[12:01] <pitti> Nafallo: if you really want to do universe updates again (YAY! :-) ), why not start with something more useful like ethereal=
[12:01] <pitti> ?
[12:01] <pitti> or clamav, spamassassin, etc.
[12:02] <pitti> Nafallo: I'd be fine with the new upstream microrelease for ethereal
[12:03] <Nafallo> pitti: mostly cause ethereal made me stop :-/. I can't get that patch down to better sizes and that made me jump off this train for a while :-/.
[12:03] <ajmitch> Nafallo: yes, ethereal is a rather evil package for security
[12:03] <pitti> Nafallo: as I said, if you package and test 0.10.11, that's fine for me in that case (the upstream release is cautious, and porting 15 patches doesn#t make sense)
[12:04] <pitti> ajmitch: that's why it is in universe, although it's highly useful
[12:04] <ajmitch> if only phpgroupware didn't take so long to download..
[12:04] <jasoncohen> pitti, why is it an evil package?
[12:04] <Nafallo> pitti: oki. I'm sorry I didn't took a chat with you at the time, but now I'm back atleast :-).
[12:05] <pitti> jasoncohen: it's just evil security-wise :-) it has a very bad vuln history
[12:05] <pitti> Nafallo: no need to be sorry :-)
[12:05] <ajmitch> jasoncohen: there are about 20 open CANs for ethereal on that list :)
[12:06] <jasoncohen> pitti, ubuntu-cve/fixed.html doesn't show CAN-2005-1921 as being fixed for hoary but it is. see USN-147-1 https://www.ubuntulinux.org/support/documentation/usn/usn-147-1
[12:06] <Nafallo> pitti: as soon as elmo add my key to the keyring I upload bugzilla for warty-security if that's okey? ;-)
[12:07] <pitti> Nafallo: well, warty's universe is a lost cause, but of course it's appreciated :-)
[12:07] <jasoncohen> ajmitch, ouch, i didnt' realize ethereal was that bad...what do the security issues involve?
[12:07] <Nafallo> pitti: the patch have been done for ages actually ;-)
[12:07] <pitti> jasoncohen: most of them are DoS only, but some are exploitable to code exection AFAIK
[12:07] <ajmitch> jasoncohen: lots of buggy plugins
[12:08] <ajmitch> format string issues, buffer overflows if you're unlucky
[12:08] <pitti> jasoncohen: http://people.ubuntu.com/~pitti/ubuntu-cve/nonvuln.html
[12:08] <pitti> CAN-2005-1921 	
[12:08] <pitti> php4	(hoary/main, breezy/main)
[12:10] <jasoncohen> but USN-147-1 shows CAN-2005-1921 for both hoary & warty
[12:10] <pitti> jasoncohen: yes, but php4-pear is in universe for hoary (from php4-universe source package)
[12:11] <pitti> jasoncohen: hoary was fixed more as a courtesy while we were at it
[12:11] <pitti> and because this was such a nasty one
[12:11] <jasoncohen> so libapache2-mod-php4 was never vulnerable in hoary?
[12:11] <pitti> nope
[12:12] <pitti> well, not against that XMLRPC issue at least :-)
[12:14] <jasoncohen> pitti, what are the hardest packages to backport security fixes for other than mozilla & firefox?
[12:14] <pitti> hmmm
[12:14] <pitti> jasoncohen: squid is relatively hard
[12:15] <pitti> and I had some pretty hairy samba patches
[12:15] <pitti> postgresql was nontrivial, too, but luckily I'm the Debian maintainer :-)
[12:16] <ajmitch> pitti: phpgroupware has 0.9.16.005-1 in hoary, 0.9.16.006-1 in debian, do you want only security fixes backported?
[12:17] <pitti> ajmitch: yes, is it so hard?
[12:17] <ajmitch> not terribly hard
[12:17] <ajmitch> since I know the upstream source
[12:17] <Amaranth> nice
[12:17] <pitti> I don't want to get more sloppy with upstream versions than now
[12:17] <ajmitch> ok
[12:17] <Amaranth> i have a 4 line python script that loads up a glade file, shouldn't have any problems right?
[12:17] <Amaranth> i managed to crash it due to the stupidity that is bonobo docks
[12:17] <jasoncohen> pitti, so, how long is too long for security updates? even on hoary, it took quite a while to get firefox updates (over 2 weeks)
[12:18] <pitti> jasoncohen: because I 
[12:18] <pitti> spent a whole week with backporting
[12:18] <jasoncohen> heh, ouch
[12:18] <pitti> that shouldn't happen any more with new upstream versions
[12:18] <Amaranth> ooh, this is even reproducable
[12:18] <pitti> jasoncohen: the minimal patches were  > 200 KB in total
[12:18] <jasoncohen> when would you have had 1.0.6 out if you just had to package the new upstream release?
[12:19] <pitti> on the very day I guess
[12:19] <pitti> updating warty took me three days
[12:19] <pitti> because 0.9.3 -> 1.0.6 is a hell of a lot of changes
[12:19] <jasoncohen> 53 CANs
[12:19] <pitti> but for hoary it was rather trivial
[12:19] <pitti> jasoncohen: also because I had to update all language packs and stuff
[12:20] <pitti> and sort out ancient patches
[12:21] <Nafallo> pitti: why does ubuntu-cve say debian's ethereal is vulnerable? :-)
[12:21] <pitti> Nafallo: because they didn't put CANs in the changelogs
[12:21] <pitti> and I didn't manually add them to my db
[12:21] <pitti> (well, I can't do that for Debian anyway)
[12:21] <Nafallo> pitti: hehe, oki :-)
[12:27] <doko> Kamion, mdz: python-qt3 was demoted to universe, although it's a build-dep for hplip
[12:28] <mdz> doko: it's been in universe since warty from what I see
[12:29] <mdz> doko: also, hplip is in universe right now (only hplip-base is in main)
[12:29] <doko> mdz: my mistake, wrong merge
[12:29] <mdz> doko: thanks for getting hplip into shape
[12:30] <doko> mdz: currently scanning doesn't work, with ort without running as root
[12:30] <mdz> oh
[12:30] <mdz> that's not very good
[12:30] <mdz> did you add hplip to scanner?
[12:31] <Mez> mdz, tseng's willing to help us backport mono stuff, is that cool with you?
[12:31] <mdz> Mez: in breezy, or as a fork?
[12:32] <doko> Mez: if that results in a buildable ironpython ...
[12:32] <Mez> mdz: from breezy
[12:32] <Mez> to hoary
[12:32] <mdz> Mez: yes, certainly
[12:32] <Mez> mdz: apparently, the current mono stuff had already been made to work in hoary (cause they wanted to release a mono livecd)
[12:32] <Mez> so, it just needs backporting in the right order and stuff
[12:33] <Mez> doko: I have no idea what that is, so I'm not gonna comment
[12:33] <ajmitch> doko: well, we don't even have ironpython in breezy yet :)
[12:34] <doko> ajmitch: yes, because it doesn't build :-/
[12:34] <ajmitch> doko: great, I see you were the ITP filer.. does it not build with the mono in sid?
[12:36] <doko> ajmitch: I'll send you and tseng my current packaging, then you test it ;-P
[12:36] <ajmitch> doko: ok, it might be a good plan to put it in pkg-mono svn :)
[12:37] <doko> mdz: adding hplip to scanner doesn't help, will look at it later
[12:42] <Nafallo> I better goto sleep now that all three gerbils sleep in the same pile :-)
[12:42] <Nafallo> see you all tomorrow! :-)
[12:49] <pitti> night everybody
[12:57] <lamont> elmo: ^^^
[12:58] <elmo> because it's a PoS that doesn't work
[12:58] <lamont> I'll pass that along to taggart
[12:58] <elmo> I've told him about it
[12:58] <lamont> did we intentionally strip it?
[12:58] <elmo> yes
[12:58] <lamont> ok
[12:58] <elmo> it false negatives
[12:59] <elmo> we can pull it into universe if you really want, I just don't want it in main
[12:59] <lamont> what's a good alternative?
[12:59] <lamont> if none, lets go for universe, I guess
[01:00] <lamont> (bdale is playing with a machine, you see...)
[01:01] <elmo> you really want the HP propreitary gunk
[01:01] <elmo> but it doesn't work asa daemon
[01:01] <elmo> or rather it does, but doesn't log to syslog
[01:02] <Mez> elmo, can you kick ff from backports, and add mono please
[01:02] <elmo> mez: pls mail me
[01:02] <Mez> elmo: kk
[01:02] <Mez> sorry
[01:16] <hughsie> ogra: ping?
[01:16] <ogra> hughsie, ?
[01:17] <hughsie> ogra: you may or may not be aware that libnotify released 0.2.1 today
[01:18] <hughsie> 0.2.0 was released yesterday, but had some dbus bugs we squashed
[01:18] <ogra> hughsie, you may or may not be aware that libnotify entered our archive today ;)
[01:18] <hughsie> okay, nice one.
[01:18] <ogra> 0.2.1-0ubuntu1
[01:18] <ogra> :)
[01:18] <hughsie> quicker than me :-)
[01:18] <ajmitch> looks like seb uploaded 0.2.1 very soon after 0.2.0
[01:19] <ogra> thast seb128 .... dont get in his way *g*
[01:19] <hughsie> you can build g-p-m with --enable-libnotify for all the new goodness
[01:19] <ogra> yeah, i'll do
[01:19] <hughsie> you'll need notification-daemon as well for it to work
[01:20] <hughsie> ogra: i've seen some of the ubuntu/debian patches that create a debian directory in the source tarball
[01:20] <hughsie> is that for every package?
[01:20] <ogra> yep
[01:20] <ogra> a debian package consists of:
[01:20] <hughsie> should it be upstream? is it easier that way
[01:21] <ogra> the orig.tar.gz file (original source )
[01:21] <hughsie> k
[01:21] <ogra> the diff.gz (all distro changes and the debian dir)
[01:21] <Mez> ogra, did youy archive php4-universe
[01:21] <ogra> and a .dsc file that describes the two...
[01:21] <ogra> Mez, nope
[01:22] <Mez> just wondering cause you mentioned about ti going into main
[01:22] <Mez> weird...
[01:22] <hughsie> ogra: cool, okay, so non of the /debian stuff is easier in package?
[01:22] <ogra> hughsie, i wouldnt mix upstream source and debian dir
[01:23] <hughsie> ogra: cool, i'm new to all this debian stuff
[01:23] <ajmitch> having the debian dir in upstream can be limiting
[01:23] <hughsie> ajmitch, ok, thanks.
[01:23] <ogra> except i develop it explicitly for this distro ... but if outher debian based distros want to package it differen mixing upstream source and packaging stuff can get odd
[01:24] <hughsie> ogra: could you (do you want to) package cvs g-p-m or shall I release 0.1.1?
[01:24] <ogra> hughsie, as you like... 
[01:24] <ogra> i can do both... 
[01:24] <hughsie> ogra: cvs is churning with new stuff at the moment
[01:24] <ogra> i think we'll have to wait a bit for dbus... 
[01:24] <hughsie> i'll let it settle then do 0.1.1?
[01:25] <hughsie> okay, define "bit"
[01:25] <ogra> daniels is the guy who modularizes X 
[01:25] <ogra> he's 4/5 throught... i dont know if he can make dbus inbetween
[01:25] <ogra> s/throught/through
[01:26] <hughsie> ogra: okay, cool. I'll just keep hacking
[01:26] <hughsie> you seen the hal patches I've sent recently?
[01:26] <ogra> i must admit i havent looked deeply...
[01:27] <ogra> but i saw you sent patches :)
[01:27] <hughsie> cool, i'm putting lots of the clever logic in hal so that other projects (like battstat-applet can use them
[01:29] <hughsie> ogra: I sleep now, I'll catch you guys tmw.
[01:29] <ogra> ah yes, i saw the request from ryan
[01:29] <ogra> hughsie, great, see you around :)
[01:29] <hughsie> cool. thanks mate
[02:08] <Mez> cause daniels = god ?
[02:09] <mrd`> Possibly.
[02:09] <mrd`> But... I didn't think I updated any packages that could have fixed it.
[02:10] <Mez> mrd`, it must have been magic(tm) then
[02:11] <mrd`> Hm, I did upgrade gnome-session... that could have done it.
[02:20] <jasoncohen> Amaranth, hey- you were right about needing gnome-menus 2.10.2
[02:21] <jasoncohen> Amaranth, i logged back in and now the multimedia menu is gone, and several menu entries that didn't show up before are now present. in addition i can delete menu entries whereas i couldn't before
[02:21] <jasoncohen> Amaranth, this is only a problem with the official backports server because the unofficial mirrors have gnome-menus and libgnome-menus 2.10.2
[02:24] <Mez> jasoncohen, I can make a special request to James to sort that out
[02:24] <jasoncohen> thanks
[02:26] <jasoncohen> Mez, can you get mozilla-mplayer 2.85 into breezy. it's already in sid
[02:28] <jasoncohen> http://packages.debian.org/unstable/misc/mozilla-mplayer
[02:28] <jasoncohen> it's been in sid for over a month
[02:28] <jasoncohen> and 2.75 was uploaded in May
[02:31] <jasoncohen> anyone here know if the official firefox build can upgrade properly using the included upgrade functionality? the last time i tried it didn't find 1.0.6 and i had to diownload it and install it manually
[02:31] <Burgundavia> jasoncohen, contat ogra about mozilla-mplayer
[02:32] <Mez> jasoncohen, ask in -motu
[02:33] <jasoncohen> ok
[02:35] <jasoncohen> Mez, i'd like to see a gtk2 version of 2.85 in breezy if possible
[02:35] <Mez> jasoncohen, tired :d talk later
[02:35] <jasoncohen> ok
[02:42] <j^> what happend to network-manager?
[02:42] <j^> 0.4.1+cvs20050618-3 is totaly broken
[02:42] <Burgundavia> yes
[02:43] <Burgundavia> the changelog says that
[02:43] <j^>  Unfortunately these changes are not properly
[02:43] <j^>     tested :-(.
[02:43] <j^> thats not is totaly broken
[02:45] <j^> any plans to unbreak it?
[02:45] <Burgundavia> ian is working on it
[02:46] <jasoncohen> now that ubuntu is packaging the latest upstream release, can they use the real firefox icon?
[02:46] <mrd`> It's still patched, isn't it?
[02:47] <jasoncohen> will they be patching it in addition to providing upstream releases?
[02:47] <Burgundavia> jasoncohen, no idea
[02:48] <jasoncohen> i have no idea why but browsing feels faster on the official build v. the ubuntu 1.0.6 build. i can't explain it
[02:48] <jasoncohen> the ubuntu version seems to freeze for a second between switching pages. the official build doesn't
[02:49] <wm_eddie> I think there might be a bug in shipit's zip code field.  Anyone know who I can e-mail to let them know?
[02:50] <wm_eddie> (My zipcode is 00976 and the label on the packaging said 976 and the USPS had a really hard time getting it to where it had to go.
[02:59] <sladen> wm_eddie: mail mako
[03:00] <wm_eddie> ok.
[03:18] <Amaranth> jasoncohen, Mez: It's amazing how much of a difference it makes, isn't it? (gnome-menus)
[03:19] <Mez> Amaranth, meh
[03:19] <Burgundavia> Amaranth, what do you mean?
[03:19] <Amaranth> Burgundavia: about an hour ago jasoncohen was telling me how much of a difference gnome-menus 2.10.2 made as far as things working
[03:20] <Burgundavia> ah
[03:20] <Amaranth> jasoncohen: btw, the patches for GNOME integration make it impossible to use the official branding
[03:20] <Amaranth> jasoncohen: for firefox, that is
[03:24] <astronut> I'm using the live cd to see how well linux handles all of my laptop's hardware....i fixed the PCMIA bus by modifying some PCI stuff, but to get the full features ofmy touchpad, i have to patch the kernel..how can i do this on a live cd? (Can i create a custem kernel on floppy, or reburn the cd, etc)
[03:26] <OddAbe19> what's the safest (or best) upgrade to breezy at the moment? Dist-upgrade or upgrade? What breakage would dist-upgrade bring versus upgrade?
