[12:01] <sivang> sorry :-/
[12:03] <qbertibur> hi
[12:04] <sivang> Keybuk: so , just drop that snippet and we got less 5 idle seconds for gnome boot no ;-) ?
[12:04] <sivang> hi qbertibur 
[12:04] <Keybuk> sivang: for us it'd be at the "start gdm" stage
[12:04] <Keybuk> ajmitch says daniels already did that
[12:05] <qbertibur> What is the ubuntu way to view dvb-s?
[12:05] <zyga> is there any #ubuntu-hardware or similar?
[12:08] <sivang> qbertibur: this is a development channel, could you please ask this in #ubuntu? this is not quite the place to ask such questions.
[12:08] <ajmitch> sivang: "  * Disable font server in dexconf for mad startup time victory.
[12:08] <ajmitch> "
[12:09] <ajmitch> so it's disabled in xorg.conf rather than in the code
[12:09] <sivang> lol
[12:09] <sivang> yes, nice
[12:09] <zyga> with very rare updates
[12:10] <qbertibur> Is anybody using a SkyStar2 with ubuntu?
[12:11] <sivang> 's late here. Good night all
[12:21] <blueyed> Could someone please bump the priority for http://bugzilla.ubuntu.com/show_bug.cgi?id=10080 (pppoeconf). There is a confirmed patch and it is critical for release! infinity said he would take a look at it.
[01:13] <Lathiat> \sh_away: bug ref for?
[01:13] <Lathiat> \sh_away: oh, 10080
[01:13] <Lathiat> \sh_away: we found the problem
[01:14] <Lathiat> \sh_away: the patch to ppp in debain which brings an interface up before using it was backed out
[01:14] <Lathiat> \sh_away: because it was rejected upstream
[01:40] <bob2>  7416 rob       16   0  217m 4124 1100 S  0.0  0.8   8:44.01 nm-applet
[01:40] <tseng> nice!
[01:42] <bob2> 217MB of virtual space is particularily special when I've disabled it
[01:44] <tseng> its angry
[01:48] <hub> bob2: it is a gnome program, right ?
[02:08] <didymo> lifeless: ping
[02:08] <didymo> lifeless: just committed bug report as you suggested. Bug 17434
[02:36] <lifeless> didymo: excellent
[02:54] <didymo> lifeless: I've joined the ubuntu-devel mailing list. Should I post the bug there as well or just bugzilla?
[02:54] <lifeless> didymo: only one place
[02:54] <lifeless> just bugzilla is sufficient.
[03:05] <didymo> lifeless: cool thanks will leave it then.
[03:25] <bob2> the new tomboy logo is way less tintinish
[03:31] <tseng> bob2: upstream hates it
[03:31] <tseng> i told him i had to replace it and there was a long string of words begining with f
[03:32] <bob2> haha
[03:33] <lifeless> I liked tintin
[03:34] <tseng> id like to see you tell that to debian-legal
[03:34] <bob2> I felt better entrusting my notes to tintin than to two squares
[03:34] <magnon> two LINKED squares, mind you
[03:34] <lifeless> that makes *all* the difference
[03:34] <daniels> bug #17394: logo contains 100% less tintin
[03:35] <magnon> bah, fooled me, daniel
[03:36] <bob2> that bug does 100% seb, tho
[03:38] <magnon> tseng: so, in essence: "Sorry, we can't replace the infringing logo, because upstream keeps swearing at us"
[03:43] <tseng> god
[03:44] <tseng> between one giggling with his gf and the other yelling at his team on battlefield 2
[03:44] <tseng> who should die first?
[03:44] <magnon> Ate, n.: "The goddess of criminal rashness and consequent punishment."
[03:45] <bob2> tseng: the battlefield guy
[03:45] <magnon> totally
[03:45] <bob2> * > gamers
[03:45] <magnon> hey, I played wow this summer
[03:45] <tseng> erm
[03:45] <tseng> welcome back?
[03:45] <bob2> magnon: be glad you only lost 3 months to it!
[03:46] <magnon> well, until I figured out that noone were up for playing socially and casually, just out for items and uh, creds
[03:46] <magnon> so I guess, * > gamers
[03:46] <magnon> (yes, that's the first mmorpg or anything big I've ever played anywaY)
[03:46] <magnon> *y
[03:46] <magnon> bob2: didn't play that much :P
[03:46] <bob2> I lost a friend to wow for about 3 months...now he's clean and his gf is hooked
[03:46] <magnon> haha
[03:46] <bob2> I've been trying to wean them off it with a nice clean coke addiction
[03:46] <magnon> jesus, gaim so needs support for MSN Spaces
[03:47] <magnon> people always go "hey, check my new picture or blog thing or whatever", I go "URL?" all I get back is "just check my space!"
[03:47] <bob2> is that MS's "I can't believe it's not livejournal!"?
[03:47] <magnon> yeah
[03:47] <tseng> uh
[03:47] <magnon> and you get new nicknames for it, unlike MSN profiles which are profiles.msn.com/addy@hotmail.com or something
[03:48] <magnon> the bastards
[03:48] <bob2> wow, that's so awesome!!1111
[03:48] <magnon> you forgot "one"
[03:48] <daniels> bob2: s/livejournal/myspace/
[03:48] <daniels> the name kind of gives it away
[03:48] <magnon> I always thought of myspace as the place people hosted warez when I was 13
[03:48] <bob2> ah
[03:48] <magnon> apparently they innovated
[03:48] <tseng> liberate ex infernis
[03:48] <daniels> magnon: no, that was master.d.o
[03:49] <magnon> :D
[03:49] <tseng> substite myspace or livejournal or whatever
[03:49] <magnon> anyway
[03:49] <magnon> apparently one of my pretty intellectual friends write on MSN Spaces, people say he writes good articles in his "blog"
[03:50] <magnon> I should propose to host it I guess, so I won't have to search, I can't fucking find it
[03:50] <tseng> give him a wordpress install
[03:50] <magnon> and when I asks for the URL, he's just saying "duh, space"
[03:50] <tseng> or leave him to rot
[03:52] <Amaranth> his "space" is based on his MSN Messenger name, iirc
[03:52] <magnon> it's not, as far as I can see
[03:52] <magnon> it has nothing to do with his addy, and people's nicknames are often 150chars+ long :P
[03:55] <magnon> Jesus christ.
[03:55] <magnon> these spaces things have support for msn smileys
[03:55] <magnon> ugh
[03:55] <magnon> it's worse than deviantart
[03:56] <HrdwrBob> mmm tarts that use debian
[04:02] <magnon> right, so
[04:02] <magnon> I'm supposed to set up this server now
[04:02] <magnon> and the guy who installed it opened ssh and gave me a user, but forgot to add me to adm
[04:02] <magnon> anyone got a hoary 0-day? :P
[04:03] <bob2> hoary's default kernel has local root exploites
[04:03] <bob2> iirc
[04:04] <ajmitch> depends if he grabbed security updates & rebooted, I guess :)
[04:04] <magnon> it has 2.6.10-5 atm
[04:04] <magnon> dunno if that's recent or not tbh
[04:04] <bob2> that's old
[04:04] <magnon> cool
[04:04] <magnon> ah, he just woke up actually, sent me an sms
[04:05] <magnon> I guess that's a better way. :)
[04:05] <mjg59> bob2: Uhm. I thought we managed not to break the ABI in Hoary security updates?
[04:05] <bob2> hrm, you're right, ignore me
[04:06] <mjg59> 2.6.10-5 could be the latest
[04:07] <magnon> never mind, got access now.
[04:09] <magnon> and yes, it's the latest
[04:09] <wasabi> Hmm so what's the best file system to use for a loopback file system?
[04:09] <wasabi> I notice cramfs.
[04:25] <Lathiat> wasabi: depends what you want to do with it
[05:06] <roxville> hello folks :)
[05:06] <roxville> ok, i'll try to make it short:
[05:07] <roxville> i read about a screen vid capture app called istanbul possibly getting incl in ubuntu or edubuntu
[05:07] <roxville> anybody know anything about it?
[05:07] <roxville> coz i can't find it anywhere
[05:13] <Burgundavia> roxville, this is really a #ubuntu question, but it is in Breezy
[05:13] <Keybuk> syndicate scott% apt-cache policy istanbul
[05:14] <Keybuk> istanbul:
[05:14] <Keybuk>   Installed: (none)
[05:14] <Keybuk>   Candidate: 0.1.1-0ubuntu2
[05:14] <Keybuk>   Version table:
[05:14] <Keybuk>      0.1.1-0ubuntu2 0
[05:14] <roxville> ok Burgundavia thank you! :)
[05:17] <Keybuk>         500 http://archive.ubuntu.com breezy/universe Packages
[06:27] <Treenaks> The danes are coming! ?
[06:29] <fabbione> yeah yeah i am around :)
[06:29] <fabbione> morning
[06:29] <fabbione> magnon: i am not danish, but i live in dk
[06:30] <jsgotangco> don't forget the pastries
[06:30] <fabbione> magnon: anything i can do?
[06:30] <magnon> fabbione: was looking for danish whitepages online, but I found them at last :)
[06:32] <Keybuk> morning bella
[06:32] <fabbione> bella Scott
[06:34] <Keybuk> ah, freeze work
[06:34] <Keybuk> week even
[06:38] <fabbione> Keybuk: how did it go the race yesterday?
[06:39] <lifeless> wow, nautilus is fragile
[06:39] <lifeless> its refusing to show my sfs mounted tree :[
[06:39] <lifeless> segsigv's R us
[06:40] <bob2> maybe it's taking a strong moral stance against SFS
[06:41] <Keybuk> fabbione: did you not watch it?!
[06:41] <fabbione> Keybuk: no
[06:41] <Keybuk> :-(
[06:41] <Keybuk> it was a brilliant one
[06:41] <fabbione> i was preparing hoary/warty security updates
[06:41] <Keybuk> Kimi and Alonso started from the back row and finished 1/2
[06:41] <fabbione> putting up cieling lists.. and whatever my wife was asking me to :)
[06:41] <fabbione> ah
[06:41] <Keybuk> some utterly fantastic moves
[06:41] <fabbione> ehhee
[06:42] <Keybuk> Alonso passed Schumacher flat out at 130R, with the UK commentator screaming "oh my god! he can't do that! he's gonna die!" or words to that effect
[06:42] <Keybuk> and Kimi only went into first on the LAST LAP
[06:42] <fabbione> ahahah
[06:42] <Keybuk> great race
[06:43] <Keybuk> hasn't sold the constructors yet either, so the last race of the season will end that
[06:43] <fabbione> did the client managed?
[06:43] <Keybuk> though Ferrari are now solidly third
[06:43] <Keybuk> worked fine, once the server started working (it was out for about the first 5 mins)
[06:43] <fabbione> oh
[06:43] <fabbione> cool
[06:43] <fabbione> we should probably consider make a package and stick it in dapper
[06:43] <fabbione> if you want to be my upstream, i can be your packager :P
[06:45] <Keybuk> yeah, need to tidy a few things with it first, etc.
[06:45] <Keybuk> but should be packaged in plenty of time for the new season <g>
[06:45] <fabbione> ehehe
[06:46] <fabbione> i need to look at the code
[06:46] <Keybuk> it's in bzr
[06:46] <fabbione> i am sure you already separated the decoding engine from the curses display.. didn't you?
[06:46] <Keybuk> http://www.netsplit.com/bzr/live-f1/
[06:46] <Keybuk> yup
[06:46] <fabbione> yes i saw that..
[06:46] <Keybuk> you can drop in a replacement display.c to do gtk+ or whatever
[06:46] <fabbione> i need to look at bzr first ;)
[06:47] <fabbione> or just make the display.c modular
[06:47] <fabbione> so we can just build all the display we want at once
[06:47] <Keybuk> yup, either works
[06:48] <Keybuk> some of it probably needs cleaning up too, the parsing engine is pretty separate but the "actually do shit with the packets" bit is pretty un-modular
[06:49] <fabbione> hmmm
[06:49] <Burgundavia> Keybuk, what does this fancy package do?
[06:49] <fabbione> what we need is the stream manager and a possible replay to feed the same kind of info to the decoding engine
[06:49] <Keybuk> Burgundavia: it's a client for the Formula 1 Live Timing stream
[06:50] <fabbione> the decoding engine to export a standard struct of into to the clients.. that could be a curses viewer or gtk or the recorder
[06:50] <fabbione> i don't think we need much more than that
[06:50] <Keybuk> yeah, I figure that packet.c should maintain a "state of the race" and call a function pointer to update the screen
[06:51] <fabbione> we might want to consider to open a 3 stream sessions to gather weather info and the last one with the positions..
[06:51] <fabbione> if they don't come all on the same stream
[06:51] <Keybuk> they all come on the same stream
[06:51] <fabbione> Keybuk: even better
[06:52] <Keybuk> the reason it doesn't display weather right now is simply that it's not useful static information -- "the track is 23C, so what?  what was it 5 minutes ago, and before that?"
[06:52] <Keybuk> it needs a graphical client to show graphs and stuff
[06:52] <fabbione> yes..
[06:52] <Keybuk> I also want the core to keep history of the race ... right now it just holds the current table, and doesn't record previous values and stuff
[06:52] <wasabi> So what would be the best way to leave a Ubuntu system in a totally unconfigured state and have d-i or debconf or whatever run at first boot?
[06:52] <Keybuk> cause I've only just really figured out what everything means
[06:52] <wasabi> d-i would need to run, right?
[06:52] <fabbione> wasabi: oem install
[06:52] <Keybuk> wasabi: you mean exactly like the Live CD? :p
[06:53] <wasabi> Yup.
[06:53] <Keybuk> or indeed the oem installer
[06:53] <Treenaks> Keybuk: you're re-implementing the formula1.com java applet in C? :)
[06:53] <Keybuk> Treenaks: implemented
[06:53] <wasabi> The oem installer? Eh?
[06:53] <Treenaks> Keybuk: cool :)
[06:53] <wasabi> New thing I've missed.
[06:53] <wasabi> (i'm working with a debootstrap, too)
[06:53] <fabbione> wasabi: use the install cd (breezy) and type oem at the install
[06:53] <fabbione> s/install/boot prompt/
[06:53] <fabbione> enjoy
[06:53] <fabbione> Treenaks: yeah it looks neat ;)
[06:54] <Treenaks> too bad the season's almost over
[06:54] <Keybuk> wasabi: https://wiki.ubuntu.com/OEMInstaller
[06:54] <fabbione> Treenaks: at least we know Keybuk can code something that works and he will be busy enough NOT to break dpkg before elease :
[06:54] <fabbione> release even ;)
[06:54] <Treenaks> fabbione: ;)
[06:54] <Keybuk> tsk, iwj's been breaking dpkg, not me
[06:54] <fabbione> Keybuk: it's to easy to blame iwj :)
[06:55] <Keybuk> I've been breaking udev and hotplug instead
[06:55] <fabbione> right.. that too
[06:55] <daniels> which aren't at all critical
[06:55] <fabbione> daniels: exactly :)
[06:55] <wasabi> interesting.
[06:55] <fabbione> wasabi: you can't
[06:56] <fabbione> there is no need to install that in a chroot
[06:56] <wasabi> Sure there is.
[06:56] <wasabi> I'm building a flash-based system, custom root image.
[06:56] <wasabi> Need to be able to have the user config at first boot. Pretty simple.
[06:56] <fabbione> oh
[06:56] <fabbione> wasabi: that's what oem offers you
[06:56] <wasabi> Yeah, just trying to figure out how to install and set it up.
[06:56] <wasabi> in a chroot. ;)
[06:57] <wasabi> So when the root is mounted/booted it runs.
[06:58] <wasabi> Cool, ya just install it.
[06:58] <wasabi> nifty.
[06:59] <wasabi> Hmm. Depends on python and glade and thus gtk.
[06:59] <wasabi> Needs to be smaller. =(
[07:00] <Treenaks> wasabi: http://lists.ubuntu.com/archives/ubuntu-devel/2005-October/011903.html :)
[07:14] <pitti> Hi
[07:14] <ajmitch> hey pitti 
[07:19] <fabbione> hey pitti
[07:19] <pitti> Hi ajmitch, hi fabbione 
[07:19] <fabbione> pitti: i did forward you a couple of mails..
[07:20] <fabbione> you might want to do some kind of CAN dance. it's all public... or at least the patch is
[07:21] <pitti> fabbione: sure; can you please send me a GIT or similar link? the mitre guys need a reference
[07:22] <fabbione> i don't have one, but i can search for it
[07:22] <fabbione> that patch just landed in my inbox
[07:23] <fabbione> just gimme a few secs
[07:24] <fabbione> http://www.kernel.org/git/?p=linux/kernel/git/davem/sparc-2.6.git;a=commit;h=ba6399334dd8a75bd295de26496196c720abae0a
[07:25] <Keybuk> I really should play with git sometime
[07:25] <Keybuk> and hg, and darcs, and monotone, and ...
[07:26] <fabbione> Keybuk: i heard that git isn't really that fantastic as it claims to be.. it tends to corrupt the archive pretty easily
[07:26] <bob2> kernel developers writing untested code?????
[07:28] <fabbione> bob2: tsk..
[07:29] <fabbione> kernel developers don't need to test code
[07:29] <dilinger> hey, testing kernel code is dangerous.  you could lose a filesystem, ya know :P
[07:29] <fabbione> dilinger: lol
[07:30] <dilinger> fabbione: i'm pretty unimpressed w/ git
[07:30] <dilinger> i use it daily, but it doesn't seem to really scale from what i've seen of kernel.org/git
[07:30] <fabbione> so are many kernel developers
[07:30] <dilinger> i still have high hopes for bzr :)
[07:31] <\sh> daniels: ping
[07:31] <dilinger> initial import/commit/diff tests are pretty impressive (on 2.4 kernel trees), even on my horribly slow laptop.  branching and merging still needs a lot of work, though :/
[07:32] <\sh> daniels: do u know something about problems with i810 driver, since friday there is no external output anymore on the laptop of a colleague, before the last xorg he had dual display on his laptop
[08:10] <wasabi> hmmm. Guess I need to create /etc/inittab to launch oem config.
[08:10] <wasabi> Doesn't seem to be documented much.
[08:17] <infinity> elmo : Can you sync expect-tcl8.3 from unstable?... It fixes the FTBFS we just saw in -autotest (and nothing else)
[08:21] <HiddenWolf> pitti, ping
[08:21] <pitti> Hi HiddenWolf 
[08:22] <pitti> HiddenWolf: I'm changing computers, back in 2 minutes
[08:23] <infinity> 0 upgraded, 0 newly installed, 882 to remove and 0 not upgraded.
[08:23] <infinity> Need to get 0B of archives.
[08:23] <infinity> After unpacking 1675MB disk space will be freed.
[08:23] <infinity> Cleaning chroots is fun...
[08:24] <fabbione> infinity: ahha where is that?
[08:24] <infinity> (a chroot that had -desktop and -live installed for some testing)
[08:26] <pitti> HiddenWolf: what's up?
[08:26] <mdke> has jeff started on his badger tour already?
[08:27] <HiddenWolf> pitti, noticed my unmounted/unused cardreader slots are showing up in gtkfilechooser suddenly. flash card, usb ports, etc.
[08:31] <HiddenWolf> pitti, doesn't show up in nautilus or the likes.
[08:33] <pitti> HiddenWolf: hmm, I'm afraid it's a bit too late for Breezy for that, unless it is a very safe patc
[08:33] <pitti> HiddenWolf: is there a bug about it?
[08:33] <HiddenWolf> pitti, I haven't posted one, just really got around to checking it.
[08:34] <HiddenWolf> pitti, it even shows my floppy drive. ;)
[08:35] <pitti> HiddenWolf: hmm, that works fine for me with an USB stick
[08:36] <fabbione> doko: oo2 works perfectly on sparc :)
[08:36] <pitti> HiddenWolf: if I unmount it, I still see it in the dialog
[08:36] <pitti> HiddenWolf: and when I click on it, it is automatically mounted again
[08:36] <pitti> HiddenWolf: just like in the computer place
[08:36] <pitti> HiddenWolf: this might even be considered as a feature :-)
[08:37] <HiddenWolf> pitti, it's rather dumbish to show an usb port without an usb-stick into it.
[08:37] <HiddenWolf> pitti, might be that my monitor is a hub/cardreader, so it's confusing the matter.
[08:38] <pitti> HiddenWolf: ah, you don't mean "unmounted", but "removed"?
[08:38] <dholbach> good morning
[08:38] <pitti> HiddenWolf: WFM - as soon as I unplug the USB stick (with an open filechooser), the drives instantly disappear
[08:38] <pitti> Hi dholbach 
[08:39] <HiddenWolf> hi
[08:39] <pitti> HiddenWolf: you seem to encounter a special case, can you please file a bug then? It does not seem to be generally broken
[08:39] <HiddenWolf> pitti, the weird thing is, nautilus does it right.
[08:40] <infinity> pitti : It shows readers with no media in them.  Is that a feature?
[08:40] <HiddenWolf> pitti, I'd think the same code would be used for nautilus/places and the file chooser.
