeless> wasabi: for instance, the local profile is named on disk with the username, not the UUID. 
[06:17] <wasabi> Yes, but it's keyed by the SID.
[06:17] <lifeless> sure.
[06:17] <wasabi> The name of the folder is immaterial.
[06:17] <wasabi> Logins to new systems will have a new properly named folder created. Logins to old ones will contiue using the old folder.
[06:17] <wasabi> If another user assumes the old name, old systems will give him a new folder, name.DOMAINNAME.
[06:18] <wasabi> And some pattern thereof to generate a unique one.
[06:18] <lifeless> I know
[06:18] <lifeless> I've spent years of my life on that platform
[06:18] <wasabi> What it comes down to though is not assuming the username will rename, at all points.
[06:18] <lifeless> I've three key comments here, then you can return your previously scheduled rant.
[06:18] <wasabi> =)
[06:19] <lifeless> 1) its not a trivial change to unix to do this 'just like AD'
[06:19] <lifeless> 2) look at what suns directory server, or NDS for Unix do, as that is more relevant prior art IMO
[06:19] <wasabi> Hey, I recognize I'm whining on IRC. ;)
[06:20] <wasabi> 1) Of course. Sucks, doesn't it?
[06:20] <lifeless> 3) spec it up, get upstream community buyin - this is not something that it makes sense to drive within a distribution - the changes are too big, and too sweeping
[06:20] <lifeless> there, I'm done :)
[06:20] <wasabi> 2) Yeah. I need to look at those. I'm curious what they've done. At the same time, I recognize they don't have much market share, maybe for a reason? This one reason? No clue.
[06:20] <maswan> If you plan for it resonably, doing a username change is not that hard, IME.
[06:21] <wasabi> 3) That's the goal. 
[06:21] <maswan> If you use the username all over dozens of different software systems as identifiers, it gets tricky, but you usually don't have to.
[06:21] <wasabi> How in the world does one get ahold of Sun directory server and NDS anyways?
[06:22] <wasabi> ahh hah there it is
[06:24] <wasabi> lifeless: As I get further and further into this it become more clear that it is less of a technical problem and more of a political problem. I suspect most people come from a large AD environment or know what to expect in one.
[06:24] <wasabi> s/come/do not come/
[06:24] <wasabi> Nor even really understand the benefits.
[06:25] <wasabi> And if told downplay said benefits. =)
[06:27] <Mithrandir> in some way, this discussion reminds me of the old "but CVS doesn't know how to rename files" discussion where most people who used CVS just didn't care that it didn't support file renames; you just ignored that missing feature and went on with your life.  Maybe not ideal, but it worked and wasn't a real pain at all.
[06:27] <wasabi> Mithrandir: I guess. It also sort of resulted in CVS being removed as a viable competitor in the commercial world, imo.
[06:28] <wasabi> Ya know, hence the invention of ClearCase and all the others.
[06:28] <wasabi> And now those are pretty well established.
[06:28] <Mithrandir> wasabi: oh, I know plenty of software companies happy with CVS.
[06:29] <wasabi> Software companies, sure.
[06:29] <Mithrandir> (and those aren't interested in free software, they just want an RCS)
[06:30] <wasabi> I guess I speak of a different set of users.
[06:31] <wasabi> Those who want an RCS, but will never in their life even comprehend editing a cvs repository to move a file.
[06:31] <wasabi> Which, in my probably naive opinion, makes up the majority.
[06:31] <wasabi> You know, all those who ended up on SourceSafe, simply because it accomplished the common case.
[06:32] <wasabi> And then all ran to Vault when those guys released that and are now being sucked back into the MS fold with Team Server.
[06:33] <wasabi> me->bed. Night.
[06:39] <lifeless> mako: ping
[06:45] <dholbach> good morning
[06:46] <towsonu2003> good morning :)
[07:07] <evilynux> I'm asking for a sync and I've been told (by Stefan aka sistpoty) to join ubuntu-universe-sponsors but I can't seem to find such a mailing list. At least not on https://lists.ubuntu.com/mailman/listinfo/.
[07:07] <evilynux> Which list did he meant?
[07:09] <LaserJock> evilynux: it isn't a list
[07:09] <LaserJock> evilynux: it's a team on launchpad
[07:09] <evilynux> oh
[07:09] <LaserJock> https://launchpad.net/people/ubuntu-universe-sponsors
[07:11] <evilynux> thanks LaserJock
[07:11] <Hobbsee> dont join it, just subscribe them to the bug
[07:12] <evilynux> ah doh
[07:12] <evilynux> i didn't get we he meant
[07:12] <Hobbsee> else i'll just deny your request :)
[07:12] <Hobbsee> seeing as you clearly cant upload if you're not a MOTU
[07:12] <evilynux> heheh, can i cancel my subscription request?
[07:12] <Hobbsee> dunno
[07:13] <Hobbsee> if not, i'll just stop it, that's fine :)
[07:13] <evilynux> oki, stop it please
[07:13] <Hobbsee> done
[07:13] <evilynux> thanks
[07:14] <evilynux> there you go
[07:14] <evilynux> Ubuntu Sponsors for universe is in Subscribers to bug 73424. So that's it?
[07:14] <Ubugtu> Malone bug 73424 in mail-notification "Request for a sync with debian's package" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73424
[07:19] <evilynux> Stefan said:  subscribe ubuntu-universe-sponsors and wait for an ack, then subscribe ubuntu-archive.
[07:19] <evilynux> Which ack did he meant? And again, subscribe ubuntu-archive to the bug or subscribe to the mailing list?
[07:20] <dholbach> https://lists.ubuntu.com/archives/ubuntu-devel/2006-August/019922.html for the sponsoring process
[07:20] <evilynux> hmmm... i don't want to be sponsored
[07:20] <evilynux> i want my debian package to be sync in ubuntu
[07:21] <evilynux> dholbach: thanks anyway :)
[07:21] <ogra> evilynux, https://wiki.ubuntu.com/DeveloperResources
[07:23] <dholbach> evilynux: you did it right - somebody of the motu team will check that the changes can indeed be dropped, then ask ubuntu-archive to do the sync
[07:23] <evilynux> dholbach: great
[07:24] <ogra> hey dholbach 
[07:24] <evilynux> ogra: hmm.. that indicates that i still need to subscribe ubuntu-archive to the bug??
[07:24] <dholbach> heya ogra
[07:25] <evilynux> or does having "Ubuntu Sponsors for universe" nullifies the need to add ubuntu-archive?
[07:25] <dholbach> evilynux: they'll take care of that
[07:25] <ogra> evilynux, well, if you are sure that everything can be dropped you can do it, else i'd go with dholbachs suggestion and leave it up to the motu to check
[07:26] <evilynux> alright
[07:26] <evilynux> thank you guys
[07:26] <evilynux> cya
[07:26] <evilynux> (that help was much appreaciated)
[08:31] <Burgundavia> hey ogra
[08:32] <highvoltage> hi ubuntuland
[08:32] <Burgundavia> hey highvoltage
[08:33] <desrt> Burgundavia; how's business?
[08:33] <Burgundavia> not bad
[08:33] <Burgundavia> about to push out another UWN
[08:33] <desrt> score
[08:33] <highvoltage> wow, the UWN's keep on rolling nicely
[08:33] <Burgundavia> sticks come in handy for this task
[08:34] <desrt> git-push over wifi feels a little bit like magic
[08:34] <Burgundavia> desrt: if you don't deliver one of 1)your gdm work 2)your applet work in the .18 cycle, I am going to not give you free beer the next time I see you
[08:34] <desrt> Burgundavia; school is priority #1
[08:35] <Burgundavia> desrt: bah, school
[08:35] <desrt> Burgundavia; plus.  my gdm work has a patch in bugzilla and my applet work is in GNOME CVS
[08:35] <Burgundavia> default branch for the latter?
[08:35] <desrt> i do not prevent people from working on either :)
[08:35] <desrt> no.  "super-sketchy" branch :)
[08:36] <Burgundavia> right
[08:36] <Burgundavia> what is needed on both?
[08:36] <desrt> but all the code is there.  i'm not hoarding anything :)
[08:36] <Burgundavia> is there a good todo somewhere?
[08:36] <desrt> no.  not really.
[08:36] <desrt> the gdm one needs me to talk with the gnome-session maintainer
[08:37] <desrt> or possibly become the gnome-session maintainer...
[08:37] <desrt> applet needs a lot of polish work
[08:37] <desrt> it's fully functional but it's rough around the edges.  probably has some leaks and crashes in corner cases.  it also totally lacks a UI.
[08:38] <Burgundavia> the session stuff is being totally rewritten by danw, from what I remember
[08:38] <desrt> well
[08:38] <desrt> perhaps, in january, i will talk to him :)
[08:38] <Burgundavia> right
[08:39] <highvoltage> (I'm sure I will when I'm broke enough though ;) )
[08:39] <Burgundavia> I still have a form from UBZ I never sent in
[08:39] <desrt> in any case, that git-push that i was just talking about is my assignment 5 for cas705 which is due in 7 hours (and i should probably sleep some time in there too)
[08:40] <Mithrandir> cjwatson: how does the installer look, merge-wise?
[08:40] <desrt> ttyl :)
[08:46] <Mithrandir> Keybuk: any particular reason why postfix hasn't been synced from Debian?
[08:47] <Mithrandir> oh, we have a newer one in feisty.
[08:47] <Mithrandir> ignore me. :-)
[08:47] <Mithrandir> (I'm unable to read)
[08:47] <Keybuk> :D
[08:52] <Burgundavia> morning jono, Keybuk
[08:53] <jono> hey Burgundavia 
[08:53] <Burgundavia> jono: unless you have not been to sleep, this means you got less than 6 hours sleep
[08:53] <jono> Burgundavia, thats right, up for work :P
[08:58] <imbrandon> moins jono \
[08:58] <ogra> tsk, slee
[08:58] <ogra> p
[08:58] <jono> heya imbrandon 
[08:58] <imbrandon> sleep is for when your dead :)
[08:59] <imbrandon> wow
[09:11] <ogra> i wonder if i can just wie the line from the code ... or if i should include it in the package, since its gone from our libc dev package
[09:11] <ogra> *wipe
[09:40] <pitti> Good morning
[09:41] <Mithrandir> fabbione: thoughts about https://bugs.launchpad.net/distros/ubuntu/+source/samba/+bug/48082 ?  It's "later"-targetted.
[09:41] <Ubugtu> Malone bug 48082 in samba "Samba fails to upgrade with dangling K09samba symlink" [High,Confirmed]  
[09:41] <fabbione> Mithrandir: looking
[09:42] <Mithrandir> hmm, doko is supposed to be back today, isn't he?
[09:43] <fabbione> Mithrandir: i can look into it... do you need to know for feisty or get it fixed for *? 
[09:43] <pitti> Mithrandir: AFAIR yes
[09:44] <Mithrandir> fabbione: technically, it's a herd 1 blocker, but given that none of us has come up with "do those steps to reproduce", I'm not sure treating it as a blocker makes sense.
[09:45] <fabbione> Mithrandir: just clean installed on feisty with no problems
[09:45] <Mithrandir> fabbione: oh sure, it's _something_ which changed the symlink at some point in the past.  The job is to find out what.