[03:27] <sladen> astronut: you shouldn't /have/ to patch your kernel;  is it a synaptics touchpad?
[03:28] <astronut> sladen, alps, see /usr/share/doc/xorg-(synaptics driver somethign)/README.alps
[03:28] <Lathiat> astronut: yeh you dont need to patch the kernel
[03:28] <Lathiat> astronut: just need to modify your X config
[03:29] <Lathiat> astronut: if you restart X with SHMConfig, you can use the 'tpconfig' to play with the values live, the default values for alps arent too great
[03:29] <astronut> Lathiat, it won't detect as a synatpics
[03:29] <Lathiat> astronut: you sure
[03:29] <astronut> it works, but won't do the scroll....i tried forcing synaptics, dmsesg said not one
[03:29] <astronut> Lathiat, oh, i've played a while
[03:29] <astronut> it uses generic drivers to act as just a mouse
[03:30] <astronut> what's SHMConfig/tpconfig?
[03:30] <astronut> not on the live cd...
[03:30] <Mez> hmm, I've just realisede how godamn hot my gf is :S
[03:30] <Mez> hehe
[03:30] <astronut> Mez, link?
[03:30] <Mez> perv astronut
[03:30] <Mez> but, one sec
[03:30] <astronut> Lathiat, i spent about 30 minutes trying to figure out WHY my mouse would work after i commented out all Mouse device sections in xorg.conf
[03:31] <astronut> Mez, you cna't make a statement like that without backing it up
[03:31] <Mez> lol :d
[03:31] <Mez> I'm just uploading astronut
[03:31] <Lathiat> haha mez
[03:31] <astronut> Lathiat: the livecd puts a synaptics section in here, but i can't force it to use synaptics
[03:32] <Mez> http://www.cheesenibbles.com/files/mez/emily.jpg
[03:32] <astronut> and i want to shoot whoever made , the default nick thing in xchat 2.x over the : in 1.x
[03:32] <astronut> Mez: girls like that don't go out with geeks...what'd you do?
[03:33] <HrdwrBoB> gave her money
[03:33] <HrdwrBoB> all girls like money
[03:33] <OddAbe19> hooker's like money too
[03:33] <HrdwrBoB> ;)
[03:33] <OddAbe19> what does that tell me
[03:33] <astronut> beat me too it OddAbe19 
[03:33] <OddAbe19> :-p
[03:33] <Mez> astronut, she's a geek too
[03:33] <astronut> Mez: liar...
[03:33] <OddAbe19> what's the safest (or best) upgrade to breezy at the moment? Dist-upgrade or upgrade? What breakage would dist-upgrade bring versus upgrade?
[03:33] <Lathiat> OddAbe19: dont :)
[03:34] <HrdwrBoB> OddAbe19:  the safest way is to not
[03:34] <OddAbe19> lol, i've been antsey for about 2 months now, but i was waiting for X and GCC changes to end
[03:34] <OddAbe19> now that they are
[03:34] <Lathiat> well, X hasn't ended yet
[03:34] <Mez> no, seriously, she is - lol :D hehe :D her dad works for apple and stuff :D and she's my lil geek (she wont stop pestering me to install kubuntu on her PC for her!)
[03:34] <HrdwrBoB> astronut: depends, geek doesn't mean you can't be a nice person :)
[03:34] <OddAbe19> it's less borked
[03:34] <Lathiat> Mez: bastard
[03:34] <astronut> or what x transition is that?
[03:34] <HrdwrBoB> Mez: kubuntu? ick
[03:34] <astronut> i'm a debian guy myself
[03:34] <Lathiat> astronut: it is
[03:34] <HrdwrBoB> my wife isn't a geek, but she uses ubuntu
[03:34] <astronut> monolithic -> modular tree?
[03:35] <Lathiat> astronut: yes
[03:35] <Mez> HrdwrBoB, my gf is and she uses kubuntu :D but each to their own
[03:35] <Lathiat> astronut: tahts wahts happening
[03:35] <Mez> and Lathiat what did I ever do to you
[03:35] <astronut> Mez: once oyu're out of the picture, he can steal her
[03:35] <Lathiat> Mez: have a good looking geek girlfriend :)
[03:35] <Lathiat> hrm.. anyone know how i can compare two struct sockaddrs
[03:35] <astronut> Lathiat: tpconfig isn't in the live cd, what package is it in?
[03:36] <Lathiat> astronut: tpconfig, oddly enough :)
[03:36] <Mez> Lathiat, and she loves same kind of music as me and loves gaming, and. ... just... *drools* every geeks dream :D
[03:36] <Mez> hehe
[03:36] <astronut> meh...i guess we'll have to do apt-get update
[03:36] <astronut> not sure how well live cd will like this
[03:36] <astronut> security/restricted???
[03:37] <Mez> anyhoo ... am off to bed :D hehe :D
[03:37] <astronut> E: Couldn't find package tpconfig
[03:38] <astronut> Lathiat: ?
[03:40] <astronut> meh...i guess cd's are weird
[03:41] <astronut> livecd*
[03:41] <astronut> found it online, installing via dpkg
[03:42] <astronut> Lathiat: ubuntu@ubuntu:~$ sudo tpconfig  -i
[03:42] <astronut> fatal:
[03:42] <astronut> No Synaptics or ALPS touchpad device found
[03:42] <astronut> ubuntu@ubuntu:~$
[03:42] <OddAbe19> hahaha, OMG, if i upgraded to breezy (not dist-upgrade) this would happen: 774 upgraded, 0 newly installed, 0 to remove and 400 not upgraded.
[03:42] <OddAbe19> Need to get 456MB of archives.
[03:42] <OddAbe19> After unpacking 41.4MB of additional disk space will be used.
[03:43] <OddAbe19> holy crap
[03:43] <OddAbe19> is mkfontsdir fixed yet?
[03:44] <Lathiat> OddAbe19: no
[03:44] <Lathiat> astronut: did you enable the SHMConfig
[03:45] <OddAbe19> how would i fix that?
[03:46] <Lathiat> OddAbe19: wait longer before upgrading :)
[03:46] <OddAbe19> no
[03:46] <jasoncohen> Amaranth, what exactly do the patches provide?
[03:46] <OddAbe19> i don't wanna
[03:46] <OddAbe19> lol
[03:46] <jasoncohen> Amaranth, the gnome integration patches in ubuntu's firefox
[03:47] <Lathiat> OddAbe19: http://blogs.gnome.org/cball has information, but dont come crying in here if its broken
[03:47] <Amaranth> jasoncohen: the make it use the file chooser and such
[03:47] <Lathiat> OddAbe19: this is a development channel not a user support channel
[03:47] <OddAbe19> i know it's development
[03:47] <OddAbe19> i was courious
[03:47] <jasoncohen> Amaranth, do you have any idea why i would be getting a slight delay between switching pages in ubuntu's firefox while the official build works fine?
[03:47] <Amaranth> jasoncohen: nope
[03:47] <jasoncohen> it only seems to be an issue on ubuntu builds. it works fine in debian
[03:48] <Amaranth> can someone tell me if they're seeing http://bugzilla.ubuntu.com/show_bug.cgi?id=13066 ?
[03:48] <jasoncohen> i've noticed in backports as well- but those are just breezy packages
[03:48] <Lathiat> jasoncohen: could be a pango thing
[03:48] <Amaranth> ah yeah, probably pango
[03:48] <astronut> Lathiat: no..where do i do that?
[03:48] <Amaranth> pango trades speed for being able to render every language ever heard of
[03:49] <jasoncohen> lol
[03:49] <jasoncohen> and i need that because....>
[03:49] <jasoncohen> ?
[03:50] <Amaranth> *shrug*
[03:50] <jasoncohen> Amaranth, what languages would i get in the ubuntu build that i don't get in the official build?
[03:50] <Lathiat> Amaranth: the bonobo docks are oh so broken
[03:50] <Lathiat> Amaranth: they arent draggable anyway ;p
[03:50] <Amaranth> Lathiat: i thought so
[03:50] <Amaranth> you can drag them, once
[03:51] <Lathiat> Amaranth: and yes happens for me
[03:51] <astronut> Lathiat: where do i enable that...SHMConfig?
[03:51] <jasoncohen> both hebrew and russian render fine on the official build
[03:51] <Lathiat> astronut: in your X config, google, and please ask in #ubuntu, this is a developemnt channel not a user support channel
[03:51] <Amaranth> Lathiat: if you have editbugs can you say that in bugzilla?
[03:51] <Lathiat> Amaranth: i dont need editbugs just to comment?
[03:52] <Amaranth> no, i meant to NEW it
[03:52] <Lathiat> oh
[03:52] <Lathiat> yeh well i cant do that
[03:52] <Amaranth> i suppose i could, but that doesn't look very good
[03:52] <Lathiat> haha
[03:52] <Amaranth> a comment will work though, thanks
[03:55] <jasoncohen> Amaranth, i guess the official firefox build is lighter then because it doesn't require pango or many other dependencies
[03:56] <Lathiat> jasoncohen: ya, i noticed that using the deer park builds
[03:56] <Amaranth> well, those are so much faster anyway...
[03:56] <jasoncohen> yeah, i'm comparing 1.0.6 official to 1.0.6 ubuntu hoary
[03:56] <Amaranth> those with the GNOME integration will be faster than official 1.0 :)
[03:56] <jasoncohen> though, doesn't debian build with the same compile options & dependencies?
[03:57] <infinity> Reasonably similar, anyway.
[03:57] <jasoncohen> Amaranth, well- doesn't hoary's have gnome integration?
[03:57] <Amaranth> yes
[03:57] <jasoncohen> but i only notice the problem in ubuntu- no such "freezing' in sarge
[03:57] <jasoncohen> or sid
[03:57] <Amaranth> the gnome integration is a seperate package
[03:57] <Amaranth> i think
[03:57] <jasoncohen> yeah
[03:58] <jasoncohen> mozilla-firefox-gnome-support
[03:59] <jasoncohen> is already instaleld
[04:00] <jasoncohen> hmm, the debian version doesn't require the bonobo or gnome dependencies
[04:02] <Amaranth> bonobo? blech
[04:02] <jasoncohen> yeah- debian doesn't require the bonobo libs or gnome libs but ubuntu does
[04:03] <mrd`> Odd, /usr/lib/mozilla-firefox/firefox-bin doesn't use any bonobo or gnome libs.
[04:03] <jasoncohen> deerpark will have automated security upgrades (and it won't upgrade if it would break one of your extensions dependeing on the setting). that'll be nice
[04:03] <Lathiat> jasoncohen: thats kinda cool
[04:04] <jasoncohen> yeah, i just tried an alpha build of deerpark today
[04:04] <Amaranth> it does binary deltas too
[04:04] <Amaranth> for the upgrades
[04:04] <jasoncohen> unfortunately the upgrade system isn't up yet- since releases are done by daily builds
[04:04] <Amaranth> yeah
[04:04] <Amaranth> the upgrade system is setup, it's just doing the 1.0 style
[04:05] <Amaranth> where it downloads the whole thing and reinstalls it
[04:05] <jasoncohen> do upgrades actually work in 1.0.x?
[04:05] <Lathiat> problem is
[04:05] <Amaranth> they do on windows
[04:05] <jasoncohen> when 1.0.6 came out it didn't show the update
[04:05] <Lathiat> how does that work on say liniux
[04:05] <Lathiat> where you cant overwrite the binaries?
[04:05] <Lathiat> (in many cases)
[04:05] <Amaranth> it doesn't work on windows when you can't overwrite the binaries either :P
[04:05] <jasoncohen> yeah
[04:05] <Amaranth> which is almost never, but they should start planning
[04:05] <jasoncohen> install in /opt and give user permission
[04:06] <mrd`> Why not just install in your home directory?
[04:06] <Amaranth> /opt is off limits too
[04:06] <Amaranth> you want ~/.local/
[04:06] <jasoncohen> well, i installed in /home but some don't like doing that
[04:06] <Amaranth> ~/.local/ should work for most things
[04:07] <jasoncohen> if the upgrade system is good on 1.5, could the apt package support it?
[04:07] <jasoncohen> or will official upgrades always be through new .deb packages?
[04:08] <jasoncohen> i.e - 1.5 is packaged by ubuntu but upgrades are handled by the automated upgrade tool which is configured pre-install
[04:09] <Amaranth> firefox is C++
[04:10] <Amaranth> unless their upgrade tool had g++ 3.3 and g++ 4.0 versions...
[04:10] <wasabi> Hey so I had this neat idea... and was expecting that somebody had already done it, since usually somebody else does my neat ideas before i Have them.
[04:10] <Lathiat> wasabi: heh
[04:11] <wasabi> I want to make a livecd that boots from a flash disk and saves changes in a seperate place.
[04:11] <wasabi> Basically for an embedded router device.
[04:11] <ajmitch> wasabi: don't exsting live cds already allow for that?
[04:11] <wasabi> LiveCD have most of the infrastructure for that?
[04:11] <ajmitch> iirc knoppix let you save settings on a usb drive
[04:11] <wasabi> I want to use Ubuntu.
[04:11] <Lathiat> knoppix basically tars up /etc and /home
[04:11] <Lathiat> sticks that on a flash drive
[04:11] <Lathiat> and untars it on next boot
[04:12] <mrd`> They support booting from a CD and saving to a usb drive... I don't see why you couldn't boot from a usb drive and save to another usb drive or somewhere else if that's what you want.
[04:12] <wasabi> Ubuntu live CD uses some unionfs or something to track mods.
[04:12] <wasabi> I want to save the mods.
[04:12] <Lathiat> wasabi: yeh what you want is like unionfs i think
[04:12] <ajmitch> Lathiat: right, so you'd want an overlay fs, I can't recall what the live cd does for ubuntu
[04:12] <wasabi> BAsically on this flash disk, I can stick a file containing the "live cd".
[04:12] <wasabi> And another file containing the user mods.
[04:13] <wasabi> And at boot just set it up exactly like hte live cd does
[04:13] <wasabi> You'd be able to upgrade the device by just replacing that one file on the flash... like a cisco or something.
[04:13] <Lathiat> well
[04:13] <Lathiat> not sure thatd work so well
[04:13] <Lathiat> depends how your 'union' fs works
[04:13] <Lathiat> a cisco works liek that because it uses a configuration file
[04:13] <wasabi> Yeah. Are there docs about how the ubuntu live cd works?
[04:13] <Lathiat> whcih does everything
[04:13] <wasabi> Oh sure, but so does Linux. It just uses a bunch of config files.
[04:13] <wasabi> Which need to be isolated. ;)
[04:14] <Lathiat> sure but
[04:14] <Lathiat> what if the config file changes format
[04:14] <Lathiat> or whatever
[04:14] <Lathiat> or you wanna add stuff to the new version
[04:14] <Lathiat> .. etc :)
[04:14] <wasabi> Good question, I'll consider it later.
[04:14] <wasabi> I'd suspect that the dpkg upgrade routines should be run in some way.
[04:14] <jasoncohen> Amaranth, so, why does ubuntu's firefox have all these extra dependencies like bonobo libraries and gnome libraries that aren't needed for the debian or official builds?
[04:14] <wasabi> Just like as if it was really being upgraded.
[04:14] <ajmitch> jasoncohen: gnome integration
[04:14] <Amaranth> jasoncohen: For the GNOME integration.
[04:14] <wasabi> And those would be responsible for making it work, however... just like they are required to do now.
[04:15] <mrd`> Amaranth: Shouldn't only the firefox-gnome-integration package have those depends though?  Not the firefox package?