[08:40] <infinity> OTOH, nauituls is showing me my readers too, so whatever.  I must be seeing a different behaviour. :)
[08:40] <pitti> HiddenWolf: I only tried a filechooser and the computer place
[08:41] <pitti> infinity: not sure whether this is a feature, probably not
[08:41] <dholbach> hey martin, HiddenWolf, infinity :)
[08:41] <pitti> infinity: but well, it shows my CD-ROM drive, which is the same case
[08:41] <infinity> Yeah, exactly.  Hence why I assumed it was a feature.
[08:42] <pitti> in any case it is nothing that should stop breezy :-)
[08:42] <HiddenWolf> pitti, yeah, but if I take out the cdrom, the entry goes away.
[08:42] <pitti> it might be a bit confusing, though
[08:42] <HiddenWolf> pitti, HOLD the presses! :)
[08:43] <HiddenWolf> There is still one annoying bug I want fixed.
[08:43] <HiddenWolf> Pretty trivial, but nobody bothered, and I can't. :(
[08:43] <pitti> $ LANG=C df
[08:43] <pitti> df: cannot read table of mounted filesystems
[08:43] <pitti> hmmmm?
[08:46] <infinity> pitti : No mtab?
[08:46] <pitti> infinity: yes, nevermind, /etc/mtab is 0 bytes
[08:46] <pitti> I was just surprised
[08:47] <infinity> That's okay, I'm more interested in why I don't have a /proc/kcore in 2.6.13...
[08:47] <infinity> Not that I want it, but some braindead debian/rules is looking for it as proof that proc is mounted.
[08:48] <HiddenWolf> infinity, give it a brain?
[08:48] <infinity> Yeah, but I'm curious about where kcore went. :)
[08:49] <HiddenWolf> http://www.google.nl/search?q=%22ohh+kcore%2C+where+are+you%3F%22&sourceid=mozilla-search&start=0&start=0&ie=utf-8&oe=utf-8&client=firefox&rls=org.mozilla:en-US:unofficial
[08:49] <HiddenWolf> ?
[08:49] <fabbione> doko: btw.. OO2 on sparc64 is faster than on amd64
[08:49] <fabbione> no kidding :)
[08:51] <infinity> Oh, it appears I explicitely disabled kcore in my config.  That was WEEKS ago, how am I supposed ot remember these things?
[08:53] <HiddenWolf> infinity, hah
[08:53] <HiddenWolf> infinity, why would you do that.
[08:54] <infinity> I don't think I did it intentionally, actually.
[08:54] <infinity> Hence why I don't remember. :)
[08:55] <infinity> Oh, cause it's disabled by default now upstream.
[08:55] <infinity> Erm, no it's not.  Oh, I have no idea.
[08:56] <infinity> I fully approve of not having kcore anyway, so whatever.  Happy accident.
[08:58] <fabbione> infinity: do you happen to know if we have some crack around that given a Sources.gz in input, it will unpack automatically (and keep updated) all the unpacked sources in a dir?
[08:58] <fabbione> remove the obsolete and stuff like that?
[08:58] <infinity> No, but I could do it in shell in about 5 minutes, I guess.
[08:59] <fabbione> well the first unpack it's easy
[08:59] <fabbione> that's not an issue
[08:59] <infinity> (Which may be why no such thing exists)
[08:59] <fabbione> but update/remove it's the pain
[09:00] <infinity> Well, are you expecting it to merge changes with unpackaed sources, or just remove the old and unpack the new?
[09:00] <infinity> The latter is easy.
[09:00] <infinity> The former would require a bit more effort.
[09:00] <fabbione> the latter
[09:00] <fabbione> no i don't expect RCS magig
[09:00] <fabbione> magic
[09:00] <fabbione> but only if there is a new one
[09:01] <infinity> Sure, so you just compare the versoin in Sources.gz with the version in your current .dsc
[09:01] <infinity> If they don't match, out goes the diff, dsc, and build-tree, and you re-update.  If you have two orig.tar.gzs after update (cause the update was a new upstream), remove the older orig.
[09:01] <fabbione> well if you need to hack it, i can do it myself :) i was just curious to know if there was something done already
[09:01] <infinity> (Do it in that order, so you can reuse the orig, if it matches)
[09:02] <`anthony> so the whois client in breezy right now doesn't know about the new .travel top level domain. Is this the sort of thing that should get fixed before the final release, or can it wait for a bugfix?
[09:02] <`anthony> hm. Or the .eu TLD
[09:02] <fabbione> `anthony: breezy will get only security updates
[09:02] <fabbione> so that's dapper stuff
[09:02] <infinity> Adding TLDs to whois is pretty trivial, I don't see why we couldn't do it pre-release without destabilising.
[09:02] <`anthony> fabbione: sigh. 
[09:02] <infinity> Doing it post-release is likely a no-no, though.
[09:03] <fabbione> infinity: i would consider them features :P
[09:03] <fabbione> but whatever ;)
[09:03] <infinity> Yeah, but tiny/trivial ones.  Worth passing by mdz anyway.
[09:04] <infinity> It's not a complete rewrite or anything (like when whois had to be hacked to handle the new world order of whois redirects)
[09:04] <`anthony> I'm mystified as to why the whois client appears to have all this stuff hardcoded into it, but whatever... ;) 
[09:04] <`anthony> whois is a disaster of a protocol, anyway.
[09:04] <infinity> `anthony : Because there's no canonical source for which whois servers to query for which TLDs that doesn't move itself every three weeks.
[09:04] <infinity> `anthony : The only way to make it work at all is to hardcode the initial discovery into whois itself.
[09:05] <`anthony> infinity: I meant "instead of having a text file with the seed info"
[09:05] <Treenaks> have one on a central server somewhere :)
[09:06] <\sh> hmmm....I should reactivate my kwhois tool ;)
[09:06] <infinity> `anthony : Oh, yes, that's just a crusty holdover of the Old World Order, where having it hardcoded made sense.  I'm sure Md would take patches to make it use a configuration in /etc (or, one in /usr/share, overridden by /etc)
[09:06] <lucas> hi
[09:08] <lucas> I'd like to point out bug #17415
[09:15] <dholbach> lucas: i just had a look at the packages, that need ruby in main, it's kdebindings (which provides libqt0-ruby1.8 - which is used by amarok) and redland-bindings which is used by redland-utils - it might help to build the new debian package and test those packages if they run perfectly well
[09:17] <lucas> note that it was been in debian for 18 days without any important bug report
[09:19] <sivang> Morning all
[09:19] <dholbach> lucas: yes, it's just that users will have to live with it for 6-18 months, so if an expert double checked it (could be somebody else too, somebody of the kubuntu crew for amarok maybe), it would give people more confidence in the sync
[09:20] <lucas> I don't think amarok is an issue : it only suggests libqt0-ruby1.8
[09:21] <dholbach> oh, i see
[09:21] <dholbach> i just flicked through the rdepends
[09:31] <\sh> dholbach: amarok uses those bindings for additional scripts...just like it's using python-kde3 
[09:31] <dholbach> i see
[09:31] <dholbach> it's just what i saw on first glance
[09:32] <\sh> dholbach: and redland-bindings (binary package librdf-ruby) is in universe (source in main right)
[09:35] <dholbach> redland-bindings is in main though
[09:35] <\sh> dholbach: the source yes, the resulting binary for ruby-bindings not...
[09:36] <\sh> dholbach: if it's only a rebuild with new ruby, I think that's fine but needs an ok by mdz/kamion ;)
[09:36] <dholbach> \sh: i hope so
[09:39] <\sh> damn..
[09:39] <\sh> xview needs X11/bitmaps/xlogo64 which is not in xbitmaps but in tendra...
[09:40] <\sh> but i think it's not the right build-dep for xlogo64...
[09:45] <infinity> mvo!
[09:45] <infinity> I've been looking for you.
[09:46] <mvo> hey infinity 
[09:46] <mvo> good morning
[09:46] <fabbione> mvo: apt sucks :
[09:46] <fabbione> P
[09:46] <infinity> mvo : Are apt's DPkg::Post-Invoke hooks just sent to "sh -c"?
[09:46] <pitti> Hi mvo, hi chmj
[09:47] <infinity> mvo : If so, you may want to change update-notifier's to have a "[ -d /var/lib/update-notifier/ ]  &&" before the touch.
[09:47] <infinity> mvo : Purging update-notifier makes apt whine about the lack of directory, and exit non-zero.
[09:48] <mvo> infinity: yes, that should work, thanks
[09:48] <mvo> pitti: hi!
[09:48] <infinity> mvo : Just discovered while cleaning a chroot that had ubuntu-desktop installed in it. :)
[09:48] <daniels> \sh: no idea, sorry
[09:48] <daniels> \sh: the i810 driver hasn't changed in that respect.  if it's terminating with a SIGILL, then his BIOS is screwed and it's a GCC bug, again.
[09:50] <\sh> daniels: I'll investigate, because I'm not sure if it's really i810 related or just a b0rked tft :(
[09:50] <\sh> daniels: thx anyways
[09:51] <daniels> \sh: np
[09:51] <fabbione> Mithrandir: ping?
[09:51] <\sh> daniels: or no...other thing...dpkg-reconfigure xserver-xorg called inside a Xorg session, during the monitor detection it was falling back to plain console and gdm doesn't come up again...
[09:52] <daniels> \sh: that sounds a lot like broken BIOS + gcc bug (the gcc bug triggers the particular broken BIOS behaviour) to me
[09:52] <\sh> daniels: hp/compaq nc6120
[09:53] <Treenaks> daniels: and what can we do about 14007 ?
[09:53] <Treenaks> daniels: (any ideas?)
[09:53] <\sh> daniels: called from console with running X session, works as expected
[09:54] <Treenaks> daniels: uh, not 14007
[09:55] <Treenaks> daniels: I meant 14043
[09:55] <daniels> Treenaks: without the hardware or co-operation from ati, I have no clue, tbh
[09:56] <Treenaks> daniels: ok, too bad :(
[09:56] <Treenaks> daniels: you'll be in Montreal, right?
[09:56] <daniels> yeah
[09:56] <chmj> hey pitti 
[09:58] <mdke> is anyone other than jdub able to add blogs to planet.u.c?
[10:00] <jsgotangco> mdke, heh you into blogs too?
[10:00] <mdke> not majorly, i just thought I would like to post things occasionally
[10:00] <dholbach> good morning mvo, seb128 
[10:00] <seb128> hey dholbach
[10:00] <pitti> Hi seb128 
[10:01] <mdke> elmo, can you add blogs to planet.u.c?
[10:01] <mdke> or any other generous soul
[10:01] <seb128> hey pitti
[10:01] <\sh> bah...real life work...brb
[10:03] <janimo> pitti, pmount no longer has async option because it's default?
[10:03] <Mithrandir> fabbione: pong
[10:04] <pitti> janimo: yes
[10:04] <fabbione> Mithrandir: yo.. did you get to try that amd64 kernel?
[10:04] <fabbione> Mithrandir: if not.. don't worry.. i just did
[10:04] <Mithrandir> fabbione: no, sorry, the network here at home decided to implode last night so I spent some time trying to track that down.  Then I managed to get a nice and evil headache which is still present.
[10:05] <Mithrandir> fabbione: thanks dude.
[10:05] <fabbione> no problem, thanks anyway
[10:07] <daniels> seb128: morning sebarino
[10:08] <seb128> hey daniels
[10:10] <daniels> could everyone on i386 please grab libxkbfile1_7.0.0-3_i386.deb and xkeyboard-config_0.6-6_all.deb from http://people.ubuntu.com/~daniels/, install them, type setxkbmap -print | xkbcomp - :0, and tell me if your keymap still works fine or not? pay particular attention to your right alt key.
[10:12] <lucas> daniels: still works fine here
[10:14] <\sh> una momenta...
[10:14] <\sh> Warning:          Type "ONE_LEVEL" has 1 levels, but <RALT> has 2 symbols Ignoring extra symbols
[10:14] <\sh> and a lot of other warnings
[10:15] <\sh> but altGr+{q,2,7,8,9,0,e} works 
[10:15] <daniels> ignore the other warnings, but the RALT bit is juicy
[10:15] <\sh> so german layout here
[10:15] <daniels> what's the symbols line from setxkbmap -print?
[10:15] <seb128> X Error of failed request:  BadValue (integer parameter out of range for operation)
[10:15] <seb128>   Major opcode of failed request:  148 (XKEYBOARD)
[10:15] <seb128>   Minor opcode of failed request:  9 (XkbSetMap)
[10:15] <\sh> xkb_symbols   { include "pc(pc105)+de"  };
[10:15] <daniels> (notably, xk-c 0.6-6 fixes grp:alts_toggle.)
[10:15] <mvo> daniels: works here too
[10:16] <seb128> daniels: score :)
[10:16] <dholbach> elmo: could you please sync xmame from sid?
[10:17] <mvo> daniels: my is xkb_symbols   { include "pc(pc102)+de+ctrl(swapcaps)"   };
[10:17] <daniels> cool, thanks
[10:17] <daniels> seb128: huh, if you do setxkbmap -print | xkbcomp - :0?
[10:17] <daniels> seb128: if so, anything interesting at the bottom of Xorg.0.log?
[10:19] <seb128> daniels: I've the same with the package before update
[10:20] <seb128> still this bug with fr, us keymaps
[10:20] <seb128> grumpf
[10:20] <lucas> I get this with an fr keymap :
[10:20] <lucas> Warning:          Type "ONE_LEVEL" has 1 levels, but <RALT> has 2 symbols
[10:20] <lucas>                   Ignoring extra symbols
[10:20] <lucas> but it still works fine
[10:20] <fabbione> seb128: no no no no.. the real bug is the country you leave in :P
[10:20] <daniels> okay, hold on a second
[10:22] <seb128> daniels: http://pastebin.com/388840
[10:23] <seb128> daniels: Xorg.0.log on chinstrap if you want it
[10:27] <daniels> hmm, okay
[10:27] <daniels> give me a second, need to look a little deeper into this
[10:29] <zoot_> hi, any LTSP devs here?
[10:34] <dholbach> zoot_: #ltsp?
[10:35] <zoot_> dholbach: am on there too, no replies, so i thought i'd checkin here, since the wiki asks to give feedback to ubuntu-devel ;)
[10:35] <dholbach> ah i see
[10:35] <dholbach> maybe ubuntu-devel@? :)
[10:36] <zoot_> yep ;)
[10:36] <janimo> zoot_, all ok with the kioks?
[10:36] <janimo> in /etc I mean
[10:36] <zoot_> hi janimo - yes, so far, thx very much!
[10:36] <janimo> cool :)
[10:37] <zoot_> tho i see there's still a mix of /etc/xdg and /etc/X11/xdg
[10:37] <zoot_> that's amongst the xfce pkgs
[10:37] <janimo> yes only the xserverrc is in /etc/X11 I think that's ok to leave for now
[10:37] <zoot_> it works, so yeah, think so to ;)
[10:37] <janimo> are there other kiosk settings besides the panel stuff?
[10:38] <janimo> I think xfce-session has some
[10:39] <zoot_> yeah, let me find the url..
[10:39] <zoot_> also for desktop menu
[10:40] <daniels> okay, please download xkeyboard-config_0.6-6_all.deb again and try that
[10:41] <seb128> dpkg: regarding xkeyboard-config_0.6-6_all.deb containing xkeyboard-config:
[10:41] <seb128>  xkeyboard-config conflicts with libxkbfile1 (<< 7.0.0-3)
[10:41] <seb128>   libxkbfile1 (version 7.0.0-2) is installed.
[10:41] <daniels> seb128: yes, so install that too :)
[10:42] <daniels> you can't compile keymaps from new xk-c with old libxkbfile1
[10:42] <seb128> from where?
[10:42] <daniels> same place
[10:42] <seb128>  libxkbfile1_7.0.0-3_i386.deb 
[10:42] <seb128> that's all
[10:42] <daniels> amd64?
[10:43] <seb128> hum?
[10:43] <seb128> no i386
[10:43] <daniels> which architecture are you on?
[10:43] <seb128> http://people.ubuntu.com/~daniels/
[10:43] <seb128> i386
[10:43] <daniels> okay ... so download and install that?
[10:43] <seb128> ups
[10:45] <mvo> Kamion: would http://people.ubuntu.com/~mvo/update-notifier_0.40.14.debdiff be ok for breezy?
[10:45] <seb128> X Error of failed request:  BadValue (integer parameter out of range for operation)
[10:45] <seb128>   Major opcode of failed request:  148 (XKEYBOARD)
[10:45] <seb128>   Minor opcode of failed request:  9 (XkbSetMap)
[10:45] <daniels> wtf
[10:46] <seb128> hum
[10:46] <seb128> now that's fine
[10:47] <seb128>         xkb_symbols   { include "pc(pc105)+fr+group(alts_toggle)"       }; 
[10:47] <seb128> changed to
[10:47] <seb128>         xkb_symbols   { include "pc(pc105)+fr"  };
[10:47] <zoot_> janimo: can't find kiosk refs to items other than -panel and -session. will let you know. the -panel url is:  http://www.loculus.nl/xfce/documentation/docs-4.2/xfce4-panel.html#panel-kiosk but i know there's a ref elsewhere with more options...
[10:48] <daniels> seb128: okay, could you run xkbcomp - :0, but add +group(alts_toggle) to the symbols line?
[10:48] <daniels> if it still gives you shit, please send me the output of xkbcomp -xkb - -, with that input
[10:51] <seb128> daniels: setxkbmap -model pc105 -layout 'fr' -option 'grp:alts_toggle'
[10:51] <seb128> daniels: with that you should get the crasher, no?
[10:52] <daniels> daniels@ephemera:~/canonical/xorg/data/xkeyboard-config% setxkbmap -model pc105 -layout fr -option grp:alts_toggle
[10:52] <seb128> daniels: xkb_file on chinstrap
[10:52] <daniels> daniels@ephemera:~/canonical/xorg/data/xkeyboard-config%
[10:52] <seb128> daniels: now | xkbcomp - :0 it
[10:52] <daniels> qrgh azerty cqn bite ,e
[10:52] <Treenaks> daniels: ;)
[10:52] <seb128> yeah
[10:53] <daniels> seb128: pleqse type setxkb,qp )lqyout us for ,e
[10:53] <HiddenWolf> who is the ubuntu-calendar artwork person?
[10:53] <seb128> daniels: ah ah
[10:53] <Treenaks> daniels: "command not found" :P
[10:53] <jsgotangco> HiddenWolf, jdub is your best best
[10:53] <daniels> seriously
[10:53] <sivang> seb128: you aware that the link for evo is broken from the panel?
[10:54] <sivang> Bon Jour BTW :)
[10:54] <seb128> it's not
[10:54] <HiddenWolf> jdub, ping
[10:54] <sivang> evolution-2.2 - "No such file or directory"
[10:54] <sivang> but that minus the version works
[10:54] <seb128> ah, that's an update
[10:54] <seb128> known "issue"
[10:55] <daniels> seb128: my testcase was layout us,de,fr option grp:alts_toggle
[10:55] <daniels> which worked fine
[10:55] <seb128> daniels: with fr alone? :)
[10:55] <daniels> seb128: with fr alone, I just make a dick of myself on IRC, as above
[10:55] <daniels> (i like testing with multiple layouts, because I'm just two keypresses away from getting myself back into a sensible keymap.  usually I test with de, and I've got that memorised now, but azerty is a whole different kettle of fish.)
[10:56] <seb128> :)
[10:57] <seb128> daniels: right, with 2 keymaps that works fine
[10:57] <seb128> daniels: the issue is if I've only fr
[10:57] <daniels> weird.  works fine with just fr here.
[10:57] <seb128> $ setxkbmap -model pc105 -layout 'fr' -option 'grp:alts_toggle'
[10:57] <seb128> $ setxkbmap -print | xkbcomp - :0
[10:57] <seb128> X Error of failed request:  BadValue (integer parameter out of range for operation)
[10:57] <seb128>   Major opcode of failed request:  148 (XKEYBOARD)
[10:57] <seb128>   Minor opcode of failed request:  9 (XkbSetMap)
[10:57] <seb128>   Value in failed request:  0x16710002
[10:57] <seb128>   Serial number of failed request:  77
[10:57] <seb128>   Current serial number in output stream:  83
[10:58] <daniels> seb128: hrm
[10:58] <daniels> seb128: okay, that happens only if I clear out my options
[11:00] <daniels> oh, duh, god I'm stupid
[11:01] <seb128> ?
[11:02] <seb128> just curious, but are you playing with some code to fix that? or with the keyboard definitions?
[11:02] <daniels> seb128: s/group(alts_toggle)/level3(ralt_switch_for_alts_toggle)/
[11:02] <daniels> seb128: does -print give you group(alts_toggle), or level3(ralt_switch_for_alts_toggle)?
[11:02] <seb128>         xkb_symbols   { include "pc(pc105)+fr+group(alts_toggle)"       };
[11:03] <daniels> okay
[11:04] <daniels> grab http://people.ubuntu.com/~daniels/xorg, stick it in /etc/X11/xkb/rules, and try again
[11:04] <daniels> this time you should get level3(ralt_switch_for_alts_toggle); seems to work here
[11:05] <daniels> make that ~daniels/xorg/xorg
[11:07] <seb128>         xkb_symbols   { include "pc(pc105)+fr+level3(ralt_switch_for_alts_toggle)+group(alts_toggle)"   };
[11:07] <seb128> and it doesn't crash
[11:07] <seb128> and keyboard works fine
[11:07] <ericf> I can repeatedly crash nautilus, don't know why, but it seems to happen when browsing a dir via a symlink. How can I make a useful backtrace? I should use gdb for that, right? I'll file this in bugzilla, if I can give some useful information. Then question 2: is this question here inappropriate? #ubuntu isn't very responsive to this question. 