[09:46] <fabbione> Mithrandir: oh i just noticed that it has been assigned to me.. 
[09:46] <fabbione> Mithrandir: ok.. i will have to dig into it
[09:46] <Mithrandir> fabbione: exactly; hence me prodding you.
[09:46] <Mithrandir> fabbione: thanks
[09:46] <fabbione> i hate random bugs reassigned without a proper ping.. they get lost in the 394u397439749302128749826459826439852645 bug emails
[09:47] <Mithrandir> hmm, I wonder if it would be worthwhile to have the personal bug pages grouped by milestone.
[09:48] <Mithrandir> it'd make it much, much easier for people to see "those bugs are high-priority ones"
[09:48] <fabbione> Mithrandir: my problem is the noise of all the other bugs
[09:48] <Mithrandir> fabbione: it wouldn't solve the mail problem.
[09:48] <fabbione> Mithrandir: yes i know
[09:48] <Mithrandir> it'd give you a more useful place in LP to look.
[09:50] <tomasz> can't you do an advanced bug search and group it by severity?
[09:51] <Mithrandir> that's not the point.  The point is visibility for all developers.
[09:53] <Mithrandir> tomasz: and severity is orthogonal to milestoning.
[09:55] <tomasz> ok
[09:57] <fabbione> Mithrandir: there is no mention of that change in the changelog... *sigh*
[10:02] <fabbione> sladen: ping?
[10:13] <jdub> Keybuk: nice entry.
[10:15] <cjwatson> Mithrandir: it's getting there, but I still have a fair bit to go ...
[10:16] <Burgundavia> Keybuk: as I pointed out in a recent conversation, since we are compromising on binary drivers, why not binary multimedia codecs?
[10:16] <cjwatson> Keybuk: how often does http://merges.ubuntu.com/main.html update? I've noticed it often being out of date (i.e. listing stuff I merged hours ago)
[10:16] <funman> hello
[10:16] <Burgundavia> that would satisfy an even larger group of users
[10:16] <funman> crimsun: are you here ?
[10:16] <Keybuk> cjwatson: hourly, unless it breaks
[10:16] <crimsun> funman: yes
[10:17] <funman> crimsun: i see ubuntu doesn't include a -dbg package for vlc
[10:17] <Keybuk> it keeps tripping over an abiword source
[10:17] <Keybuk> ValueError: process failed 17: dpkg-source -x abiword_2.4.5-0ubuntu2.dsc
[10:17] <Keybuk> +/srv/patches.ubuntu.com/unpacked/a/abiword/2.4.5-0ubuntu2
[10:17] <funman> does these debug packages only exist for important software ?
[10:18] <funman> it's quite helpful
[10:18] <crimsun> funman: let's migrate to #ubuntu-motu, please, since it's a universe package
[10:18] <funman> ok
[10:18] <Keybuk> cjwatson: tbh, I suspect casey is running out of disk space
[10:19] <cjwatson> ah
[10:19] <cjwatson> it updated for your LP link addition, so I assumed it must need manual prodding :)
[10:19] <Keybuk> looks like it's failed all weekend
[10:23] <seb128> funman: since edgy there is "-dbgsym" packages on "deb http://people.ubuntu.com/~pitti/ddebs edgy main universe"
[10:23] <seb128> funman: dbgsym packages are available for everything which has been rebuilt since the service is running
[10:24] <funman> seb128: yeah crimsun told me that
[10:25] <funman> thanks
[10:25] <seb128> np
[10:25] <funman> vlc is in there
[10:25] <funman> now i must convince people to use these debs ;)
[10:25] <funman> it'll be the most complex part of the trick ;)
[10:27] <fabbione> dholbach: what's the source of the package that offers the System->Administration->Share Folders ?
[10:27] <dholbach> fabbione: gnome-system-tools
[10:28] <seb128> which is the UI part
[10:28] <seb128> fabbione: what are you trying to do?
[10:28] <fabbione> dholbach: thanks
[10:28] <fabbione> seb128: i am trying to figure what hell is triggering #48082
[10:28] <seb128> bug #48082
[10:28] <Ubugtu> Malone bug 48082 in samba "Samba fails to upgrade with dangling K09samba symlink" [Medium,Confirmed]  http://launchpad.net/bugs/48082
[10:28] <seb128> ah, k
[10:29] <fabbione> it looks to me that users did try to share the folder, got a "there is no samba" and tried to install samba with that error
[10:29] <fabbione> because there is no way that samba does that
[10:29] <fabbione> so it is an external entity 
[10:29] <fabbione> now.. which one is what gives me some pain
[10:36] <tepsipakki> cjwatson: good morning. I tried to netboot feisty on Friday, but it installed edgy instead. Is Herd 1 going to have netboot-capability?
[10:36] <fabbione> Mithrandir: well i have no clue. samba is not at fault. let's ignore it for Herd 1
[10:37] <fabbione> tepsipakki: there is no feisty d-i yet
[10:37] <tepsipakki> so it doesn't matter if I preseed the suite?
[10:37] <tepsipakki> bummer
[10:37] <fabbione> no it's still too early. please be patiente
[10:37] <pitti> Mithrandir: btw, it seems that NetworkManager will not even need any ZeroConfigNetworking spec related changes
[10:37] <tepsipakki> sure ;)
[10:37] <cjwatson> tepsipakki: as Fabio says, wait.
[10:38] <tepsipakki> just asking
[10:38] <cjwatson> tepsipakki: it's in progress.
[10:38] <tepsipakki> of course
[10:38] <cjwatson> and yes, netboot will work before CDs, I imagine
[10:38] <cjwatson> I've already merged/uploaded choose-mirror
[10:38] <pitti> Mithrandir: avahi-autoipd is invoked as a dhclient-enter-hooks.d script from dhcdbd
[10:38] <cjwatson> but seeing as Herd 1 is a *CD* release, it does not make sense to talk about it having netboot capability as such
[10:39] <pitti> Mithrandir: and the route changes are done when calling the if-up.d scripts
[10:39] <tepsipakki> cjwatson: ok, thanks for the update
[10:40] <imbrandon> jono: ping ( see query )
[10:43] <Mithrandir> pitti: yay, goodie.
[10:43] <Mithrandir> cjwatson: ok, do you think it'll be working for herd 1 or will that be hard to achieve?
[10:47] <Mithrandir> Riddell: https://bugs.launchpad.net/distros/ubuntu/+source/python-qt4/+bug/67300 ; any idea about this?
[10:47] <Ubugtu> Malone bug 67300 in python-qt4 "python-qt4 FTBFS in edgy, broken QT environment?" [High,Confirmed]  
[10:47] <cjwatson> Mithrandir: I don't see how you can do Herd 1 without the installer being merged enough to work ...
[10:48] <cjwatson> I strongly suspect it's in a half-working state at the moment
[10:49] <Mithrandir> cjwatson: well, herd 1 would have to wait until the installer works most of the time at least.
[10:49] <Mithrandir> cjwatson: my poorly-phrased question was intended to find out if the original timeline was still feasible.
[10:51] <pitti> Mithrandir: ... except that it doesn't actually work; dhcdbd doesn't seem to call the dhclient enter/exit hook scripts; I'll figure that out, though
[10:52] <cjwatson> Mithrandir: mm. hard to tell yet. I hope to have a better idea by the end of today
[10:54] <ogra> :)
[10:55] <cjwatson> it doesn't matter if big chunks of the system are still unmerged, as long as it works at all
[10:55] <ogra> i didnt have the time to do more than merges yet, i dont even know if my seeds are fine ... (but worst case i'd skip herd1)
[10:55] <ogra> i skipped the first milestone for edgy as welll ...
[10:55] <Mithrandir> cjwatson: ok, thanks.
[10:56] <Mithrandir> ogra: kubuntu is probably skipping; this is more to see if we have our shit somewhat holding together than anything else, really.
[10:56] <ogra> yep, as it was in edgy ... 
[10:56] <cjwatson> traditionally we have done the first milestone as soon as the installer works.
[10:56] <ogra> lets see, if i arrive home in 4-5h i'll check the seeds ...
[10:57] <cjwatson> what's to check?
[10:57] <ogra> i anyway have no rea idea how i will split them for the two CD thing ...
[10:57] <cjwatson> updating seeds is an ongoing process
[10:57] <ogra> *real
[10:57] <Mithrandir> ogra: I suggest you don't try to do that for herd1.
[10:57] <ogra> (but thats indeed nothing for herd1)
[10:57] <cjwatson> the feisty seeds should basically work OK as they stand
[10:57] <ogra> right
[10:58] <Mithrandir> would anybody mind if I unseeded ndiswrapper-utils (without version)?
[10:58] <jdub> http://www.wazobialinux.com/
[10:58] <jdub> ^ awesome -> check the graphic for "news"
[10:58] <Mithrandir> it just gives us 1.1 now
[10:58] <ogra> i'd like to at least have an updated set of metapackages before building any CD, but that should work in time before thursday
[10:58] <cjwatson> IIRC ndiswrapper-utils was eventually supposed to be dropped in favour of the versioned packages
[10:58] <cjwatson> ogra: do that now, then. There should be nothing blocking it.
[10:59] <ogra> me boarding a plane in 10-15min blocks it from doing it *now* but i'll care tonight ...
[10:59] <ogra> ;)
[10:59] <cjwatson> s/do/delegate/
[11:00] <ogra> i'm fine doing it myself ... no worries ... 
[11:00] <Mithrandir> fabbione: another fun one for you ; bug 59983
[11:00] <Ubugtu> Malone bug 59983 in ndiswrapper "ndiswrapper in edgy broken" [Medium,Confirmed]  http://launchpad.net/bugs/59983
[11:00] <fabbione> Mithrandir: i don't have anything to test ndiswrapper...
[11:00] <ogra> die ndiswrapper die !
[11:01] <fabbione> and why people keep reassigning me bugs...
[11:01] <Mithrandir> fabbione: I think "people" in this case is your manager. ;-)
[11:01] <fabbione> Mithrandir: yes i know... i need to talk to him about it
[11:02] <fabbione> i have my work load already
[11:02] <cjwatson> (so does everyone else ...)
[11:02] <fabbione> cjwatson: lack of hw to test makes it more difficult
[11:02] <Mithrandir> fabbione: I'm sure you do, but those are herd 1 bugs, not random ones.
[11:02] <cjwatson> sure
[11:03] <jdub> http://www.geticeweasel.org/
[11:03] <highvoltage> geez, strange that they have to have a different marketing campaign
[11:03] <_ion> I have a faint feeling i've seen this logo somewhere before. http://elektrostore.se/Bilder/electro_loga.gif
[11:04] <Mithrandir> mvo: if you could tackle 66783 and the other "later" targetted bugs, I'd appreciate.  If you're already on it, ignore me.
[11:04] <fabbione> Mithrandir: yes, that means that fix a bug i get to have crap in Herd 1 :P
[11:04] <cjwatson> _ion: that's been known for almost as long as Ubuntu has existed
[11:04] <jdub> _ion: report to SPI
[11:05] <jdub> highvoltage: read it.
[11:05] <Mithrandir> _ion: that has been discussed numerous times in the past.