[04:15] <jasoncohen> Amaranth, but gnome-integration is a seperate pacakge
[04:15] <Amaranth> *shrug*
[04:15] <jasoncohen> apt-cache show mozilla-firefox-gnome-support --> it has a bunch of dependencies as well
[04:26] <wasabi> Doesnt' look like much of hte livecd documentation has been changed since warty
[06:10] <Burgundavia> wasabi, what do you mean?
[06:12] <Burgundavia> jasoncohen, the ubuntu version of FF has the upgrade notifier working. That FF stuff is a hack for systems that don't have a proper package management system (windows, primarily)
[06:12] <wasabi> hmm?
[06:13] <Burgundavia> wasabi, wht about the livecd stuff needs to be updated?
[06:13] <wasabi> LiveCDDesign
[06:13] <wasabi> Still talks about Warty, etc.
[06:13] <Burgundavia> ok
[06:13] <wasabi> Very little info on Casper.
[06:13] <wasabi> I'm trying to put together my own custom livecd, and not having an easy go at it. ;)
[06:13] <Burgundavia> I thought there were some other livecd docs lying around
[06:14] <wasabi> There are some for modifying the existing live cd
[06:14] <Burgundavia> ask Luis Villa (lu on irc) about it
[06:14] <wasabi> but none for creating one from scratch
[06:14] <Burgundavia> ah, ok
[06:14] <wasabi> I'm basically trying to create a Really Small one.
[06:15] <fabbione> morning
[06:21] <bddebian> Hello fabbione 
[06:26] <jasoncohen> Burgundavia, you sure? it didn't properly notify me when 1.0.6 was out. how does firefox learn of the new version?
[06:26] <jasoncohen> Burgundavia, i checked for updates using the official build and none were shown
[06:27] <Burgundavia> jasoncohen, the Ubuntu build disables the auto checking
[06:27] <Burgundavia> because we have our own package management system
[06:27] <jasoncohen> Burgundavia, i know- i was using the official build
[06:27] <Burgundavia> ok
[06:28] <jasoncohen> anyways 1.5 will have a much more advanced upgrade system that will allow you to upgrade without breaking extensions
[06:28] <jasoncohen> and do so automatically
[06:53] <wasabi> I am so close to having my custom live cd working
[06:53] <wasabi> except I can't get the new kernel installed somehow.
[08:03] <pitti> Good morning
[08:04] <\sh> doko: ping
[08:04] <\sh> morning :)
[08:04] <\sh> well...I didn't shower so far....so I have to rush
[08:05] <fabbione> hey pitti
[08:06] <tseng> some guy named ian of whom ive never heard trashed network-manager
[08:06] <tseng> (yay)
[08:07] <Lathiat> im over network-manager
[08:07] <Lathiat> its useless
[08:08] <infinity> I'm supposed to be putting some effort into whipping n-m into shape in the next week.
[08:08] <infinity> Anyone have any input more useful than "it's useless" or "egads, Ian broke it"?
[08:08] <Lathiat> infinity: please make it really easy for me to right click it and say "Stop network-manager from managing my devices" and then after that have it have a "Start managing my devices again"
[08:08] <tseng> infinity: well, the applet doesnt start
[08:08] <Lathiat> if you did that, i'd love you forever
[08:09] <tseng> infinity: so i cant much test it anymore
[08:09] <jdub> infinity: thom and i had a long discussion about integration with ifupdown and resolvconf a while back; pretty sure he wrote notes about it
[08:09] <infinity> jdub : I'll pester him.
[08:09] <jdub> infinity: oh, and it was related to making n-m work sanely :)
[08:09] <tseng> infinity: just sortof extra boggled as there are people who upload to ubuntu who never show their face in the community
[08:10] <infinity> jdub : Speaking of working sanely, why has no one made the obvious leap to use nscd, instead of switching from one nameserver to another?
[08:10] <jdub> tseng: ian has just started; he's a well known debian developer :)|
[08:10] <infinity> tseng : He's a new hire.  He's also the original dpkg author.
[08:10] <tseng> jdub: cheers.
[08:10] <\sh> doko: is hplip  build-depending on python-qt3? or on python-sip4-qt3?
[08:10] <Lathiat> infinity: im told nscd is evil, but i have nothing to base that on
[08:11] <jdub> infinity: dunno, though there are some strong technical/emotional opinions about nscd.
[08:11] <infinity> Lathiat : It's not perfect, but neither is running a local nameserver.
[08:11] <Lathiat> \sh: i saw an upload with a changelog sayign that was removed
[08:11] <jdub> infinity: (the original reason for using bind was that it was very configurable wrt multiple networks and name lookups)
[08:11] <Lathiat> infinity: but it affects mroe than just dns?
[08:11] <jdub> infinity: (nscd and resolv.conf doesn't quite solve that)
[08:11] <infinity> jdub : I don't dig the idea that a machine with NetworkManager installed can't have a nameserver installed (well, unless you ask the nameserver not to bind to localhost)... Kinda ick.
[08:12] <Lathiat> infinity: it can
[08:12] <Lathiat> infinity: network-manager starts its own bind9
[08:12] <\sh> Lathiat: thx...I should read first -changes then irclogs ,-)
[08:12] <Lathiat> with its own config file
[08:12] <Lathiat> the problem is
[08:12] <Lathiat> it just blanket installs bind9
[08:12] <Lathiat> and so bind9 runs as normal as well
[08:12] <infinity> Lathiat : Yes, but still on port 53.
[08:12] <infinity> Lathiat : Run your own instance, boom.
[08:12] <Lathiat> that needs to be sorted somehow if we're gona use bind9, tho i hear theres plans to use dnsmasq
[08:12] <Lathiat> infinity: but it binds to localhost
[08:12] <jdub> infinity: are you on the networkmanager list?
[08:12] <tseng> ian attempted to make it use resolvconf
[08:12] <Lathiat> infinity: so i suppose thats a mild issue
[08:12] <infinity> Lathiat : the latest packages swapped to dnsmasq, but the same issue survives.  You're eating port 53.
[08:12] <jdub> infinity: it's worth reading back about this stuff
[08:12] <Lathiat> infinity: but it would still listen on your real interfaces which is what you really care about
[08:13] <Lathiat> infinity: s/would/wouldnt
[08:13] <infinity> Lathiat : You;'d have to hand-configure bind not to bind to localhost, or it'll probably just die on invocation.
[08:13] <Lathiat> resolvconf woudl be nice
[08:13] <Lathiat> infinity: bind9 will start fine
[08:13] <Lathiat> infinity: by default, it binds to * which grabs all it can
[08:13] <Lathiat> so if it cant get localhost it wont care
[08:13] <jdub> brb
[08:13] <Lathiat> resolvconf woudl be nice
[08:13] <Lathiat> cus then it can use avahi-dnsconfd :)
[08:14] <infinity> resolvconf is what the current (broken) upload was meant to integrate.
[08:14] <Lathiat> right
[08:14] <Lathiat> so it just needs to be unbroken
[08:14] <Lathiat> :)
[08:14] <infinity> That doesn't change anything, though.
[08:15] <Lathiat> whats the issue with just changing /etc/resolv.conf? thats always seemed to work for me
[08:15] <\sh> but a running bind9 for a laptop install with NetworkManager? 
[08:15] <infinity> Many applications cache resolver requests, so changing resolv.conf on the fly doesn't buy you much, which is why we're still running a local nameserver.
[08:15] <infinity> So, resolv.conf always points to 127.0.0.1, and we only update it to update our search lists.
[08:16] <infinity> (Which is still not ideal, cause we could be cacheing the search lists)
[08:18] <tseng> its polling the hell out of something, likely dbus
[08:19] <tseng> but it never draws the little icon bit
[08:19] <infinity> Little icon bits are overrated.
[08:19] <Lathiat> heh
[08:20] <\sh> rushing to work....bbl
[08:20] <jdub> tseng: how's beagle faring?
[08:21] <tseng> jdub: 0.12 is pretty solid here
[08:21] <tseng> jdub: and joe is coming up w/ some more big performance boosts in cvs
[08:21] <tseng> we just need to get some crap moved to main so evo-sharp can build
[08:21] <tseng> and it will actually work for people
[08:22] <bob2> Lathiat: mozilla, e.g. seems to cache /etc/resolv.conf forever
[08:22] <tseng> jdub: if you get the evo-sharp source yourself, youll be in business
[08:22] <infinity> bob2 : Which, really, is something we should be fixing in mozilla, not working around it with hideous hacks... :/
[08:22] <bob2> well, yeah
[08:23] <tseng> hopefully the fixer isnt @redhat.com
[08:24] <tseng> or rather, they make it poll the file ocassionally instead of binding more of the desktop to dbus
[08:24] <infinity> It'll be a cold day in hell before mozilla.org accepts patches to make mozilla depend on dbus, I think.
[08:24] <Amaranth> dbus is our CORBA or XPCOM
[08:25] <infinity> I could be a crack monkey, but isn't CORBA our CORBA? ;)
[08:25] <Amaranth> haha
[08:25] <Amaranth> dbus is the new one
[08:25] <Amaranth> something that sounds cool and gets used way to much
[08:25] <tseng> infinity: we are quick to replace it with something more crackful
[08:25] <bob2> make it interpret SIGPOWER as "uncache the dns server addresses, idiot"
[08:26] <bob2> or some other pointless signal
[08:26] <Amaranth> that we have to spend years ripping out of things later
[08:28] <jdub> yay for breezy kernels on hoary
[08:28] <tseng> we should get rml's networkmanager patch
[08:28] <Lathiat> tseng: which does what?
[08:28] <tseng> makes it a real tray icon
[08:28] <Lathiat> how is it not a real tray icon?
[08:28] <tseng> instead of a proper applet forced badly into the tray
[08:28] <Lathiat> oh
[08:28] <Lathiat> heh
[08:29] <tseng> notice the funny gradient in clearlooks
[08:29] <tseng> and other subtle badness
[08:29] <Lathiat> yeh
[08:29] <Lathiat> and its extra whide
[08:29] <Lathiat> *wide
[08:29] <bob2> rml?
[08:29] <Lathiat> bob2: robert love
[08:29] <bob2> novell dropped netapplet?
[08:29] <tseng> dude rml <3 nm
[08:29] <bob2> I know wh orml is :)
[08:30] <bob2> but he wrote netapplet to begin with
[08:30] <infinity> tseng : I'm going to be talking to rml about his patches RSN.
[08:30] <infinity> tseng : He fixed up some VPN stuff too.
[08:30] <tseng> infinity: yay.
[08:31] <Lathiat> i wish vpnc worked at my uni
[08:31] <Lathiat> it seems to send the same packet the cisco client does to start with, but then nothing happens
[08:56] <infinity> s/sued/used/
[08:56] <infinity> -EWIN
[09:07] <lool> quitte: the configure script first checks for /etc/xml/catalog (which you can override) and then searches for particular catalogs with JH_CHECK_XML_CATALOG calls
[09:09] <lool> quitte: try xmlcatalog --noout /etc/xml/catalog '-//OASIS//DTD DocBook XML V4.1.2//EN'
[09:10] <lool> quitte: be sure to have xml-core, docbook-xml, and that docbook-xml's did register
[09:24] <pitti> Hi mvo
[09:27] <pitti> morning seb128
[09:28] <seb128> hey pitti
[09:29] <pitti> ogra: so mdz ack'ed the new hal upstream version?
[09:29] <seb128> pitti: have you read the bug I reassigned to you this night about this?
[09:29] <seb128> pitti: the battstat one
[09:30] <pitti> seb128: not yet, I'm still catching up with bug triage
[09:30] <seb128> pitti: http://bugzilla.ubuntu.com/show_bug.cgi?id=12909
[09:30] <seb128> pitti: current comment on it
[09:31] <pitti> cool, I wanted to prepare the update anyway
[09:31] <pitti> it also fixes various issues with udev
[09:32] <seb128> nice
[09:32] <seb128> pitti: and about dbus? going to update too?
[09:32] <pitti> seb128: it seems reasonably safe and some guys need it
[09:33] <pitti> daniels: do you know whether dbus 0.35 has a new API or ABI?
[09:34] <seb128> when we spoke about it the other day some guys said only the glib binding stuff changed API wise
[09:43] <Amaranth> seb128: Does gnome_vfs_monitor_add() use gamin to watch the uri?
[09:43] <seb128> elmo: glitz (incoming) intltool gdesklets gdesklets-data (the current gdesklets-data ships a pile of b0rked desklets due to API changes and a Debian guys has just cleaned all this)
[09:43] <seb128> Amaranth: yep
[09:43] <Amaranth> seb128: trying to find out why my menus don't update :)
[09:44] <Amaranth> ok, then something in gnome-vfs is broken
[09:44] <seb128> Amaranth: gamin is b0rked
[09:44] <seb128> I would bet on gamin
[09:44] <Treenaks> seb128: will it be fixed before breezy-final?
[09:44] <Amaranth> gamin debug log says it never got a request to watch the things libmenu says it's setting up monitors for
[09:44] <seb128> Treenaks: if you send a patch for sure
[09:44] <Amaranth> the only thing inbetween them is gnome-vfs
[09:44] <Treenaks> seb128: ;) of course.. 
[09:45] <seb128> Amaranth: good, debug is welcome too ... I don't really trust gamin and would bet on it anyway though
[09:45] <Amaranth> it'll have to wait until morning though, it's 3am and things are starting to blur :)
[09:45] <Amaranth> night all
[09:45] <seb128> good night
[09:45] <seb128> oh
[09:45] <seb128> and how is smeg going?
[09:46] <Amaranth> i lost all my work :/
[09:46] <seb128> still an option for 5.10 or you trashed all your changes with your hdd and we should better delay it?
[09:46] <Amaranth> well, if 0.7.5 is usable it's an option
[09:47] <Amaranth> otherwise i don't think i'll be finishing 0.8 until late august or so
[09:47] <seb128> you are the best placed to know what issues 0.7.5 has
[09:47] <Amaranth> well, as long as you have the fixes from pyxdg CVS 0.7.5 has no issues that i know of
[09:47] <seb128> Amaranth: k, good, thanks
[09:47] <Amaranth> of course it dies on badly encoded filenames but the spec doesn't define what to do with those and no one on the xdg list seems to want to decide
[09:48] <seb128> you had a patch for this, no?
[09:48] <seb128> can't you just ignore those?
[09:48] <Amaranth> the patch silently dropped chars that couldn't be represented in UTF-8
[09:48] <Amaranth> gnome-menus doesn't ignore them
[09:48] <Amaranth> it shows them then fails to find them to launch them
[09:48] <Amaranth> so no one really knows what to do with them
[09:48] <seb128> I would just ignore these files
[09:49] <seb128> better than crashing
[09:49] <Amaranth> i'll work on a patch for pyxdg tomorrow
[09:50] <seb128> cool, thanks
[09:50] <seb128> 'night
[09:56] <\sh> pitti: thx..the you ff from yesterday doesn't crash anymore with flash enabled
[09:57] <pitti> \sh: I didn't upload ffox yesterday
[09:58] <\sh> pitti: or the day before yesterday
[09:58] <\sh> the last upload 
[10:26] <\sh> doko: ping
[10:27] <doko> pong
[10:27] <\sh> doko: hplip...u included a build-dep on python-sip4-qt3? or doesn't it use python qt bindings anymore? :)
[10:31] <\sh> grmpf..who is setting the bugzilla entries to pending upload and is not assigning them to his account...
[10:36] <doko> \sh: no, not needed
[10:37] <seb128> doko: cairo 0.6 uploaded
[10:38] <doko> seb128: ok, will test ... glitz enabled?
[10:39] <seb128> doko: nop, I've asked a sync from incoming to elmo for that ... 