[11:09] <seb128> ericf: https://wiki.ubuntu.com/DebuggingProgramCrash
[11:09] <seb128> ericf: install libgnomevfs2-0-dbg nautilus-dbg and describe the version/arch you use and what to do exactly to get the issue
[11:10] <seb128> ericf: and this chan is not to speak about how to get a backtrace :)
[11:10] <ericf> seb128: Thank you. I'll have a look.
[11:10] <seb128> np
[11:11] <daniels> seb128: rockin, thanks a lot
[11:11] <seb128> np, thank *you* :)
[11:12] <daniels> i've pinged svu too, so we can shove this hack of a fix through
[11:12] <daniels> i started working on the real fix on the tram
[11:13] <seb128> what is difference between "group(alts_toggle)" and "level3(ralt_switch_for_alts_toggle)+group(alts_toggle)" exactly?
[11:14] <daniels> well, +group(alts_toggle) shouldn't be there
[11:14] <daniels> but, basically, group(alts_toggle), does this: key <RALT> { (some stuff blah blah) [ NoSymbol, ISO_Next_Group ]  }
[11:14] <seb128> that's to switch between layouts?
[11:14] <daniels> and level3(ralt_switch) does key <RALT> { (stuff) [ ISO_Level3_Shift ]  }
[11:15] <daniels> yeah
[11:15] <seb128> how do I use it?
[11:15] <daniels> so ralt_switch_for_alts_toggle does key <RALT> { [ ISO_Level3_Shift, ISO_Next_Group ]  }
[11:15] <daniels> setxkbmap -layout fr,us -option grp:alts_toggle
[11:15] <daniels> just press both alt keys at the same time, and you will have switched between fr and us
[11:15] <HiddenWolf> jdub, seb128, some patches attached to bug 8721, fairly trivial and obvious, is it possible to fix this?
[11:16] <seb128> daniels: right :)
[11:17] <seb128> HiddenWolf: a real patch instead of changing dir as a workaround would be nice
[11:18] <daniels> seb128: so the real fix is to let the xkb parser know that you just want to skip a specific level, so we don't have these stupid collisions
[11:18] <daniels> also there's a bit of handwavy alt-handling stuff that needs to get fixed
[11:18] <HiddenWolf> seb128, not my patch btw, but i'll look into it.
[11:19] <seb128> daniels: k
[11:23] <daniels> http://www.icedoutgear.com/DT39.php
[11:28] <tepsipakki> kamion: does the installer remove etc/dhclient.conf?
[11:29] <tepsipakki> kamion: from the image
[11:33] <sivang> seb128: re 14910, I managed to reproduce that on 2 machines of mine, but the bug report says it was fixed upstream.
[11:34] <sivang> seb128: (after upgrades, that is)
[11:37] <Lathiat> mjg59: about?
[11:38] <seb128> sivang: have you actually read it?
[11:46] <tepsipakki> kamion: yes it seems to do that, read the netcfg sources... bugger
[11:48] <carlos> pitti, could you give me an URL where I could get the tarball with your hoary's translations? so I can debug the missing strings?
[11:48] <pitti> Hi carlos 
[11:49] <seb128> hey carlos
[11:49] <carlos> pitti, btw, hi :-P
[11:49] <carlos> seb128, hi
[11:49] <pitti> carlos: http://people.ubuntu.com/~pitti/langpacks/hoary-final.tar.gz
[11:49] <pitti> carlos: did you get my email with the bugs I found?
[11:49] <mvo> pitti: when did you build it?
[11:50] <pitti> mvo: that tarball? right before hoary was released
[11:50] <Kamion> mvo: please either (a) mail me or (b) just upload and let it sit in the approvals queue - if you ask me on IRC, I'll lose stuff, amid the two-and-a-half days of #ubuntu-devel I've just had to trawl through
[11:50] <carlos> pitti, yeah, I'm debugging it atm
[11:50] <mvo> pitti: sorry, me should read more carefull :)
[11:50] <Kamion> tepsipakki: the installer doesn't remove anything, it either installs stuff or doesn't install it
[11:51] <pitti> mvo: actually I built that tarball at Friday, but it's contents is the same as in the zip file I built for haory
[11:51] <janimo> seb128, do you have suggestions how I should package a modified evince-0.4.0 for universe so as it doesn't conflict with the original
[11:51] <janimo> it is a diff to take out gnome dependencies from it
[11:51] <seb128> janimo: DON'T
[11:51] <mvo> pitti: I was wondering when you will build the breezy langpack (IIRC it was scheduled for today?)
[11:51] <seb128> janimo: no no no
[11:51] <pitti> mvo: it will be today, yes
[11:52] <janimo> would you take a patch for the original package to build two binaries?
[11:52] <seb128> janimo: we have enough packages without starting to have 2 packages for the same stuff
[11:52] <pitti> mvo: I'm waiting for carlos' tarball to arrive and my cronjob to finish
[11:52] <seb128> janimo: no
[11:52] <janimo> seb128, any other suggestion to have evince with gtk only for lighter desktops?
[11:52] <mvo> pitti: ok, thanks
[11:52] <janimo> in universe?
[11:52] <seb128> janimo: don't? :)
[11:52] <sivang> lol
[11:52] <janimo> seb128, that's not an option many people will like :)
[11:52] <carlos> pitti, the tarball is there already
[11:52] <seb128> janimo: seriously, having 2 source package for the same stuff is not a good idea
[11:53] <sivang> evince rulez
[11:53] <sivang> :)
[11:53] <tepsipakki> kamion: it does remove for example etc/dhclient.conf and etc/dhcp3/dhclient.conf if they exist (from the ramdisk, that is), see netcfg-common.c
[11:53] <janimo> seb128, xpdf, gpdf, kpdf, why evince then ;)?
[11:53] <carlos> pitti, I got the email an hour ago
[11:53] <pitti> -rw-r--r--  1 langpack langpack 278376521 Oct 10 09:37 buildd-rosetta-merged.tar.gz
[11:53] <janimo> not the same stuff, it's a friendly fork
[11:53] <seb128> janimo: that's the GNOME official pdf/ps viewer
[11:53] <pitti> oh, yay
[11:53] <azeem> janimo: why do you want to strip down evince anyway?
[11:53] <pitti> carlos: so my cronjob worked as well :)
[11:53] <tepsipakki> kamion: but I can change the dhclient-script which is enough, I think..
[11:53] <janimo> azeem, for xubuntu so it doesn not bring in gnoem stuff
[11:53] <seb128> janimo: it'll Depends on libgsf and GNOME office stuff anyway
[11:53] <mvo> janimo: I was under the impression that evince makes very heavy use of various gnome-libs, do you have a url for the patch?
[11:54] <carlos> pitti, cool!
[11:54] <janimo> seb128, what do you mean depends on gnome stuff office?now?
[11:54] <martink> mvo, it's in the evince-list archives at mail.gnome.org
[11:54] <azeem> janimo: why not use xpdf for that?
[11:54] <janimo> mvo, I sent it to the evince list last Thursday
[11:54] <janimo> azeem, it's ugly and not nearly as nice as evince
[11:54] <seb128> janimo: no, but they plan to open ppt files, etc
[11:55] <janimo> evinces niceties are mostly non-gnome related
[11:55] <Kamion> tepsipakki: ok, sorry, I'm really busy at the moment so if you can read the source and work it out yourself that's far preferable to asking me
[11:55] <janimo> seb128, ok but for now and the xubuntu releas it'd be nice. People seem excitded about it and they haven't even tested it :)
[11:55] <seb128> janimo: I'm against it
[11:55] <janimo> mvo, beware the pacth is ugly
[11:55] <seb128> we have double bugs, double upload, double package to patch, etc
[11:56] <tepsipakki> kamion: yes, sorry for the noise ;)
[11:56] <dholbach> janimo: might be a target for dapper?
[11:56] <seb128> janimo: did you get any reply upstream?
[11:56] <janimo> seb128, you are aginst me forking an open source project and providing it to ubuntu users on older machines? that's not nice of you ;)
[11:56] <dholbach> janimo: don't you go and blame seb :)
[11:56] <janimo> seb128, the evincelist seems dead
[11:56] <seb128> janimo: no, I'm against you putting extra work and duplicating the same code 
[11:57] <janimo> on IRC J blandford seemed skeptic a couple weeks ago of a gtk only evince, but avoiding to tell me what it would take to change it if I really wanted
[11:57] <seb128> janimo: a 1500 lines diff is not something we want to push the week of a stable
[11:57] <janimo> seb128, not really extra code, I mostly deleted stuff :)
[11:57] <seb128> still 1500 lines of changes
[11:58] <janimo> seb128, of course not in main, that's why I asked of advice under what name shoudl  I package it in universe so that it does _not_ conflict with evince
[11:58] <seb128> with some FIXME comments
[11:58] <Kamion> janimo: it's not possible to have two source/binary packages of the same name in main and universe; therefore they'd have to have different names and provide the same files etc., which is very nasty and difficult to do right, especially at short notice
[11:58] <seb128> janimo: having the same source package to main and universe with different names is not nice
[11:58] <Kamion> janimo: it would cause support problems in main
[11:59] <janimo> Kamion, seb128, of course different names, that's what I was asking
[11:59] <Kamion> janimo: yes, that doesn't make it non-nasty
[11:59] <seb128> and different path to install files, etc
[11:59] <janimo> seb128, there's lots of resued code in different packages
[11:59] <seb128> and we will get bugs about yours
[11:59] <Kamion> in fact it makes it worse if you do it wrong
[11:59] <Kamion> janimo: that's not an excuse for more
[11:59] <janimo> no, it can be renamed
[11:59] <daniels> does anyone here have a powerpc that can test XKB stuff?
[11:59] <janimo> I was going to call it evionce-gtk to give credit for upsteram but it can be called gtk-docview or piglet or anything
[12:00] <janimo> you won't get bugreports fro evince
[12:00] <daniels> pitti: word
[12:00] <seb128> you are willing to maintain the work yourself?
[12:00] <janimo> seb128, Conflicts: evince
[12:00] <seb128> s/work/fork/
[12:00] <janimo> seb128, yes I am 
[12:00] <mjg59> Lathiat: Hi
[12:00] <carlos> pitti, hmmm, did you patched Python to use language packs?
[12:00] <seb128> janimo: they have planned quite a lot of changes, you are going to keep the amount of work?
[12:00] <Lathiat> mjg59: is there any reason "Loading modules" would be different in kubuntu
[12:01] <Lathiat> mjg59: in kubuntu, it doesn't time out usplash
[12:01] <Lathiat> in ubuntu, it did every time
[12:01] <daniels> pitti: could you please get the libxkbfile sources from p.u.c/~daniels/, compile and install libxkbfile1, then install xkeyboard-config from p.u.c/~daniels?
[12:01] <mjg59> None that I can think of
[12:01] <seb128> janimo: I would say after 5.10 is better than to fork it now, 3 days before 5.10 with 0 feedback
[12:01] <janimo> seb128, not necessarily. It can remain at this stage as evince 0.4 I find it already excellent for pdf and ps
[12:01] <pitti> carlos: I didn't, but since it obviously works, somebody apparently did it
[12:01] <Kamion> janimo: Conflicts: evince is not acceptable at this stage
[12:01] <Lathiat> maybe something new went into kubuntus RC iso that changes on install but not upgrade or something
[12:01] <seb128> janimo: and poppler will change and you will have to follow
[12:01] <pitti> daniels: how urgent is this? 
[12:01] <daniels> pitti: i'd really like to get it in breezy
[12:02] <pitti> daniels: i. e. would doing this in 1 hour block you?
[12:02] <janimo> seb128, for the first release of xubuntu, for dapper we may have something cleaner
[12:02] <carlos> pitti, your hoary tarball has sane-backends.LANG directories...
[12:02] <daniels> pitti: oh, not at all; need to go make dinner and do something that doesn't involve a computer for a while
[12:02] <janimo> so poppler changing after breezy does not affect me 
[12:02] <sivang> be back later, going to eat
[12:02] <carlos> that looks like the python layout but are not exactly the same...
[12:02] <pitti> carlos: these will just be ignored
[12:02] <pitti> carlos: the breezy importer knows how to handle these
[12:03] <pitti> carlos: but not the hoary branch
[12:03] <carlos> pitti, what's that?
[12:03] <pitti> carlos: there was a reason why I branched langpack-o-matic for hoary, but I don't know it currently
[12:03] <janimo> seb128, as for conflicting paths I can install elseweher
[12:04] <janimo> most files are nautilus.gconf/gnome related anyway, thiose won;t be in anyway
[12:04] <carlos> pitti, should I ignore it then?
[12:04] <pitti> carlos: mainly because I want to use the same scripts that I used for the initial langpacks, I guess
[12:04] <pitti> carlos: yes
[12:05] <seb128> janimo: I'm against pushing a new code untested 
[12:05] <seb128> 3 days before a stable
[12:05] <seb128> but that's my opinion
[12:06] <Kamion> janimo: standard release management practices suggest that anything that xubuntu requires for release should really have been in the archive for some time, not be shoved in at the last minute
[12:06] <janimo> seb128, but it's universe, and does not affect reputation of gnome/evince
[12:06] <seb128> still
[12:06] <Kamion> janimo: I'd strongly, strongly suggest using a different PDF viewer, even if it's not as pretty.
[12:06] <janimo> Kamion, there are still xubuntu packages in the NEW queue (artwork etc)
[12:06] <Kamion> janimo: no there aren't
[12:07] <Kamion> xubuntu-artwork was rejected; you should have mail
[12:07] <Kamion> (licensing)
[12:07] <janimo> Kamion, I answered to elmo but no respponse
[12:07] <janimo> you saw those mails too?
[12:07] <Kamion> janimo: you're supposed to reupload, not argue :)
[12:07] <janimo> I copy pasted kubuntu license
[12:08] <Kamion> janimo: I didn't see the mails originally, but I can see the file in the reject queue with a copy of the mail
[12:08] <janimo> but I wasn;t arguing but asking why the kubuntu art license isn;t approproate for xu
[12:08] <Kamion> The Ubuntu artwork is licensed under the CC-BY-SA 2.5.  Given the Xubuntu artwork
[12:08] <Kamion> seems to be derived from the Ubuntu artwork, it seems only reasonable it should follow
[12:08] <Kamion> the Ubuntu artwork's licensing rather than relicensing it?
[12:08] <janimo> I happened to copy from kubuntu-artwork
[12:08] <janimo> not edubuntu
[12:08] <Kamion> if it's derived from the Ubuntu artwork, I don't see that "but Kubuntu gets it wrong" is a justification for you getting it wrong too?
[12:08] <janimo> I did not know kubuntu got it wrong
[12:09] <janimo> and after elmo's mail I asked
[12:09] <Kamion> janimo: in any case, nothing will happen until you reupload
[12:09] <janimo> ok, thanks
[12:09] <janimo> I thought I'd wait till elmo sees my mail
[12:09] <Kamion> the file is not in the new queue, it's rejected; that's permanent for that upload
[12:09] <Kamion> janimo: that would seem wise, certainly
[12:09] <Kamion> janimo: but alternatively, is it so hard to licence it under the union of the licences of anything you've copied from?
[12:09] <janimo> so as a curiosity is kubuntu art license wrong then?
[12:09] <Kamion> you could just do that, trivially
[12:10] <Kamion> janimo: I have no idea; I have not checked it
[12:10] <janimo> no it;s not hard
[12:10] <janimo> I made the pa ckage based on edubuntu artwork. I am inncocent :)
[12:10] <mjg59> Derived works of CC-SA have to be CC-SA, from what I can remember
[12:10] <Kamion> janimo: I think the term is "complicit", technically ;)
[12:10] <janimo> oh kubuntu, sorry
[12:11] <janimo> oh well I reupload with CC license, sigh
[12:11] <Kamion> anyhow, this argument is pointless. why not just get it right once you've been informed of the problem?
[12:11] <janimo> I thought that by asking I may raise and lead to solving another prob, the possible kubuntu license one
[12:11] <Kamion> that's fine, but in the meantime you should resolve your own release blockers
[12:11] <janimo> I want to learn something from what I do not just comply :)
[12:12] <janimo> and is xfce4-taskmanager not in NEW either?
[12:12] <janimo> another MOTU supposedly uploaded it
[12:13] <janimo> ivoks to be precise
[12:15] <mvo> ping carlos 
[12:15] <Kamion> janimo: there is nothing matching x* in NEW
[12:15] <carlos> mvo, pong
[12:15] <janimo> Kamion, I'll upload it myself then, thanks
[12:15] <segfault> ogra: ping
[12:16] <Kamion> janimo: the person who uploaded xfce4-taskmanager forgot to build with -sa
[12:16] <segfault> ogra_: looks like the desktop fix is likely to have something wrong with its charset
[12:16] <Kamion> cjwatson@jackass:~/queue$ cat reject/xfce4-taskmanager_0.3.1-0ubuntu1_source.reason
[12:16] <Kamion> Rejected: xfce4-taskmanager_0.3.1-0ubuntu1.dsc refers to xfce4-taskmanager_0.3.1.orig.tar.gz, but I can't find it in the queue or in the pool.
[12:16] <Kamion> they should have got mail about that
[12:16] <bob2> jackass is the most awesome ftpmaster. name, ever.
[12:16] <segfault> ogra_: looks like the file is in ISO-8859-1 and the string i wrote is in UTF-8
[12:17] <janimo> Kamion, thanks
[12:17] <ogra_> hmm, shouldnt be.... my vi is working in utf8 normally... i'll check and fix it :)
[12:17] <ogra_> thanks for reporting
[12:17] <segfault> ogra_: thanks
[12:29] <seb128> What package is to blame for a "eth0 no set up during the boot"?
[12:30] <seb128> I've to run "sudo dhclient" 
[12:30] <seb128> hoary sets it correctly on the same box
[12:30] <mjg59> seb128: Is eth0 wireless?
[12:30] <ogra_> seb128, hotplug/dhclient i'll look up the bug...
[12:30] <seb128> yeah, ipw2200
[12:31] <mjg59> Wireless devices don't seem to get set up during the install
[12:31] <mjg59> Kamion: ^ ?
[12:31] <ogra_> seb128, 13497
[12:32] <seb128> nop
[12:32] <seb128> sudo ifup eth0 works fine here
[12:32] <ogra_> seb128, as it does for all others in the bug
[12:32] <ogra_> the hotplug script doesnt work right...
[12:32] <ogra_> it will also work if you add "auto eth0" to your interfaces file...
[12:33] <seb128> I should not have to
[12:33] <seb128> hotplug is supposed to get it
[12:34] <mvo> ogra_: do you got the /msg from me?
[12:34] <infinity> seb128 : Does /etc/init.d/hotplug restart get it after you're done booting?
[12:36] <ogra> mvo, in fact i was answering since 20min already :/ damn freenode
[12:36] <mvo> ogra: sorry, I should change this "unfiltered" thing (or whatever it is called)
[12:37] <seb128> infinity: yep, it does
[12:37] <ogra> mvo, nope, i should notice if i'm not logged in :)
[12:37] <infinity> seb128 : Yell at Keybuk, really loudly, then.
[12:37] <seb128> infinity: will do, thanks
[12:37] <daniels> infinity: i find that works for most things
[12:37] <daniels> well, s/works/applies/, I suppose
[12:38] <ogra> yelling at Keybuk or restarting hotplug ? 
[12:40] <pef> hello
[12:41] <bob2> is "buch" a german word?
[12:41] <zoot_> bob2: yes, book
[12:41] <Treenaks> bob2: Buch, yes. it means "book"
[12:41] <bob2> ah
[12:41] <Kamion> mjg59: atheros?
[12:42] <pef> does every FTBS has a bug  opened ?
[12:42] <daniels> mostly
[12:42] <mjg59> Kamion: No - plausibly something that was fixed
[12:42] <infinity> pef : Probably not.
[12:42] <mjg59> Kamion: But what does happen on Atheros systems?
[12:43] <pef> mmm and if I fix a FTBS which hasn't one bug report opened, should I open one ?
[12:43] <Kamion> mjg59: not supported on breezy; the card won't be recognised
[12:43] <mjg59> Kamion: Once the system is installed?
[12:43] <infinity> pef : In which package?
[12:43] <pef> infinity: it's just an example :)
[12:43] <Kamion> mjg59: once the system is installed, it should be fine, although you'll probably have to munge /e/n/i
[12:43] <mjg59> Kamion: Oh, FFS.
[12:43] <infinity> pef : Oh, sure, if you find a bug, file a bug and attach a patch.
[12:44] <infinity> pef : bugzilla for main, malone for universe.
[12:44] <mjg59> Kamion: (Yes, I know it's not your fault)
[12:44] <pef> infinity: ok, thank you
[12:44] <mjg59> Kamion: Do you know if that's been release-noted?
[12:45] <Kamion> mjg59: no, sorry
[12:46] <Kamion> Riddell: kubuntu MD5SUMS fixed, sorry about that
[12:47] <Kamion> elmo: I think I'll move the source images back to cdimage, which will free up 4GB or so
[12:47] <Kamion> DVDs, yeah, I guess
[12:57] <daniels> i thought l-r-m had properly-sized udebs now?