[11:05] <pitti> mvo: ping
[11:06] <Keybuk> _ion: it's a relatively standard photoshop brush :)
[11:06] <highvoltage> jdub: oic
[11:12] <sladen> fabbione: I'm always here
[11:13] <fabbione> sladen: 48082
[11:13] <sladen> bug #48082
[11:13] <desrt> Burgundavia; done :)
[11:13] <dholbach> Gloubiboulga: there's a new gnumeric release - in case you want to have it
[11:13] <Ubugtu> Malone bug 48082 in samba "Samba fails to upgrade with dangling K09samba symlink" [Medium,Confirmed]  http://launchpad.net/bugs/48082
[11:13] <fabbione> sladen: i can't reproduce it here and no matter what i do, i can't find anything creating those symlinks
[11:13] <mvo> Mithrandir: will do,
[11:13] <mvo> pitti: pong
[11:13] <Mithrandir> mvo: thank
[11:13] <Mithrandir> +s
[11:13] <fabbione> sladen: they are definitely not from samba pkgs
[11:13] <pitti> mvo: hil
[11:13] <pitti> mvo: hi even
[11:14] <fabbione> sladen: so either you can give me more info.. or i am not sure how am i supposed to fix that
[11:14] <sladen> bug #9208
[11:14] <Ubugtu> Malone bug 9208 in gnome-system-tools "Samba upgrade failure due to broken rc.d symlinks" [High,Fix released]  http://launchpad.net/bugs/9208
[11:15] <pitti> mvo: for ZeroConfNetworking it is necessary to inform the user about an incompatible network configuration for ipv4ll (existence of a .local network in unicast DNS); i. e. a dhcp client script needs to send out a dbus message which something in the desktop needs to pick up and show a notification
[11:15] <sladen> interesting, this looks like it's been around since 2004.  And still not reckoned to be from samba
[11:15] <mvo> pitti: and you have a paricular application in mind ,) ?
[11:15] <pitti> mvo: and I wondered what this 'something' could be; (event-notifier!!)
[11:15] <pitti> mvo: we should really do a proper event-notifier spec for feisty+1
[11:16] <mvo> pitti: totally. it really gets out of hand :)
[11:16] <pitti> mvo: I don't really mind whether we do that in u-n or in g-v-m; I just like to hear your opinion
[11:17] <fabbione> sladen: i did check the sources from warty to feisty
[11:17] <fabbione> sladen: none of the dh_installinit or even samba EXAMPLES use K09
[11:17] <sladen> fabbione: the only hint in those other dups I've just found is tht it might be g-s-t (!?!?!?! crack)
[11:18] <fabbione> sladen: i was looking at that source too
[11:18] <fabbione> sladen: but no luck yet
[11:18] <mvo> pitti: I don't really mind either way, it seems like its not a perfect fit for both of them
[11:18] <sladen> fabbione: 09 =  100-91, wonder if something is getting inverted
[11:19] <fabbione> sladen: samba has been using 20 19 forever
[11:19] <fabbione> sladen: ok.. can you try to grep trough your entire system for anything that does reference samba?
[11:20] <fabbione> i really don't know what could do that
[11:20] <fabbione> other than a broken update-rc.d
[11:20] <fabbione> (that is well .. extremely unlikely)
[11:25] <sladen> fabbione: doing
[11:25] <fabbione> sladen: thanks
[11:40] <pitti> mvo: oh, does u-n provide any dbus interface at all? g-v-m doesn't
[11:42] <mvo> pitti: not right now. it only uses dbus indirectly via talking to libhal
[11:42] <sladen> fabbione: http://ubuntuforums.org/archive/index.php/t-6318.html "I just noticed that I didn't originally install samba, it was smbclient"
[11:43] <pitti> mvo: that's what I suspected; ok, I'll figure out something else
[11:43] <fabbione> sladen: smbclient is the same source as samba... oh meh
[11:45] <sladen> fabbione: that the links are to  /samba   makes me think it didn't come from update-rc.d
[11:46] <fabbione> sladen: yes, i am grepping trough the entire code to find references to init.d
[11:46] <pitti> mvo: hm, one evil thing would be: for u in `who | cut -f 1 -d\ |sort -u`; do notify-send bla blubb; done
[11:47] <pitti> mvo: with the appropriate 'su -c $u' of course
[11:47] <pitti> would it be too evil to execute something like this in the system-wide DHCP script?
[11:48] <mvo> pitti: what about times when the user is not logged in? is it ok if the message is lost?
[11:48] <pitti> mvo: yes, it is
[11:49] <pitti> mvo: it's mainly useful for dynamic network roaming
[11:49] <mvo> ok
[11:49] <mvo> I would rather want to teach u-n about dbus than having this su -c solution
[11:50] <pitti> mvo: in the end it wouldn't make much of a difference
[11:50] <pitti> mvo: notify-send calls a org.freedesktop.Notification function
[11:50] <pitti> mvo: and with u-n it would just call a different function, but still something on the session dbus
[11:51] <pitti> mvo: if we want a better solution, then we should have something asynchronous that picks up avahi disabling at system bootup time, too
[11:51] <pitti> i. e. if avahi is disabled at the time of /etc/init.d/networking start, then there will not be an user session yet
[11:53] <fabbione> sladen: sbmclient has no maintainer scripts
[11:54] <sladen> fabbione: I'm beginning to think that it's some "helpful" admin script (eg. g-s-t) that likes to help the user start/stop "file sharing".  But the  /samba   bit I don't understand (it's the type of thing I'd expect to see on a trojaned box)
[11:55] <fabbione> sladen: i am thinking of /samba as a missing --sysconfig-dir = /etc or something like that
[11:55] <fabbione> i am trying to be positive
[11:56] <fabbione> dholbach, seb128: ping?
[11:56] <seb128> fabbione: pong
[11:56] <fabbione> seb128: we need your help to investigate that samba bug. 
[11:56] <dholbach> fabbione: pong
[11:56] <seb128> that bug is there since warty or hoary
[11:56] <seb128> and we have no clue about it
[11:56] <fabbione> seb128: do you think you or dholbach can check g-s-t for anything that could restart samba or add fake symlinks?
[11:57] <seb128> services-admin
[11:57] <fabbione> s/fake/broken/
[11:57] <fabbione> seb128: what's the source package?
[11:57] <fabbione> gnome-services-admin?
[11:57] <seb128> no, it's a part of gnome-system-tools
[11:57] <dholbach> gnome-system-tools (UI), system-tools-backends (backend)
[11:57] <seb128> the work might be done by system-tools-backend though
[11:57] <dholbach> see you
[11:57] <seb128> dholbach: have fun
[11:58] <fabbione> dholbach: AH don't run away when you get to debug the fun :)
[11:58] <dholbach> I'm debugging something else funny already
[11:58] <fabbione> ChangeLog:      * network.pl.in: in debian, made it to enable/disable the samba
[11:59] <StevenK> ... it does the dance? :-P
[12:00] <seb128> fabbione: services-admin is made to disable/enable any service
[12:00] <fabbione> seb128: yes.. i am trying to understand where it fails
[12:01] <seb128> are you positive it's g-s-t?
[12:03] <cjwatson> it seems like the sort of thing that could happen if /usr/share/system-tools-backends-2.0/scripts/Init/Services.pm:get_sysv_paths doesn't have an entry for the current distribution?
[12:03] <dholbach> fabbione: http://webcvs.freedesktop.org/system-tools-backends/system-tools-backends/network.pl.in?hideattic=0&r1=1.165&r2=1.166 is the change you're referring to - I think it just makes sure that it operates on "samba", not "smb" - however, the file doesn't exist anymore today. it's http://webcvs.freedesktop.org/system-tools-backends/system-tools-backends/Init/ServicesList.pm?hideattic=0&revision=1.5&view=markup and http://webcvs.freedeskt
[12:03] <dholbach> op.org/system-tools-backends/system-tools-backends/Init/Services.pm?hideattic=0&revision=1.14&view=markup today
[12:03] <dholbach> oops, a bit long the links
[12:04] <fabbione> cjwatson: that's exactly what i am looking at
[12:04] <cjwatson> I think the bug is that run_sysv_initd_script and (relevantly) set_sysv_service do not check that $rcd_path and $initd_path are defined
[12:04] <sladen> my auth.log only goes back to October... and these were created in September
[12:05] <cjwatson> oh, and set_filerc_services and get_suse_services too
[12:05] <fabbione> so let see if i try to use that thing, i can reproduce the problem
[12:07] <cjwatson> $ perl -Mlib=/usr/share/system-tools-backends-2.0/scripts -MUtils::Platform -le 'Utils::Platform::guess();'
[12:07] <cjwatson> platform_unsup::Your platform []  is not supported.
[12:08] <cjwatson> hmm, odd, I might be doing something wrong there
[12:12] <cjwatson> ok, I'll have a reproduction recipe in a minute, I think
[12:13] <fernando> moin all
[12:19] <cjwatson> no worries, I speak perl and I think I'm on it
[12:21] <fabbione> cjwatson: ok thanks
[12:32] <Keybuk> pitti: will that upgrade nsswitch.conf properly?
[12:32] <bhale> pradeep++
[12:32] <bhale> uh wtf
[12:33] <pitti> Keybuk: yes
[12:33] <bhale> pitti not pradeep 
[12:33] <pitti> hosts:          files mdns4_minimal [NOTFOUND=return]  dns mdns4
[12:33] <Keybuk> pitti: would you like me to blacklist libnss-mdns from the sync process?
[12:33] <pitti> that's the upstream recommended config
[12:33] <pitti> Keybuk: I don't think it's necessary; I made the version 0.8-6ubuntu1 (although I based it on -5)
[12:34] <Keybuk> pitti: right, but it'll continue to show up, etc. --  people might forget we've forked an NMU, rather than keeping up to date ?
[12:34] <pitti> Keybuk: but right, we would never want to synchronize to current Debian version
[12:34] <Keybuk> right
[12:34] <sjoerd> pitti: not protection in place for not breaking sites using .local ?
[12:34] <Keybuk> btw, pmount to be removed? :(
[12:34] <highvoltage> what would replace pmount?
[12:34] <pitti> sjoerd: I added a check for that to avahi-autoipd dhclient script
[12:34] <pitti> Keybuk: there are still rdependencies
[12:35] <sjoerd> pitti: ?
[12:35] <pitti> highvoltage: in Ubuntu, the hal backend scripts
[12:35] <Keybuk> pitti: there aren't
[12:35] <Keybuk> pitti: aptitude wants to remove it
[12:35] <highvoltage> pitti: aah
[12:35] <sjoerd> pitti: the avahi-autoipd script ensure that mdns is disabled when it detects a .local domain coming from dhcp ?
[12:35] <pitti> sorry, phone
[12:36] <pitti> -- feisty/main amd64 deps on pmount:
[12:36] <pitti> kdebase
[12:36] <pitti> kubuntu-desktop
[12:36] <pitti> thunar
[12:39] <pitti> re
[12:39] <pitti> /etc/dhcp3/dhclient-exit-hooks.d/zzz_avahi-autoipd:
[12:39] <pitti> if host -t soa local. >/dev/null 2>&1; then
[12:39] <pitti>     if [ -x /usr/share/avahi/enable_avahi ] ; then
[12:39] <pitti>         /usr/share/avahi/enable_avahi 0
[12:39] <pitti>     fi
[12:39] <pitti>     exit 0
[12:39] <pitti> fi
[12:39] <pitti> sjoerd: ^
[12:40] <sjoerd> that'll disable avahi
[12:40] <pitti> right, Lennart advised us to do so
[12:40] <sjoerd> you compile mdns without legacy report, so it won't actually resolve anything in that case right ?