[10:39] <seb128> doko: I'll redo a build with glitz when elmo has synced it
[10:44] <Kamion> infinity: I guess you didn't spot my seed breakage comment from yesterday
[10:45] <Kamion> infinity: I've unbroken the archive the hard way, by copying patch-80 to a fresh directory that I own, fixing up the permissions, and killing your old patch-80 directory
[10:45] <Kamion> infinity: please move umask 002 to the top of ~/.bashrc, before [ -z "$PS1" ]  && return
[10:46] <fabbione> Kamion: how badly would you feel if i upload another kernel today?
[10:46] <Kamion> fabbione: oh well
[10:47] <fabbione> Kamion: i did all the "heavy lifting" job before VAC
[10:47] <fabbione> so that i left only a few bits for infinity 
[10:48] <fabbione> seb128: no i don't have it right now, but i can grab it for you in a minute or two
[10:48] <highvoltage> hi guys, everyone involved in edubuntu, this is a notice that there will be a meeting this afternoon in #edubuntu to discuss current status. time: 12:00 UTC/GMT.
[10:48] <seb128> fabbione: would be nice, thanks
[10:48] <fabbione> seb128: thanks to you!
[10:49] <fabbione> Kamion: btw.. did you ever get around to fix debootstrap for sparc or is it still untested?
[10:50] <Kamion> fabbione: still untested probably, I've entirely forgotten what bugs you reported in it, if any ...
[10:50] <Kamion> ogra: do you use PHP in Edubuntu?
[10:51] <ogra> Kamion, ha ha ha...
[10:51] <fabbione> Kamion: no, no bugs.. once you told me that debootstrap was going to be broken for sparc and i agreed on it.. i just don't recall us talking about it anymore.. if it needs testing i can do it :)
[10:51] <ogra> Kamion, i guess there is not much that more used in edubuntu
[10:51] <Kamion> ogra: would it break anything if I merged infinity's change to the Ubuntu seeds that switches from PHP4 to PHP5?
[10:51] <ogra> Kamion, yes
[10:51] <ogra> :(
[10:51] <ogra> many of the packages i use arent switched to 5 yet.... 
[10:52] <pitti> I thought it was supposed to be backwards compatible?
[10:52] <Kamion> fabbione: oh, that needs arch-specific overrides, I can't do that
[10:52] <Kamion> ogra: they should depend on php4 then
[10:52] <ogra> (moodle, mediawiki (which just gets packaged in debian, but they told me these pacages wont be php5 compatible))
[10:52] <fabbione> Kamion: so what do we need to do exactly? (other than giving it a shot)
[10:52] <pitti> Kamion: the problem is that we actually intended to demote php4 to universe once we have php5 in main
[10:52] <Kamion> ogra: if those depend on php4, it'll be fine
[10:52] <Kamion> pitti: not my problem
[10:53] <ogra> Kamion, but those will have to move to main
[10:53] <Kamion> fabbione: try it, report a bug with what breaks
[10:53] <fabbione> Kamion: make sense :)
[10:53] <ogra> Kamion, which will keep php4 there too
[10:53] <Kamion> ogra: I think the seed merge will be fine actually, as long as the dependencies are correct it won't break anything in itself
[10:53] <ogra> ok
[10:53] <Kamion> it's only changing the supported seed
[10:54] <ogra> fine then :)
[10:54] <Kamion> I need to merge the installer and casper seeds which is a later patch to ubuntu-devel@lists.ubuntu.com/seeds--breezy--0, which is why I ask :)
[10:54] <Kamion> ok, done
[10:55] <fabbione> seb128: never mind.. it's still the binutils problems that reflects into the gtk test of lp-integration...
[10:56] <fabbione> seb128: nothing you can do about it.. thanks tho.
[10:56] <Kamion> ok, this morning's daily CD build will probably be fucked, but I'm rebuilding it now
[10:56] <Kamion> "can't find kernel modules" style of breakage
[10:56] <seb128> fabbione: k
[10:56] <fabbione> Kamion: uh.. udebs are there :)
[10:56] <Kamion> fabbione: seeds weren't up to date, I've fixed them
[10:57] <fabbione> oh ok...
[10:57] <Kamion> the seeds matter, otherwise cdimage will leave all the kernel module udebs off the CD :-P
[10:57] <fabbione> didn't we want more space on CD anyway? :P
[10:57] <Kamion> heh
[11:24] <seb128> JaneW: YOUR MAIL TITLE LOOKS LIKE SPAM WITH ALL THE CAPS :p
[11:25] <fabbione> pitti: ping?
[11:26] <pitti> I'm right here
[11:26] <fabbione> pitti: before suggesting users to file bugs because /dev/inotify is not there.. do you mind to ask me..?
[11:26] <fabbione> it's normal
[11:26] <fabbione> the device is not required anymore
[11:26] <JaneW> seb128: It's a notice :P
[11:26] <fabbione> now i am getting hammered by people with "OMG THE DEVICE IS NOT THERE ANYMORE! IT MUST BE BROKEN"
[11:27] <fabbione> when instead it works perfectly on i386/amd64
[11:27] <seb128> JaneW: yeah, but you probably want to people to read it and not to trash it with the spam? :)
[11:27] <fabbione> ia64 and ppc...
[11:27] <JaneW> seb128: I use caps at times, when I am trying to be LOUD
[11:27] <fabbione> meh s/ppc/sparc
[11:27] <pitti> fabbione: sorry, there was no other reply on the udevel list, so I just took seb128's approach of "use bz for bugs"
[11:27] <JaneW> seb128: sorry, won't do that again then, I guess I aussmed people could read...
[11:27] <jsgotangco> SURE WE CAN :)
[11:28] <Treenaks> JaneW: at least the subject wasn't "Congratulations" :)
[11:28] <fabbione> pitti: ok.. don't worry.. just next time let's avoid panic ;)
[11:28] <pitti> sorry, dude! :-)
[11:28] <pitti> ok
[11:29] <seb128> JaneW: no need to be sorry, but I'm used to categorize CAPS SUBJECT == spam and almost trashed this one ... so I just pointed it may have this effect for some people :)
[11:30] <JaneW> Treenaks: no it;s ok, I am keeping all my lotery winning to myself, I have won at least 20 times now, I'm extremely lucky...
[11:30] <jsgotangco> Treenaks, lol
[11:30] <JaneW> seb128: point taken, thanks
[11:30] <seb128> np :)
[11:31] <seb128> pitti: thanks for pointing bugzilla to users abusing the list for bugs ;)
[11:35] <pitti> fabbione: out of interest, what is now used instead of /dev/inotify?
[11:36] <pitti> this worked with ioctl()s formerly, right?
[11:36] <Treenaks> pitti: syscalls
[11:36] <fabbione> pitti: as Treenaks said.. syscalls
[11:36] <pitti> ah, ok
[11:37] <fabbione> pitti: same reason why we have coverage for 4 out of 6 arches for now
[11:37] <fabbione> pitti: ppc and hppa don't have syscalls assigned yet
[11:40] <sivang> Morning all
[11:41] <sivang> he JaneW 
[11:41] <mvo> am I blind? or is there no search in malone?
[11:43] <Kamion> fabbione: at least powerpc is going to be fixed before breezy, I hope?
[11:43] <JaneW> hi sivang
[11:44] <fabbione> Kamion: yes yes.. they are fixing them on a daily base
[11:44] <fabbione> Kamion: but they are asking the differnet $arch maintainers to assign the proper numbers...
[11:44] <fabbione> Kamion:   Package lib64gcc1 is not installed.
[11:44] <fabbione> Kamion: that's needed for sparc bootstrapping..
[11:44] <Kamion> fabbione: right, that's part of the arch-specific overrides problem
[11:45] <fabbione> Kamion: ok.. do i need to ask elmo to do something?
[11:45] <Kamion> because powerpc also has lib64gcc1, but it isn't needed in base there
[11:45] <fabbione> nope.. but we have nothing in 64bit userland for ppc :)
[11:45] <Kamion> fabbione: technically yes, but it's "implement a bunch of hard stuff in katie", it's not "flip this switch"
[11:46] <sivang> Kamion: does that mean that bootstrapping ppc always occurs in 32bit mode ?
[11:46] <Kamion> I can just bump lib64gcc1's priority and accept the bloat on powerpc I guess
[11:46] <Kamion> sivang: of course
[11:46] <fabbione> Kamion: how much will ppc bloat?
[11:46] <Kamion> debootstrap has no concept of "mode"
[11:47] <Kamion> it's not a lot, 100KB or so
[11:47] <Kamion> seems like a DOIT I guess
[11:47] <fabbione> Kamion: well we are like talking 100KB on ppc or an entire arch out of the game
[11:47] <fabbione> 100KB are doable...
[11:47] <Kamion> yeah, one second while I run germinate for powerpc
[11:47] <fabbione> thanks
[11:48] <sivang> fabbione: do you recall I asked you about the bu that  rpevented booting on pSeries, where can I read about what it was and how it got fixed?
[11:48] <sivang> fabbione: (botting from install image, that is)
[11:51] <fabbione> bu ? 
[11:51] <sivang> fabbione: s/botting/booting/
[11:51] <Kamion> s/bu/bug/
[11:51] <sivang> heh
[11:51] <sivang> Kamion: thx
[11:52] <fabbione> sivang: the installer problem we had was missing drivers in the .udeb
[11:52] <sivang> fabbione: a couple of weeks ago I tried booting the install image under one of my LPARs, and was unable to get into the grub prompt
[11:52] <fabbione> if there are other bugs i don't know
[11:52] <fabbione> there is no grub on ppc...
[11:52] <sivang> fabbione: ah ok, do you have it in debian bts or something?
[11:52] <Kamion> fabbione: actually, I don't get it, lib64gcc1 doesn't show up in jessica output
[11:52] <sivang> fabbione: sorry, yaboot :)
[11:52] <Kamion> fabbione: can I demote libc6-sparc64 to priority optional? what breaks if I do that?
[11:52] <sivang> fabbione: still need to adjust to the arch specifics 
[11:53] <fabbione> sivang: the bug was ubuntu specific.. we readded the drivers to the udeb.. that's it
[11:53] <sivang> fabbione: ah cool, which drivers were those?
[11:53] <fabbione> Kamion: one second...
[11:53] <fabbione> sivang: almost all of them.
[11:53] <fabbione> Kamion:  libc6-sparc64 depends on lib64gcc1; however:
[11:54] <fabbione> Kamion: i think there is quite a bunch of stuff that Depends: libc6-sparc64
[11:54] <Kamion> fabbione: right, libc6-sparc64 is the only thing that needs it, and nothing else in base Depends: libc6-sparc64
[11:54] <Kamion> fabbione: nope, hardly anything
[11:54] <Kamion> fabbione: does anything need it but not state a Depends?
[11:54] <fabbione> Kamion: i am checking...
[11:54] <fabbione>   libdlm0
[11:55] <fabbione> and we don't care
[11:55] <fabbione> ok.. demote it my friend
[12:01] <fabbione> and see what else breaks :)
[12:01] <Kamion> $ sudo -u katie ./alicia.new libc6-sparc64 optional
[12:01] <Kamion> I: Will change priority from important to optional
[12:01] <Kamion> Continue (y/N)? y
[12:01] <Kamion> Done
[12:01] <fabbione> for desktop that stuff is ok to stay out.. but they are almost required for the buildd...
[12:01] <fabbione> but let's solve one problem at a time
[12:01] <Kamion> needs to be added to the explicit build-essential list, I'm fixing that now
[12:01] <Kamion> why does the buildd need it?
[12:01] <fabbione> Kamion: i don
[12:02] <fabbione> Kamion: i don't recall the details..
[12:02] <fabbione> i will check it again later..
[12:02] <fabbione> i still need to build vim before i can bootstrap properly
[12:02] <fabbione> and i will test buildd later..
[12:02] <fabbione> sparc is anyway a bit of a special case...
[12:02] <fabbione> Kamion: we just got unionfs to build on ppc..
[12:03] <fabbione> i am definetely going to upload today
[12:08] <Kamion> fabbione: ok
[12:09] <fabbione> JaneW: i need a name please...
[12:10] <JaneW> fabbione:  Scintilating Sesame
[12:11] <JaneW> :)
[12:11] <Lathiat> mdz: ping
[12:13] <Mez> yes?
[12:13] <Mez> oh 
[12:13] <Mez> sorry
[12:13] <Mez> :P
[12:13] <Mez> read that as mez
[12:14] <Lathiat> heh
[12:25] <pitti> seb128: here?
[12:25] <seb128> pitti: pong
[12:26] <cartman> 50 nicely packaged Hoary Cds
[12:26] <cartman> found its way to my home
[12:26] <cartman> thank you guys! :)
[12:26] <pitti> seb128: wrt #11486 (gstreamer sounds bad with alsasink), can we try to apply the patch in http://bugzilla.gnome.org/show_bug.cgi?id=305186?
[12:28] <pitti> seb128: oh, it's already applied, sorry
[12:29] <seb128> pitti: we have 0.8.10
[12:30] <seb128> pitti: yeah
[12:30] <pitti> darn
[12:32] <JaneW> Hi all, could anybody who has not updated their BreezyGoals in the last week, please do so.
[12:33] <JaneW> we need a clear indication of where each goal is, and whether it is going to be finished in time for Breezy
[12:33] <JaneW> Thanks
[12:35] <JaneW> I will start contacting anyone who has no update in their goal in the next few hours
[12:35] <bob2> bear in mind it's well after work hours for .auers
[12:36] <tseng> yay for firefox crashing
[12:38] <tseng> JaneW: ok, i cant update anything because firefox wont stay open for 12 seconds
[12:38] <daniels> pitti: yeah, API and ABI both chagngedc, but its' reasonably trivial; we can easily 
[12:38] <daniels> pitti: patch the apps using it.  it's a better long-term solution.  i'll knock up a new set of packages on monday for you.
[12:39] <pitti> daniels: cool; well, it's not really "for me", but for some other spec a guy works on
[12:39] <Mez> hmm
[12:39] <Mez> anything that's very likely to break if I install breezy ...?
[12:39] <Mez> and upgrade it to current)
[12:40] <tseng> plenty
[12:40] <Mez> like .. ?
[12:40] <Lathiat> Mez: X amon g other things
[12:40] <tseng> like X not starting
[12:40] <Mez> (I'm just burning a colony 2 cd)
[12:40] <Mez> I thought someoen said the other day they got X working?
[12:40] <tseng> BWAR
[12:40] <daniels> pitti: yeah, we need 0.35
[12:40] <tseng> firefox must die
[12:41] <Mez> well, I shall see eh?
[12:41] <Mez> It cant do much harm as long as I make sure I do the boot partition righ
[12:41] <Lathiat> http://blogs.gnome.org/callum
[12:41] <Lathiat> see whats linked from there
[12:42] <Mez> Error: No such notebook: callum
[12:42] <tseng> one l
[12:42] <Mez> http://blogs.gnome.org/calum
[12:42] <Lathiat> my bad
[12:43] <tseng> JaneW: can i please email you a status update
[12:43] <tseng> JaneW: i have tried to login at least 12 times now, firefox refuses to do anything but crash
[12:44] <JaneW> tseng: absolutely, I can do the wiki part.
[12:44] <JaneW> tseng: thanks
[12:44] <JaneW> that applies to everyone
[12:46] <Kamion> so that I can think about them without evicting useful state from my brain
[12:47] <pitti> tseng: I always use ffox for editing the page, how on earth do you manage to crash it so often?
[12:47] <pitti> tseng: do you have some weird extensions installed?
[12:47] <tseng> pitti: its segfaulting
[12:47] <tseng> pitti: all over the udu wiki
[12:47] <pitti> tseng: besides, mozilla should work :-)
[12:48] <Lathiat> bah firefox is crashing isnt it
[12:48] <Lathiat> i only just noticed now
[12:48] <tseng> you too?