[12:57] <infinity> It does.
[12:57] <daniels> Kamion: btw, you're going to get a bunch of CD space back with -77
[12:57] <infinity> We don't, however have a binutils udeb (that change was vetoed), nor d-i hooks to link ath_hal.
[12:58] <daniels> ah, rockin.
[12:59] <Kamion> daniels: oh, driver stripping. cool
[12:59] <daniels> Kamion: novel, innit?
[01:00] <segfault> ogra: http://www.prognus.com.br/~carlos/xscreensaver_4.21-4ubuntu16.diff.gz
[01:00] <ogra> segfault, thanks a lot :)
[01:01] <Kamion> daniels: -77 looks OK to me but I'd rather leave xorg to mdz
[01:01] <daniels> Kamion: okay
[01:03] <sivang> back
[01:04] <Diziet> Any kubuntu people here ?  I'd like to talk to someone about 17390.
[01:06] <\sh> when riddell is not answering ;) 
[01:06] <\sh> Diziet: how can I help you
[01:07] <sabdfl> fabbione: ok, the "official" tag is added to my branch
[01:07] <sabdfl> should land this week
[01:08] <sabdfl> and be in production next week
[01:08] <fabbione> sabdfl: cool! thanks a lot
[01:08] <Diziet> Hello.  Could you take a look at 17390 and tell me if you agree with dennis@kaarsemaker.net ?  If so I'll reassign the bug.  Otherwise let's talk about what to do about the problem.
[01:08] <sabdfl> np
[01:08] <sabdfl> kinnison will be doing the import of breezy into launchpad at some stage
[01:08] <fabbione> sabdfl: i will still need your suppah powha to add the info tho...
[01:08] <sabdfl> discuss with him whether to reflect those architectures
[01:08] <sabdfl> hmm...
[01:08] <sabdfl> they are on our servers, right?
[01:09] <fabbione> sabdfl: the .deb's yes
[01:09] <\sh> Diziet: well...actually I don't know what riddell thinks about it...i'll discuss this with him
[01:09] <fabbione> they are all on jackass and splitted with rsync magic to archive and ports
[01:11] <\sh> Diziet: the default homepage for ff is in ubuntu-artwork? 
[01:11] <ogra> \sh, -docs
[01:11] <Diziet> The default homepage is a file which is currently provided by -docs, yes.
[01:12] <Diziet> The file is in .../share/.../ubuntu-artwork because that's the package it used to be in.
[01:12] <\sh> Diziet: symlink is as well in /usr/share/ubuntu-artwork/home/index.html
[01:13] <Diziet> Oh, they've made it a symlink ?  OK.  That symlink is what Firefox is looking for.
[01:13] <Diziet> NB, see also https://wiki.ubuntu.com/BreezyFirefoxStartPageTranslation
[01:14] <Diziet> I think you'll find that you must perform the role quoted there as ubuntu-doc'.s
[01:14] <Diziet> s/\.s/s.
[01:14] <mjg59> Kamion: hotkey-setup-0.9ubuntu2 - can you approve that?
[01:15] <\sh> Diziet: so the link is provided by ff?
[01:17] <Kamion> mjg59: done
[01:17] <daniels> mdz: ping
[01:17] <Riddell> Diziet: it could be a link to alternatives and have an alternative provided by ubuntu-doc and kubuntu-doc
[01:18] <daniels> (or kubuntu-doc could divert.)
[01:18] <Riddell> yes
[01:21] <mjg59> Kamion: Thanks
[01:22] <Riddell> Kamion: is the kubuntu live CD set up to use the new winfoss?
[01:23] <Kamion> Riddell: yes
[01:23] <Kamion> as stated in #17125
[01:26] <Diziet> Alternatives are for when the situation depends on local admin configuration decisions, not when it depends on which packages are installed.  If kubuntu includes ubuntu-doc then yes, divert it.
[01:27] <Riddell> Kamion: great, thanks
[01:27] <\sh> Riddell: do we such a page like the index page of ubuntu-docs? we could make a link in kubuntu-docs and the problem would be solved at least until a better solution is found
[01:27] <\sh> +have 
[01:28] <Riddell> \sh: we have aboutkubuntu (which I havn't look at yet but will do today)
[01:28] <\sh> Riddell: so adding debian/links with the correct locations in kubuntu-docs, will solve it...well, bug is pendingupload  I would say ;)
[01:29] <daniels> \sh: then you'll need a Conflicts
[01:30] <\sh> daniels: yepp
[01:31] <\sh> a nicer solution would be that ff package is detecting the (k)ubuntu-desktop postinst and should set the link correctly to ubuntu/kubuntu-docs dir (including the locale for dapper)
[01:32] <carlos> pitti, you have an update about Hoary's language packs in your mailbox
[01:32] <carlos> pitti, thanks for the report
[01:32] <Kamion> \sh: a diversion would avoid the need for a conflict
[01:32] <Kamion> firefox should not be getting into detection of {ubuntu,kubuntu}-desktop; that way clearly lies madness
[01:32] <seb128> carlos: is there anything to mask all the xxx-... from the rosetta list? It makes the list and the stats crappy
[01:33] <carlos> seb128, will fix that today and ask for a fast cherry pick
[01:35] <torkel> dholbach: thanks for taking care of #17465 (e-d-s)
[01:35] <dholbach> torkel: thanks for fixing it :)
[01:37] <torkel> dholbach: well, as I was the only one suffering from it, I guess I had to fix it :-)
[01:37] <dholbach> you won't have been the only one :)
[01:38] <ogra> segfault, ping
[01:40] <seb128> carlos: thanks
[01:43] <seb128> carlos: same for "review-...." items?
[01:53] <Kamion> mvo: does your update-notifier change actually make any difference? note that '[ -d /nonexistent ]  && true' exits 1
[01:54] <Kamion> mvo: I would prefer either '[ ! -d /var/lib/update-notifier ]  || touch /var/lib/update-notifier/dpkg-run-stamp' or (perhaps clearer) 'if [ -d /var/lib/update-notifier ] ; then touch /var/lib/update-notifier/dpkg-run-stamp; fi'
[01:58] <mvo> Kamion: right. can you reject the upload?
[02:00] <bob2> holy god the monospace font in breezy is unreadable
[02:01] <ogra> bob2, thats your eyes :)
[02:02] <smurf> ... or your settings for font smoothing
[02:02] <Kamion> mvo: no, due to the way approvals work; just upload 0.40.15 superseding it
[02:07] <mvo> Kamion: thanks for the review. I uploaded a new version
[02:09] <pitti> carlos: thanks for the mail
[02:20] <dholbach> bbl
[02:23] <ogra> Kamion, any hint about http://cdimage.ubuntu.com/edubuntu/daily/current/report.html ? it doesnt have an OVERSIZED tag in the dir...so i wonder what broke...
[02:31] <j^> why is it that i need mp4 to watch go-opensource videos [http://www.go-opensource.org/go_open/news/download_go_open/]  but can watch aplle advertisement in Ogg Theora [http://theora.watchmactv.com/] ?
[02:38] <Kamion> ogra: hmm, I'm not sure, nothing jumps out at me as obviously problematic there
[02:38] <ogra> for me neither... i'll rip out kstars from ppc... probably it just overflowed silently ? 
[02:40] <Kamion> ogra: ah, yes, it overflowed
[02:40] <ogra> ok
[02:40] <Kamion> ogra: the .OVERSIZED thing isn't reliable once you get beyond 700MB, at the moment
[02:40] <ogra> ah, ok
[02:40] <Kamion> ogra: so it probably means your CDs are just ridiculously big and you need to sort that out ...
[02:40] <Kamion> CD 2 will only be filled with 17508534 bytes ...
[02:41] <ogra> Kamion, another thing, if you find a minute, could you wipe the default IP from the installer ? the install can manage now without a special range
[02:41] <ogra> is there a public place where i can see the second CD data ? 
[02:41] <Kamion> it can? I thought I tried that and you told me it still broke after the fix
[02:41] <Kamion> ogra: no
[02:42] <Kamion> but "reduce what you're putting on the CD by 20MB or so" should be enough information; it doesn't matter exactly what randomly happened to go on CD2
[02:42] <ogra> Kamion, ltsp can install now without braking the install.. that was the reason for putting it there first place
[02:42] <Kamion> -EPARSE, sorry, try again
[02:42] <ogra> the IP
[02:42] <Kamion> it wasn't that that broke; it was a bizarre message in netcfg
[02:43] <Kamion> which I thought I'd fixed, but IIRC you told me was still broken
[02:43] <ogra> i still have the message and i think its unlikely we'll solve it now... nd the installer doesnt need a default IP anymore:)
[02:43] <ogra> so just rip out the default and the message should be fine for now
[02:44] <ogra> (trying to go with the least amount of work for you here)
[02:45] <pitti> Mithrandir: we currently have cfengine-doc in main and cfengine in universe - should we unseed the former or seed the latter?
[02:45] <Kamion> or do I totally misremember?
[02:45] <pitti> Mithrandir: only having the documentation in main as the sole reason for keeping the source package in main seems silly
[02:45] <Kamion> ogra: I thought you asked me to put the IP address in because you didn't want to use DHCP and you thought the static address question was unreasonably complicated or something
[02:45] <ogra> Kamion, nope, you are right... but the reason for having a default there is gone... having it fixed with default would be nicer, but there is no urgent need for a default setting... 
[02:45] <Kamion> if I remove the preseed then you won't see the "you should never see this message" thing, certainly
[02:46] <Kamion> since that's triggered by some corner case in preseeding
[02:46] <ogra> exactly
[02:46] <ogra> but the IP should still be static...
[02:47] <ogra> ltsp-server broke in base-config when the dhcpd range didnt match the configured IP, that was the main reason for having a default... this part is fixed....
[02:47] <Kamion> ah, I see
[02:47] <Kamion> although I think you're editing history a bit, because the IP address thing went into debian-cd long before any of the LTSP stuff did
[02:47] <ogra> it would be nicer to have the right ip though... but i can live with a note in the install notes :)
[02:48] <Kamion> ogra: anyway, done
[02:48] <ogra> ta :)
[02:48] <j^> pitti wrt /dev/raw1394 not being accessible for plugdev users by default. any chance to work that out for breezy?
[02:49] <pitti> j^: sorry, I'm afraid not
[02:49] <pitti> j^: we can't fix upstream software to use /dev/video1394 that quickly
[02:49] <ogra> Kamion, the IP thing was planned from the beginning... but being forced to have a default was the thing that came later with ltsp...
[02:49] <j^> pitti you never will
[02:50] <ogra> (i'm probably just expressing myself wrongly :) )
[02:51] <j^> pitti as i said before, raw1394 is the way to go, with the new libiec61883. its the only way to i.e. get HDV(aka MPEG2-TS over FireWire) working
[02:51] <pitti> j^: ah, I remember that mess - this is such a broken design, there is no hope to get this right for breezy, I'm afraid
[02:52] <pitti> j^: this needs to be discussed with upstream and kernel guys
[02:52] <pitti> j^: we can't just allow all users to write to all Firewire devices, at least not without having evaluated the consequences properly
[02:53] <Kamion> ogra: ok
[02:53] <pitti> j^: there are more firewire devices than just video cams; Imagine Firewire hard disks, ethernet cards, and the like
[02:53] <j^> pitti will you do this, talk with upstream/kernel guys? because so far i am not aware of someone complaining about the "broken design", besides you
[02:54] <pitti> j^: if you are interested in that topic, can you do it yourself?
[02:54] <j^> pitti i see 4 firewire hard disks next to my camera, and i am dont care if i can access them via software
[02:54] <pitti> j^: maybe *you* don't care
[02:54] <j^> i can just drag and drop them on the floor
[02:54] <pitti> j^: I will care if a random user can write anythign to my firewire ethernet card
[02:54] <pitti> j^: so we can't open raw device access to all users *by default*
[02:55] <pitti> j^: if you change this locally for you in /etc/udev/permissions.rules, that's of course fine
[02:55] <j^> pitti, i am fine chaning that for myself
[02:55] <pitti> j^: but it is by far not a proper solution for the default
[02:55] <infinity> pitti : Good thing we just hired a Linux ieee1394 hacker, eh?
[02:55] <pitti> j^: getting DV cams work OOTB would of course rock, but we should not sacrifice security for it IMHO
[02:56] <pitti> infinity: we did?
[02:56] <j^> pitti but thats not why i am using ubuntu, its because i just had to "fix" a friends lapotp so i could capture video
[02:56] <infinity> pitti : BenC.
[02:56] <pitti> infinity: oh, cool; I didn't know that he hacks on firewire stuff
[02:57] <pitti> j^: I understand that it might be frustrating, but there apparently is no easy fix for breezy (or even a patch at all)
[02:57] <pitti> j^: from a design perspective it is silly for an user app to assume that it can write to raw devices
[02:57] <j^> pitti its also frustrating on the dev side, what way should i work on HDV support for gstreamer? and yes my goal would be that it works out of the box 
[02:58] <pitti> j^: so for me this mainly looks like a driver/user app interaction problem
[02:58] <pitti> j^: so the problem is that the /dev/dv1394 and /dev/video1394 do not offer an adequate interface?
[02:59] <j^> pitti no they dont, also not for AMDTP (audio and music)
[02:59] <pitti> j^: then it seems that this is the real bug
[03:00] <j^> pitti have you filed a bug report?
[03:00] <pitti> j^: no
[03:01] <pitti> j^: I'm not a good person to push this; I don't have video cams and the like
[03:01] <pitti> j^: talking with BenC could be enlightening, though
[03:01] <bddebian> Hello
[03:02] <j^> pitti if ben collins works for canonical now, you should talk to him
[03:02] <pitti> j^: I can do that, yes
[03:02] <pitti> j^: but I can't tell him more than "it does not appear to work, please fix it"
[03:02] <pitti> j^: which packages are affected by this?
[03:03] <pitti> j^: we should contact their upstreams and have them find a proper interface design with the kernel guys, instead of relaying this discussion over people who have NFC about the details
[03:04] <j^> pitti kino, gstreamer, dvgrab, ffmpeg
[03:04] <j^> pitti so far there pov is change permissions of /dev/raw1394
[03:13] <\sh> guys...dpkg-divert...when I set up a divertion of file X to be renamed to file Y, is it only done once, or will it be regocnized even in future...let's say, kubuntu-docs is installed by default during a kubuntu-installation. now the user installs ubuntu-desktop with ubuntu-docs as dep...will it rename the ubuntu-docs link to the mentioned name in the dpkg-divert list?
[03:14] <\sh> (regarding the ff default homepage problem) 
[03:14] <Kamion> diversions stay in effect
[03:14] <Kamion> dpkg honours existing diversions when installing new packages
[03:15] <Kamion> you should make sure to remove the diversion when the package is removed, of course
[03:16] <\sh> Kamion: sure..it's done..but if I install the kubuntu-docs package now, on a mix system, it divert the default ubuntu homepage to lets say index-ubuntu.html ...
[03:16] <Diziet> When _your_ package is removed, that is.
[03:17] <\sh> Kamion: means, when I'm starting up firefox, even in gnome, i have a kubuntu default homepage ;)
[03:17] <doko> Kamion: I added openoffice.org2-help-en-us to the desktop seed
[03:17] <Kamion> \sh: not much you can do about that
[03:17] <Diziet> Are you saying you want the startup page to depend on what desktop the particular user happens to be running ?
[03:17] <Kamion> doko: how big is it?
[03:17] <doko> 12MB
[03:17] <ogra> doko, how big is that ? 
[03:18] <Kamion> doko: er ... please revert that
[03:18] <Kamion> doko: it must be a dependency of language-support-en instead
[03:18] <doko> Kamion: hmm, ok
[03:18] <Kamion> doko: talk to pitti to make that happen
[03:18] <\sh> Kamion: ok..I wanted to be sure, that this way is ok
[03:18] <Kamion> I think this might screw the live CDs, incidentally :(
[03:19] <Kamion> only the install CDs currently have that kind of space available
[03:19] <Diziet> \sh: The diversion will do what you might hope.  But it's not possible to override it eg by the syadmin.
[03:19] <doko> Kamion: done
[03:19] <doko> pitti: ^^^
[03:20] <doko> could you add the openoffice.org2-help-* packages?
[03:21] <\sh> Diziet: well..then it's fixed...I'll let riddell test it on a pure kubuntu install and I think then we're ready to upload the package..let's see
[03:21] <\sh> Kamion: thx for the hint...it's a really handy tool ;)
[03:24] <pitti> doko: yes, I'll add the help files to l-support-*
[03:24] <pitti> doko: but 12 MB? that's nasty
[03:24] <doko> pitti: thanks
[03:24] <doko> pitti: why?
[03:24] <carlos> seb128, no, sorry, the review-* will take more time as we need to fix them one by one by hand
[03:25] <pitti> doko: because we already have overfull CDs
[03:25] <pitti> Kamion: I still fail to see how replacing langpacks by English support packages is beneficial for the majority of users... *sigh*
[03:25] <doko> pitti: no, mdz did remove thunderbird and ephipjany
[03:25] <pitti> doko: ah, fine
[03:25] <seb128> carlos: can I help on them?
[03:25] <seb128> grrrrrrr
[03:25] <pitti> doko: I will build new langpacks today anyway, in that course I will build new suport packages, too
[03:26] <seb128> epiphany-browser got dropped? :(
[03:26] <doko> pitti: for -en, you cannot depend on -help-en-us and -help-en-gb, that would be too big
[03:26] <pitti> doko: humm, right
[03:26] <pitti> doko: so which one should I pick?
[03:26] <Kamion> \sh: use it with care
[03:27] <doko> pitti: openoffice.org2-help-en-us
[03:27] <Kamion> doko: (it's nasty anyway, because of the live CD problem)
[03:27] <pitti> Kamion: should l-support-en only depend on -en-us, and we seed -en-gb? 
[03:27] <Kamion> doko: mdz's removals only affected the install CD
[03:27] <Kamion> pitti: yes
[03:27] <Kamion> it'll have to do
[03:27] <pitti> alright
[03:28] <Kamion> for dapper I'd like us to consider installing something that pops up in place of the oo.o2 help content and points people to online help
[03:28] <doko> Kamion: hmm, anything, I can do now?
[03:28] <Kamion> this sort of 12MB growth isn't sustainable
[03:28] <Kamion> doko: unless you can figure out how to do the above now ...
[03:29] <doko> Kamion: we don't have online help
[03:29] <Kamion> bah, y'all suck
[03:29] <pitti> or, pops up and tells to install language-support-foo
[03:29] <pitti> (or the help packages itself)
[03:29] <Kamion> pitti: language-support-en is installed by default; I doubt that will change
[03:29] <Kamion> but help packages, sure
[03:29] <pitti> Kamion: we could seed a subset of l-s-en to desktop and enrich l-support-en, too
[03:29] <doko> well, I can look at it, but I doubt, that I'll have sucess ...
[03:30] <pitti> Kamion: (not for breezy, of course)
[03:30] <pitti> I guess for breezy we just need to live with the current situation
[03:30] <\sh> Kamion: for sure :)
[03:30] <Kamion> pitti: that makes it unremovable by tools like language-selector, which seems like a regression
[03:31] <pitti> hmm, right
[03:31] <Kamion> pitti: we can't entirely live with it. #17123 is a blocker bug
[03:31] <Kamion> which I had almost got fixed before the oo.o2 help thing came along
[03:33] <sabdfl> jbailey: is your nightly bzr build up to date?
[03:34] <j^> pitti ah ok, so there is some discussion about raw1394 #10517 and #14330
[03:36] <jbailey> sabdfl: No.  Lifeless asked me to stop producing them while the shift to weaves happens.
[03:36] <sabdfl> ok thanks, hope to see it again soon now that weave landed
[03:37] <Diziet> Quiz question: what demented character set transformation (or sequence of transformations) attempts to represent U+00E9 (e', UTF-8 C3 A9) as C3 C2 A9 ?
[03:37] <pitti> sabdfl: weave is fully there now? great, can't wait to try it out :)
[03:37] <jbailey> sabdfl: Yup.  Basically it should be back when 0.1 has landed.
[03:39] <sabdfl> which is tomorrow am aussie time, tonight your time
[03:39] <janimo> Kamion, ping
[03:39] <jbailey> Diziet: Is that still unicode?  There's a whole compose section that I suspect nothing honours.
[03:39] <pitti> carloooooooos
[03:39] <pitti> carlos: missing translations in the breezy tarball alert!
[03:39] <Kamion> Diziet: UTF-8 C3 A9 run through iconv -f ISO-8859-1 -t UTF-8 gives C3 83 C2 A9, so maybe that's the start of it
[03:39] <Kamion> janimo: pong
[03:40] <janimo> Kamion. let me know when you have time for talking to me about seeds/buildd
[03:40] <Kamion> jbailey: hmm, ubuntu-docs is hyooooooge. Anything we can do to reduce the amount of it that we have to ship on the live CD?
[03:40] <carlos> pitti, ?
[03:40] <carlos> pitti, which module?
[03:40] <jbailey> sabdfl: Yup!  It will make it tonight if I'm awake when the aussies come on and give the go ahead, otherwise I'll email and look for the answer when I wake up.
[03:40] <pitti> carlos: I /msg
[03:40] <jbailey> Kamion: Yeah.  I'm trying to prune out unused graphics.