[12:40] <pitti> sjoerd: right
[12:40] <sjoerd> k
[12:40] <sjoerd> imho avahi-autoipd is the wrong place, but the solution works
[12:40] <pitti> sjoerd: well, it already ships this dhclient script
[12:41] <pitti> sjoerd: and I'd like to keep the bits in one source package
[12:41] <sjoerd> sure, but you not guaranteed have -autoipd installed if -daemon is installed :)
[12:41] <Keybuk> pitti: I just meant getting removed from my machine
[12:42] <pitti> sjoerd: if we don't have autoipd installed, then this is not an issue anyway?
[12:42] <pitti> Keybuk: ah, right
[12:42] <sjoerd> anyaway, need to retry getting some sense into the debian nss-mdns guy...
[12:42] <pitti> sjoerd: yeah, -6 seems to be pretty f**ed
[12:43] <sjoerd> yup, he fucked up my and joeyh's changes.. Sent some private mail later offering his appolagies and then he stopped answering again :(
[12:43] <sjoerd> you still can have nss-mdns installed when you don't have -autoipd right ?
[12:43] <gnomefreak> mvo: if your around i can really use your opinion on an apt issue
[12:44] <sjoerd> nss-mdns isn't just for when you've got a LL ip
[12:44] <pitti> sjoerd: right, it just won't do anything
[12:44] <pitti> (with the ministack disabled)
[12:44] <sjoerd> why won't it do anything ? you don't enable avahi if you haven't got an LL ip ?
[12:45] <sjoerd> nss-mdns will nicely resolve ip's for you when avahi is enabled.. It doesn't matter what kind of address you have
[12:45] <pitti> sjoerd: if avahi-autoipd is not running, ipv4ll addresses aren't resolved (in my tests at least)
[12:45] <sjoerd> .local addresses are
[12:45] <sjoerd> ipv4ll address should be too
[12:45] <pitti> sjoerd: but you are right, putting this check into libnss-mdns might be more robust
[12:46] <pitti> sjoerd: that didn't work for me
[12:46] <sjoerd> well, avahi-daemon if the ministack is disabled anyway
[12:46] <pitti> avahi-autoipd seems to 'inject' the ipv4ll dns knowledge into avahi daemon, or so
[12:46] <sjoerd> neh
[12:46] <pitti> without it, I didn't get either avahi services nor .local name resolution
[12:46] <gnomefreak> pitti: i added you to a bug because i coulsnt find you. i think i am able to remove you from it because i had it happen without your repo yesterday.
[12:47] <sjoerd> If you resolve an ipv4ll ip without actually having one avahi-daemon will probably return right away because it's got no usefull interface to resolv on
[12:47] <gnomefreak> s/coulsnt/couldnt
[12:47] <pitti> gnomefreak: context?
[12:47] <gnomefreak> https://launchpad.net/distros/ubuntu/+source/apt/+bug/73062
[12:47] <Ubugtu> Malone bug 73062 in apt "[feisty]  apt and aptitude crashing" [High,Confirmed]  
[12:47] <sjoerd> pitti: i'm using .local name resolution daily, so i'm pretty sure it should work :)
[12:47] <sjoerd> on a normal 192.168.x.x range, not a ipv4ll range
[12:47] <pitti> sjoerd: heh, ok :) I'll move the script around then
[12:48] <sjoerd> :)
[12:48] <gnomefreak> i think it was just crappy luck that i addded your repo and apt started seg. faulting
[12:48] <sjoerd> I'll go beat the nss-mdns guy again, so hopefully we can have the same setup for debian..
[12:48] <pitti> gnomefreak: ah, this reminds me of a similar bug mvo and I discussed the other day, apt-index-watcher messing up the indices
[12:49] <pitti> sjoerd: but if I put the check into a different script, then I cannot stop avahi-autoipd from being spawned from the interface any more
[12:49] <gnomefreak> could very well be but that would happen more often for the apt-index-watcher no?
[12:49] <gnomefreak> this happens as i found out yesterday just for the hell of it :(
[12:49] <sjoerd> pitti: should -autoipd be spawned if the dhclient got an address ?
[12:50] <pitti> sjoerd: no, of course not; it's just spawned on timeout
[12:50] <cjwatson> seb128: do you know when the unsupported-system dialog in g-s-t was introduced? is it new, or since the dawn of Ubuntu time?
[12:50] <pitti> sjoerd: (on EXPIRE|FAIL|RELEASE|TIMEOUT)
[12:50] <seb128> cjwatson: the later I think
[12:51] <cjwatson> hmm, ok, there goes my hypothesis
[12:51] <cjwatson> I know that ServicesConfig does bad things if it can't figure out the system, but presumably it can never actually get there ...
[12:51] <seb128> " gnome-system-tools (0.90.0-2ubuntu1) warty; urgency=low
[12:51] <seb128>  .
[12:51] <seb128>    * debian/patches/05_ubuntu-no-warning.dpatch:
[12:51] <seb128>      - don't display a warning about warty not supported (Warty #461)."
[12:51] <seb128> right, that was already there on warty :)
[12:52] <cjwatson> seb128: right now, it's fatal and you can't get past it without picking a system. Was there a point when it was just a warning you could ignore?
[12:52] <sjoerd> pitti: hmmm, thus only when you didn't get a new lease you can meaningfully disable -autoipd.. 
[12:53] <seb128> cjwatson: I *think* it's the same since warty
[12:53] <cjwatson> ok
[12:53] <fabbione> # This function gets the state of the service along the runlevels,
[12:53] <fabbione> # it also returns the average priority
[12:53] <fabbione> sub get_sysv_runlevels_status
[12:53] <fabbione> could that return K09 somehow?
[12:53] <fabbione> if samba is not installed that might explain the K09 symlink
[12:53] <seb128> hey garnacho
[12:53] <garnacho> hi seb128 :)
[12:53] <seb128> cjwatson, fabbione: garnacho is gst upstream
[12:54] <fabbione> hi garnacho 
[12:54] <garnacho> hi fabbione :)
[12:54] <fabbione> seb128: thanks.. but i am heading for food right now... pasta is almost ready ;)
[12:54] <sladen> fabbione: average of 19 and zero ?
[12:54] <seb128> garnacho: cjwatson and fabbione are trying to figure what creates a broken K09samba symlinks for some users
[12:54] <cjwatson> garnacho: looking at https://launchpad.net/distros/ubuntu/+source/gnome-system-tools/+bug/9208, I've been noting that there are several functions in Init::Services that call get_sysv_paths but don't check that the resulting paths are defined
[12:54] <Ubugtu> Malone bug 9208 in gnome-system-tools "Samba upgrade failure due to broken rc.d symlinks" [High,Fix released]  
[12:55] <sladen> garnacho: the result is symlinks end up    /etc/rc?.d/K09samba -> /samba
[12:55] <cjwatson> garnacho: I can't see a way to actually get there, though, so I'm not sure if that's the thing at fault; but it would certainly be a way for symlinks to "/samba" to be created if you could
[12:55] <garnacho> cjwatson: hmmm, for a supported distribution with sysV init, those should be defined
[12:56] <garnacho> weird...
[12:56] <cjwatson> garnacho: right, but it's possible to end up with an unsupported distribution on Ubuntu easily enough if e.g. (a) you think it's a good idea to remove /etc/debian_version and (b) you upgrade base-files without also upgrading system-tools-backends
[12:56] <sladen> could this occur in the middle of an upgrade when the particular variation of distribution might be "unsupported"
[12:56] <cjwatson> sladen: did you remove /etc/debian_version, by any chance?
[12:56] <cjwatson> sladen: only if /etc/debian_version is removed
[12:56] <sladen> cjwatson: testing/unstable
[12:57] <garnacho> cjwatson: there's /etc/lsb_release too, g-s-t would detect that first
[12:57] <sladen> cjwatson: however, *during upgrade*, is that file removed at any point?
[12:57] <cjwatson> garnacho: upgrading base-files means that /etc/lsb-release is changed, and would have a value that system-tools-backends won't recognise until upgraded.
[12:57] <cjwatson> garnacho: /me <- person who tends to upload base-files for this ;-)
[12:58] <sladen> garnacho: it's not calling 'lsb_release' is it.  That will fail as we don't install an LSB compliant system by default
[12:58] <bhale> Mithrandir: could you please give archive love to beagle? i am trying to get people to test it
[12:58] <cjwatson> sladen: lsb_release> no, and it doesn't fail
[12:58] <bhale> Mithrandir: on a bug
[12:58] <cjwatson> at least not if you call it with the relevant options
[12:58] <Mithrandir> bhale: sure, wait a second.
[12:58] <cjwatson> $ lsb_release -rs
[12:58] <cjwatson> 7.04
[12:58] <Treenaks> grand_ubuntu_poo: I think you hit a size limit there
[12:59] <cjwatson> sladen: removed> shouldn't be, it's a conffile. I doubt that's it
[12:59] <bhale> Mithrandir: the split -dev might be confusing, libbeagle-dev is a C lib, beagle-dev is mono development stuffs
[12:59] <sladen> cjwatson: nods
[12:59] <Mithrandir> bhale: shouldn't the latter be libbeagle-cil or something then?
[01:00] <cjwatson> oh, /etc/debian_version could have been "3.1" at one point in the past, and perhaps system-tools-backends didn't know about Debian 3.1 then
[01:00] <cjwatson> in whatever version we happened to have at the time, or that users happened to have installd
[01:00] <cjwatson> installed
[01:01] <bhale> Mithrandir: possibly, but it went to Debian like this
[01:01] <cjwatson> /etc/debian_version said 3.1 in hoary
[01:01] <seb128> cjwatson: might be a good idea to break debian_version on purpose and note if that lead to a breakage
[01:01] <garnacho> cjwatson: however, seems an unlikely "race condition", would require to run s-t-b (detects distro at init) while base-files is being reinstalled
[01:01] <bhale> Mithrandir: and it isnt really a -cil I think because it doesnt follow GAC policy (globally installed, api versioned)
[01:01] <Vollstrecker> Anyone here that can help me with this: http://forum.amule.org/thread.php?threadid=11561&sid= ?
[01:01] <seb128> Vollstrecker: #ubuntu
[01:01] <cjwatson> garnacho: we support partial upgrades; you can upgrade base-files without upgrading anything else
[01:02] <Mithrandir> bhale: ack, I'll new it now I found the power to my laptop
[01:02] <Vollstrecker> seb128: It's not a bug report, I'm searching people that can translated the mans
[01:02] <Mithrandir> cjwatson: and somebody is _bound_ to have done that.
[01:02] <bhale> Mithrandir: thanks! big hugs
[01:02] <seb128> Vollstrecker: still not the right chan :)
[01:02] <seb128> Vollstrecker: maybe mail rosetta list
[01:03] <Vollstrecker> rosetta?