[12:48] <pitti> tseng: yes, but it doesn't happen without extensions
[12:48] <tseng> pitti: hm well
[12:48] <Lathiat> not on the udu wiki 
[12:48] <Lathiat> just in general
[12:48] <Lathiat> and i have no extensions
[12:48] <JaneW> Kamion: thanks, but mdz is pushing quite hard for updates on all goals now, so I can;t promise to leave you alone until it's done :)
[12:48] <tseng> pitti: my extensions window has "English (GB) Pack"
[12:48] <Kamion> JaneW: sure, just saying that's all I can do for now
[12:48] <pitti> hm, I even have some more
[12:48] <Lathiat> tseng: hm, so does mine
[12:48] <tseng> pitti: i have one theme installed (and using)
[12:49] <Lathiat> tseng: the version on the language pack is 1.0..4
[12:49] <JaneW> pitti: yes I had that problem until I uninstalled all flash extensions
[12:49] <tseng> ohh
[12:49] <Lathiat> could be why
[12:49] <tseng> i do have flash
[12:49] <pitti> JaneW: I do have flash installed, works fine
[12:49] <JaneW> I had to chose between flash and being able to work, was a tough choice ;)
[12:49] <JaneW> pitti: yeah but you have breezy not hoary...
[12:49] <sivang> fabbione, Kamion : well, thinking about that you uncovered a missing udeb problem, I think that's not the actual cause of what I was experiencing unde the LPAR, since it never even got to the yaboot screen, so that's probably have to do something with the image layout , or something else more low level
[12:50] <pitti> JaneW: it still makes a difference? it's the same upstream version now
[12:50] <davyd> does anyone know about Breezy/AMD64 something in GTK+/GLib being broken?
[12:50] <fabbione> sivang: can you please file a bug with details?
[12:50] <pitti> GTK BOOG! :-)
[12:50] <fabbione> sivang: i don't even understand what you mean by not arriving at yaboot.. did the installer actually worked? or it didn't do anything?
[12:50] <Kamion> sivang: the missing udeb thing was just this morning's CDs
[12:50] <Kamion> and not getting to yaboot has zero to do with udebs
[12:50] <fabbione> Kamion: meh well not really...
[12:51] <fabbione> Kamion: the i told sivang about the missing drivers inside the udebs
[12:51] <sivang> fabbione: yes, when I booted the LPAR, I just got straight away a msg from the firmware "no operating system installed" 
[12:51] <sivang> Kamion: right
[12:51] <sivang> Kamion: that's why my conclusion
[12:51] <Kamion> fabbione: that can't possibly be sivang's problem either, his machine almost certainly just doesn't support yaboot
[12:51] <davyd> pitti: what sort of bug?
[12:51] <fabbione> sivang: before or after the installer?????
[12:51] <sivang> Kamion: no it does, SLES9 and RHEL3 do boot
[12:51] <sivang> fabbione: before the installer, sorry
[12:52] <pitti> davyd: never mind, that was just a reflex to poke seb :-)
[12:52] <sivang> Kamion: that is , each LPAR is a box in its own regards, at teh hardware level. 
[12:52] <Kamion> sivang: CD boot or netboot?
[12:52] <sivang> Kamion: cd boot
[12:52] <davyd> this is a little distressing
[12:53] <fabbione> sivang: try a netboot...
[12:53] <pitti> davyd: sorry
[12:53] <davyd> nah, it's ok
[12:53] <davyd> I was just trying to finish some GNOME screenshots
[12:53] <pitti> davyd: it's a long story
[12:53] <davyd> so I thought I should grab the latest versions of some things
[12:53] <seb128> davyd: what is the issue?
[12:53] <davyd> seb128: try to start a GTK+ app and it segvs
[12:53] <tseng> JaneW: sent.
[12:53] <Kamion> sivang: CD booting can be finicky, I may well be missing stuff in cdimage
[12:54] <seb128> davyd: utch
[12:54] <seb128> davyd: backtrace ?
[12:54] <davyd> seb128: I'm seeing what I can do in that department
[12:54] <Kamion> sivang: unfortunately I don't have time to investigate at the moment, but if you can find out what mkisofs flags are used for the SLES9 or RHEL3 CDs then that would be useful
[12:54] <seb128> davyd: what/when have you updated?
[12:54] <sivang> Kamion: cool, that's one lead I could use :)
[12:54] <davyd> seb128: about 5 minutes ago
[12:54] <davyd> on amd64
[12:55] <seb128> davyd: cairo 0.6 ?
[12:55] <davyd> gconf is running, so I must have glib
[12:55] <davyd> seb128: seems that way
[12:55] <seb128> pango 1.9.1 ?
[12:55] <davyd> 1.9.0
[12:55] <sivang> Kamion: I'm already in contact with a couple of fedora guys so it won't be hard to find that out - futhermore, they have also told me that they use anaconda to create those boot disks, and I might find more stuff to "borrow" from there, does that make sense? (I wasn't aware anaconda produces install disks)
[12:56] <davyd> hmm, that sounds suss
[12:56] <davyd> like the buildds are behind
[12:56] <Lathiat> or it ftbfs
[12:56] <seb128> davyd: bah, that's random guess ...  a backtrace would be nice
[12:56] <seb128> davyd: updating just cairo works fine here on x86
[12:56] <davyd> seb128: gdb doesn't like something...
[12:57] <seb128> maybe try downgrading cairo to 0.5.2?
[12:57] <Kamion> sivang: ah, I see the relevant anaconda code, yes
[12:57] <davyd> cairo_scaled_font_create
[12:57] <seb128> gar
[12:57] <seb128> what about it?
[12:57] <sivang> Kamion: nice, I'm very interested in this, should I just get anaconda's source to read and find out?
[12:57] <davyd> that's where the segv is occurring, hang on, I'll need to install symbols to get more
[12:58] <Kamion> sivang: hmm, they invoke mkisofs in a totally different way from the way we do, though
[12:58] <seb128> davyd: would be nice to try with pango 1.9.1 when your mirror get it
[12:58] <Kamion> sivang: sure, 'cvs -d :pserver:anonymous@rhlinux.redhat.com:/usr/local/CVS co anaconda', it's in scripts/mk-images.ppc
[12:58] <davyd> seb128: I'm using the main mirror
[12:58] <sivang> Kamion: thanks, I will fetch that now
[12:59] <seb128> davyd: http://archive.ubuntu.com/ubuntu/pool/main/p/pango1.0/libpango1.0-0_1.9.1-0ubuntu1_amd64.deb
[12:59] <seb128> davyd: it has it
[12:59] <Kamion> meh, they build pseries ISOs in a totally different way to mac ones
[12:59] <davyd> seb128: seems like the Packages file isn't updated yet
[01:00] <JaneW> tseng: many thanks
[01:00] <Kamion> I don't want to have to build two ISO images ...
[01:00] <sivang> Kamion: I had a feeling :)
[01:00] <seb128> davyd: right, you can get this one and the -common and dpkg -i or wait some minutes for the update
[01:00] <sivang> Kamion: why don't I try to build that image like they do, see if it works at all and then will think about it?
[01:00] <Kamion> sivang: you probably won't be able to build it literally like they do
[01:00] <davyd> seb128: downloading now
[01:00] <sivang> Kamion: ah, how so?
[01:00] <JaneW> tseng: I haven;t receievd it yet... ps did you ever get e-mail from me? I mist have tried about 7 addresses for you now...
[01:01] <tseng> JaneW: i replied to that too
[01:01] <tseng> uhm
[01:01] <tseng> JaneW: Unfrgiven (Ankur Kotwal) has owned that for months, ive asked him to update it
[01:01] <tseng> write(2, "firefox-bin: cairo.c:86: _cairo_"..., 130firefox-bin: cairo.c:86: _cairo_error: Assertion `status > CAIRO_STATUS_SUCCESS && status <= CAIRO_STATUS_FILE_NOT_FOUND' failed.
[01:01] <Kamion> sivang: because anaconda != d-i :)
[01:01] <tseng> ) = 130
[01:02] <tseng> does this look horribly bad seb128 ?
[01:02] <sivang> Kamion: ah :)
[01:02] <Kamion> sivang: you should be able to try cherry-picking mkisofs flags though
[01:02] <JaneW> tseng: oh that's right... sorry now I recall
[01:03] <Kamion> sivang: our script is in arch: baz register-archive http://people.ubuntu.com/~cjwatson/archives/colin.watson@canonical.com--2005; baz get colin.watson@canonical.com--2005/debian-cd--ubuntu--0 debian-cd; look in debian-cd/tools/boot/breezy/boot-powerpc
[01:03] <seb128> tseng: does every gtk app crash? what arch are you using? what did you update?
[01:03] <tseng> seb128: i updated everything an hour ago
[01:03] <davyd> haha!
[01:03] <Kamion> sivang: could be worth trying the -J and/or -T flags
[01:03] <tseng> oh there goes Evo
[01:03] <seb128> tseng: and other questions?
[01:03] <Kamion> as a first guess
[01:03] <tseng> seb128: x86, and yes
[01:03] <sivang> Kamion: Ok, I will start experimenting and tell you if anything works
[01:04] <seb128> tseng: every gtk app crash?
[01:04] <tseng> moz and evo so far
[01:04] <tseng> s/moz/ff
[01:04] <seb128> start gedit?
[01:04] <tseng> open file in gedit = crash
[01:04] <Kamion> it may well just need the ISO to have Joliet extensions
[01:04] <seb128> that's a fileselector crash
[01:04] <seb128> it's known and not new
[01:05] <seb128> ok, so that's not every app :p
[01:05] <sivang> seb128: crashes for me as well
[01:05] <davyd> hmm, seems there are some XKB issues
[01:05] <sivang> seb128: my gedit patch is not yet in right?
[01:05] <seb128> sivang: not, it's not correct
[01:05] <tseng> seb128: muine isnt crashing either
[01:05] <tseng> (yet)
[01:05] <seb128> tseng: getting a debug backtrace would be nice
[01:05] <sivang> seb128: I will fix it (I already know what's wrong) and ping you when I have refreshed pkgs
[01:06] <seb128> k
[01:07] <tseng> seb128: does that mean i have to build all this crap with symbols?
[01:08] <tseng> hm the cairo stuff resolves
[01:08] <seb128> tseng: all this crap being libcairo which is like a 600k package
[01:08] <seb128> libpango1.0-dbg 
[01:08] <tseng> i have libcairo symbols somehow
[01:08] <seb128> and use LD_LIBRARY_PATH
[01:08] <tseng> set to?
[01:08] <seb128> /usr/lib/debug
[01:11] <tseng> seb128: tseng.ath.cx/gdb-backtrace is a start
[01:11] <tseng> but it doesnt resolve everything stiff
[01:11] <tseng> still*
[01:12] <tseng> or did i need debug symbols when i did the coredump too
[01:12] <sivang> Kamion: I'm not sure I understand the cherry-picking mode, what does it entails?
[01:13] <Kamion> sivang: do you know the term "cherry-picking"?
[01:13] <Kamion> sivang: it means to select and try things that look good, one by one
[01:13] <seb128> tseng, davyd: I'm updating gtk, maybe that will fix it 
[01:13] <tseng> seb128: k
[01:13] <sivang> Kamion: ah ok, I wasn't aware of that nice english term :)
[01:14] <tseng> oh blah theres no function names there
[01:14] <seb128> tseng: just wait an hour for the gtk update
[01:15] <tseng> ok.
[01:15] <seb128> downgrade cairo to 0.5.2 for the moment if you want
[01:15] <tseng> thanks seb
[01:15] <seb128> np
[01:17] <sivang> Kamion: can I produce the test images under i386? or must I produce them under the same arch on which I intend to use them?
[01:17] <doko> seb128: anything wrong with libcairo?
[01:17] <seb128> doko: the new cairo/pango seems to have issues
[01:18] <doko> ok, libgcj doesn't use pango ...
[01:19] <tseng> JaneW: how about now
[01:21] <sivang> seb128: (just to be syncd) what issues did you spot with the gedit patch?
[01:21] <seb128> Build-Depends not updated
[01:21] <sivang> seb128: ah right :-/
[01:24] <JaneW> tseng: got it
[01:24] <tseng> JaneW: cheers.
[01:26] <Kamion> sivang: i386 is fine, we don't have a separate cdimage machine for each architecture :)
[01:27] <pef> hello
[01:27] <ogra> Kamion, how hard would it be to make the network setup in the installer default to manual in the edubuntu installer, assuming i install a standalone ltsp server there is no dhcp around
[01:28] <sivang> Kamion: cool
[01:31] <pitti> trulux: ping
[01:34] <Kamion> ogra: it's one preseed, netcfg/disable_dhcp=true, which is documented in the installer boot screens, press F7 at the isolinux prompt
[01:34] <Kamion> ogra: want that switched on for Edubuntu then?
[01:35] <ogra> Kamion, yep, i assume having the machine standalone is the more common case... and i'd like to avoid seeing a errormessage
[01:35] <Kamion> ogra: surely this is only for the server?
[01:35] <Kamion> not clients?
[01:35] <ogra> Kamion, the ltsp server is the default...
[01:36] <Kamion> ogra: yes, and debian-cd has a non-server config
[01:36] <ogra> not for the standalone workstation install though
[01:36] <ogra> oh, that needs to change then too
[01:36] <Kamion> why? the workstations should default to DHCP I'd've thought
[01:36] <ogra> but in other news edubuntu daily from yesterday seems to install fine :)
[01:37] <ogra> Kamion, the Ws should use dhcp, the server not... but the server should be the default
[01:38] <ogra> currently the WS install is only thought for teachers to run the setuo at home...
[01:38] <ogra> setup
[01:39] <ogra> also the hostname should default to edubuntu, but thats marginal
[01:39] <Kamion> ogra: right - I've changed the server config to default to manual network configuration; please file a Bugzilla bug against debian-cd to get the Edubuntu default changed to server
[01:39] <Kamion> ogra: for server or workstation?
[01:40] <ogra> Kamion, for both ? 
[01:40] <Kamion> ok
[01:40] <ogra> if i install in my network i'll change it anyway... but the default shouldnt be ubuntu...
[01:41] <Kamion> ogra: done
[01:41] <ogra> yay, thanks...
[01:42] <ogra> from a installer perspective the CD seems to be in a awesome condition... (i'm not at X yet ;) ) kudos Kamion  :)
[01:42] <Kamion> glad that works at least
[01:42] <pitti> seb128: I can't reach neither daf nor carlos :-/
[01:43] <seb128> pitti: maybe they are travelling to go to brazil
[01:43] <Mez> mako: ping
[01:43] <seb128> pitti: bah, a few days will not matter, update next week ... 
[01:43] <mako> Mez: yes
[01:44] <Mez> With Regards to your email: what is there left pending?
[01:45] <mako> Mez: umm.. remind me which email i'm talkinb about.. i've sent like 150 today already :)
[01:46] <Mez> Backports Team formation
[01:46] <pitti> ah, main network back, brb
[01:46] <Mez> I asked whether anything was pending, and if so what it was and if I should add it to CC agenda
[01:46] <Mez> you replied "yes"
[01:46] <Mez> lol
[01:46] <Mez> (near enough)
[01:46] <mako> yes, add it to the agenda
[01:46] <mako> no, i think that's it
[01:47] <Mez> what is still pending
[01:47] <mako> i've talked to matt briefly about it.. if things are rolling, we'll just stamp it at the next meetting
[01:47] <mako> sorry that it was unclear.. i'm sort of bulldozing through this mail right now :)
[01:47] <Mez> so, what should I add ?
[01:47] <mako> creation of backports team
[01:48] <Mez> kk
[02:01] <infinity> Kamion : Shit, dude, sorry about the umask.
[02:02] <Kamion> infinity: no problem, fixed now :)
[02:02] <daniels> infinity: maybe you should read up on permissions.
[02:02] <Kamion> haha
[02:02] <infinity> daniels : Cock.
[02:03] <fabbione> infinity: yo....
[02:04] <fabbione> infinity: eheh up to you :)
[02:04] <fabbione> i had good news ;)
[02:04] <fabbione> i guess i will keep them for me
[02:05] <fabbione> infinity: let's move to u-k
[02:05] <infinity> ?