[03:40] <sabdfl> ok, cool thanks
[03:41] <carlos> ok
[03:41] <Kamion> janimo: perhaps it would be better if you sent me mail with what you need
[03:41] <sabdfl> actually, you know, a current build should be fine since it's post-rc2 which seems to have been well received
[03:41] <janimo> Kamion,ok
[03:41] <Kamion> janimo: cjwatson@ubuntu.com
[03:41] <Diziet> jbailey: No, the result isn't UTF-8.
[03:41] <Kamion> janimo: then I can deal with it regardless of whether we're both simultaneously online
[03:42] <Diziet> Maybe something once more thinks it's ISO-8859-1, and decides to drop the invalid 83.
[03:42] <Kamion> jbailey: ok, please do that urgently if you can - are there multiple documents in there at the moment that could perhaps be split out?
[03:44] <Diziet> kamion: I'm about to upload new firefox and mozilla to add -fomit-frame-pointer re 17276.  I hope that's OK.
[03:44] <Kamion> Diziet: yes, that's fine
[03:45] <infinity> Kamion : I'm about to upload Thunderbird 1.0.7 (security fixes ahoy), do you want me to walk through the debdiff or anything with you first, or is "I've audited it, and the changes are all small and sane" good enough for you?
[03:46] <infinity> (I've also been abusing it for the last while to make sure it's okay)
[03:46] <Kamion> infinity: new upstreams need special exception; I'd prefer to leave this one to mdz
[03:47] <Kamion> are *all* the changes security fixes?
[03:47] <pitti> Kamion: yes
[03:47] <pitti> Kamion: this time they stayed sane and did not add new features
[03:47] <pitti> Kamion: well, sorry, not all fixes are
[03:47] <pitti> Kamion: some patches fix regressions from previous security patches
[03:47] <infinity> It's not all security, a few small bugfixes.
[03:47] <Kamion> have the dependencies of the various locale packages been checked?
[03:47] <infinity> Nothing even close to a feature, though.
[03:48] <pitti> Kamion: yes
[03:48] <infinity> Kamion : Locales packages are fine, the only thing that needs rebuilding is enigmail, which I'll upload at the same time.
[03:48] <Kamion> ok, mdz should be up in a few hours, please run it past him
[03:48] <infinity> Alright..
[03:48] <Kamion> (or send mail and leave the packages in a location where somebody else can upload them later, if that's getting late for you)
[03:49] <infinity> Nah, I suck at sleeping anyway, I'll wait up.
[03:50] <Diziet> kamion: I already reviewed the diff between upstream firefox 1.0.6 and 1.0.7 and they seemed very good about it there.  Obviously I don't know for sure about Thunderbird.
[03:50] <hno73> Kamion: are you now pulling the Ubuntu-AMD winfoss from here?: http://www.theopencd.org/winfoss/ubuntu-AMD/current/ (it's about 7MB smaller than the old one)
[03:50] <Kamion> sabdfl: I'd like to consider taking python2.4-samba, python-opengl, and perhaps python-twisted out of the desktop seed. They (or their dependencies) are all large, and we're under serious space pressure on the live CD.
[03:51] <Kamion> hno73: yes
[03:51] <pitti> Kamion, mdz: for the records, I cannot use the current Rosetta tarball for building the gold langpacks, it has regressions again
[03:51] <sabdfl> Kamion: is the livecd an exact desktop clone?
[03:51] <Kamion> sabdfl: live CD includes all of ubuntu-desktop
[03:51] <pitti> Kamion, mdz: let's hope that these can be fixed by tomorrow
[03:51] <Kamion> and a few other bits, but only 8MB or so worth of other bits
[03:52] <pitti> Hi jdthood 
[03:52] <jdthood> hi ho
[03:52] <sabdfl> Kamion: can the livecd be different to desktop?
[03:52] <Kamion> we've squeezed pretty much all we can out without touching desktop
[03:52] <Kamion> sabdfl: it's theoretically possible, but there's no infrastructure for doing it at the moment; I'm not sure I want to try to create that infrastructure three days before release
[03:53] <Kamion> we thought we'd be OK for breezy without that
[03:53] <sabdfl> i'm not sure i want to weaken our python position in favour of winfoss
[03:53] <sabdfl> we already dropped thunderbird, right?
[03:54] <hno73> heh, looks like I'm against the well
[03:54] <Kamion> yes, but that only affected the install CD; thunderbird and epiphany were never on the live CD in the first place
[03:54] <Kamion> is python2.4-samba really a core part of our Python position? at 5MB+ it's an awfully large piece
[03:55] <infinity> Nothing depends on it, except the metapackages..
[03:55] <Kamion> it's about as big as the OO.o2 database and spreadsheet components put together
[03:55] <infinity> And, despite years of samba usage, I can't claim to have ever used it.
[03:56] <infinity> Take that for what it's worth.
[03:56] <hno73> The Ubuntu winfoss i386 is currently: OpenOffice2, Firefox and gaim
[03:56] <doko> I think we don't loose anything removing python2.4-samba
[03:57] <hno73> Firefox is 5mb, gaim is 6. We can't drop firefox though. the poor w32 users needs it :)
[03:57] <infinity> (I don't suppose we could sneak PuTTY on the WinFOSS side?... Then I'd always have a copy everywhere I went..)
[03:57] <infinity> :)
[03:58] <infinity> elmo : ping ... Can I get a sync of expect-tcl8.3 from Debian?... The only thing is changes is to fix the FTBFS we just saw in -autotest.
[03:58] <infinity> elmo : s/is changes/it changes/
[03:59] <\sh> well...I would remove gaim, cause there are other, better GPL alternattives for gaim on windows, e.g. mirandaIM
[03:59] <infinity> To each their own, I've used GAIM on Win32 for ages.
[04:00] <Kamion> python2.4-samba accounts for 30% of the total space consumed by all python packages in desktop; there are 126 such packages in total
[04:00] <Kamion> sabdfl: in any case, powerpc is also under pressure, and dropping WinFOSS won't help there
[04:00] <smurf> IMHO dropping python-twisted won't hurt too many people (if any)
[04:01] <sivang> lol, but really firefox on that cd is a must have, everytime someone cmoplains about viruses coming through it's explorer, I give them the WinFOSS cd :)
[04:01] <Kamion> smurf: I'm reluctant to push for that one
[04:01] <doko> smurf: on the desktop, it's not a loss
[04:01] <smurf> Kamion: people who really want to use twisted need the docs anyway
[04:02] <fabbione> Diziet: ping?
[04:02] <infinity> Diziet : BTW, I'll apply the no-strict-aliasing fix to Thunderbird as soon as I get approval for the 1.0.7 upload.
[04:02] <Kamion> samba's the big killer, at >5MB. opengl is 2.8MB or so, twisted is 1.2MB.
[04:02] <fabbione> infinity: thanks.. that's what i was going to ask to Diziet 
[04:03] <smurf> 5 MB is a lot of Python :-/
[04:03] <Diziet> fabb: Yes ?
[04:03] <Diziet> Oh, right.
[04:03] <fabbione> Diziet: moz-thunderbird :)
[04:04] <Kamion> it's almost all enormous .so files which are apparently big chunks of samba
[04:04] <Diziet> The change is trivial.  OPTFLAGS += -fno-strict-aliasing in the right point in debian/rules.
[04:04] <Kamion> or massive autogenerated bindings, I can't tell
[04:05] <Diziet> Near where it says -DDEBIAN.
[04:05] <Diziet> HTH.
[04:05] <Kamion> hno73: what are the applications in the amd64 tarball at the moment?
[04:06] <sabdfl> Kamion: ok, go ahead and drop them
[04:07] <sabdfl> infinity: is 1.0.7 coming into breezy?
[04:07] <infinity> sabdfl : That's the plan.
[04:07] <infinity> sabdfl : It's security fixes, plus a few conservative bugfixes, just waiting on mdz to wake up for approval.
[04:08] <pitti> sabdfl: if that happens, we will also put it into hoary and warty
[04:08] <ogra> Kamion, * python-imaging-sane # pulls in big packages, can cut it if needed
[04:08] <ogra> what about that one ? 
[04:10] <hno73> Kamion: firefox, abiword, gaim, thunderbird
[04:11] <doko> ogra: what do you save with python-imaging-sane
[04:11] <ogra> doko, no idea, i didnt make that comment in the seeds....
[04:11] <sabdfl> pitti: awesome, folks will be very happy i think
[04:11] <ogra> but looking at it i dont see the mentioned big deps...
[04:11] <doko> afaik libsane is needed by other packages as well
[04:12] <pitti> Kamion: since I don't add ooo-help-en-gb to l-support-en, shall I also remove ooo-l10n-en-gb from it? it would save 2.5 MB
[04:12] <hno73> Kamion: if you need to drop one I would say thunderbird
[04:12] <sabdfl> python-imaging-sane can go
[04:12] <smurf> doko: like OOo2 ;-)
[04:13] <doko> smurf: ?
[04:13] <pitti> sabdfl: hm, that's only 25 kB
[04:13] <segfault> ogra: pong
[04:13] <smurf> doko: wrt libsane
[04:13] <pitti> sabdfl: sorry, didn't look at the dependencies; python2.4-imaging is big
[04:14] <ogra> segfault, what arches do you have around there ? i'd like to prepare a xss package for testing for you... it seems the whole pt_BR.po file was set to iso-8859-1 
[04:14] <ogra> segfault, i'd like to change that... all others are utf-8
[04:14] <doko> pitti: python2.4-imaging is 300k
[04:14] <segfault> ogra: just x86
[04:15] <Kamion> sabdfl: thanks
[04:15] <smurf> pitti: I'd rather keep python-imaging, it does have a couple of desktop-ish scripting uses
[04:15] <pitti> doko: yes, it's still tiny comared against twisted and such, but more than 25 kB :-)
[04:15] <pitti> smurf: agreed
[04:15] <Kamion> ogra: I think that comment is out of date; there are no packages in desktop that are pulled in only by {python,python2.4}-imaging-sane
[04:15] <ogra> segfault, ok, i'll prepare a package for testing later today... i'll ping you or add a comment to the bug
[04:15] <pitti> but before we start to drop packages to free 70 kB (python-imaging) we should identify the bigger ones which can go
[04:15] <ogra> Kamion, yes, thats what i figured now... i was just alerted by the comment :)
[04:16] <segfault> ogra: nice, thanks.
[04:17] <Kamion> pitti: maybe that should be the other way round; at the moment the only OO.o2 locale we have on the CD is en_GB
[04:17] <Kamion> pitti: so perhaps add openoffice.org2-help-en-gb instead?
[04:18] <doko> Kamion: no
[04:18] <pitti> Kamion: WFM
[04:18] <Kamion> doko: ?
[04:18] <pitti> Kamion: and same for -locale?
[04:18] <infinity> Kamion : Approval for pppoeconf, svp.  According to feedback from people at bug 10080, this should fix it once and for all.
[04:18] <smurf> Hmm, UBZ's FirstAgainstTheWall session will be interesting ;-)
[04:18] <doko> Kamion: en-US is needed for OOo2
[04:18] <doko> for all languages
[04:18] <Kamion> doko: only help, or l10n too?
[04:19] <doko> only -l10n, as as a fallback help, when the locale package is not found
[04:19] <doko> so I would prefer -en-us over -en-gb
[04:19] <fabbione> hey doko
[04:19] <Kamion> doko: well, germinate output suggests that it is not that way at present
[04:19] <Kamion> I agree with pitti; if we're taking en-us we should drop en-gb
[04:20] <doko> Kamion: openoffice.org2-core depends on openoffice.org2-l10n-en-us
[04:20] <Lathiat> infinity: i'll try out pppoeconf
[04:20] <pitti> doko, Kamion: ok, so I seed l10n/help en-gb and add en-us to the l-support-en package?
[04:20] <jbailey> It's funny to see a whole bunch of people who don't speak American English proposing that it should be the default.
[04:20] <Lathiat> infinity: not sure this fixes upgrades tho
[04:20] <Kamion> doko: oh, crap, ok
[04:20] <Kamion> you're right; I hadn't checked deskop
[04:20] <Kamion> desktop
[04:21] <Kamion> pitti: yes, please do
[04:21] <pitti> jbailey: few people would agree to put l-support-de onto the CD instead :-)
[04:21] <infinity> Lathiat : I think we're too late to try to sort out upgrades on this one. :/
[04:21] <doko> pitti: please test before, that en-gb is used as fallback help, if the language specific help is not installed
[04:21] <Lathiat> infinity: heh
[04:21] <pitti> doko: you mean en-us?
[04:21] <Lathiat> i don't knwo why that patch was rejected upstream, its a perfectly sane thing to do
[04:21] <Lathiat> i should ask the debian maintainer
[04:22] <doko> pitti: AFAIU you do want to put help-en-gb on the cd, which IMO is wrong
[04:22] <jbailey> pitti: I imagine.  But AFAIK, The US is the only place that uses US spellings in English.  I thought everywhere else used UK spellings.
[04:22] <pitti> doko: no: <pitti> doko, Kamion: ok, so I seed l10n/help en-gb and add en-us to the l-support-en package?
[04:22] <pitti> doko: I have to explicitly put -en-gb into supported since l-support-en does not depend on them any more
[04:22] <Kamion> so, I've moved python-twisted and python2.4-samba out to supported; not sure about python-opengl, since it seems more obviously desktop-ish. we'll see how it goes
[04:22] <doko> pitti: yes, you are reight
[04:22] <pitti> doko: is that fine?
[04:22] <doko> pitti: yes
[04:23] <doko> fabbione: hi
[04:23] <fabbione> doko: did you read the scrollback from this morning?
[04:24] <doko> fabbione: nothing highlighted ...
[04:24] <fabbione> doko: hmmm strange.. anyway OO2 works perfectly on sparc64.. it's way faster than the amd64 
[04:24] <fabbione> but faster a lot...
[04:25] <doko> cool :)
[04:25] <fabbione> yup :)
[04:25] <doko> fabbione: wait, until it's built natively for amd64
[04:25] <doko> anyway, need to run, see you later
[04:25] <fabbione> doko: right.. later..
[04:28] <tepsipakki> kamion: sorry to bother you, but on the d-i preconfig-example it says that from kernel 2.6.9-> the number of environment and kernel variables is 32+32, but my breezy netboot won't accept more than 8 (+?)
[04:28] <tepsipakki> kamion: so is this a bug or a feature?
[04:28] <fabbione> tepsipakki: it depends how long they are
[04:28] <fabbione> tepsipakki: there is also a limit in size
[04:29] <tepsipakki> so a static network is basically not possible to put there?
[04:29] <fabbione> tepsipakki: it depends.. it might be
[04:30] <Kamion> yeah, I think the size limit is something like 256 characters - can't remember for sure
[04:30] <tepsipakki> fabbione: I also need to preseed console-keymap and the preseed-file... so a full "append" line contains some 12 variables
[04:31] <Kamion> tepsipakki: use kbd-chooser/method=us (or similar) instead of console-keymap-blah
[04:31] <Kamion> that's usually shorter
[04:31] <tepsipakki> ok, thanks
[04:31] <Kamion> you still have the option of rebuilding the initrd to put a preseed file in there if necessary
[04:31] <fabbione> tepsipakki: yeah, you are over the max lenght. I suggest you to do the minimum you need on the cmd line required to pull in the preseed file
[04:31] <fabbione> and manage as much as you can from the preseed
[04:32] <jdub> mjg59: ping
[04:32] <tepsipakki> I already tried to work with the initrd, but the problem is that you need one initrd per machine, if they use static ip:s
[04:33] <Kamion> you could possibly hack around that with variables set on the command line
[04:33] <Kamion> (kernel command line)
[04:33] <mjg59> jdub: Hi
[04:33] <Kamion> i.e. NETCFG_IP=blah, then use a preseed command to fill in the actual preseed from that
[04:34] <Kamion> or alternatively use /preseed.cfg in the initrd for everything except the IP address
[04:34] <jdub> mjg59: Robot101 asked me about something you wanted to do on the 12th - is that in addition to the formal hall?
[04:34] <Kamion> or use kickstart (which brings the network up earlier) and supplement the result with preseeds for finer control
[04:34] <tepsipakki> uh, kickfart
[04:35] <Kamion> it's just a layer over preseeding in Ubuntu
[04:35] <Kamion> pejorative names notwithstanding
[04:35] <tepsipakki> but I've just learned to love d-i ;)
[04:35] <tepsipakki> (and hate)
[04:35] <Kamion> kickstart in Ubuntu *is* d-i
[04:36] <mjg59> jdub: Yup
[04:36] <Kamion> we translate kickstart directives into d-i preseeding
[04:36] <ogra> tepsipakki, its the "Kamion is a magician" ubuntu kickstart :)
[04:36] <jdub> mjg59: what are the timings?
[04:36] <tepsipakki> ogra, kamion: ok ok I'll look into it ;)
[04:36] <mjg59> The meal will be 7:30ish
[04:36] <ogra> tepsipakki, and it has a nice gu to ease the work
[04:36] <mjg59> If you wanted to give a talk beforehand, that could be arranged
[04:37] <ogra> s/gu/gui
[04:37] <pitti> Kamion, doko: new support packages uploaded, seed adjustments done
[04:37] <tepsipakki> ogra: my gui is the cmdline and cron
[04:37] <Kamion> ogra: IME people doing lots of funky installer deployments don't really care about GUIs, but your mileage may vary
[04:38] <ogra> i found the gui very helpful...
[04:38] <Kamion> pitti: d'oh, I'd just uploaded ubuntu-meta
[04:38] <pitti> Kamion: and that hurts?
[04:38] <Kamion> oh, good, I don't need to reupload it
[04:39] <Kamion> nah, just mistakenly thought you'd have had to change the live seed
[04:39] <pitti> Kamion: the seeding was just to make anastacia quiet
[04:39] <pitti> ah, alright
[04:40] <tepsipakki> ogra: the server I work with is a woody
[04:41] <Mithrandir> pitti: (re cfengine and -doc), yes that sounds silly.  If anything, we should rather have cfengine2 in main, not cf1.
[04:41] <pitti> Mithrandir: I just did a security fix to cf1, cf2 also needs one
[04:41] <pitti> Mithrandir: can you please discuss this with mdz/Kamion and unseed it perhaps?
[04:42] <segfault> ogra: IIRC, you can use "recode" to convert everything from ISO-8859-1 to UTF-8
[04:43] <Kamion> segfault: or just iconv, which comes with glibc
[04:43] <ogra> segfault, that should be done on build depending on the charset selected in the po file ...
[04:43] <ogra> segfault, there are no special chars in the pt_BR file, except this one line
[04:44] <segfault> humm, ok
[04:44] <Mithrandir> pitti: yes, once my body stops melting and my head stops wanting to explode, I can.
[04:44] <\sh> ogra: do u need a german sun usb keyboard including mouse?
[04:45] <\sh> ogra: new one
[04:45] <ogra> segfault, in fact, if you look a bit closer, all strings that are translated have the translation: "Propriedades da proteo de tela
[04:45] <ogra> "
[04:45] <ogra> \sh, absolutely !
[04:45] <\sh> ogra: ok...u sure they're running on a bloody x86 laptop?
[04:45] <ogra> why not...
[04:46] <\sh> ok...grab one for u as well...ish throws out...NEW SUN USB keyboards +lol*
[04:46] <ogra> i used susus's ppc keyboard when i had to work around the breakage on my laptop...
[04:46] <Kamion> Riddell, ogra: I'm going to do {kubuntu,edubuntu}-meta uploads to sync up with Ubuntu seed changes, if that's OK
[04:46] <Riddell> Kamion: ok with me
[04:46] <ogra> Kamion, will you trigger new CD builds after that ?
[04:46] <\sh> ogra: delivery will be friday after work ;)
[04:47] <Kamion> ogra: no, I'll probably just wait for the dailies
[04:47] <ogra> Kamion, else leave it to me, i'll have to find another 20MB to drop on ppc anyway
[04:47] <Kamion> ogra: ok, fine
[04:47] <segfault> ogra: yeap... xscreensaver is messed up. Should i work on it or gnome-screensaver will be the default on dapper?
[04:47] <ogra> segfault, gnome-screensaver will be the default on dapper... but xscreensaver is the supported one fr breezy...
[04:48] <ogra> Kamion, any opinion ? the po file for pt_BR is totally messed apparently... would it be ok to replace the complete translation at this point ? 
[04:48] <segfault> ogra: no problem then. i'll try to work on it.
[04:49] <Kamion> ogra: I've done a seed merge, which should help
[04:49] <Kamion> (slightly)
[04:49] <ogra> ok, thanks :)
[04:49] <Kamion> ogra: what package is this?
[04:49] <\sh> ok..going home....
[04:49] <\sh> cu around 
[04:49] <ogra> Kamion, xscreensaver
[04:50] <Kamion> ogra: yes, if you do it very soon
[04:50] <ogra> Kamion, it would else pick up pt.po 
[04:50] <ogra> segfault, so if you volunteer, id replace the whole traslation...
[04:51] <segfault> ogra: but please change the desktop description. :-)
[04:51] <segfault> ogra: i can work on it, when is the deadline?
[04:51] <ogra> segfault, in pt you mean ? would that be ok for normal pt translations ? 
[04:51] <ogra> segfault, last month ? *g*
[04:52] <ogra> segfault, no, to be serious, asap...