[01:03] <seb128> Vollstrecker: rosetta-users or ubuntu-translators maybe
[01:03] <cjwatson> garnacho: so my question is whether there's ever been a version of g-s-t that might have allowed you to proceed with an unsupported distribution, perhaps with a warning
[01:03] <Vollstrecker> k
[01:03] <cjwatson> garnacho: as opposed to the current code that forces you to pick a similar distribution
[01:04] <garnacho> no, g-s-t always forced you to pick a distribution
[01:04] <cjwatson> hmm, ok
[01:04] <cjwatson> garnacho: can I persuade you to add sanity checks for those paths being undefined anyway, for my peace of mind? :-)
[01:04] <garnacho> even more, now s-t-b caches the detected platform, so if you decide to remove lsb_release or debian_version, should work anyways
[01:05] <garnacho> cjwatson: it's in my TODO :)
[01:05] <cjwatson> thanks
[01:05] <sladen> cjwatson: interesting thing is that until today, I don't think I've ever run Services Admin (so hopefully unless it was init scripts, then it would not have had root-level sudo foo at any point to create the symlink)
[01:05] <cjwatson> frustrating though, I thought I'd at last found out how that was happening to people
[01:06] <cjwatson> sladen: how about any of the other g-s-t programs? somebody in the bug reported the symlink timestamp being the same as network-admin being run
[01:06] <sladen> Mithrandir: when short on disk-space I've often had to  apt-get install <100MB of packages manually at a time
[01:07] <Hobbsee> fabbione: ping?
[01:07] <sladen> cjwatson: interesting.  I'm certainly a very regular network-admin user
[01:08] <seb128> shares-admin touch that too
[01:08] <seb128> it has an option to install samba
[01:08] <seb128> I doubt network-admin does anything with samba
[01:09] <seb128> (note that the "install samba" patch is a distro one)
[01:10] <sladen> Ubuntu one, or one ported from Debian?
[01:12] <Keybuk> who merged sysklogd?
[01:12] <garnacho> yes, as seb128 says, shares-admin is also able to start smb and nfs services...
[01:12] <Keybuk> ogra!
[01:12] <bhale> orga
[01:12] <sladen> garnacho: did you say that the OS detection was done once, during postinst?
[01:13] <sladen> garnacho: or on every run?
[01:13] <fabbione> Hobbsee: pong?
[01:14] <garnacho> sladen: when system-tools-backends starts, if the platform is cached (in /var/cache/system-tools-backends-2.0/detected_platform IIRC) then uses that, if not, parses /etc/lsb_release, /etc/debian_version, etc...
[01:14] <garnacho> and s-t-b starts during init, when DBus starts
[01:15] <Hobbsee> fabbione: can we get ubuntulog into #ubuntu-classroom for the open week please?  it'd be useful to have the logs all in the same place as usual
[01:15] <fabbione> Hobbsee: yes we can.. but nobody requested it... gimme a few minutes
[01:15] <Hobbsee> fabbione: i just thought of it, and i just did :)
[01:15] <Hobbsee> sure
[01:16] <fabbione> Hobbsee: last minute = BAD.. i was about to head offline
[01:16] <Hobbsee> fabbione: sorry about that :(
[01:16] <Hobbsee> fabbione: as i say, i only just thought of it.  i'm not involved with the openweek at all
[01:16] <sladen> garnacho: how does it know when the cache is out of date?  timestamp on /etc/lsb_release ?
[01:20] <garnacho> sladen: until now there has been no reason to deprecate that cache, unless you begin to install a fedora on top of ubuntu :)
[01:21] <cjwatson> if ubuntu-$later_version ever needed to have different handling for something or other from ubuntu-$earlier_version ...
[01:21] <garnacho> cjwatson: hmm, and is the program guaranteed to be present?
[01:21] <cjwatson> garnacho: yes
[01:22] <cjwatson> the program is the thing that's specified in the LSB, not the file
[01:22] <garnacho> cjwatson: oh, cool, will have to make s-t-b use it :)
[01:23] <cjwatson> http://refspecs.freestandards.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/lsbrelease.html
[01:23] <garnacho> cjwatson: regarding the cache, until now there haven't been any cases of distros being so incompatible with the previous versions
[01:24] <cjwatson> lsb_release -d -s and lsb_release -r -s are the ones you probably want
[01:24] <cjwatson> er, -i, not -d
[01:24] <garnacho> but may be implemented if we find out there's a need :)
[01:24] <cjwatson> heh, fair enough
[01:27] <garnacho> hmm, lsb_release returns a python traceback for me, not good :(
[01:27] <sladen> garnacho: what's this on?  Debian?
[01:28] <garnacho> edgy here
[01:29] <sladen> garnacho: what's the backtrace, can you paste it somewhere?
[01:30] <sladen> caching could/should be moved to lsb_release and lsb_release made a little faster
[01:30] <sladen> (0.4 seconds per innovcation)
[01:32] <garnacho>   File "/bin/lsb_release", line 154, in guess_debian_release
[01:32] <garnacho>     distinfo['DESCRIPTION']  = '%(ID)s %(OS)s %(RELEASE)s (%(CODENAME)s)' % distinfo
[01:32] <garnacho> KeyError: 'RELEASE'
[01:32] <garnacho> that's the last line, pastebin sucks for me today...
[01:33] <cjwatson> I tweaked that a bit in the feisty merge (that line was actually there by accident due to a bodged previous merge), but I'm sure it worked in edgy
[01:35] <garnacho> cjwatson: I'm not sure, maybe I broke something
[01:36] <pitti> ajmitch, slomo: did you happen to see bug 70957? can we do something similar for mono?
[01:36] <Ubugtu> Malone bug 70957 in python2.5 "support apport reporting for python programs" [Wishlist,Confirmed]  http://launchpad.net/bugs/70957
[01:37] <cjwatson> garnacho: do you have /etc/debian_version? what's in it?
[01:38] <Mithrandir> bhale: are you aware that beagle blew up on sparc? (mono seems mostly busted on sparc, tbh)
[01:38] <garnacho> cjwatson: ooh, that's it, I moved debian_version away
[01:39] <cjwatson> yeah, that's a bug we fixed and that got accidentally broken again in a merge
[01:39] <bhale> Mithrandir: yep, all mono things seem to be broken on sparc
[01:39] <bhale> Mithrandir: uninstallable packages
[01:40] <cjwatson> garnacho: confirmed that that bug is fixed in feisty
[01:40] <bhale> Mithrandir: when mono FTBFS on an arch you still have the arch indep parts, so it becomes a mess
[01:41] <garnacho> cjwatson: nice :)
[01:41] <hunger>  /etc/init.d/sysklogd contains diff-markers.
[01:42] <Keybuk> hunger: already fixed
[01:42] <hunger> Keybuk: Oh, sorry.
[01:42] <Mithrandir> bhale: I presume you want the -dev in main, or?
[01:42] <garnacho> cjwatson: will have a look today at adding sanity checks at that portion of the code in s-t-b, I'm not sure whether/how s-t-b causes that, but will prevent it to do any changes if there's any inconsistency
[01:42] <bhale> Mithrandir: the native lib, yes
[01:43] <bhale> Mithrandir: well, all
[01:43] <cjwatson> garnacho: thanks
[01:43] <Mithrandir> bhale: beagle-dev isn't NEW.
[01:44] <bhale> nod
[01:44] <bhale> the NEW one will be a b-d for nautilus
[01:44] <bhale> and yelp
[01:44] <Mithrandir> stuff. in. new.  must. process.
[01:44] <fabbione> bhale: why did mono blewup on sparc? what about the patches we had beforE?
[01:44] <Keybuk> heh
[01:45] <bhale> fabbione: it is a few versions newer, maybe they broke something more.. it did a SIGSEGV
[01:46] <fabbione> bhale: and did you ask somebody to give it a retry?
[02:03] <bhale> fabbione: nope
[02:04] <fabbione> Mithrandir: thanks
[02:04] <Mithrandir> not that I think it'll make a difference, but we can try.
[02:05] <fabbione> Mithrandir: i agree, it will exclude the random lunar ray hitting the buildd console cable because SUN hates elmo
[02:12] <bhale> fabbione, Mithrandir, i am off to work, will check on sparc later
[02:12] <bhale> probably needs a bug upstream
[02:13] <seb128> Keybuk: how often is http://merges.ubuntu.com/main.html updated?
[02:16] <Keybuk> seb128: see scrollback
[02:17] <seb128> Keybuk: ah, ok
[02:17] <seb128> Keybuk: any ETA for fixing?
[02:18] <Keybuk> got to figure out what's wrong with it first
[02:18] <Keybuk> it only breaks sometimes
[02:32] <geser> libgnomevfs2-dev in feisty doesn't depend on libdbus-glib-1-dev anymore. is this on purpose or on error?
[02:39] <Keybuk> cjwatson: ping
[02:39] <cjwatson> Keybuk: Please tell me what you want and I'll reply when I'm around.
[02:40] <dholbach> geser: today I added a dbus-glib buid-dep to evolution-jescs, coming to think of it, I think that's a dependency missing somewhere on the way and not a gnome-vfs problem
[02:41] <cjwatson> Keybuk: ...
[02:42] <geser> dholbach: I looked at the pages on packages.u.c for libgnomevfs2-dev for edgy and feisty
[02:43] <geser> as one of my uploads has the same problem of missing dbus-glib-dev, should it be added to the package itself?
[02:44] <dholbach> no, we have to find out where it's missing
[02:44] <dholbach> oh, you're probably right about libgnomevfs2-dev, hmhmhmhm
[02:47] <Mithrandir> pitti: shouldn't python-apport really be using python-central or python-support?
[02:47] <pitti> Mithrandir: the source package uses python-support
[02:48] <pitti> Mithrandir: and that didn't add any binary dependency, nor postinst scripts
[02:49] <Mithrandir> pitti: hmm, that sounds strange, doesn't it?
[02:49] <Mithrandir> pitti: given that there are nothing byte-compiling the .py files now?
[02:49] <Keybuk> cjwatson: I assume you'll want to merge pcmciautils, seeing as you have it in bzr?
[02:49] <pitti> Mithrandir: hm, no idea
[02:50] <pitti> Mithrandir: I do have .pyc files for the apport package
[02:50] <pitti> black magic...
[02:50] <cjwatson> Keybuk: yeah
[02:51] <Mithrandir> pitti: oh well, I guess I'll just NEW it.  If you could investigate, that'd be good.
[02:51] <Mithrandir> pitti: also, NEW?
[02:51] <Mithrandir> uh
[02:51] <Mithrandir> main?
[02:51] <pitti> Mithrandir: yeah, python-apport is new; should go into main
[02:51] <cjwatson> pitti: didn't you use Depends: ${python:Depends}?
[02:51] <pitti> cjwatson: I did
[02:51] <cjwatson> oh
[02:51] <pitti> cjwatson: but p-support didn't create any maintianer scripts
[02:52] <cjwatson> Keybuk: what changes are required for udev rules in feisty? I should do those at the same time
[02:52] <Keybuk> cjwatson: none, assuming they haven't changed from edgy
[02:53] <cjwatson> Keybuk: don't think so. I thought there was an incompatible BUS/SUBSYSTEM change?