[02:05] <tseng> #ubuntu-kernel
[02:06] <infinity> Right , somehow I thought he meant "let's move to the UK".
[02:06] <fabbione> dude.. i like beer.. even after 11
[02:06] <pitti> seb128: so the polypaudio sink is broken? since we will keep esd in breezy, would you mind if I unseed polypaudio?
[02:08] <daniels> i thought polypaudio was the way of the future
[02:08] <Kamion> fabbione: my fiancee asks "if you like beer, what are you doing living in Denmark?"
[02:08] <daniels> fabbione: beer is good, but infinity doesn't like beer.  only alcoholic lemonade.
[02:09] <daniels> Kamion: i think canonical should pay me relocation to belgoiubgleium.
[02:09] <fabbione> Kamion: well denmark still produces beer.. i didn't say GOOD.. but well. i am only 6 km from Tuborg :P
[02:09] <daniels> the second attempt was even worse than the first.
[02:09] <fabbione> daniels: yeah yeah
[02:09] <daniels> fabbione: the julebryg isn't bad.
[02:10] <daniels> not as good as, say, a bottle of chimay blue, though ...
[02:10] <Kamion> fabbione: maybe you should move to Scotland, it's about as cold as Denmark, they don't close at 11, and you can actually buy decent English beer ...
[02:11] <jdub> i'm sure they also serve perfectly respectable scottish beer too
[02:11] <daniels> Kamion: what, the Scottish climate cools the beer down? :)
[02:11] <jdub> royalist!
[02:11] <Kamion> the Scottish beer's OK but not as good
[02:11] <Kamion> daniels: thpppppt
[02:11] <Kamion> anyway, work ;)
[02:12] <infinity> Work, a 10pm?
[02:12] <infinity> s/a/at/
[02:13] <seb128> pitti: not at all
[02:13] <seb128> infinity: can you give a retry to epiphany's build
[02:13] <daniels> infinity: next he'll try to tell us that UTC is the one true timezone.
[02:13] <seb128> jdub: hey :)
[02:13] <jdub> yo seb128!
[02:13] <seb128> pitti: so we stay on esd?
[02:14] <pitti> yes, polypaudio is still too buggy (some crashes and lack of sound events), ENOTIME to work on it, and we are too close to the freeze
[02:14] <tseng> jdub: oh, other beagle issue
[02:14] <pitti> seb128: however, Erik said that upstream is back :-)
[02:14] <pitti> seb128: so maybe there's again hope for breezy+1 :-)
[02:14] <seb128> pitti: nice, for 6.04 we will use it maybe :)
[02:14] <seb128> yeah
[02:14] <tseng> jdub: 0.12 still expects /dev/inotify to do things normally
[02:15] <tseng> jdub: we need cvs/next release for the syscall crap
[02:15] <jdub> ahr
[02:15] <daniels> seb128: maybe we could use howl for 6.04 also
[02:15] <Kamion> infinity: no, I mean me ...
[02:15] <tseng> i think ill test that now, in fact
[02:15] <jdub> daniels: has howl finally killed off the APSL bits?
[02:16] <seb128> daniels: you better start to run now
[02:16] <daniels> jdub: don't think so, I just like giving you shit.
[02:16] <daniels> seb128: HowlRoadMap, SebastienBacherLead
[02:16] <jdub> it's on its way, don't know if it's going to be in a sane timeframe
[02:16] <jdub> i should bug Lathiat about avahi
[02:17] <seb128> jdub: somebody sent a patch for gnome-vfs bonjour support yesteday (#311882)
[02:17] <jdub> aha, and bonjour is now bsd, right?
[02:18] <daniels> only seb128 should be allowed to say bonjour.
[02:20] <jdub> heh, it's in a tarballs/apsl/ directory
[02:20] <jdub> seb128: close the bug and tell them the apsl is evil ;)
[02:20] <seb128> bah
[02:21] <seb128> all that sucks
[02:22] <jdub> yeah
[02:51] <pitti> alsa-utils changes changelog, debian/rules:
[02:51] <pitti> +       # do not install alsaconf for Ubuntu
[02:51] <pitti> +       rm -f debian/alsa-utils/usr/{sbin/alsaconf,share/man/man8/alsaconf.8.gz,share/locale/ja/LC_MESSAGES/alsaconf.mo}
[02:51] <pitti> and control:
[02:51] <pitti> -Depends: ${misc:Depends}, ${shlibs:Depends}, whiptail | dialog, modutils (>= 2.3.5-1) | module-init-tools, pciutils (>= 1:2.1.11-4), python, linux-sound-base (>= 1.0.9b-1)
[02:51] <pitti> +Depends: ${misc:Depends}, ${shlibs:Depends}, whiptail | dialog, modutils (>= 2.3.5-1) | module-init-tools, pciutils (>= 1:2.1.11-4), python, linux-sound-base (>= 1.0.9b-1), lsb-base (>= 1.3-9ubuntu2)
[02:51] <pitti>  Recommends: alsa-base (>= 1.0.9b-3)
[02:51] <pitti> -Suggests: udev (>= 0.063), lsb-base
[02:51] <pitti> +Suggests: udev (>= 0.063)
[02:52] <pitti> really cool now :)
[02:52] <pitti> I still remember the pain when merging a previous version
[02:52] <jordi> pitti: hmm, the rm is a bashism. :)
[02:52] <pitti> argh, sorry, ECHAN
[02:52] <pitti> sorry for the flood
[02:53] <tseng> infinity: blargh, tomboy fails on sbuild now because of unseeded stuff now also
[02:53] <tseng> infinity: is that on its way now?
[02:59] <aigarius> fabbione, ping
[03:02] <fabbione> aigarius: pong?
[03:03] <aigarius> fabbione, I am thinking about when to make full and when to make incremental backups
[03:03] <fabbione> aigarius: i am officially in holidays since 3 minutes :)
[03:03] <aigarius> ok
[03:03] <fabbione> aigarius: but ok :)
[03:04] <fabbione> hmm
[03:04] <fabbione> how much does it on average an incremental backup?
[03:04] <aigarius> 5-10% of a full backup
[03:05] <aigarius> I am thinking that I could just define a max_days_between_full_backups config option
[03:05] <fabbione> i would say an incremental to run at night via crontab or on request
[03:06] <fabbione> and a weekly full as default
[03:06] <aigarius> and do incremental backups withing that period
[03:06] <fabbione> but yeah.. configurable it's even better
[03:07] <aigarius> I want this behaviour to be consistent regardless of use of automatic/manual backup runs
[03:08] <aigarius> I have got full backups running both locally and trough ssh
[03:08] <aigarius> now working on incremental backups :)
[03:08] <fabbione> aigarius: yes, but you did ask me how often... right?
[03:09] <fabbione> so my suggestion is daily incremental and weekly full
[03:09] <fabbione> but if you can make it configurable it's even better...
[03:09] <aigarius> yes, right. good. happy holidays then. when do you plan to come back?
[03:10] <fabbione> as i just wrote in the mail
[03:10] <fabbione> aigarius: thanks :)
[03:10] <aigarius> ah, ok
[03:11] <pvanhoof> will Ubuntu breezy be xen enabled in it's kernel?
[03:11] <pvanhoof> and/or is it already? (the current kernel package)
[03:11] <Treenaks> xen: the hype of te moment
[03:13] <pvanhoof> Treenaks, but will it be supported by the breezy kernel? :)
[03:13] <carstenh> aigarius: i guess you don't backup things like /bin and /usr, because they can be restored from the packages that are installed?
[03:14] <aigarius> carstenh, yes I don't do that by default
[03:15] <carstenh> aigarius: how do you ensure that the same package-version are installed after restoring a full backup?
[03:15] <aigarius> carstenh, you can check out my progress at http://koyanet.lv/soc/  . It is a bzr repository, but you can also get the latest version with you browser.
[03:16] <aigarius> carstenh, I backup "dpkg --get-selections"
[03:16] <carstenh> aigarius: i think this is nessessary, because new package-versions may have new configuration-files
[03:16] <carstenh> aigarius: this does von include package-versions
[03:16] <hughsie> ogra: can i ask a newbie question
[03:17] <ogra> hughsie, hehe, sure :) these are my favorites ;)
[03:17] <carstenh> aigarius: i guess you have well thought about the backup :) just my 5 cent ;)
[03:17] <aigarius> carstenh, a full restore is kind of a special case scenario. the usual use is backup/restore of a single file or a subdir.
[03:17] <carstenh> aigarius: ah, ok
[03:18] <hughsie> ogra:I've installed 5.04, and want to upgrade to latest and greatest so I can try our the new dbus and hal and libnotify stuff for g-p-m
[03:18] <hughsie> how ould i do this the easiest?
[03:18] <aigarius> carstenh, it would be pretty hard to restore to a specific version if that is not the latest in the repository
[03:19] <carstenh> aigarius: yes, ubuntu is iirc missing something like snbapshot.d.n
[03:19] <ogra> hughsie, edit all occurences of hoary to breezy in the file /etc/apt/sources.list...
[03:19] <hughsie> is that it? then do an apt upgrade?
[03:19] <aigarius> hughsie, X is quite broken in breezy, so better don't :)
[03:19] <ogra> hughsie, then run "sudo apt-get update" to get the new package lists
[03:20] <ogra> hughsie, after this "sudo apt-get dist-upgrade"
[03:20] <aigarius> or be prepared to use a text mode IRC client for debuging :)
[03:20] <ogra> hughsie, anfter that have fun in getting X working
[03:20] <carstenh> aigarius: thanks for the information so far, i will have a look at it if i have time this weekend :)
[03:20] <hughsie> okay, lol, thanks. how long till x being fixed? hours or days?
[03:21] <ogra> hughsie, i wouldnt suggest a upgrade to breezy if you are not familiar with working around that X breakage
[03:21] <Kamion> not hours
[03:21] <aigarius> IIRC it has been broken for weeks and is improving very slowly
[03:21] <ogra> hughsie, most likely days... daniels said something about next week...
[03:22] <hughsie> ogra: okay, thanks. is apt clever, so i did sudo apt upgrade libnotify if would do all the rest? 
[03:22] <hughsie> of is breezy a development snapshot
[03:22] <hughsie> (shouldn't be asking q's like this in devel, sorry)
[03:22] <ogra> hughsie, on breezy, yes... if you do such things in hoary, it will pull in the dependencys
[03:22] <Kamion> (a) the command is apt-get not apt (b) you use 'install' to upgrade single packages, not 'upgrade'
[03:24] <hughsie> Kamion, thanks. I want to make sure g-p-m works in breezy with a manual dbus and hal install
[03:24] <carstenh> pitti: will you be here later to talk with jeff an me about wheter our files shuold be marked as conffiles and about the script that will change the configfile?
[03:24] <pitti> carstenh: I should still be here for ~ 2 hours
[03:25] <ogra> hughsie, hal is already updated :)
[03:25] <hughsie> 0.5.3?
[03:25] <ogra> hughsie, pitti is fast ;)
[03:25] <hughsie> pitti sounds like a legend
[03:25] <hughsie> :-)
[03:25] <hughsie> dbus?
[03:25] <ogra> nope :/
[03:25] <carstenh> pitti: ok, thanks. i hope jeff will be here in the next 2 hours and have some time :)
[03:26] <ogra> hughsie, thats the one that will take time i guess...
[03:26] <seb128> tseng: new gtk is apt-gettable if you want to give it a try
[03:26] <hughsie> ogra: i'll have a go at --prefix'ing it locally
[03:26] <ogra> oki
[03:26] <hughsie> thanks for the pointers, i'll five it a go now.
[03:27] <ogra> yay... so we made you a ubuntite :)
[03:27] <hughsie> ogra: only on my 2nd pc, laptop if firmly fc4 i'm afraid :-(
[03:27] <tseng> seb128: got it
[03:27] <hughsie> i figured i might as well take the plunge
[03:28] <tseng> seb128: crash
[03:28] <ogra> hughsie, ah well.... we all started with only one ubuntu install.... they tend to grow through your network... ITS A VIRUS watch out !! ;)
[03:29] <hughsie> ogra: what can breezy do for a AGP card that gets so hot it's untouchable. I would be inpressed if it would work on X with that.
[03:29] <ogra> *grin*
[03:29] <hughsie> ogra: great. just as i though linux didn;t need antivirus
[03:29] <ogra> it should work... if not, kick daniels to fix it :)
[03:29] <hughsie> ogra: cool :-)
[03:30] <ogra> there is no distro that ships a newer X then breezy... daniels is working also upstream....
[03:31] <seb128> tseng: crash what/when?
[03:31] <hughsie> i was listening in on your firefox discussion yesterday. whats ubuntus policy on backporting? fc seems to be "upgrade to latest" as it's simplest
[03:31] <seb128> tseng: update cairo/pango/gtk and restart your app ... if that still crash backtrace is welcome
[03:31] <tseng> seb128: firefox
[03:31] <tseng> seb128: i restarted all of X
[03:31] <tseng> after updating
[03:31] <tseng> ill get a backtrace later with cairo debug
[03:31] <seb128> do you have the backtrace and the assertion of what you get on the crash?
[03:32] <seb128> k
[03:32] <tseng> the assertion is the same
[03:32] <tseng> firefox-bin: cairo.c:86: _cairo_error: Assertion `status > CAIRO_STATUS_SUCCESS && status <= CAIRO_STATUS_FILE_NOT_FOUND' failed.
[03:32] <tseng> Aborted
[03:32] <ogra> hughsie, yes, we do this too now, we just changed the policy, since the upstream patches for FF are not easy backportable...
[03:32] <hughsie> ff tend to update other stuff in a release i've noticed
[03:32] <tseng> consider that to be a rare exception
[03:33] <tseng> not a precedent to changing the rules
[03:33] <ogra> hughsie, before we had a policy to just backport security related stuff to avoid new bugs through new features... 
[03:33] <hughsie> gotcha, thanks.
[03:33] <seb128> tseng: k, let me know if you get a backtrace
[03:33] <tseng> seb128: ok.
[03:33] <hughsie> so you guys are going to want me to branch CVS ideally when I release 0.2.0 g-p-m?
[03:33] <hughsie> at the moment i tent to just have one release (CVS!)
[03:34] <ogra> hughsie, when wil that be ? 
[03:34] <ogra> 0.2.0 i mean
[03:34] <seb128> tseng: does anything out of firefox crash?
[03:34] <hughsie> few weeks at this rate
[03:34] <tseng> seb128: not yet
[03:34] <seb128> k, so maybe a firefox issue
[03:34] <tseng> seb128: im using alot of other stuff
[03:35] <pitti> hughsie: daniels wanted to update dbus on monday
[03:35] <ogra> its unlikely we wont change g-p-m after my next package... since the freeze dates are near...
[03:35] <ogra> s/we wont/we will
[03:35] <hughsie> ogra: thats cool with me.
[03:35] <hughsie> i can stick to 1.1
[03:35] <ogra> so you dont need to branch
[03:35] <hughsie> you'll package cvs tho?
[03:36] <ogra> i think so... except you want to roll out a tarball release during the next days while we wait for dbus
[03:36] <ogra> then i'd take the tarball
[03:36] <hughsie> ogra: okay, give me a few hours to finish the part-glib-ification in cvs, and i'll do 0.1.2 for you
[03:36] <ogra> yay
[03:42] <ogra> seb128, do you have a bug about missing transparency for png's in firefox already ?
[03:43] <ogra> seb128, ah, nm, its not ff, its the screenshooter that produces them like that already
[03:44] <seb128> ogra: I don't maintain firefox
[03:44] <seb128> I don't have bugs about firefox
[03:44] <ogra> seb128, isnt the maintainer the last one who touched it ? ;)
[03:45] <Treenaks> ogra: so now IE7beta supports transparent PNGs, but firefox doesn't anymore? :P
[03:45] <\sh> lol
[03:45] <ogra> seb128, but the prob isnt ff as i said... my gnome-screenshooter seems to produce them without transparency
[03:45] <seb128> k
[03:45] <ogra> i.e. a big black border
[03:45] <\sh> ogra: use gimp
[03:46] <ogra> \sh, mm, wil that fix the bug... ?