[04:52] <segfault> ogra: heh. pt should be ok, and at least change the desktop file, because the "screen" description in pt is "ecr", and in pt_BR is "tela"...
[04:53] <ogra> ah
[04:53] <segfault> ogra: i'll try to work today, changing in rosetta would be sufficient?
[04:53] <ogra> hmm, i thought you couldnt fint it in rosetta
[04:53] <ogra> *find
[04:53] <Kamion> hno73: how much space would we save on amd64 by dropping Thunderbird?
[04:54] <segfault> ogra: its there, but looks like to be empty
[04:54] <segfault> dunno why.
[04:54] <hno73> Kamion: 5.75 MB
[04:55] <segfault> kamion: are you guys fighting to get all the langpacks in the livecd? :P
[04:55] <Kamion> segfault: not a hope. I'm fighting to get everything *else* in the live CD that has to go in there
[04:55] <segfault> kamion: the live dvds should contain them, right?
[04:56] <Kamion> hno73: let me get back to you on that. I think it'll be necessary, but don't have the figures right now
[04:56] <hno73> Kamion: make that 6mb (some screenshots get removed too)
[04:56] <Kamion> segfault: at the moment I'm afraid the live DVD contains the same filesystem image as the live CD
[04:56] <Kamion> we should probably fix that eventually
[04:56] <hno73> Kamion: OK
[04:56] <Kamion> or else make it know how to install language packs from the archive on the DVD, which contains the .debs
[04:57] <ogra> pitti, a reboot is required if i update xscreensaver ???
[04:58] <segfault> probably not, just kill the daemon
[04:58] <da_bon_bon> does the hibernate of ubuntu unmount all disks before hibernating ?
[04:59] <ogra> pitti, it will break if it doesnt reread the X defaults from /etc/X11/app-defaults ... but asking for a reboot is heavy... this code was there since warty and is also in debian
[04:59] <ogra> segfault, thats exactly what it does since ages
[05:03] <Kamion> -Depends: myspell-ga, aspell-ga, mozilla-firefox-locale-ga-ie,
[05:03] <Kamion> +Depends: myspell-ga, aspell-ga, mozilla-firefox-locale-ga-ie, , openoffice.org2-help-ga
[05:03] <Kamion> pitti: looks funny
[05:04] <pitti> ogra: no, of course a reboot is not required
[05:04] <Keybuk> seb128: ping
[05:05] <seb128> Keybuk: pong
[05:05] <pitti> ogra: for my sake you could just drop the kill command completely
[05:05] <ogra> pitti, this will break
[05:05] <Keybuk> seb128: could you add "exec 2>/tmp/network.log" to the top of /etc/hotplug/net.agent for me and attach that file to your bug
[05:06] <ogra> pitti, the daemon will just silently quit... 
[05:06] <pitti> Kamion: oh, I think I know what's wrong - I should filter out empty lines in the db lists
[05:06] <seb128> Keybuk: sure, doing that now
[05:06] <pitti> Kamion: shall I upload a new one?
[05:07] <Kamion> pitti: does it actually break anything? if not, no need
[05:07] <Kamion> if so, feel free
[05:07] <Keybuk> uh, sorry net.rc!
[05:07] <Keybuk> actually, can you add it (with 2>> instead) to the tops of net.agent, net.rc and net.ifup
[05:07] <Keybuk> and stick "set -x" after it <g>
[05:10] <pitti> Kamion: dpkg-buildpackage seems to sanitize it anyway
[05:11] <ogra> pitti, i tae back the "since ages" :)
[05:11] <ogra>   * debian/xscreensaver.postinst:
[05:11] <ogra>     + send SIGHUP to all running xscreensaver processes.
[05:11] <ogra> [....]    -- Josselin Mouette <joss@debian.org>  Thu, 24 Mar 2005 00:06:43 +0100
[05:11] <ogra> *take
[05:11] <pitti> ogra: I just generally think it is evilish
[05:12] <pitti> ogra: especially if it breaks screen locks
[05:12] <ogra> pitti, but it will have evil side effects if it doesnt reread the config
[05:12] <pitti> ogra: but if it breaks otherwise, I'll shut up :-)
[05:12] <ogra> i.e. it might just stop working the next time xscreensaver-command acts
[05:12] <pitti> ogra: why should it? once it runs, it shouldn't re-read its config, or does it?
[05:13] <pitti> ogra: but if it stumbles over a new config later, why should it work if you forcefully restart it?
[05:13] <ogra> if the config changed on upgrades it needs to reread the changes... to not call non existing screensavers etc
[05:13] <ogra> and there were a lot changes hoary->breezy
[05:13] <ogra> i.e. the lockscreen will not work if the new config isnt in place...
[05:14] <ogra> (the handling of the new login button done by upstream is through the new config file... the lockwindow drawing depends on it for example)
[05:16] <Kamion> pitti: ok, fine, can just be left alone then
[05:16] <Keybuk> hmm, you know, sometimes I think my brain has a flappy-paddle gearbox
[05:17] <pitti> ogra: not having the new login button until the user reboots seems fine to me, but of course it should not just die
[05:17] <ogra> pitti, the real bug is that xss doesnt act rigth on HUP
[05:17] <pitti> ogra: still this is confusing to me
[05:17] <ogra> pitti, you dont understand...
[05:17] <pitti> ogra: either xscreensaver restarts automatically (i. e. after killing it) or it doesn't
[05:17] <ogra> pitti, the whole window and all text will be gone ...
[05:18] <ogra> its not that easy...
[05:19] <pitti> ogra: I didn't claim that it was easy, I just want to understand it :-)
[05:19] <pitti> ogra: if it breaks that hard, then the current way needs to stay for breezy anyway
[05:19] <ogra> pitti, exactly...
[05:19] <seb128> Keybuk: 17472  updated
[05:20] <Keybuk> elmo: could you PLEASE buy us a real SSL cert
[05:20] <Keybuk> or just change to one that's not EXPIRED :p
[05:20] <ogra> pitti, and the right fix is to move on to a better screensaver implementation... i'm fine with having the same behavior sarge and sid have now...
[05:20] <ogra> ...and to move to gnome-screensaver next week :)
[05:21] <Znarl> Keybuk : I know, I know.  I will get to it soon.
[05:21] <Keybuk> seb128: hmm, your /etc/iftab is trying to swap your ethernet cards around?
[05:21] <Keybuk> (it's plugging eth0, and getting it renamed to eth1, and getting all confused)
[05:23] <seb128> Keybuk: that's a stock install, iftab has eth0 / eth1 lines with the corresponding mac
[05:23] <Keybuk> right, try swapping them around <g>
[05:23] <Keybuk> so what was eth0 is now eth1, and vice-versa
[05:24] <Keybuk> fix network/interfaces to match
[05:24] <ogra> lol
[05:25] <Keybuk> hotplug can't deal with interface swapsies because it's stupid
[05:27] <Keybuk> it decides it needs to coldplug eth0 and eth1, starts eth0 which has to be renamed to eth1, so moves eth1 to eth2 and eth0 to eth1, then ifup's eth1; then starts on eth1 (not eth2)
[05:29] <azeem> Md agreed to first look at /etc/modules in the future, so you can force an order through that, btw
[05:29] <mjg59> This would be easier if we just put mac addresses in e/n/i
[05:30] <Keybuk> mjg59: that's my dapper plan
[05:30] <mjg59> Hurrah
[05:30] <Keybuk> to remove all vestiges of eth0/eth1 etc. so they just don't fucking matter
[05:30] <Keybuk> the user will see "Wireless card" and "Wired card" or whatever
[05:30] <azeem> what about NetworkManager?
[05:30] <azeem> does that play with e/n/i at all?
[05:30] <Keybuk> basically same as we do for sda* -- doesn't matter what the * is as long as the mount point is sane
[05:31] <seb128> Keybuk: switching eth0/eth1 from iftab fixes the issue
[05:31] <Keybuk> seb128: I thought it might; thanks
[05:31] <seb128> np
[05:31] <Keybuk> I had a fix for that, but the fix breaks more than it solved
[05:31] <seb128> any chance it got fixed for 5.10?
[05:31] <Keybuk> no, it's just too damned low-level
[05:31] <seb128> k
[05:32] <seb128> what component decide the switch the cards?
[05:32] <Keybuk> and completely broke pcmcia network cards, etc.
[05:32] <Keybuk> ifrename
[05:32] <Keybuk> I want to kill ifrename, and just make it so we never care which card is eth0 and eth1, because we configure by MAC, rather than kernel-assigned-random-number
[05:32] <mjg59> Keybuk: Are we still using ifrename?
[05:32] <mjg59> (since it gets called on resume)
[05:32] <Keybuk> yes, in breezy
[05:32] <seb128> hum, k
[05:33] <mjg59> Ok
[05:33] <Keybuk> breezy's hotplug is a mess, sadly
[05:36] <j^> Kaloz configuration based on MAC is not so nice on a server, where you just want to replace the first network adapter and be done with it
[05:36] <Keybuk> j^: define "first"
[05:37] <Keybuk> is that the first by PCI order, the first by module load order, or the first by entropy of whichever one completes PCI initialisation first?
[05:37] <Lathiat> Keybuk: any idea if hotplug-ng is at all sane?
[05:37] <j^> Keybuk i know thats right now order of modules, pci order
[05:37] <Keybuk> Lathiat: "replacements for hotplug" are very sane
[05:37] <Keybuk> j^: actually, wrong
[05:37] <Lathiat> Keybuk: i more meant hotplug-ng specifically
[05:37] <Lathiat> but ok
[05:37] <Keybuk> modules now don't block during loading
[05:37] <Keybuk> load one with firmware before one without
[05:38] <Keybuk> and guess which way they end up plugged
[05:38] <Keybuk> (tip: it's pretty much 50/50)
[05:38] <Keybuk> you can even get a cute effect where loading them means the second one wins because it's got better PCI karma
[05:38] <Keybuk> even without firmware
[05:38] <Keybuk> it's just too unpredictable these days
[05:38] <Keybuk> I think it's _FAR_ better for you as a sysadmin to assign names to them (if you don't like the ones the Installer picked)
[05:39] <Keybuk> so you could replace "Wired Ethernet Card #1"
[05:39] <Keybuk> and not care which eth? that was
[05:39] <Lathiat> i think you more need reliable actual names, for scripts etc
[05:40] <jdthood> Lathiat: There has been discussion of this in #290
[05:41] <Lathiat> malone?
[05:41] <Lathiat> hm, no
[05:41] <Lathiat> ah i was using the wrong url to bugz
[05:41] <Keybuk> which scripts?
[05:42] <Lathiat> Keybuk: anything use interface names
[05:42] <Lathiat> say, iptables
[05:42] <Lathiat> pppoe
[05:42] <Keybuk> we'll find a way to fix those
[05:42] <Keybuk> the other option is just not using eth* and assigning different names to them with ifrename
[05:42] <Keybuk> which is even ickier
[05:42] <Keybuk> and actually has exactly the same damned bug
[05:43] <Keybuk> ifconfig net0 would annoy people
[05:43] <Lathiat> cant we just assign an ethX to each new card we find
[05:43] <Lathiat> and keep it only for that
[05:43] <Keybuk> no
[05:43] <Lathiat> why not?
[05:43] <Keybuk> because the kernel assigns those
[05:43] <Lathiat> yeh but you can rename them?
[05:43] <Lathiat> if we make issues with that not suck
[05:43] <Keybuk> right, but what if the X is already taken by another card?
[05:44] <Keybuk> which you're also _simultaneously_ hotplugging?
[05:44] <Lathiat> my point being, it shouldnt
[05:44] <Keybuk> but it will be
[05:44] <Lathiat> so the first time you hotplug
[05:44] <Keybuk> two network cards, the first is eth0, the second is eth1
[05:44] <Lathiat> Keybuk: oh so your saying
[05:44] <Lathiat> at the split second you get both
[05:44] <Keybuk> you want to swap them
[05:44] <Keybuk> excatly
[05:44] <Keybuk> you can't, because you're trying to swap them both simultaneously
[05:44] <hughsie> hey, desrt?
[05:44] <Lathiat> could hack the drivers to initially name them netX and rename to ethX ;p
[05:44] <Keybuk> that's just as bad
[05:45] <Lathiat> and then first time you find a card, assign ethX, and then rename to that each time you see it, and nothing else wil el ever use it
[05:45] <Lathiat> heh
[05:45] <Lathiat> yeh
[05:45] <jdthood> eth_X would be better than netX
[05:45] <mae> If something is MIT licensed, is it legal to take that code and incorporate it into gpl code and then release it as gpl code?
[05:45] <jdthood> You just need to use a separate namespace.
[05:46] <\sh> does anyone have a sun keyboard map for x86 xorg with german layout?
[05:46] <Keybuk> mae: no.  you must release it as MIT code
[05:46] <Keybuk> but you can integrate MIT code into GPL code, and release the GPL bits as GPL, and the MIT bits as MIT
[05:46] <Keybuk> which is what you intended to mean, you just need to be carefully specific
[05:46] <mae> Keybuk, but the mit license says basically, you are free to do whatever you want to it.
[05:46] <Keybuk> except relicence
[05:46] <Keybuk> the MIT licence does not grant you that permission
[05:46] <mae> ah.
[05:47] <Keybuk> it is however compatible with the GPL, and vice-versa, so code licenced under both can be combined into an aggregate work (with each bit licenced appropriately)
[05:47] <Lathiat> mae: you could also  ask the author if theyd relicense it to GPL
[05:47] <Keybuk> jdthood: that's just a kludge though, in the long term
[05:47] <Keybuk> I think removing reliance on the kernel-assigned names is the right solution
[05:47] <Lathiat> Keybuk: but thats hard to do nicely
[05:47] <Keybuk> it's _SO_ easy for a firewall script to get the name of the device from the address
[05:47] <Keybuk> Lathiat: isn't, it's pretty easy actually
[05:47] <Lathiat> Keybuk: sure but addresses arent always static
[05:48] <Lathiat> many ISPs (here at least) assign dynamic IPs say
[05:48] <jdthood> Well, using ifrename to map to eth_X eliminates dependency on what the kernel decides to name interfaces originally.
[05:48] <Keybuk> address == MAC address
[05:48] <Lathiat> Keybuk: oh, right
[05:48] <jdthood> Other solutions are just going to reproduce what ifrename already does, in other tools.
[05:48] <Lathiat> still makes it hassly for lots of existing stuff
[05:48] <Keybuk> jdthood: not at all
[05:48] <Keybuk> I'm saying we configure network interfaces based on details other than their kernel-assigned name
[05:49] <jdthood> ifrename looks at the MAC address and assigns a name, say eth_0.  This is a stable name that can be used by everything else.
[05:49] <Keybuk> Lathiat: the existing stuff will work as well as it does now, because it will use the kernel-assigned names which we won't change
[05:49] <Keybuk> the existing stuff is _currently_ sensitive to kernel-assigned name changes
[05:49] <jdthood> The alternative suggestion, I take it, is to make everything else look at the MAC addresss.
[05:50] <Keybuk> jdthood: where "everything else" is basically just network-manager
[05:50] <Lathiat> Keybuk: doesn't solve reliability for other thigns tho, thats what im trying to say
[05:50] <Lathiat> but yes, that would be a start
[05:50] <Lathiat> whats plans for dapper in hotplug/udev/etc land
[05:51] <jdthood> Well, reinventing the wheel seems to be the name of the network-manager game.
[05:54] <sivang> Lathiat: see the first groupe of BOFs suggestions :) (most of them but scott I think)
[05:54] <jdthood> In Debian I take it Md has replaced hotplug with udev and coldplug.
[05:54] <j^> sometimes its better to start over than to try to fix things
[05:54] <Keybuk> Lathiat: not for "existing" stuff, but we'll provide an easy way for those to be updated to work
[05:54] <j^> jdthood but i do not agree with your that network-manager is about reinventing the wheel, it actually brings network configuration to the desktop, something i did not see before on linux.
[05:55] <Keybuk> firewall configuration in a "network up" event that confirms the card is the right one -- rather than a boot-time script run at a hopeful point in the sequence
[05:58] <Keybuk> new world order should be an /etc/scripts/firewall-dmz which has at the top "run when interface up 12:34:56:78:9a" or something
[05:59] <mdz> seb128: what is the reason for this evolution-2.2.desktop addition?  it seems safe of course, but I don't understand why it is necessary
[05:59] <Keybuk> you could even add "hourly, boot, ipod plugged in" type triggers
[05:59] <Keybuk> but that's a bit past dapper, which has to be supportable for a while <g>
[05:59] <infinity> Ahh, mdz, ahoy!
[05:59] <Keybuk> morning, boss
[05:59] <sivang> mdz: there's an issue with that one , after an upgrade the panle's link/launcher stops to work
[06:01] <seb128> mdz: gnome-panel writtes its configuration on first startup. You install hoary, login evolution-2.2.desktop is used. You upgrade to evolution2.4, the package ships evolution-2.4.desktop ... 
[06:01] <infinity> mdz : I've reviewed the upstream thunderbird 1.0.6 -> 1.0.7 diff, and it's all very small/safe security and bug fixes.
[06:02] <ogra> mdz, do you have any suggestion about my ldm change i mailed you on friday ? i'd like to see that in, its already available in my bazaar repo
[06:02] <infinity> mdz : I've been running test packages here for the last 12+ hours, stressing them a afair bit, and all is well.
[06:02] <seb128> mdz: did you get my mail about it ?
[06:02] <infinity> mdz : Do I have approval to shove it into breezy (and do the security updates with 1.0.7 to hoary/warty)?
[06:02] <HiddenWolf> hm, firefox is slow
[06:02] <dholbach_> re
[06:02] <infinity> mdz : As well as do the -fno-strict-aliasing fix (in breezy only)
[06:03] <pitti> Hi mdz 
[06:04] <mdz> seb128: yes, package is approved now
[06:04] <mdz> infinity: thunderbird 1.0.7 OK
[06:05] <mdz> ogra: did you talk to Kamion about fixing this it on the ssh side instead?
[06:05] <mdz> I don't know where ssh gets its default PATH from
[06:05] <infinity> mdz : Thanks.
[06:05] <mdz> probably hardcoded, knowing those guys
[06:06] <ogra> mdz, all DMs do it this way, i dont think setting /usr/games for ssh at complie time would be right... Kamion ?
[06:06] <infinity> mdz : Can I get you to approve the pppoeconf in the queue too?
[06:06] <infinity> mdz : The patch applied there has had some positive feedback from a few users.
[06:06] <ogra> mdz, its hardcoded
[06:07] <ogra> mdz, i checed gdm, wdm and xdm, they all set the path themselves before establishing a session...
[06:07] <mdz> infinity: already done
[06:07] <infinity> mdz : Cheers.
[06:07] <mdz> ogra: but we don't use gdm, wdm or xdm
[06:07] <mdz> ogra: everything should use the same default path, including ssh
[06:07] <ogra> mdz, but we establish a session after login... 
[06:07] <ogra> mdz, root shouldnt have /usr/games in the path...
[06:08] <mdz> ogra: why?
[06:08] <ogra> heh, no idea, beacause its traditional ? 
[06:09] <mdz> Kamion: what's the latest on the live CD squeeze?
[06:09] <ogra> Kamion, any opinion? ltsp needs /usr/games in PATH ... ssh has set a default path without it at compile time, if i understand it right, ~/.ssh/environment is considered evil 
[06:10] <wasabi_> hey so, need some oem-config advice. Installing it on a chrooted system, want to configure it so it pops up on the first boot of the chroot.
[06:10] <wasabi_> Not finding what needs to be set in inittab documented.
[06:10] <ogra> Kamion, my guess would have been to set the path from ldm bfore the session is established
[06:10] <ogra> Kamion, mdz thinks that should be done in ssh
[06:11] <seb128> Keybuk: are you sure than the eth bug affect "few unfortunate people"
[06:12] <seb128> Keybuk: my laptop has nothing special, it just has a wireless eth0 and an eth1 (and it works fine with hoary)
[06:12] <Keybuk> tbh, I really don't know
[06:13] <Keybuk> the installer seems to sometimes randomly plug them the other way to the boot sequence
[06:13] <Keybuk> which is why you end up with a backwards iftab in the first place
[06:13] <seb128> so there is an installer bug too?
[06:14] <Keybuk> the installer detects hardware differently
[06:14] <Keybuk> for whatever reason it saw your wireless card first, so named that eth0 and your wired eth1; if I'm reading the runes right
[06:14] <Keybuk> where the normal boot sequence sees your wired card first and your wireless seconds
[06:14] <Keybuk> and ifrename tries to swap them, but hotplug fucks it up
[06:15] <seb128> hum, I though that the wireless had been eth0 
[06:15] <seb128> with hoary too
[06:15] <seb128> not sure right now though
[06:15] <seb128> I'll play with that
[06:16] <Keybuk> I'd love to know why the installer behaves like that
[06:16] <Keybuk> it might have something to do with when the restricted modules get added to the pudding I guess
[06:16] <Keybuk> or even just the wireless ones
[06:17] <Keybuk> I couldn't get it to do anything here
[06:17] <Keybuk> the explanation how to fix it ("swap the lines in iftab") was far less scary than a ground-up rewrite of the hotplug net.* stuff right before release
[06:18] <seb128> sure
[06:18] <seb128> but it makes than the laptop works out of the box with hoary and not now
[06:18] <mdz> my desktop hung overnight, second time this week
[06:18] <seb128> I've to "sudo dhclient" before doing anything
[06:18] <mdz> s/this/& past/
[06:18] <mdz> not very confidence-inspiring
[06:18] <seb128> I don't expect any user to go edit /etc/iftab
[06:19] <mdz> something X-related, the box was still alive and on the network
[06:19] <Keybuk> seb128: the reports are incredibly rare, yours is actually the first since well before preview
[06:19] <jdthood> mdz: What video driver?