[02:53] <Keybuk> BUS has to become SUBSYSTEMS
[02:53] <Keybuk> but you don't use BUS
[02:55] <pitti> Mithrandir: right, I don't get modules for python2.5, I'll check this out and fix it
[02:55] <Mithrandir> pitti: thanks.
[02:55] <cjwatson> Keybuk: ok, thanks
[02:55] <Mithrandir> pitti: I accepted the package anyway, so next upload should wander through without delay.
[02:55] <pitti> great, thanks
[03:14] <cjwatson> Keybuk: pcmciautils merged
[03:15] <cjwatson> rodarvus: did you get my mail about libx11? that chain is the last thing blocking a debian-installer upload now, I think
[03:17] <Mithrandir> cjwatson: yay, so d-i is almost there?
[03:19] <cjwatson> Keybuk: how come merges.u.c thinks console-setup is at 1.7ubuntu19 (and updated, no less, despite that being the edgy version) when I uploaded 1.13ubuntu1 days ago?
[03:19] <Keybuk> it doesn't ...
[03:19] <Keybuk> reload
[03:19] <cjwatson> aha, it did this morning; thanks
[03:20] <Keybuk> yeah
[03:20] <Keybuk> it stuck over the weekend
[03:20] <Keybuk> it keeps tripping over an abiword source
[03:20] <Keybuk> dunno why, it works fine 7/10 times
[03:20] <Keybuk> but once it sticks, it needs manual fixing
[03:20] <cjwatson> Mithrandir: oh, there's cdebconf to go as well
[03:20] <Keybuk> http://merges.ubuntu.com/main-trend.png
[03:20] <cjwatson> Mithrandir: just waiting for the import to catch up on that
[03:20] <Keybuk> ^ I like the decreasing red there
[03:21] <Hobbsee> Keybuk: what do we have to put in for removing packages from the archive? just what it is, why, and if it has any rdeps?
[03:22] <Mithrandir> cjwatson: ok, so still not horrible.
[03:23] <Keybuk> Hobbsee: yup
[03:23] <pitti> BenC, kylem: if you have some minutes, could you please take a look at https://wiki.ubuntu.com/ApportImprovements and give me an opinion about the required kernel changes?
[03:23] <Hobbsee> Keybuk: okay, cool.  i'll double check that i'm not reading this wrong first though :P
[03:23] <kylem> certainly.
[03:23] <pitti> kylem: thanks
[03:28] <BenC> pitti: The kernel side will be pretty easy
[03:29] <BenC> pitti: I'm thinking that if we set core_pattern to a pipe, it should be safe to assume we can ignore ulimit -c
[03:29] <pitti> BenC: the field splitting is certainly just a SMOP, but the ulimit -c setting?
[03:29] <pitti> BenC: ah, right, then the core_pattern process could decide itself when to stop reading from the pipe
[03:30] <pitti> BenC: that would work for me; do you think that upstream would approve this?
[03:30] <BenC> pitti: Right, no need to be concerned with a file size, when we aren't writing files
[03:30] <BenC> pitti: I can argue the case and see what happens
[03:30] <Hobbsee> Keybuk: um, if a package has a differnet source package name, but the same binary name, that does need a removal, then a sync, right?  you cant just sync over the top?
[03:31] <Keybuk> Hobbsee: right, we still need to remove the old source
[03:31] <Hobbsee> Keybuk: thought so.  done :)
[03:31] <Hobbsee> Keybuk: just requestsync appeared to actually find the old package, even with a different source name
[03:32] <Hobbsee> just double checking, as my brain doesnt work too well at this time of night :P
[03:32] <pitti> BenC: ok, I update the spec accordingly; thanks!
[03:35] <Keybuk> Hobbsee: what's the package?
[03:35] <Hobbsee> Keybuk: https://bugs.launchpad.net/distros/ubuntu/+source/icemon-kde3/+bug/73469
[03:35] <Ubugtu> Malone bug 73469 in icemon-kde3 "please remove this from the archive" [Undecided,Unconfirmed]  
[03:36] <Keybuk> what builds its binaries?
[03:37] <Hobbsee> sorry?
[03:38] <Hobbsee> the buildds build its binaries, surely?  :P
[03:39] <elmo> Hobbsee: => what source package builds those binaries
[03:39] <Hobbsee> aww crap
[03:39] <Keybuk> icecc apparently
[03:40] <Hobbsee> excuse my idiocy, i'll have to have *more* of a careful look at this
[03:41] <Keybuk> amusingly
[03:41] <Keybuk> in EDGY it doesn't build anything, because they were built by icecc
[03:41] <Keybuk> but in FEISTY, they're built by icecc-monitor
[03:41] <Keybuk> which was split from icecc
[03:41] <Hobbsee> what the?
[03:42] <Keybuk> I: icecc-monitor [universe]  -> icecc-monitor_0.6svn20060501-0ubuntu2 [universe] .
[03:42] <Hobbsee> now i'm majorly confused.
[03:42] <Keybuk> there you go :p
[03:43] <Hobbsee> but icecc has a totally different version to icecc-monitor, which appears to be in the source package icecc?
[03:43] <Keybuk>      icecc | 0.6svn20060501-0ubuntu2 | feisty/universe | hppa
[03:43] <Keybuk>      icecc |   0.7.14-4 | feisty/universe | source, amd64, i386, ia64, powerpc, sparc
[03:43] <Keybuk> icecc-monitor | 0.6svn20060501-0ubuntu2 | feisty/universe | amd64, hppa, i386, ia64, powerpc, sparc
[03:43] <Keybuk> libicecc-dev |   0.7.14-4 | feisty/universe | amd64, i386, ia64, powerpc, sparc
[03:43] <Keybuk> l
[03:43] <Keybuk> (same for edgy)
[03:43] <Keybuk>      icecc | 0.6svn20060501-0ubuntu2 | dapper/universe | source, amd64, hppa, i386, ia64, powerpc, sparc
[03:43] <Keybuk> icecc-monitor | 0.6svn20060501-0ubuntu2 | dapper/universe | amd64, hppa, i386, ia64, powerpc, sparc
[03:44] <Keybuk> so in dapper, icecc-monitor was built by icecc ;P
[03:44] <Hobbsee> yeah, i see
[03:44] <Keybuk> we picked up icemon-kde3 in edgy somehow
[03:44] <Hobbsee> yeah, it's in debian
[03:44] <Keybuk> but it never did anything, because its version was always lower than icecc's
[03:44] <Hobbsee> oh wait, yeah, no idea how that got in there
[03:45] <Keybuk> it's not in Debian
[03:45] <Hobbsee> can you remove that please?  the icemon-kde3?
[03:45] <Keybuk> done already
[03:45] <Hobbsee> way cool :)
[03:45] <Keybuk> and I sync'd icecc-monitor (source) in your name
[03:46] <Hobbsee> yeah, that's right
[03:54] <bddebian> Heya
[03:59] <lamont> Mithrandir: thanks for dealing with postfix-cdb
[04:09] <BenC> Keybuk: ping
[04:10] <Keybuk> BenC: yo
[04:11] <kylem> pitti, i think that it shouldn't be a problem to do a pipe even with ulimit -c 0
[04:11] <pitti> kylem: great, this will make things easy
[04:14] <kylem> pitti, i think the macro stuff is simply a matter of strsep on corename+1 and adding to the call_usermode_helper fn
[04:15] <pitti> kylem: well, macro expansion needs to be done in a loop then, right?
[04:15] <pitti> kylem: since it's more robust to split first, expand afterwards
[04:16] <kylem> pitti, agreed.
[04:16] <kylem> the macro expansion should already be done in format_corename
[04:24] <BenC> Keybuk: I need to talk to you about how best to handle ide_all_generic
[04:24] <Keybuk> BenC: right
[04:25] <BenC> Keybuk: right now, it doesn't even work in edgy because it's only available when built-in
[04:25] <BenC> but 2.6.19, and 2.6.17 when I stick a patch in, will allow it as a module param
[04:29] <Keybuk> you mean all_generic_ide ?
[04:29] <BenC> yeah, that's the one
[04:30] <Keybuk> it worked for me ...
[04:30] <Keybuk> I did modprobe ata_generic all_generic_ide=1
[04:30] <Keybuk> and that found my ata card
[04:30] <BenC> it works in 2.6.19
[04:30] <BenC> or 2.6.17 with ata_generic
[04:31] <Keybuk> right
[04:31] <Keybuk> we only care about 2.6.19 though, yes?
[04:31] <Keybuk> (feisty)
[04:31] <BenC> ata_generic is blacklisted in edgy, right?
[04:32] <BenC> no, I care about dapper and edgy too, actually
[04:32] <Keybuk> in dapper and edgy we have ide-generic, no?
[04:32] <BenC> we fix this, and it will keep us from chasing new IDE chipsets
[04:32] <Keybuk> which supports it as a module param, iirc
[04:32] <BenC> not in edgy it doesn't
[04:32] <BenC> err ide-generic isn't what we are talking about
[04:32] <BenC> it's generic
[04:33] <Keybuk> ahh
[04:33] <Keybuk> that's what I meant, sorry
[04:33] <Keybuk> do we really chase new IDE chipsets that often?
[04:33] <BenC> but generic doesn't have the module param in 2.6.17 and prior
[04:33] <mjg59> BenC: The right way is definitely not to default to all_ide_generic (or whatever it's called)
[04:33] <BenC> Keybuk: Often enough for LTS
[04:34] <elmo> Keybuk: remember that white box at allhands?  the IDE in that doesn't work :-P
[04:34] <elmo> in _Edgy_
[04:34] <Keybuk> mjg59: we're just talking about having a kernel option to switch that on
[04:34] <Keybuk> not by default
[04:34] <mjg59> Well, it needs two components
[04:34] <mjg59> Adding the option to the kernel is a trivial and non-invasive patch
[04:34] <mjg59> But it also needs initramfs-tools support
[04:35] <Keybuk> right, it'd need a script that detected the option, and if there force-loaded generic all_generic_ide=1
[04:35] <Keybuk> and for feisty, would force-load ata_generic instead
[04:35] <mjg59> Yes
[04:35] <Keybuk> if grep -q -w pleasegodmakemyidecardwork /proc/cmdline; then
[04:36] <Keybuk>         modprobe -Qb generic all_generic_ide=1
[04:36] <Keybuk> fi
[04:36] <BenC> I say just check for all_generic_ide in /rpco/cmdline
[04:36] <Keybuk> right  :)
[04:36] <BenC> just to make it compatible
[04:42] <Keybuk> http://people.ubuntu.com/~scott/edgy-all_generic_ide.patch
[04:42] <Keybuk> err
[04:42] <Keybuk> http://people.ubuntu.com/~scott/edgy-all_generic_ide.debdiff
[04:43] <BenC> Keybuk: Great, now I just need to add the module_param() to the edgy modue
[04:56] <Mithrandir> lamont: happy to help. :-)
[05:02] <cjwatson> wow, a debootstrap merge to do
[05:05] <cjwatson> p/wg 45
[05:05] <Ng> cjwatson: talking of deb* things, could we ubuntuify debmirror? (bug #64345)
[05:05] <Ubugtu> Malone bug 64345 in debmirror "defaults are very debiany" [Wishlist,Unconfirmed]  http://launchpad.net/bugs/64345
[05:06] <Ng> nm
[05:11] <cjwatson> Ng: I'm sure somebody !me could :-)
[05:11] <Ng> yeah, sorry, I realise now it's not your problem :)
[05:13] <abattoir> cjwatson: hi :)
[05:13] <cjwatson> abattoir: hello
[05:13] <abattoir> cjwatson: are you free now?