[03:46] <ogra> let me first check how outdated my libpng is ;)
[03:46] <daniels> iz gtk boog
[03:47] <pitti> elmo: please sync pmount from Debian incoming
[03:47] <\sh> ogra: no..but will help u until the bug is fixed ,-)
[03:47] <ogra> \sh, i can live with black borders for now :) its not this important
[03:48] <ogra> i just wanted to know if its a known bug already....
[03:48] <daniels> Kamion: can we please move xterm to universe?  i can't be arsed maintaining it.
[03:48] <seb128> ogra: how outdated your gnome-utils is?
[03:49] <\sh> daniels: no ways...xterm belongs to X like steven jobs belongs to apple
[03:49] <\sh> steve even
[03:49] <daniels> \sh: so find someone to maintain it ... ?
[03:49] <daniels> \sh: i don't use it myself and we already have a terminal emulator in main
[03:49] <ogra> seb128, a bit :) 2.11.1-0ubuntu1 
[03:49] <seb128> ogra: that's probably it
[03:50] <ogra> yep
[03:50] <\sh> daniels: what must be done for xterm?
[03:50] <seb128> try with current one
[03:50] <ogra> i was suspecting that
[03:50] <ogra> seb128, i'll do, thanks
[03:50] <daniels> \sh: someone to keep on top of the bugs, to deal with new usptream versions, and to deal with ishikawa mutsumi in debian and get the debian and ubuntu packages synchronised.
[03:50] <daniels> \sh: although ishikawa might be handing over to branden robinson, so you may have to deal with it.
[03:51] <\sh> daniels: hmmm...
[03:51] <daniels> s/it/him/
[03:53] <daniels> which other people are far better-qualified to do than me.
[03:53] <\sh> daniels: it sounds like a nice job *sigh* is it ok for u when I'm doing it after breezy release? cause right now I want to get finished with the universe work and some kde stuff
[03:53] <Kamion> daniels: :P
[03:53] <daniels> \sh: sure.
[03:53] <daniels> Kamion: um, this isn't an xlibs-dev situation.  this is a serious question.
[03:54] <daniels> Kamion: (and gravity really wants to kill xlibs-dev ...)
[03:54] <Kamion> daniels: we clearly shouldn't move it to universe; it is widely used
[03:54] <Kamion> and definitely in the "traditional" category
[03:54] <daniels> Kamion: even in the presence of g-t and stuff like rxvt?
[03:54] <\sh> daniels: ok..setteled...I will do it after breezy release in october
[03:54] <daniels> Kamion: dude, xedit is traditional :P
[03:54] <Kamion> yes, rxvt is quirky
[03:55] <Kamion> not nearly so widely used as xterm
[03:55] <daniels> i haven't even packaged that, and don't intend to.
[03:55] <daniels> Kamion: not used at all AFAICT.  given that it doesn't work with UTF-8 at all.
[03:55] <Kamion> and g-t is slow and bulky and really not what some people want
[03:55] <\sh> daniels: xterm works everywhere (normally) and UTF-8 is sometimes not wanted ,-)
[03:55] <daniels> to be fair, g-t is actually a lot faster these days.
[03:56] <daniels> \sh: uhm, if a text editor refuses to even start if your locale is UTF-8, it's complete crap.
[03:56] <ogra> daniels, youre vouching for moving xedit to main ?
[03:56] <crimsun> (xterm, I think)
[03:57] <ogra> crimsun, i know ;)
[03:57] <\sh> daniels: vi is starting in xterm nicely
[03:57] <\sh> daniels: anyways...give me a pointer...let me have a look on it...
[03:57] <daniels> ogra: no, I'm vouching for xedit ceasing to be distributed at all.
[03:57] <daniels> \sh: um, apt-get source xterm?
[03:57] <ogra> heh
[03:58] <ogra> doko, where is wxgtk2.6 ? i need a working audacity for edubuntu
[03:59] <wasabi> My X colors are all messed up. How can I reset them?
[03:59] <wasabi> vmware did it. =/
[03:59] <crimsun> heh, I was just going to ping doko about that for vlc ;)
[03:59] <ogra> heh
[03:59] <daniels> \sh: and restore lxterm, koi8rterm, and uxterm (or whatever) from xterm 6.8.2-10, becuase I think they went missing
[03:59] <daniels> wasabi: try doing a VT swtich
[03:59] <doko> ogra: needs some fixing, to have both 2.4 and 2.6 installed at the same time
[03:59] <wasabi> daniels: can't.
[03:59] <ogra> doko, any ETA ?
[03:59] <daniels> wasabi: can't?
[03:59] <wasabi> My GF can't come back if I do so.
[04:00] <wasabi> Once I go to console, I have to reboot. ;)
[04:00] <daniels> really?  what the bloody hell sort of driver are you running?
[04:00] <doko> this weekend
[04:00] <wasabi> the binary one.
[04:00] <ogra> yay
[04:00] <wasabi> DVI out.
[04:00] <wasabi> Works fine with VGA out.
[04:00] <daniels> wasabi: matrox?
[04:00] <ogra> doko, thanks a lot :)
[04:00] <wasabi> geforce.
[04:00] <daniels> wasabi: heh.  both pieces.
[04:00] <\sh> DAMN why can't I SHUT MY MOUTH ,->
[04:01] <daniels> AHR IT'S THE RETURN OF THE ONE-EYED EMOTICON
[04:01] <wasabi> Yeah, well, it sucks. But that's the way it has been for about a year now.
[04:01] <daniels> accompanied by ZIPPY the PINHEAD
[04:01] <ogra> daniels, mga is totally fucked by intention ? 
[04:02] <\sh> daniels: uxterm should ne in xterm source + a special lib
[04:02] <daniels> ogra: nope, it's just that mga only allows DVI through their binary driver because they're a completely closed shop at the moment
[04:03] <daniels> \sh: um, isn't it just a shell script?
[04:03] <ogra> daniels, i have a fresh install here, that hardlocks after trying to start X
[04:03] <\sh> daniels: a wrapper yes..let me have a look first...*grrr*
[04:03] <ogra> daniels, the console gets a green border and doenst show anything anymore
[04:04] <\sh> I think I have to marry daniels...sometimes I think it good to have him as wife ;-) he will tell me when to shut up *harhar*
[04:04] <daniels> ogra: yeah, gcc4 sucks.  the next xorg upload (monday?) will be compiled against a fixed version of gcc4.  in the mean time, try grabbing libvgwha.a from a hoary system.
[04:04] <daniels> \sh: at least you have two eyes now
[04:05] <ogra> daniels, i'm fine to run this system in console mode for now... thanks for the info
[04:05] <daniels> ogra: no worries
[04:06] <\sh> daniels: I'm not a crippled guy...dodn't u know..two eyes, two arms, two legs, two..nevermind and I'm working...in germany this is really hard to get...so last chance *evilgrin*
[04:14] <pitti> Hi lamont
[04:15] <lamont> mornign pitti
[04:15] <JaneW> hi lamont
[04:15] <lamont> morning JaneW 
[04:16] <seb128> Hi lamont
[04:16] <lamont> hi seb128 
[04:16] <lamont> and everyone else
[04:16] <seb128> lamont: could you put a retry on epiphany-browser ?
[04:16] <lamont> sure
[04:16] <seb128> thanks
[04:19] <lamont> seb128: done
[04:27] <hughsie> ogra: well impressed with your updates tool
[04:27] <ogra> hughsie, apt ?
[04:28] <hughsie> no, i'm just updating hoary with all the updates
[04:28] <hughsie> the visual update program
[04:28] <ogra> ah
[04:28] <hughsie> much better than fc4
[04:28] <ogra> kudos for that one go to mvo :)
[04:28] <hughsie> but fc4's networking tool is better
[04:28] <hughsie> so it's one all :-)
[04:29] <ogra> we'll have NM in breezy (hopefully)
[04:29] <hughsie> ogra: thats cool. where's your services tool?
[04:29] <tseng> system - admin - services
[04:30] <ogra> gnome will have the g-s-t tool upstream... we'll use it
[04:30] <tseng> we have it
[04:30] <ogra> tseng, not in hoary
[04:30] <tseng> buh
[04:30] <tseng> who uses that :P
[04:30] <hughsie> tseng :-)
[04:30] <doko> seb128: evolution doesn't work anymore reading from an IMAP server ... started after the latest upload
[04:30] <hughsie> i'll update soon, but i want x!
[04:30] <ogra> tseng, people that use X
[04:30] <hughsie> heh
[04:30] <daniels> um, X works fine in breezy
[04:31] <daniels> latest X works fine in breezy
[04:31] <daniels> i'm even running it now
[04:31] <\sh> yes...I have a german keyboard again :)
[04:31] <ogra> daniels, haha
[04:31] <tseng> i think we need mkfontdir
[04:31] <tseng> yet
[04:31] <seb128> doko: what does it say?
[04:31] <daniels> tseng: yeah, mkfontdir is needed, and I have its required library in NEW now
[04:31] <baggins> hi danies
[04:31] <baggins> +l
[04:32] <daniels> tseng: but it won't matter on upgrades from hoary
[04:32] <daniels> baggins: yo
[04:32] <baggins> how's it going?
[04:32] <tseng> daniels: huh, my upgrade from hoary the other day broke pretty bad
[04:32] <tseng> cant find fixed font
[04:32] <tseng> and yes i fixed the paths
[04:32] <daniels> tseng: oh well
 tseng: yeah, mkfontdir is needed, and I have its required library in NEW now
[04:32] <daniels> baggins: not too badly
[04:32] <tseng> yes
[04:32] <ogra> ;)
[04:32] <doko> seb128: Loading ...
[04:33] <seb128> doko: with the update from this afternoon?
[04:33] <doko> ye
[04:33] <\sh> daniels: u want to have a new xterm-203 in breezy?
[04:34] <daniels> \sh: sure, if it works
[04:34] <\sh> lets see :)
[04:34] <daniels> \sh: it needs --sysconfdir=/etc in debian/rules
[04:34] <daniels> apparently it doesn't load its app-defaults right now
[04:37] <\sh> daniels: it has..
[04:38] <daniels> \sh: hm, ok
[04:44] <mvo> I need a good name for the NetwidePackageUpdate program. it's called auto-pkg-update-{server,client} right now and that name sucks :/
[04:44] <bddebian> Heya
[04:44] <mdz> Lathiat: yes?
[04:44] <mdz> pitti: yes, I acked the new hal
[04:45] <pitti> mdz: Good morning! Thanks, ogra and seb128 confirmed, so I already uploaded it :-) It works just fine and fixes a couple of bugs
[04:45] <\sh> mvo: apucs 
[04:45] <mvo> \sh: nice and short, but a bit cryptic. what does it mean?
[04:45] <mdz> pitti: great
[04:45] <ogra> mdz, first edubuntu testinstall up and running here.... no X though...
[04:46] <hughsie> ogra: in sources.list, deb http://archive.ubuntu.com/ubuntu/ breezy main restraicted universe
[04:46] <ogra> hughsie, looks ok
[04:46] <mdz> ogra: aren't all of the edubuntu apps X apps?
[04:46] <hughsie> ogra: then sudo apt-get install libnotify doesn't work
[04:46] <ogra> mdz, not the server part
[04:46] <ogra> hughsie, sudo apt-get update
[04:46] <\sh> auto-pkg-update-client'n'server 
[04:46] <ogra> hughsie, before the install command
[04:47] <seb128> hi mdz
[04:47] <hughsie> ogra: i have to update the whole lot?
[04:47] <carstenh> is it possible to limit the output of packages.ubuntu.org to show only packages in main?
[04:47] <ogra> hughsie, the package list
[04:47] <ogra> hughsie, apt-get update only updates the info about available packages
[04:47] <hughsie> ohh, sorry, yes i did this
[04:47] <\sh> mvo: it's an abbrev. of your package name...:)
[04:48] <\sh> mvo: in alphabetic order :)
[04:48] <lamont> (if it's marked 'Building', and it's in main, it's almost certainly a failure)
[04:48] <ogra> hughsie, and s/restraicted/restricted i hope
[04:49] <hughsie> yes, sorry, transcribing all this from an vnc window
[04:49] <carstenh> and if not, has anybody a idea to get a list of packages in main that provide a file in /etc/init.d/?
[04:49] <hughsie> i can do sudo apt-get update libnotify
[04:49] <ogra> hughsie, just to be sure :)
[04:49] <hughsie> but that want to download 200mb of packages
[04:50] <ogra> hughsie, yep... dependency management.... libnotify depends on the stuff it was built against (breezy)
[04:50] <hughsie> okay, gotcha. I'll give that a go.
[04:51] <hughsie> my server kicks arse.
[04:51] <ogra> hughsie, look that there gets no X stuff upgraded... it could break badly
[04:51] <mvo> \sh: I think apu is a good name, so it will be apu-client and apu-server
[04:51] <hughsie> i get to keep both bits.
[04:51] <hughsie> presumably I can just apt-get from non-x terminal in a few days?
[04:51] <ogra> yep
[04:52] <ogra> hughsie, daniels said something about monday... lets see
[04:52] <hughsie> ogra: cool
[04:52] <ogra> hughsie, for X  (not for dbus)
[04:53] <daniels> x will hopefully be monday also
[04:53] <ogra> daniels, also ? 
[04:53] <Keybuk> daniels: you an I need a word, outside ... :)
[04:53] <\sh> mvo: yeah :) apu sounds good :) 
[04:53] <ogra> daniels, does that mean you already prepared a new dbus ?
[04:53] <daniels> Keybuk: not until you fix dpkg.  it's screwing us up trying to build mesa.
[04:54] <Keybuk> "fix dpkg" ?
[04:54] <daniels> Keybuk: actually, that might make the word outside even more interesting ...
[04:54] <daniels> Keybuk: apt-get install leading to dpkg segfaults are bad, ok?
[04:54] <Keybuk> bug#?
[04:54] <daniels> ogra: largely, yeah
[04:54] <ogra> daniels, yay....
[04:54] <ogra> hughsie, ^^^
[04:54] <ogra> :)
[04:54] <hughsie> sweet as
[04:55] <hughsie> nice one daniels
[05:00] <\sh> ok..going home..and if everything works fine, daniels will have his xterm tonight
[05:00] <\sh> bbl
[05:00] <daniels> seb128: cairo.c, line 86.  dunno if it's in firefox or pango.
[05:00] <hughsie> anyone want to buy a server?
[05:04] <Keybuk> so, who wants to help me debug dpkg? :p
[05:04] <daniels> iwj volunteers
[05:05] <Keybuk> s/iwj/Diziet/, non? :p
[05:08] <seb128> daniels: what cairo.c ?
[05:08] <daniels> seb128: dunno.  that's the fun part. :)
[05:08] <seb128> daniels: where/how did you get that?
[05:08] <seb128> daniels: cairo.c is a cairo source file
[05:09] <daniels> 13:03 < thaytan> firefox-bin: cairo.c:86: _cairo_error: Assertion `status > CAIRO_STATUS_SUCCESS &&
[05:09] <daniels>                  status <= CAIRO_STATUS_FILE_NOT_FOUND' failed.
[05:10] <daniels> i don't have copies of any source trees with me, so I can't check it out at the moment.  but you sure it's not in pango or anything?
[05:10] <lamont> Keybuk: what did you do to dpkg?
[05:10] <daniels> although _cairo_error does look a little sus
[05:10] <Keybuk> lamont: nothing
[05:10] <Keybuk> quite literally
[05:10] <daniels> lamont: broke it on hppa, and only hppa
[05:10] <Keybuk> I'm wondering whether the zlib security upload broke it
[05:10] <lamont> hrm... gonna need a really really really big trebuchet to reach .au
[05:11] <lamont> what's the failure mechanism?