[06:19] <mdz> jdthood: ati
[06:20] <Keybuk> and the last person did a reinstall and it didn't happen that time :-/
[06:20] <mdz> was very stable for most of breezy
[06:20] <seb128> Keybuk: crispin just pinged me on #epiphany when you dupped my bug from 290 saying the bug is really annoying, but yeah, that does only 2 people :)
[06:20] <jdthood> mdz: https://bugs.freedesktop.org/show_bug.cgi?id=4486 ?
[06:20] <Kamion> ogra: at this stage, I'd much rather see the change happen in ldm
[06:20] <mdz> there haven't been ati changes in breezy for quite a while, I don't think
[06:20] <seb128> Keybuk: I've done random candidate/daily install for 10 days, that happens 100% of the time here
[06:21] <infinity> mdz : If you can approve expect-tcl8.3 too, that'd be great (just an FTBFS fix)
[06:21] <mdz> jdthood: that's vague enough to match my symptoms
[06:21] <Keybuk> it's a bit confusing because #290 was originally a different bug, but it's become the home to this problem now <g>
[06:21] <ogra> Kamion, me  too... do you think doing it in shh would be right at all ?
[06:21] <mdz> infinity: done
[06:21] <seb128> Keybuk: just booted hoary on the laptop, eth0 is ipw2200, eth1 is 8139 ... ie: the same
[06:21] <ogra> *ssh
[06:21] <Kamion> ogra: .ssh/environment is security-broken and will go away
[06:21] <infinity> mdz : The only radeon changes I know of are blacklisting (and unblacklisting) renderaccel on various PCI IDs.
[06:21] <Keybuk> oh bugger
[06:21] <Keybuk> you know what I think it is
[06:21] <infinity> mdz : If you haven't noticed X getting noticeably slower and/or faster, that hasn't hit your card.
[06:21] <Keybuk> in breezy, firmware for the ipw2200 isn't loaded synchronously
[06:21] <ogra> Kamion, yup, thats how i understood it
[06:22] <mdz> infinity: and that only on 20 Sep according to the changelog
[06:22] <hub> seb128: the abiword package in main does not have equation and grammar support :-(
[06:22] <hub> seb128: FYI
[06:22] <Kamion> ogra: not really; I'm fairly sure that if I put /usr/games in root's $PATH I'd get bugs saying "why is all this random crap in my system administration path?"
[06:22] <seb128> hub: what is needed for that? build-depends?
[06:22] <Keybuk> so ipw2200 is loading firmware while 8139 is loading and beats it to getting eth0
[06:22] <Keybuk> so at boot, they get swapped
[06:22] <mdz> I'm unable to get the video card back into a working state by any means other than a hard reset
[06:22] <seb128> Keybuk: better explanation imho
[06:22] <mdz> it's good and wedged
[06:22] <hub> seb128: gtkmathview and link-grammar respectively
[06:22] <Kamion> ogra: shouldn't you be starting a user session in some way rather than relying on root's path? I'd have expected root's path to be essentially irrelevant for desktops
[06:23] <Keybuk> random, could you try booting with the hoary kernel on the breezy userland? :p
[06:23] <hub> seb128: I should check the package: we had a talk with joshk last night
[06:23] <ogra> Kamion, ldm runs ssh -X <server> Xsession
[06:23] <Keybuk> actually, that almost certainly won't work :-/
[06:23] <Keybuk> udev won't start
[06:23] <jdthood> mdz: Search bugs.freedesktop.org under the ati and radeon driver components.  There are lots of reports of X lock ups, most that are reproducible under particular conditions (such as: only with acceleration)
[06:23] <mdz> jdthood: I have no explanation for why it only started happening recently, though
[06:23] <ogra> Kamion, so there are three points where we could set it, ldm, ssh or Xsession...
[06:24] <Keybuk> does crispin also have an ipw/non-firmware combination?
[06:24] <seb128> hub: not main packages, this is not going to be changed for 5.10 ...
[06:24] <Keybuk> and does the ipw win eth0 in the installer (and iftab?)
[06:24] <seb128> Keybuk: asking
[06:24] <mdz> ogra: Xsession doesn't and shouldn't set a path
[06:24] <ogra> Kamion, by looking at all other DMs i found they set it before establishing a session...
[06:24] <ogra> mdz, yup... 
[06:24] <hub> seb128: fine with me. I just thought you should know
[06:24] <Kamion> ogra: surely you should be running the remote session as non-root
[06:24] <seb128> hub: yeah, thanks
[06:24] <ogra> Kamion, yes
[06:24] <Kamion> i.e. ssh -X nonprivilegeduser@<server> Xsession
[06:24] <infinity> mdz : Even more curious, when you realise that 99% of video card lockups wedge the hardware so badly that the wohle machine freezes, so your bug is SPECIAL.
[06:25] <infinity> mdz : Then again, that also makes youre a tiny bit debuggable, perhaps.
[06:25] <infinity> s/youre/yours/
[06:25] <ogra> Kamion, err, yes, i omitted that bit, sorry... indeed we give the username to ssh
[06:25] <hub> seb128: equation in a plugin. grammar too. so we can make universe packages for that:-) and push them to the next release in main :-))
[06:25] <Kamion> ogra: ah, I see, it's not about root at all - something in scrollback confused me
[06:25] <mdz> yeah, unfortunately I was in no position to debug it the last time
[06:25] <seb128> hub: no, because abiword source package is main and the Build-Depends have to be main
[06:25] <Kamion> ogra: I'll fix ssh to include /usr/games in the default non-root $PATH
[06:26] <ogra> Kamion, ok, thanks :) that was my last bad edubuntu bug
[06:26] <Kamion> ogra: go ahead and reassign the bug to me
[06:26] <ogra> Kamion, ok
[06:26] <hub> seb128: we'll talk about that at ubz :-) no urge
[06:27] <mdz> Kamion: are we still overflowing after removing -twisted and -samba?
[06:28] <mjg59> infinity: Has fglrx changed at all lately?
[06:28] <elmo> Riddell: ?
[06:28] <Kamion> mdz: I don't know yet - sorry, I was at the shops and am still working my way through all the questions people asked me in the meantime :)
[06:28] <infinity> mjg59 : Define "lately"... Hasn't for a month or so, at least.
[06:28] <jdthood> mdz: Is X using 100% CPU?
[06:28] <Kamion> mdz: FWIW the reason that the default paths are built into sshd is that there is no stable API for extracting the values from /etc/login.defs
[06:28] <mjg59> infinity: Ok. Suspend/resume seems to have broken for a load of fglrx people over the past week or so (not that I actually care, but still)
[06:28] <Kamion> (no, grep doesn't count)
[06:29] <mdz> Kamion: perhaps we should have a SetDefaultPathInOneGodDamnPlaceSpec at UBZ
[06:29] <infinity> mjg59 : Ask them what framebuffers they're running, that's more likely a culprit than fglrx itself.
[06:29] <Riddell> elmo: hi
[06:29] <mdz> jdthood: dunno, this was last night and I reset it
[06:29] <Keybuk> could we have a SetDefaultLang/Umask one too?
[06:29] <infinity> mjg59 : The revelation that usplash now supports vesafb may have caused a mass migration or something.
[06:29] <mjg59> Ah
[06:29] <mjg59> True
[06:30] <Kamion> mdz: I'd be happy to talk about it with shadow upstream at some point if I get the time
[06:30] <infinity> mjg59 : Also, radeonfb is known buggy (read: effin' broken) on most setups, so if anyone's using that, rap them on the knuckles.
[06:30] <Kamion> it's certainly been a long-standing annoyance
[06:30] <ogra> Kamion, err, i have no bugreport about it, only user reports about non working games entrys in the manu... want to fill me a new one? 
[06:30] <Kamion> ogra: never mind then
[06:31] <ogra> ok
[06:32] <jdthood> "System locks up w/ radeonfb" https://bugs.freedesktop.org/show_bug.cgi?id=197
[06:33] <Kamion> mdz: so I doubt -twisted and -samba will free up enough space, but they will get us a good chunk of the way there; the questions are (a) whether removing Thunderbird from the amd64 WinFOSS tarball will be enough after that, (b) whether powerpc will fit after that
[06:33] <Kamion> mdz: what do you think about python-opengl? it does seem moderately desktop-ish
[06:34] <mdz> Kamion: as far as I'm concerned all of the python-* in the desktop seed is superfluous, but sabdfl has been known to disagree ;-)
[06:34] <Kamion> mdz: sabdfl OKed removing python-opengl if we needed to
[06:34] <janimo> elmo, is there an xfce4-taskmanager in the NEW queue? I uploaded one today but got no mail not even reject
[06:35] <Kamion> wasabi_: I don't use inittab to start oem-config any more; it caused a variety of tedious problems
[06:35] <janimo> I just signed it the changed-by field is not whitelisted
[06:35] <Kamion> wasabi_: see /usr/lib/base-config/menu/oem-config, which does 'update-rc.d oem-config start 12 2 3 4 5 .'
[06:36] <Kamion> janimo: it's been processed already
[06:36] <pitti> janimo: xfce4-taskmanager_0.3.1-0ubuntu1_source.changes was accepted
[06:36] <pitti> janimo: at 2005-10-10 17:25
[06:36] <janimo> hmm I wonder why I got no mail
[06:36] <janimo> thanks
[06:37] <pitti> janimo: yep, and it is in nEW
[06:37] <seb128> Keybuk: <crispin> seb128: yeah, an ipw2200
[06:38] <janimo> can the NEW queue be seen from the outside?
[06:38] <Kamion> janimo: no
[06:38] <Keybuk> ok, so I understand why the bug affects you now
[06:38] <Keybuk> the ipw driver used to block for firmware, but that was crap because if you didn't find it in time, the driver sulked
[06:39] <Keybuk> they changed it to request firmware and get called back, which means the kernel is free to load other things while it's waiting
[06:48] <Kamion> mdz: daniels tells me that xorg -77 fixes a mistaken lack of stripping of drivers, thereby saving a fair amount of space; it seems to revert the problematic untested stuff from -76. What do you think of it?
[06:49] <mdz> Kamion: oh, I didn't notice that -76 had been replaced by -77 in helena output.  I'll have a look
[06:52] <seb128> "<crispin> seb128: yeah, but I don't have the 290 problem on my laptop, just a desktop machine" ... grumpf, no clear
[06:58] <Keybuk> seb128: his desktop, I think, has an odd card which d-i doesn't recognise but lrm does
[06:58] <seb128> may be
[06:59] <Keybuk> and it's at a higher pci number, so gets plugged on boot t'other way round
[06:59] <mjg59> Keybuk: Like any Atheros?
[07:01] <Keybuk> atheros tend to be later in the pci numbering than the internal
[07:01] <Keybuk> the problems are:
[07:01] <Keybuk> 1. the kernel plugs devices in a random order
[07:01] <Keybuk> well, pseudo-random
[07:01] <mjg59> Keybuk: Atheros stuff isn't supported in d-i
[07:01] <Keybuk> which is in their view notabug, because you shouldn't rely on the order; and they go into details about schroedinger's network card which is a bit like schroedinger's freedrom toaster, but without the plane crash
[07:02] <Keybuk> 2. the events arrive at userspace in a random order
[07:02] <Keybuk> which is pretty much just #1 again
[07:02] <Keybuk> 3. userspace takes a random amount of time to react to the event
[07:02] <Keybuk> 4. the event takes a random amount of time to finish
[07:03] <Keybuk> fortunately we live in a universe of malignancy, so for most people the first one in the list ends up as eth0; just like their bread lands butter-side-down
[07:03] <Keybuk> but for some people, it just doesn't
[07:04] <ficusplanet> Hey everyone.  With the Tango Project (http://tango-project.org/Tango_Desktop_Project) announced, I was wondering if Ubuntu will be supporting/using this icon theme?
[07:04] <smurf> Keybuk: so the solution seems to be to rename the interfaces based on their MAC address, or something
[07:04] <Keybuk> smurf: which we do
[07:04] <mjg59> I sense that we're approaching the beginning of this conversation
[07:04] <mdz> Kamion: kelly'd
[07:04] <Keybuk> but the trouble is you end up trying to rename eth0 to eth1 and eth1 to eth0 at the same time
[07:05] <Keybuk> and something goes ping and one of your network cards (which, sods law states, is usually the one you actually wanted) ends up flying across the room and sulking in the corner
[07:06] <smurf> Keybuk: so think of entirely new names, and use these instead when you find a new card ..?
[07:06] <Keybuk> smurf: so Ubuntu wouldn't use eth0, eth1, et. al. except for some cards, which it would
[07:06] <Keybuk> this would be upsetting
[07:06] <Keybuk> if I wanted to have to brute-force my network device names, I'd use FreeBSD
[07:06] <smurf> Well, having cards randomly fly into a corner and sulk isn't entirely optimal either
[07:07] <Keybuk> very true; what I'd like to be doing next week is configuring and activating cards based on their MAC address, or perhaps device path
[07:07] <Keybuk> so we wouldn't need to _care_ what random kernel name got assigned to it
[07:08] <smurf> Yeah, right now is probably the worst possible time to attempt stuff like that ;-)
[07:09] <Keybuk> yes
[07:09] <smurf> I'd use MAC addresses, BTW; device paths can change randomly too ...
[07:09] <Keybuk> this isn't a regression on hoary, afaict ... hoary would fail to swap the network devices and ifup the wrong one
[07:09] <Keybuk> ie. it'd configure your wired for dhcp instead of your wireless
[07:10] <Keybuk> which is basically what breezy is doing
[07:12] <Kamion> mdz: running a live filesystem build now so that I can see what's happening size-wise
[07:12] <mdz> Kamion: thanks
[07:13] <wasabi_> Does oem-config require X?
[07:13] <wasabi_> Seems to.
[07:13] <wasabi_> I'd be happy with my app for some ncurses debconf questions
[07:14] <infinity> mdz : Can we get Tbird out of the queue and building, so I can babysit it and go to bed? :)
[07:16] <HiddenWolf> Configuring postgresql.conf to use port 5432...
[07:16] <HiddenWolf> Warning: The socket directory for owners other than 'postgres'
[07:16] <HiddenWolf> defaults to /tmp. You might want to change the unix_socket_directory parameter
[07:16] <HiddenWolf> in postgresql.conf to a more secure directory.
[07:16] <HiddenWolf> w t f!
[07:16] <HiddenWolf> It should run under it's own user.
[07:27] <Kamion> wasabi_: yes, sorry, at the moment it does; that can probably be fixed later
[07:27] <HiddenWolf> infinity, new TB build?
[07:28] <wasabi_> Kamion, cool! Also, serial support. ;)
[07:29] <wasabi_> And "first time SSH login"
[07:37] <elvirolo> hi all
[07:38] <elvirolo> breezy still suffers from very critical bugs ...
[07:39] <elvirolo> like this one http://bugzilla.ubuntu.com/show_bug.cgi?id=14566 and this one http://bugzilla.ubuntu.com/show_bug.cgi?id=16066
[07:39] <dhonn> damn hp
[07:40] <HiddenWolf> elvirolo, there are sure to be more than a few of those.
[07:40] <elvirolo> the spca5xx is absolutely critical ... it completely freezes my (and others') system, even the magic sys req keys don't work
[07:40] <elvirolo> HiddenWolf, I imagine so yeah
[07:41] <elvirolo> will they be fixed though ?
[07:41] <HiddenWolf> elvirolo, sorry to say, but at this piont, I doubt it.
[07:41] <HiddenWolf> elvirolo, best you can do is ask pitti or mdz
[07:42] <elvirolo> sad :(
[07:43] <dhonn> im sure eventually it will get fixed its probably not just ubuntu users that have this problem
[07:43] <HiddenWolf> dhonn, most likely not.
[07:44] <dhonn> hasnt hp been helping 
[07:45] <elvirolo> my pinter did work under hoary though
[07:45] <elvirolo> how annoying
[07:47] <dhonn> remove hp printing services 
[07:47] <dhonn> that might help
[07:48] <dhonn> chmod -x /etc/init.d/hplip
[07:48] <dhonn> it never was in hoary
[07:49] <elvirolo> god of course i'm using windows right now
[07:49] <elvirolo> i'll write this down
[07:49] <elvirolo> thanks
[07:49] <Belutz> can i make ubuntu fit enough into a 512mb flash drive?
[07:50] <Belutz> or should I ask in #ubuntu?
[07:50] <tseng> I dont think so
[07:50] <tseng>  /usr on my server install is at 717 atm
[07:50] <mdz> infinity: I had already approved it before you said anything
[07:51] <dhonn> recompile the livecd without a few packages
[07:51] <Belutz> hmmm i see
[07:51] <dhonn> dump that to your flash drive
[07:52] <Belutz> dhonn, any howto how to do that?
[07:52] <infinity> mdz : Cool.  An enigmail rebuild to go with it just got tested locally and uploaded.
[07:52] <infinity> And now I'm off ot bed.
[07:52] <\sh> infinity: good night
[07:53] <dhonn> just the knoppix documentation
[07:54] <dhonn> get a gig flash drive
[07:54] <dhonn> install grub boot loader
[07:54] <ogra> Belutz, dhonn, there is plenty of docs in the wiki about remastering the live CD and this is rather appropriate for #ubuntu
[07:55] <dhonn> there you go Belutz
[07:55] <Belutz> ogra, dhonn, thanks :)
[07:55] <Belutz> someone ask me how to do that for government project
[07:56] <sivang> ogra: or #ubuntu-remastering :)
[07:56] <dhonn> i learned how to make knoppix run directly from a ISO without the use of the live dvd, and just just grub
[07:57] <dhonn> im sure it works the same way for a flash drive as it did for my hd
[08:00] <dhonn> i wish the ubuntu live cd had the toram option again
[08:19] <wasabi_> unionfs is pretty cool, except it doesnt' support moving files
[08:26] <kbrooks> Hey all.
[08:26] <kbrooks> I have a e-mail to send to -- shall I send it to ubuntu-users mailing list
[08:53] <fabbione> evening
[08:58] <kbrooks> anyone here subscribed to ubuntu-devel?
[08:59] <kbrooks> Please read the post I just sent
[09:00] <mirak> hhhelp
[09:00] <kbrooks> mirak: no support here. #ubuntu
[09:00] <mirak> kbrooks: I am there
[09:03] <sivang> mirak: I'm with you on #ubuntu
[09:03] <mirak> ok
[09:04] <Keybuk> kbrooks: the plan for the next release is to use network-manager
[09:04] <Keybuk> http://people.redhat.com/dcbw/NetworkManager/
[09:04] <kbrooks> Keybuk: what is that
[09:05] <Keybuk> it handles which network card in your system is active, and which wireless network you connect to
[09:05] <Keybuk> just about everything else in your mail is already handled by udev
[09:06] <Keybuk> I chatted to the ndiswrapper developers the other week about getting proper support for it
[09:06] <mirak> hem what fs is used for the initrd ?
[09:07] <kbrooks> mirak: cramfs i think
[09:07] <Keybuk> mirak: breezy doesn't use initrd
[09:07] <kbrooks> not sure
[09:07] <Keybuk> we use initramfs, which is a compressed cpio file
[09:07] <Keybuk> rather than a filesystem
[09:07] <kbrooks> Keybuk: i see
[09:07] <mirak> Keybuk: what does it use then ?
[09:08] <kbrooks> mirak: he said so altrady
[09:08] <siretart> Keybuk: will it be possible to configure NetworkManager without X? Current implementation seems to lack this :(
[09:08] <mirak> ok but the concept of an external file with module is the same 
[09:08] <mirak> ?
[09:08] <Keybuk> siretart: I should think so
[09:09] <Keybuk> mirak: I don't understand your question
[09:09] <mirak> Keybuk: what don't you understand ?
[09:09] <Keybuk> what you're asking
[09:09] <mirak> I need to see what is in the file along the kernel
[09:09] <Keybuk> right
[09:09] <mirak> don't know how to call it since it's not an inirt
[09:09] <mirak> rd
[09:10] <mirak> initrd
[09:10] <Keybuk> "call it" ?
[09:10] <mirak> name it
[09:10] <Keybuk> do you mean how to see what's inside it?
[09:10] <mirak> you said it's not an initrd so it might be nammed someting else isn't it ?
[09:10] <Keybuk> no, named the same thing
[09:10] <mirak> ok
[09:10] <mirak> so how to see what is in it ?
[09:11] <Keybuk> gunzip -c initrd... | cpio -t
[09:11] <mirak> Keybuk: thank you
[09:11] <siretart> Keybuk: btw, is the installers initrd an real initrd or an cpio archive, too?