[05:13] <cjwatson> ish
[05:13] <abattoir> cjwatson: ok, just wanted to ask about the dir. oem-config is heading for feisty
[05:13] <abattoir> cjwatson: is it going to be merged w/ Ubiquity?
[05:13] <abattoir> *direction
[05:14] <cjwatson> abattoir: slightly unclear. that's where I'd like it to go but I don't know how much effort I'll be able to put into that
[05:14] <cjwatson> abattoir: we haven't had much in the way of concrete requirements from OEMs
[05:14] <cjwatson> I would like at the very least to port over ubiquity's timezone page
[05:14] <cjwatson> and the four or five bug-fixes at the bottom of the braindump in SimplifyOEMInstallation
[05:25] <lamont> pitti
[05:25] <lamont> er, pitti ping
[05:25] <lamont> 09:23:41.316 [I]  blockdev.c:141: Add callouts completed udi=/org/freedesktop/Hal/devices/volume_uuid_eca295e2_7840_4259_b1da_359435cdb21e
[05:25] <lamont> 09:23:41.316 [I]  hald.c:82: Added device to GDL; udi=/org/freedesktop/Hal/devices/volume_uuid_eca295e2_7840_4259_b1da_359435cdb21e
[05:25] <lamont> 09:23:41.399 [W]  hald_dbus.c:87: No property storage.partition_table_changed on device with id /org/freedesktop/Hal/devices/storage_serial_ST316002_1A_502010008842
[05:25] <lamont> 09:23:41.420 [W]  hald_dbus.c:87: No property info.callouts.add on device with id /org/freedesktop/Hal/devices/volume_uuid_eca295e2_7840_4259_b1da_359435cdb21e
[05:25] <pitti> cjwatson: if neither d-i nor ubiquity use shadow any more for user setup, does that mean we can drop this delta? (passwd/root-password-empty debconf question and special sudo/group handling in user setup)?
[05:26] <cjwatson> pitti: yes
[05:26] <cjwatson> haven't used it since dapper, which should be enough leeway :-)
[05:34] <jayteeuk> pitti: I suspect you didn't mean right away, but when you've got a minute. :)
[05:35] <pitti> cjwatson: so we can also strip shadow translations again and use langpacks?
[05:35] <pitti> jayteeuk: just ask away
[05:35] <cjwatson> pitti: yeah, I believe so
[05:37] <pitti> cjwatson: oh, oops, it already was :)
[05:37] <jayteeuk> OK -- I'd like to see NM as a default component, I think it's a massive plus for usability.  I've forsaken my Windows (wash my mouth out with soap and water) install in favour of Ubuntu, but find the network management a little hard to get my head around at the moment.
[05:37] <pitti> cjwatson: we already did that in Jan 2006
[05:37] <cjwatson> pitti: heh. initial-passwd-udeb maybe, or is the blacklist organised by source package?
[05:37] <pitti> cjwatson: no, binary packages, but no udebs there
[05:37] <cjwatson> pitti: ok
[05:38] <cjwatson> cool
[05:38] <cjwatson> jayteeuk: we did try to include it by default in the dapper cycle, but ran into fairly massive breakage, which is why we don't
[05:38] <abattoir> cjwatson: ok, thanks, i was thinking of porting oem-config-kde to qt4
[05:38] <elmo> can NM do static addresses yet?
[05:39] <cjwatson> IIRC it's basically that it works like a charm on some hardware and is a hideous mess on others
[05:39] <cjwatson> abattoir: sounds like a reasonable project
[05:39] <jayteeuk> cjwatson: Understood, and I support that decision.  I use Ubuntu on a laptop on a daily basis.  I work from home, where I have complete control over the network etc, but make occasional trips to the office where I don't. :)
[05:40] <jayteeuk> It would be nice to be able to just switch across to the office WLAN with minimal user intervention -- and I think that's the intention with NM?
[05:41] <pitti> jayteeuk: right
[05:42] <pitti> jayteeuk: please see the https://wiki.ubuntu.com/NetworkRoaming spec for our current plans
[05:44] <jayteeuk> Yep, currently reading that -- I notice some of the items I'm inerested in are planned for inclusion, VPN management being a major one.
[05:44] <pitti> jayteeuk: we won't do any VPN related implementation on our own, just take what upstream does
[05:46] <jayteeuk> Ah, OK.  So if I wanted to try and help out there (try being the important word -- I'd have to fit a lot of learning time into my non-Ubuntu-related day job) I'd be better speaking to upstream?
[05:46] <jayteeuk> I thought NM was Ubuntu-specific. :)
[05:46] <jayteeuk> Shows what I know. :-P
[05:47] <Ng> the network manager openvpn stuff is pretty cool
[05:47] <lamont> pitti: you want the whole hal debug log, or is just the pasted stuff sufficient?
[05:48] <jayteeuk> Ng: I haven't seen that -- URL?
[05:48] <pitti> lamont: can you please send me the whole log? the snippets above don't give any clue
[05:48] <jayteeuk> Although I'm personally looking for vpnc integration, since our VPNs at work are Cisco.
[05:48] <Ng> jayteeuk: no idea if it has a homepage, but it's in REVU http://revu.tauware.de/details.py?upid=3297
[05:48] <jayteeuk> And a CheckPoint client would also be useful. :)
[05:49] <cjwatson> jayteeuk: it's originally from Red Hat, actually
[05:49] <lamont> pitti: incoming
[05:49] <pitti> jayteeuk: indeed it would be really useful to have someone who would test, search for bugs, and report them upstream (i.e. doing coordination)
[05:50] <pitti> lamont: however, in the end this could require some more in-depth debugging; is it possible to get a (non-root) ssh on that box?
[05:50] <lamont> pitti: not trivially, no
[05:50] <pitti> lamont: ok, I'll look at the log and reply with some further questions
[05:50] <jayteeuk> pitti: I'd like to put myself forward for the role, but as I said, I'd have to fit it into my day job. :)  I don't have much spare time outside that to contribute I'm afraid. :S
[05:51] <lucas> wow, according to bug #66702, gcc generates opteron-specific code on edgy/amd64
[05:51] <Ubugtu> Malone bug 66702 in gcc "GCC emits 3DNow!-specific instruction for __builtin_prefetch" [Unknown,Rejected]  http://launchpad.net/bugs/66702
[05:51] <lucas> that seems a huge problem
[05:51] <pitti> jayteeuk: right, you should only do what you can and like, of course; after all, it's supposed to be fun :)
[05:52] <jayteeuk> pitti: Oh absolutely -- the enjoyment part certainly isn't an issue!  If I didn't have bills to pay I'd be spending a significant amount of my time doing more open source work.
[06:07] <pitti> fabbione, dholbach: you talked earlier about a samba bug; samba is on my merges list, anything I should look at while merging?
[06:07] <dholbach> pitti: dunno, giving a talk
[06:12] <seb128> pitti: no
[06:12] <seb128> pitti: that's the "something create a broken K09samba" link bug
[06:12] <seb128> pitti: that's happening since warty or hoary
[06:12] <seb128> and that's likely to be a g-s-t bug
[06:14] <pitti> seb128: alright, thansk
[06:17] <seb128> pitti: np
[06:26] <fabbione> pitti: no it's not a samba bug.
[06:28] <BenC> Keybuk: Are you doing the initramfs update for all_generic_ide, and if so could you also do one for edgy/dapper proposed?
[06:28] <BenC> Keybuk: if not, I can do it
[07:35] <BenC> Keybuk: Hey, that's the other thing I need to know about. I think we need to make the /dev/{s,h}d* conversion to UUID happen again
[07:50] <stgraber> Anyone know a good doc of howto write the USAGE part of a man page ? (to tell that this parameter is optional, this one is not, ...)
[07:54] <bronson> stgraber: no...  personally, I just crib from an existing command.
[07:54] <bronson> Find a well-documented command that does something similar and copy that.
[07:54] <Treenaks> bronson: that's what people usually do :)
[07:54] <Treenaks> just copy from a known-good source :)
[07:55] <stgraber> ok :)
[08:16] <mdz> mvo: why can't dist-upgrade figure out nfs-common and nfs-kernel-server? it looks like a straightforward soname change in library dependencies
[08:29] <cjwatson> pitti: would you mind looking over tipa for main? it's a dependency of jadetex, so a fair bit of stuff is failing to build due to it being in universe
[08:30] <mvo> mdz: dapper->edgy? edgy->feisty? do you have a bugnumber? I remember a postinst problem in nfs-kernel-server, but not a library dependency
[08:30] <pitti> cjwatson: will do right now
[08:30] <cjwatson> thanks!
[08:31] <cjwatson> don't spoil your evening though
[08:50] <pitti> cjwatson: package looks fine and works OOTB for me; I'll write an approved report now
[08:50] <cjwatson> thanks
[08:50] <cjwatson> I'll promote
[08:52] <ajmitch> pitti: oh, are you merging samba now?
[08:52] <pitti> ajmitch: oh, right, you wanted to, right?
[08:53] <pitti> ajmitch: I didn't start yet, I just walked through my list and took a look at the patch
[08:53] <pitti> cjwatson: alright, then I'll wikify it directly to promoted
[08:53] <ajmitch> pitti: yeah, it's mostly a few ubuntu patches to look over, I've just been slacking with getting it finished
[08:53] <pitti> ajmitch: did you add an additional conflicts/replaces for python2.4-samba?
[08:54] <ajmitch> yeah, I believe so
[08:59] <the_mug> any idea why Apache 2.2.3 is not included in Edgy?
[09:03] <ajmitch> night cjwatson 
[09:11] <aleka>  <jordi> 21:01 < aleka> QUESTION: where can I get help in aquiring fonts that I need for translating (Amharic - ethiopic fonts) that work in Linux. This question emailed to team leader twice and no response..
[09:11] <aleka> [15:07]  <jordi> aleka: you might want to ask in the ubuntu-devel list
 mark shuttleworth has big interest in getting ubuntu working out of the box for special script like yours
[09:12] <aleka> If anyone knows of any work that has been done in this area/that is being done...
[09:13] <jordi> aleka: first, do you know if there are free amharic fonts available?
[09:13] <jordi> aleka: I can see ethiopian just perfectly in my system
[09:13] <aleka> jordi, I have downloaded a couple of .ttfs
[09:13] <aleka> have added these fonts under fonts:///
[09:14] <aleka> but most of the Amharic script fonts out there are written for windows environment..
[09:14] <aleka> I haven't had any luck yet of getting one to work on Ubuntu
[09:15] <aleka> I can view it too in a web browser,
[09:15] <aleka> on the launchpad pages, the Amharic pages... i can read all of these perfectly..
[09:16] <aleka> problem is, writting them....
[09:16] <aleka> specially since i am interested in helping with translation... i want to be able to type and get those fonts...
[09:16] <sladen> aleka: if we're not shipping fonts by default for a particular language (and you know why to find free+distributable fonts), then that's a bug and we'd like to make that work out of the box
[09:16] <Amaranth> if you can read them without those windows-only fonts that part is fine
[09:17] <sladen> aleka: if you can view Amharic via the web-browser, then the fonts are there and working
[09:18] <aleka> I thought FF handled Unicode fonts on the fly, evenif you don't have those fonts downloaded
[09:18] <sladen> aleka: TTFs are TTFs, Linux can cope with them either way
[09:18] <Amaranth> aleka: Nope, you have to have a font to render them
[09:18] <aleka> How could I get this fonts active to write (for example in Openoffice)
[09:18] <sladen> any programs displaying text needs to be able to get the fonts from somewhere.  normally locally
[09:18] <Amaranth> aleka: If they worked before you added those extra fonts then you don't need the extras
[09:18] <Amaranth> aleka: You need to type something in that language
[09:18] <sladen> aleka: it should be just the case of starting to type!
[09:19] <sladen> aleka: missing characters are fetching (replaced) from other fonts
[09:20] <mdz> mvo: edgy->feisty, no bug number, I just noticed it
[09:20] <mdz> mvo: The following packages have been kept back:
[09:20] <mdz>   nfs-common nfs-kernel-server python-egenix-mxdatetime
[09:20] <aleka> Ok.. say I downloaded one ttf called "jiret", under windows, this self instals and opening MS word and typing produces the Amharic fonts, In ubuntu, I added this .ttf under my fonts dir, and chose this font in OO, when typing, i still get english characters
[09:21] <sladen> aleka: is the keyboard set to Amharic?
[09:21] <aleka> how/where would I set that?
[09:21] <Amaranth> aleka: You need to type Amharic letters, not English letters
[09:21] <aleka> I do apologize guys, as I understand this is not the right room to be asking for support for such a silly feature...
[09:22] <Amaranth> hmm
[09:22] <sladen> aleka: when you install you'd normally select the keyboard type.  Or  System->Preferences->Keyboard after installation
[09:22] <Amaranth> System->Preferences->Keyboard does not have an Amharic setting
[09:22] <aleka> most of these fonts are phonetic... you type the english character and it produces the corresponding phonetic character in amharic...
[09:23] <sladen> aleka: ah okay, I understand the concept, but it seems kludgey when we have Unicode
[09:24] <sladen> aleka: can you try copying and pasting some text from Firefox into OpenOffice.  Does the text display fine
[09:25] <aleka> I can not do this now as I am at work on an Xp machine...but that is an idea
[09:26] <Burgwork> cjwatson: there needs to be a conversation about turning off hardware intensive on clients that can't handle that
[09:26] <mvo> mdz: ok, thanks. I have  a look now
[09:26] <Burgwork> I believe LTSP sets an env variable
[09:26] <aleka> http://www.alanwood.net/unicode/ethiopic.html sort of gives some explanation on what I am talking about
[09:28] <sladen> aleka: I can't see Amharic under Ethiopia/Africa/etc  sounds like it might be a bug if the keyboard type is not selectable
[09:29] <sladen> aleka: would you normally enter text using a meta-key such as Alt-Gr ?
[09:30] <sladen> aleka: have you installed   'language-support-am' 
[09:34] <aleka> sladen, I have not installed that pakacge
[09:34] <aleka> package**
[09:35] <aleka> Setting up language-support-am (6.06+20060413) ...
[09:35] <aleka> Error: 'am' is not a supported language or locale
[09:36] <aleka> This is the error I got installing language-support-am
[09:38] <sladen> aleka: it's depending on im-switcher and SCIM, so presumbly there's some special way of typing
[09:39] <ogra> Burgwork, there shouldnt be probs with LTSP ... dont worry ... 
[09:40] <ogra> Burgwork, l-r-m isnt included in ltsp, it grabs a lot of mem through the tmpfs for the modules
[09:41] <sladen> aleka: can you install  language-pack-am  aswell
[09:41] <Burgwork> ogra: ah, but the context I was speaking of was the use of opengl screensavers on thin clients
[09:42] <ogra> ah, yeah thats something the LTSP_CLIENT variable is for, youre right
[09:42] <aleka> sladen...ma not sure I understand..
[09:43] <sladen> aleka: can you please *both*  'language-support-am' and 'language-pack-am'
[09:44] <aleka> sorry...
[09:44] <aleka> didn see the different b/n support and package
[09:44] <aleka> Setting up language-pack-am-base (6.10+20061019) ...
[09:44] <aleka> Generating locales...
[09:44] <aleka>   am_ET.UTF-8... done
[09:44] <aleka> Generation complete.
[09:44] <sladen> aleka: and file a bug report at  https://launchpad.net/distros/ubuntu/+source/language-support-am/+filebug  with the error message that you pasted above
[09:44] <sladen> aleka: please don't paste large amounts
[09:49] <aleka> sladen.. reported the bug  ( Bug #73530), apologies if I didn;t follow standard bug reporting procedures
[09:49] <Ubugtu> Malone bug 73530 in language-support-am "error at the end of installation, language support am" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73530
[09:52] <gnomefreak> what country is am?
[09:53] <crimsun> 15:52 <@feh> rumour has it, .am is Armenia (+374)
[09:53] <lehaid> i'm wondering, if i want to compile the ubuntu kernel to an image with the same settings the offical one gets, where can i find information on that ?
[09:55] <lehaid> any dev here compiles the kernel ?
[09:56] <jdong> lehaid: there's wiki articles on doing that
[09:57] <sladen> crimsun: funky.  The Description: says Amharic
[09:57] <sladen> crimsun: maybe the packages are wrong
[09:57] <lehaid> jdong: any recommended?
[09:58] <jdong> lehaid: searching...
[09:58] <jdong> lehaid: https://help.ubuntu.com/community/Kernel/Compile?highlight=%28Kernel%29
[09:58] <lehaid> searching != recommended, but ok
[09:58] <jdong> that's a really good one for Edgy
[09:58] <lehaid> actually looking for 6.06
[09:58] <lehaid> it is diffrent?
[09:59] <jdong> lehaid: it actually is, unfortunately
[09:59] <jdong> lehaid: only in 6.10 were the build scripts made such that it's easy to customize and build your own variants
[09:59] <jdong> lehaid: might wanna ask in #ubuntu-kernel about 6.06
[09:59] <jdong> lehaid: but basically you'll be getting the "linux-source-2.6.15" source package and tweaking it
[10:00] <aleka> sladen am is for amharic and the country would be .et, ethiopia..
[10:05] <sladen> aleka: yup, that's what I thought
[10:05] <sladen> aleka: you'd probably have  am_ET for your locale
[10:08] <jdong> any mono guys around?
[10:08] <lehaid> there is a mono dev channel
[10:08] <lehaid> altough on a diffrent irc network
[10:09] <jdong> lehaid: I was looking for Ubuntu mono packages
[10:09] <jdong> packagers*
[10:09] <ajmitch> jdong: yes?
[10:09] <slomo> jdong: yeah?
[10:09] <jdong> ajmitch: can you look at bug 73512 for me and comment on the sanity?
[10:09] <Ubugtu> Malone bug 73512 in edgy-backports "Please backport beagle from feisty to edgy" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73512
[10:09] <jdong> the state-of-mono is outta my scope :)
[10:11] <slomo> jdong: should be fine... backporting cli-common is probably a sane idea anyway for any other mono backports you want to do
[10:12] <jdong> slomo: ok, thanks
[10:13] <jdong> when's next archive day, btw?
[10:19] <lehaid> i want to compile an kernel module which packages should i have besides the linux-headers or is it enough ?
[10:20] <jdong> lehaid: apart from the linux-headers that matches your running kernel, build-essential...
[10:20] <jdong> lehaid: and that should do the trick, unless your module in particular demands more
[10:20] <jdong> (for example, madwifi wants uuencode, etc)
[10:20] <lehaid> it's an iptables one, so i put iptables-dev as well
[10:29] <jdong> is there anythign worthwhile new in the new dvd+rw-tools release?
[10:37] <lehaid> jdong: shouldn't /usr/include/linux hold the kernel headers?
[10:38] <jdong> lehaid: no....
[10:38] <jdong> lehaid: /usr/src does
[10:38] <lehaid> hmm
[10:38] <lehaid> yeah, it installs them there
[10:38] <jdong> lehaid: /usr/include/linux are the linux headers that are used to build the userland
[10:38] <lehaid> but modules look for them there
[10:38] <jdong> which do not have to match up with your kernel version
[10:39] <jdong> modules look in /lib/modules/`uname -r`/build
[10:39] <lehaid> which does not exist
[10:39] <lehaid> without having the sources
[10:39] <jdong> then you didn't install the right headers
[10:39] <jdong> they certainly do exist when linux-headers-`uname -r` is installed
[10:39] <lehaid> it's installed /lib/modules/`uname -r` exists
[10:39] <lehaid> the build doesnt
[10:39] <lehaid> (didn't install sources, only headers)
[10:40] <lehaid> ahh
[10:40] <lehaid> blah
[10:40] <lehaid> my bad
[10:40] <lehaid> didnt have the up2date linux-headers for the current one
[10:43] <lehaid> ok, works, thanks again ! :)
[10:45] <lehaid> hmm.. in what package does 'stdio.h' exist?
[10:47] <geser> libc6-dev
[10:48] <jdong> hmm, can qemu allow me to build amd64 packages on i386?
[10:48] <jdong> (albeit very very slowly)
[10:50] <HrdwrBoB> what's wrong with cross compiling?
[10:51] <jdong> HrdwrBoB: <ignorant_hat>how would I go about doing that?</ignorant_hat>
[10:54] <lehaid> how can i get Makefile to show me compilation flags isntead of CC <filena,e> ?
[10:54] <HrdwrBoB> depends on what you're doing.. google is your friend
[10:55] <lehaid> compiling an module 
[10:55] <lehaid> ?
[10:58] <lehaid> dont know what to look for
[11:02] <lehaid> nobody knows how to make make show me the compile flag 
[11:02] <lehaid> ?
[11:02] <lehaid> not as verbose as make -n 
[11:05] <ogra> pitti, argh, looks like we will need cmake in main for the new cdrecord
[11:06] <ogra> (cdrkit/wodim)
[11:26] <pitti> ogra: that doesn't sound too scary
[11:27] <ogra> well 2 "makes" ? do we need that ?
[11:28] <pitti> ogra: 2? I bet we have at least 5 build systems in main, probably more
[11:28] <ogra> hmm, and a transition from cdrecord to wodim will be needed
[11:30] <lehaid> thanks for all the help, got what i wanted to work, good day (especially jdong) !
[11:30] <pitti> ogra: make, autotools, ant, scons, etc.
[11:30] <ogra> yep, i just thought there were security reasons or duplication reasons why we didnt have it in main yet ...
[11:33] <ogra> how did cdrkit get to main anyway, that looks strange ...
[11:57] <keescook> say... can an ubuntu-core-dev sponsor some main uploads for me?  http://people.ubuntu.com/~kees/feisty-uploads/
[11:58] <keescook> I can publish main security updates everywhere but feisty.  ;)
[11:58] <keescook> libmusicbrainz is just a merge, tar is a security update to match the one about to be published for the other releases.
[12:03] <crimsun> keescook: done.