[09:12] <Kamion> siretart: it's a real initrd for now
[09:12] <siretart> okay
[09:12] <mirak> lib/modules/2.6.12-9-powerpc/kernel/drivers/ide/pci/cmd64x.ko
[09:12] <Kamion> it'll switch to initramfs eventually, probably
[09:12] <mirak> so this module is inside but it's not loaded at boot time
[09:12] <kbrooks> Kamion: why is it a real initrd
[09:12] <kbrooks> ?
[09:12] <Kamion> kbrooks: no compelling reason to change
[09:13] <kbrooks> Kamion: what would be?
[09:13] <mirak> how can I be sure the initrd is correctly loaded ?
[09:13] <Kamion> why do you care? :)
[09:13] <Keybuk> mirak: do you see the graphical splash?
[09:13] <fabbione> mirak: if it's not loaded, your machine won't boot
[09:13] <Kamion> the reason the main system moved from initrd-tools to initramfs-tools was that we needed to get away from devfs and initramfs-tools was easier to handle
[09:13] <mirak> fabbione: thanks for your intervention LOL
[09:13] <Kamion> the other reasons for moving to initramfs are more nebulous, at least as far as the installer is concerned
[09:14] <Kamion> anyway, got to go
[09:14] <mirak> fabbione: it doesn't boot
[09:14] <Keybuk> Kamion: ah, marriage
[09:14] <kbrooks> nebulous?
[09:14] <Keybuk> mirak: are you using a stock ubuntu kernel?
[09:14] <kbrooks> define?
[09:14] <mirak> Keybuk: I don't see the graphical splash
[09:14] <Kamion> Keybuk: no, hitting things
[09:14] <mirak> Keybuk: I use the ubuntu ppc image
[09:14] <Kamion> (i.e. karate)
[09:14] <fabbione> Kamion: have fun :)
[09:14] <dholbach> i'm out for a walk, bbl
[09:15] <mirak> Keybuk: vmlinux-2.6.12-9-powerpc
[09:15] <\sh> yeah baby yeah....one romanian leaded company will switch from suse to ubuntu....tomorrow i know more :)
[09:16] <kbrooks> \sh: so.
[09:16] <kbrooks> you're famous eh
[09:16] <Keybuk> mirak: in what way doesn't it boot?  how far does it get?  if you remove "quiet splash" from the kernel command-line, what do you see?
[09:17] <mirak> Keybuk: ubuntu ppc uses yaboot
[09:17] <\sh> kbrooks: no...I just evangelise all my colleagues to use ubuntu...and they like it...and just now I had a phone call of one romanian guy who is administrating some servers for a romanian company...he will switch them all to ubuntu
[09:17] <mirak> Keybuk: I can't tell if there is a splash at all, because my last working kernel is 2.6.10-5
[09:17] <mirak> and the yaboot config file is not automatically updated
[09:17] <kbrooks> \sh: no, i'm talking about you being op in a channel...
[09:19] <mirak> Keybuk: can I flood 8 lines in this channel ?
[09:19] <mirak> my yaboot.conf file
[09:19] <\sh> kbrooks: ???? i don't understand
[09:19] <mirak> hum kernel entry
[09:19] <Keybuk> I don't know yaboot, sorry
[09:19] <fabbione> mirak: no flood in here
[09:19] <Keybuk> you may have more luck in #ubuntu
[09:19] <mirak> there is no ppc support
[09:19] <mirak> that's horrible
[09:20] <mirak> Keybuk: maybe yaboot doesn't support this cpio file format
[09:20] <mirak> this would be my bet
[09:20] <Keybuk> this isn't a support channel
[09:20] <fabbione> Keybuk: did you ever build a deb in your life? because i did build hundreds of them...
[09:20] <mirak> I get no help in the other channel
[09:20] <\sh> kbrooks: I was never an op in one channel
[09:20] <mirak> fabbione: yes I did
[09:20] <mirak> however I just need to know what is different
[09:20] <mirak> and why it doesn't work
[09:21] <fabbione> mirak: ?
[09:21] <kbrooks> \sh: never mind, not you
[09:21] <fabbione> mirak: i did say something to Keybuk ...
[09:21] <fabbione> not to you
[09:21] <mirak> if the automated deb file creates a initrd with cpio it's maybe not gona work
[09:21] <mirak> fabbione: yes I didn't seen it
[09:22] <mirak> I can ask in the other place, but there is no answer
[09:22] <mirak> people that knows this kind of thing are only lurking here it seems
[09:25] <mirak> nobody knows that in the #ubuntu channel
[09:25] <mirak> if I ask here it's because here people knows. I wouldn't bother you otherwise
[09:25] <Keybuk> this still isn't a support channel; this channel is for discussing development
[09:26] <mirak> I got an answer
[09:26] <mirak> I might want to look on #ubuntu then :D
[09:33] <speel> hey did you guys fix the spca5xx issues? ( just curious )
[09:34] <fabbione> speel: bug number?
[09:34] <speel> hang on let me get it 
[09:36] <rob^^^> is Wed still looking good for 5.10?
[09:37] <speel> #14566
[09:37] <fabbione> rob^^^: any reasons why it shouldn't be?
[09:38] <rob^^^> no, just wanted to know whether to clear my calendar ;)
[09:38] <elmo> umm, Thursday, not Wednesday
[09:38] <fabbione> release is the 13.. 
[09:38] <fabbione> elmo: exactly :)
[09:38] <rob^^^> ok, I'm mixed up then
[09:39] <fabbione> speel: -> dapper.
[09:39] <fabbione> speel: it won't be fixed for breezy
[09:39] <speel> ah ok
[09:40] <speel> How come?
[09:40] <fabbione> speel: because it's a minor issue and breezy will release in 3 days
[09:40] <fabbione> the kernel is frozen and only security updates will go in
[09:41] <speel> oo ok. heh sorry I dont know how development works and etc :P
[09:42] <fabbione> speel: nothing to be sorry about.. next time we will hunt you down and burn you alive :P
[09:42] <speel> Lol
[09:42] <Keybuk> hey, don't forget the flaying
[09:42] <fabbione> eheh
[09:46] <kbrooks> fabbione: not "the kernel"
[09:46] <kbrooks> ubuntu
[09:47] <fabbione> kbrooks: sorry?
[09:47] <kbrooks> Ubuntu is frozen
[09:47] <kbrooks> for release
[09:47] <chip> elmo: You live!
[09:47] <slomo> mdz: why did you assign #17359 to me? ;) but i'll take a look at it... seems to be a non-bug
[09:48] <mdz> slomo: weren't you working on things mono-related?  perhaps I confused you with someone else
[09:48] <fabbione> hey mdz
[09:49] <mdz> fabbione: sshh, I am not here
[09:49] <fabbione> kbrooks: we are still allowed to do RC bug fixes
[09:49] <chip> elmo: What say you update my key?
[09:49] <slomo> mdz: no that was me... np :) i'll find someone to sponsor a fix for that for main... if this is really a bug ;)
[09:49] <kbrooks> fabbione: i mean after release
[09:49] <fabbione> mdz: i didn't see you.. don't worry
[09:50] <fabbione> mdz: hide! hide!
[09:50] <fabbione> kbrooks: oh.. really?
[09:50] <elmo> chip: what does this have to do with Ubuntu development?
[09:50] <fabbione> kbrooks: but can we get the latest firefox in?
[09:50] <kbrooks> fabbione: i'm saying ubuntu will be frozen -- no new apps only sec updates
[09:51] <kbrooks> on october 13th
[09:51] <chip> elmo: I'll be happy to discuss it in #debian-devel if you will respond there
[09:51] <kbrooks> breezy release
[09:51] <fabbione> kbrooks: ah.. i must have missed that
[09:51] <elmo> chip: I'm not in #debian-devel, so how exactly do you expect me to respond?
[09:51] <kbrooks> i have 1.0.7
[09:51] <chip> elmo: /join, I suppose
[09:51] <elmo> chip: so, please take off-topic talk out of this channel
[09:52] <elmo> chip: no
[09:53] <fabbione> kbrooks: so what will happen with 1.0.8??
[09:53] <kbrooks> fabbione: i'm not a ubuntu developer
[09:53] <kbrooks> i was just corecting you
[09:53] <kbrooks> thats all
[09:54] <fabbione> kbrooks: hmmm.. ok.. i guess i will need to read up on that.. afterall i was only the Xfree86/Xorg and kernel maintainer for Ubuntu
[09:55] <fabbione> i guess i got too old for that :)
[09:55] <Keybuk> right, am gonna crash and get an early night
[09:55] <fabbione> night Keybuk 
[09:55] <Keybuk> elmo stole my timezone, so now I'm taking it back, or something
[09:55] <fabbione> Keybuk: lol
[09:55] <\sh> Keybuk: btw...can we talk at ubz about planet and s9y problems? ,-)
[09:56] <Keybuk> no, because I don't maintain Planet
[09:56] <Keybuk> you can talk to jdub though :p
[09:56] <\sh> Keybuk: software side...not administrating ;)
[09:56] <Keybuk> that's what I mean
[09:56] <\sh> actually...I would like to fix the issues in planet ;)
[09:57] <Keybuk> go for it :p
[09:57] <\sh> Keybuk: bah ;) solution: install s9y with planet plugin ;)
[09:58] <\sh> hmmm...if i didn't fix universe until 22:00 UTC .. i'm lame
[09:59] <fabbione> \sh: you have less than 60 secs..
[09:59] <\sh> fabbione: that's 22:00 uhr ;) german time...
[09:59] <fabbione> it's still 21:59
[09:59] <fabbione> you have a few secs left
[10:00] <\sh> fabbione: it's still 19:59 UTC ;)
[10:00] <fabbione> oh right.. UTC
[10:00] <\sh> HAR HAR
[10:00] <fabbione> dong
[10:00] <fabbione> ahaha
[10:00] <\sh> I'll fix it ;)
[10:00] <fabbione> you got me :)
[10:03] <\sh> fabbione: ;)
[10:07] <moyogo> is there any reason why we have freetype6-2.1.7 when 2.1.10 is out since june?
[10:07] <moyogo> 2.1.7 is from 2004-02
[10:08] <kent> moyogo, 2004 was perhaps a very good year?
[10:08] <moyogo> hehe
[10:21] <dieman> of course! i try seeing how fast i can download from a gigE machine from ftp.heanet.ie and sure!
[10:21] <Mithrandir> elmo (and maswan): any thoughts on pointing no.archive.ubuntu.com to ftp.acc.umu.se instead of level3?  The latter is often faster.
[10:21] <dieman> it dies!
[10:21] <dieman> (playing with mirroring)
[10:21] <dieman> ftp.heanet.ie should be the default for .edu network users though, holy crap is it fast.
[10:21] <dieman> from the usa even
[10:22] <rob^^^> dieman: "how fast is it ;)"
[10:22] <dieman> heh
[10:23] <dieman> i was doing 100mbps to the mirror ive got
[10:23] <dieman> but another box i was testing with was 200mbps+
[10:24] <rob^^^> dieman: I'm bottlenecked by my 10/100 switch to my campus mirror
[10:24] <dieman> heh
[10:24] <rob^^^> ooh
[10:24] <rob^^^> idea ;)
[10:24] <dieman> ive got a mirror up here at cs.umn.edu too
[10:24] <dieman> but its only 100mbps
[10:24] <dieman> i need to get my hands on a gigE card for it perhaps though
[10:24] <dieman> yeah
[10:24] <dieman> got it up to 445mbps for 3 tcp streams
[10:40] <\sh> g'evening _mvo_ 
[10:41] <rob^^^> boo, only 5.5 megs a second on VirtualPC :(
[10:41] <\sh> buh...mvo@ping.de and _mvo_@ish.de both are well known provider *grin*
[10:41] <rob^^^> ooh up to 7.5 now;)
[10:42] <fabbione> \sh: dude.. you blog too much
[10:42] <\sh> fabbione: it's planet 
[10:42] <\sh> fabbione: there is a bug
[10:42] <_mvo_> hey \sh!
[10:42] <fabbione> ah ok
[10:42] <\sh> fabbione: and I want to fix it with keybuk...but he refused ... brrr
[10:42] <haggai> what's the best tool to import CVS into a new baz archive?
[10:43] <_mvo_> \sh: my ish was down for a couple of minutes, but it's back apparently
[10:43] <fabbione> haggai: probably tailor
[10:43] <\sh> _mvo_: well..that's ish ;)
[10:43] <haggai> fabbione: thanks, I started looking at that and was wondering if it is any good. Have you used it?
[10:43] <_mvo_> \sh: actually I'm pretty happy with it
[10:43] <fabbione> haggai: approx the time to apt-get install it, see it failing on svn and purge it
[10:44] <haggai> fabbione: uhh :)
[10:44] <fabbione> haggai: get the latest from upstream if you want to play
[10:44] <janimo> fabbione, do you know if xdm worked when you uploaded it last time?
[10:44] <fabbione> the package is old and buggy
[10:44] <fabbione> janimo: yes it did
[10:44] <haggai> hmmkay
[10:44] <janimo> fabbione, it does not start here and I recall someone saying the same thing last week
[10:44] <\sh> bah
[10:44] <\sh> _mvo_: ish is ok ..
[10:44] <janimo> I found some things needed changing in order for it to start
[10:45] <fabbione> janimo: it's universe :)
[10:45] <janimo> it looked for config file in /usr/lib/X11 instead of /etc
[10:45] <janimo> and tried to load a greeter .so which was not built
[10:45] <janimo> so can I upload my change then?
[10:45] <fabbione> janimo: sure.. i am not jalous of my pkgs
[10:46] <\sh> janimo: what is it?
[10:46] <fabbione> not the one in universe at least
[10:46] <janimo> fabbione, ok I was just trying to make sure it is not broken beyond hope and I waste time on it
[10:46] <janimo> \sh xdm
[10:46] <janimo> or you mean the .so?
[10:46] <janimo> Some libGreeter.so
[10:47] <janimo> I got it to start eventually 
[10:47] <\sh> janimo: hmmmm....
[10:47] <\sh> janimo: I think ogra was working on it..and something is really broken
[10:47] <janimo> \sh what is it working for you?
[10:47] <\sh> janimo: I'm using gdm or kdm
[10:48] <janimo> \sh well it seems to work now
[10:48] <\sh> hu_
[10:48] <janimo> well those are out for xubuntu
[10:48] <\sh> ?
[10:48] <ogra> \sh, i had a look at it and had no clue whats broken was what i said ;)
[10:48] <\sh> ogra: ok...but it was somehow broken
[10:48] <ogra> i think it still is
[10:49] <\sh> well..1:11 still left until I fixed the universe *lol*
[10:52] <zyga> hello overworked good people
[10:54] <\sh> hey zyga 
[10:56] <zyga> hello, how is ubuntu doing today?
[11:05] <zyga> new kernel package
[11:05] <dholbach> re
[11:16] <dieman> heh
[11:16] <dieman> fix my cd images ehre, and start doing 40-50mbps out
[11:22] <pvanhoof> dudes. by the time I can upgrade my breezy, you have have 106 new packages. You guys are working to hard
[11:22] <maswan> dieman: heh, yeah, ftp.heanet.ie is fast. :)
[11:23] <dieman> they have a *scsi cache*
[11:23] <dieman> raid-0
[11:23] <dieman> because the 12tb or raid is 'too slow'
[11:23] <dieman> of, rather
[11:23] <maswan> dieman: :)
[11:23] <Kamion> silly ssh's connection sharing support is a bit racy ...
[11:23] <dieman> i liked the graphs in their apachecon presentation (i think thats the right one)
[11:23] <maswan> dieman: I'm syncing up some test data for the offload host
[11:23] <dieman> showing that 4gb -> 12gb with PAE is worthless.
[11:24] <dieman> but 32GB works great!
[11:24] <maswan> hehe
[11:24] <dieman> hrm, their last ubuntu sync took 2hrs
[11:24] <dieman> im going to probally have to resync my ubuntu here and push back the interval
[11:25] <maswan> .pool/ubuntu-5.04-install-powerpc.iso
[11:25] <maswan>    654131200 100%   19.54MB/s    0:00:31  (10, 5.4% of 479)
[11:25] <maswan> :)
[11:25] <dieman> yeah
[11:25] <dieman> i did 445 to a gigE connected machine
[11:25] <dieman> (mbps)
[11:25] <dieman> i was impressed
[11:25] <maswan> ftp.sunet.se has bigger hardware though. I wonder what kind of rates I'd get from them
[11:26] <dieman> heh
[11:26] <dieman> only 2 of the mirrors on their box take 2 hours (that they report)
[11:26] <dieman> gentoo-portage
[11:26] <dieman> and ubuntu
[11:27] <maswan> hmm.. only about the same rates though
[11:27] <dieman> yay, nearly through universe
[11:27] <maswan> http://ftp.sunet.se/about.html
[11:28] <notmdz> seb128: which package is responsible for the "Failed to find mountpoint for device <foo> in /etc/fstab" error that I get from totem?
[11:28] <dieman> we generally can't get funding like that for mirrors
[11:28] <dieman> since we dont have this whole 'transit is expensive' problem much.
[11:28] <maswan> yeah, those are exceptions
[11:28] <dieman> and never really did
[11:29] <maswan> a few NRENs do
[11:29] <notmdz> oh, it's totem itself
[11:29] <seb128> notmdz: I would say totem, we have upstream and downstream bugs about it
[11:29] <notmdz> odd that it doesn't even open /etc/fstab before giving that error
[11:29] <notmdz> which is why I assumed it came from another place
[11:30] <maswan> dieman: in the sunet case it is considered a strategic project, to provide a good general and "close" mirror for the swedish academia and the genreal public
[11:30] <dieman> yeah
[11:30] <maswan> dieman: also, back in the days, it helped to even out the balance on the transit links.
[11:30] <dieman> heh
[11:30] <dieman> most spots in the US never really had that sort of pressing issue
[11:31] <notmdz> hmm, it's doing a gnomevfs call
[11:31] <maswan> yeah, back in the 90:ies the transatlantic link situation wasn't as good as it is today
[11:31] <seb128> notmdz: it uses gnomevfs yep, the message has not been updated since gnomevfs has switched to hal
[11:32] <seb128> notmdz: the code is totem-1.2.0/src/plparse/totem-disc.c: cd_cache_get_dev_from_* if you want to debug it
[11:32] <notmdz> seb128: I get that error if g-v-m is not running
[11:33] <seb128> hum
[11:33] <seb128> lemme try
[11:33] <notmdz> I was trying to use totem on a system which doesn't run GNOME
[11:36] <notmdz> starting g-v-m got me past that error, but now it segfaults when I try to play the DVD ;-)
[11:36] <seb128> notmdz: right, same here. I've to mount the DVD before using totem
[11:39] <seb128> grumpf
[11:39] <seb128> does somebody knows about a "running dhclient breaks the loopback" bug?
[11:39] <notmdz> seb128: yes
[11:40] <notmdz> seb128: http://bugzilla.ubuntu.com/show_bug.cgi?id=10174
[11:40] <seb128> which is great because GNOME waits on some timeout when you try to log out or log in
[11:40] <seb128> thanks
[11:41] <notmdz> dhclient should probably just abort if you don't tell it which interface to use
[11:42] <seb128> yeah
[11:42] <\sh> I'm off, elmo: you have mail :) and the rest: have a nice night, morning, afternoon, whereever you are, whatever you do...
[11:44] <notmdz> seb128: would a backtrace for this crash be at all interesting?  seems to be gst-ffmpeg related
[11:44] <seb128> notmdz: if the backtrace is correct yep
[11:46] <dholbach> slomo will be delighted to hear :)
[11:46] <seb128> notmdz: I've uploaded gst-plugins0.8_0.8.11-0ubuntu5. I've fixed a Depends on ${misc:Depends} not listed (and useful for dh_gconf) and dropped the autosinks which are bugged (crash totem by example), upstream recommand not shipping them before 0.9/0.10
[11:47] <seb128> notmdz: should I mail you about it?
[11:47] <notmdz> seb128: approved
[11:47] <seb128> thanks
[11:47] <slomo> dholbach: what gst-ffmpeg crash?
[11:47] <seb128> slomo: notmdz's one :)
[11:47] <notmdz> building it with debug now
[11:48] <slomo> ok, i can take a look at it... i've worked with gdb for hours today so some more wouldn't hurt ;) how can it be reproduced?
[11:50] <notmdz> slomo: happens when I try to play a DVD in totem
[11:54] <notmdz> hmm, I built with nostrip and it's still stripped
[11:56] <Kamion> notmdz: http://people.ubuntu.com/~cjwatson/germinate-output/ubuntu-server-breezy/, FYI
[11:57] <Kamion> notmdz: (ubuntu-devel@lists.ubuntu.com/seeds--breezy-server--0)
[11:58] <Kamion> notmdz: it looks like it should come out smaller than our normal images, so at least that's one place where we won't be under space pressure for now
[11:58] <notmdz> Kamion: I should hope it does
[11:59] <Kamion> well, one never knows
[11:59] <Kamion> it's not a lot smaller, only 15-20MB or so
[11:59] <notmdz> slomo: I have a trace for you
[12:00] <notmdz> not sure how valid it is, but it's something
[12:00] <slomo> notmdz: and my dvd drive is busy for the next 12 hours... i'm currently testing thoggen and theora encoding is so damn slow :(
[12:01] <notmdz> slomo: what's your email address?
[12:01] <slomo> notmdz: slomo@ubuntu.com
[12:02] <slomo> notmdz: hey, you assigned a bug to this address a few hours ago ;)