[05:11] <Keybuk> infinity: *poke*
[05:12] <lamont> Kamion: ping
[05:14] <Keybuk> sweet, metacity segfaulted again
[05:14] <lamont> Kamion: nm
[05:14] <Keybuk> I didn't need a window manager anyway
[05:15] <daniels> gtk bug
[05:24] <infinity> Keybuk : You're not going to like it.
[05:25] <hughsie> ogra: ubuntu supports rpm?
[05:26] <ogra> hughsie, through alien, yes... but i'd prefer a .deb from the package if i were you 
[05:26] <ogra> s/a .deb/a real .deb/
[05:26] <hughsie> ogra: okay, just seemed like the coke c.e.o saying "is pepsi okay?"
[05:27] <ogra> heh
[05:27] <mjg59> So I have a couple of packages that could do with uploading for Breezy
[05:27] <mjg59> But I don't have upload rights
[05:27] <mjg59> What's the best approach?
[05:27] <ogra> alien is a great tool if you cant find any .deb.... but with more then 16000 packages in the archive its seldom that you dont find what youre looking for
[05:28] <mjg59> (Other than to bully James into adding me to the keyring...)
[05:28] <daniels> mjg59: step one: buy me beer
[05:28] <daniels> mjg59: step two: profit
[05:28] <ogra> mjg59, main ? universe ? for universe, just thriw them at anybody in #ubuntu-motu
[05:28] <ogra> throw even
[05:29] <mjg59> Probably universe for now, but targetted at main
[05:29] <mjg59> seb128: Did you get the bug I filed for hotkey support?
[05:30] <seb128> yep, I'm busy with cairo/gtk update today though
[05:30] <hughsie> ogra: bah, real men don'e need X.
[05:30] <mjg59> seb128: No problem
[05:31] <ogra> hughsie, true... except when they have to build a distro alone that has two critical places that depend on X ;)
[05:31] <ogra> (talking about edubuntu)
[05:32] <hughsie> ogra: pahh, all i need is a curses xchat replacement !
[05:33] <ogra> hughsie, i doubt i'd make the teachers happy with that ... especially a ltsp install without X is funny :)
[05:33] <ogra> mdz, ltsp breaks in xlibs installation currently... 
[05:33] <mdz> ogra: did you file a bug?
[05:34] <ogra> mdz, nope, because next X is due for monday anyway...
[05:34] <ogra> mdz, it worked on monday though...
[05:34] <daniels> how does it break?
[05:35] <ogra> daniels, non informative with "Errors were encountered while processing: ... " no error above that...
[05:35] <ogra> i'll investigate a bit more in about 30min...
[05:35] <daniels> ogra: should work in -43, but -44 will definitely fix
[05:36] <ogra> it might also be the us mirror... the ltsp script defaults to that... i'll try my next ltsp install with archive.ubuntu.com
[05:38] <mdz> ogra: there was surely an error above that
[05:38] <ogra> wait a sec... its in the other room
[05:39] <daniels> mdz: no, not necessarily
[05:39] <daniels> mdz: there was a problem (which was only in -42, IIRC), where the last line of xlibs.preinst was test -d foo && bar
[05:39] <ogra> nope, no other error...
[05:39] <daniels> mdz: test -d foo fails, and the last exit status is 1
[05:39] <mdz> daniels: in which case dpkg would have printed a message saying that the exit status was nonzero
[05:39] <mdz> ogra: please mail me an unabridged copy of the log
[05:40] <ogra> mdz, ok
[05:44] <\sh> re
[05:47] <bddebian> wb \sh
[05:54] <\sh> daniels: xterm-203 works for me on i386
[05:55] <mdz> ogra: I don't see it yet
[05:55] <ogra> mdz, i just reran the install wait a sec
[05:55] <daniels> \sh: cool
[05:57] <\sh> u need it now? or in 2h..so I can go and take a nap first ,-)
[05:58] <daniels> \sh: um, whenever.  i'm not desperate for it.  might be worth emailing ishikawa@debian.org to see what he thinks first.
[05:59] <ogra> now, could someone tell my why "2&>1 > ltsp.log" swallows the error ?
[05:59] <bddebian>  2>&1
[05:59] <ogra> thats what i meant... sorry typo here
[05:59] <bddebian>  :-)
[05:59] <\sh> daniels: debian is using the bundeled xterm...hmmm
[06:00] <daniels> \sh: yeah, but ishikawa -- who did all the work on xfree86 4.2 -- did xterm packages for when it gets split out
[06:00] <daniels> and it's in our interest to not deviate from debian if we can avoid it
[06:00] <daniels> so if you chat to him and see if you can get the packages as close together as possible ...
[06:00] <\sh> k..will email him :)
[06:03] <\sh> k..mailed him...
[06:04] <daniels> cool
[06:09] <ogra> mdz, dpkg: error processing /var/cache/apt/archives/xlibs_6.8.2-43_all.deb (--unpack):
[06:09] <ogra>  subprocess pre-installation script returned error exit status 1
[06:09] <ogra> it occurs way earlier in the process and was hidden
[06:10] <daniels> right, that's probably one of the ones I fixed for -44
[06:10] <daniels> which will land on Monday
[06:20] <ogra> daniels, rmdir: `/etc/X11/xkb/geometry/digital_vndr': Directory not empty
[06:20] <ogra>  seems to be the original reason for the error
[06:21] <Amaranth> *groan*
[06:22] <Amaranth> it's something that's supposed to fail on breezy but not make the script exit
[06:22] <Amaranth> because it needs to be done for a clean upgrade from hoary
[06:22] <Amaranth> or something like that
[06:22] <daniels> yeah, I fixed the bug that tanked that
[06:22] <ogra> oki
[06:23] <bddebian> Nighty night :-)
[06:23] <\sh> laters
[06:34] <Kamion> mdz: I've just uploaded oem-config. Can I seed it for main? Do I need pitti to look over it for security first?
[06:40] <HiddenWolf> kamion, any other devel, is that ubuntu-lite project official, will it ever be?
[06:41] <ogra> HiddenWolf, vedran is working out the seeds... i'm not sure if a iso image is planned, but you will find a ubuntu-lite metapackage
[06:42] <Kamion> http://udu.wiki.ubuntu.com/LightweightDesktop
[06:42] <HiddenWolf> good. My roommate just came running up to me with an ancient win2k laptop bsod-ing on boot. 
[06:42] <HiddenWolf> And I'd like to stick er on ubuntu.
[06:43] <ogra> HiddenWolf, i'm talking about breezy
[06:44] <ogra> so you probably should take hoary with a server install and install the desktop manually
[06:44] <HiddenWolf> ogra, I've given her a windows cd for now, but just figuring out the options.
[06:44] <ogra> ah
[06:44] <HiddenWolf> I'm not going to fiddle with a 300mhz/64meg laptop on a saturday night. ;)
[06:44] <HiddenWolf> nor a friday.
[06:44] <ogra> heh
[06:45] <HiddenWolf> would hoary/xfce do for now?
[06:45] <ogra> sure
[06:46] <HiddenWolf> ogra, once again, you rock.
[06:46] <ogra> thanks :)
[06:48] <vedran> HiddenWolf: try the iso here www.ubuntulite.org it uses a different installer that wipes out your hard drive ;) but otherwise is pretty much the future lightweight desktop seed
[06:48] <HiddenWolf> vedran: how usable is it, atm?
[06:49] <vedran> v1.0 works but you have to login as root and configure x.org manually
[06:49] <HiddenWolf> ah.
[06:49] <vedran> v1.1 will fix that and other issues
[06:49] <HiddenWolf> what's the desktop used?
[06:50] <vedran> icewm
[06:50] <HiddenWolf> I need something that won't scare away a teenager.
[06:50] <HiddenWolf> any screenshots anywhere?
[06:51] <vedran> umm no
[06:51] <vedran> http://www.linuxcdmall.com/ubuntulite-screenshots.html
[06:52] <HiddenWolf> lol
[06:52] <HiddenWolf> that's a ripped off winxp install. :P
[06:52] <mdz> Kamion: OK to seed
[06:53] <vedran> HiddenWolf: heh it's just an icewm theme but official ubuntulite will not use that ofcourse
[06:53] <HiddenWolf> vedran: the site says it's not an official project. Is it not a project yet, or will it never be?
[06:54] <vedran> long story short - the guy started it independently and we are now trying to merge
[06:55] <HiddenWolf> Good.
[06:55] <Kamion> mdz: done
[06:56] <HiddenWolf> I'm not going to go anywhere else, not for a crappy laptop owned by the dumbest of girls I've ever known. :P
[06:56] <shackan> vedran, does it have anything good for merging ?
[06:57] <Amaranth> does icewm support all the fd.o specs?
[06:57] <HiddenWolf> shackan: if nothing else, a config that runs on "anything"
[06:57] <vedran> shackan: well mostly package list, some config hacks and the installer has the same goals as oem-installer
[06:58] <shackan> uhm, the installer seems still the old stile textmode stuf
[06:58] <shackan> *stuff
[06:58] <doko> seb128: imap doesn't work with evolution 2.3.6.1 as well
[06:58] <shackan> are we getting a graphical installer some day ?
[06:58] <seb128> doko: you already said so
[06:59] <HiddenWolf> shackan: count on it, if ubuntu is planning to distribute with HP. ;)
[06:59] <doko> seb128: no, that was 2.3.6
[06:59] <seb128> doko: ah
[06:59] <shackan> uh, didn't know that
[06:59] <seb128> doko: it does nothing?
[06:59] <vedran> shackan: why graphical? it isn't even interactive :P
[06:59] <seb128> doko: try to evolution --force-shutdown && evolution
[06:59] <vedran> guess that beats microsoft for usability...
[07:00] <shackan> lol
[07:01] <doko> seb128: same behaviour
[07:01] <Kamion> shackan: see http://udu.wiki.ubuntu.com/GraphicalInstaller, http://udu.wiki.ubuntu.com/UbuntuExpress
[07:01] <seb128> doko: it hangs?
[07:02] <doko> seb128: Clicking on the Tab in the left pane, it says "Loading ..." (the mailbox), then nothing happens. I'm able to click on other things, so the application itself doesn't hang
[07:04] <seb128> doko: that's using IMAP or IMAP4v1?
[07:05] <doko> imap4
[07:07] <doko> seb128: wait, the preferences overview says imap4, but the dialog imap4v1
[07:14] <ogra> Amaranth, icewm has a own menu system :/
[07:18] <jasoncohen> ogra, hey, i asked yesterday if mozilla-mplayer 2.85 would get into breeezy. 2.85 has been in sid for over a month and it works here. The releases since 2.7 have made signifigant aesthetic changes to the UI and there have also been some features added I believe
[07:18] <ogra> jasoncohen, i saw the bug already... 
[07:18] <Amaranth> what UI?
[07:18] <Amaranth> there's a UI?
[07:19] <jasoncohen> Amaranth, yes, if you compile with --enable-gtk2
[07:19] <daniels> http://thedailywtf.com/forums/39325/ShowPost.aspx
[07:19] <daniels> argh
[07:19] <jasoncohen> let me get the bug
[07:19] <ogra> Amaranth, yes after the totem plugin mplaye now stars thattoo
[07:19] <Amaranth> Can someone translate for me? :)
[07:20] <Amaranth> they're doing it because totem got a plugin?
[07:20] <ogra> Amaranth, totem plugin has a pause and a fullscreen button
[07:20] <jasoncohen> Amaranth, basically, the debian mozilla-mplayer is compiled without gtk2 support so there are no menus or buttons
[07:20] <ogra> Amaranth, i guess thats what they added to mplayer too now...
[07:20] <jasoncohen> if you compile with gtk2 (the default for mozilla-mplayer), you get a more attractive interferace, buttons, and a menu which you can download the stream
[07:21] <jasoncohen> Amaranth, https://launchpad.ubuntu.com/malone/bugs/1598
[07:22] <jasoncohen> i've already packaged a new 2.70 deb with the trivial changes
[07:22] <jasoncohen> ogra, would you be opposed to making the change?
[07:23] <ogra> jasoncohen, i currently dont have the time for it... but will look into it.. i'm not sure if the current version in multiverse is marillat or sid
[07:24] <jasoncohen> does marillat even have mozilla-mplayer?
[07:24] <Amaranth> daniels: that is sad
[07:24] <jasoncohen> i thought they used debian's version
[07:24] <jasoncohen> Maintainer: Rene Engelhard <rene@debian.org>
[07:25] <jasoncohen> so, i guess it's sid
[07:25] <shaya> anyone having firefox crashes due to cairo?
[07:25] <shaya> firefox-bin: cairo.c:86: _cairo_error: Assertion `status > CAIRO_STATUS_SUCCESS && status <= CAIRO_STATUS_FILE_NOT_FOUND' failed.
[07:26] <shaya> do I have to rebuild firefox due to cairo 0.6?
[07:32] <jasoncohen> i just receivedc DSA-769-1 for gaim. It fixes CAN-2005-2370 but this is not shown on http://gaim.sourceforge.net/security/. Also, CAN-2005-1934 AND CAN-2005-1269 remain unfixed. anyone know where this CAN came from or if it affects hoary?
[07:33] <Kamion> mdz: I'm thinking that OEMRescue wants to be deferred for breezy
[07:34] <mdz> Kamion: given that it's always said that it wants OEMInstaller in place first, and OEMInstaller is getting down to the wire, yeah, I agree
[07:34] <Kamion> mdz: OEMInstaller is in plausible shape now (the main problems are ropey UI, not major lacks of functionality), but OEMRescue is going to take a fair bit more development
[07:34] <Treenaks> jasoncohen: http://cve.mitre.org/cgi-bin/cvename.cgi?name=CAN-2005-2370
[07:35] <Kamion> so yeah, I'd like to do it right for breezy+1 rather than do a rushed half-arsed job and neglect other higher-priority things
[07:35] <jasoncohen> thanks Treenaks 
[07:37] <jasoncohen> heh, i don't get the debian security team. they fix a big in gadu which affects a miniscule number of people but leave firefox unpatched for weeks
[07:37] <Kamion> mdz: sorry the dependent task took so long
[07:37] <mdz> Kamion: no problem; rescue was always lower priority
[07:38] <Kamion> rescue's pretty heavily debated as well, judging from its comments section ...
[07:40] <mjg59> What C library do we use in the initrd? Any?
[07:41] <HiddenWolf> jasoncohen: did they ever make sense? perhaps that gadu bug was just easier.
[07:41] <jbailey> mjg59: glibc and klibc right now.
[07:41] <mjg59> jbailey: Both? Eww.
[07:41] <jbailey> mjg59: Likely to stay with glibc for breezy.
[07:41] <mjg59> Ah, fine.
[07:41] <mjg59> So usplash doesn't have to be statically linked?
[07:41] <jbailey> mjg59: I had hoped to do klibc.  It has certain deficiancies.  I got a chance to talk to HPA about them at OLS, and they won't be solved in the timeframe we need.
[07:41] <jbailey> mjg59: Right.
[07:41] <jasoncohen> HiddenWolf, they're probably trying to figure out a way to backport the security fixes from 1.0.5 without breaking firefox which neither Ubuntu or Red Hat could do
[07:42] <mjg59> jbailey: We still need to work out some way to keep the fifo available to it
[07:42] <HiddenWolf> jasoncohen: well, good for them. ;)
[07:42] <jasoncohen> Debian would never just package the upstream version. Rather than package a new mozilla in woody, they left mozilla 1.0.0 unpatched for over a year
[07:42] <HiddenWolf> imho they're just destroying their legitimacy as a rock-solid safe system.
[07:42] <jbailey> mjg59: We've been talking about move_mount'ing /dev to the new root partition.  If you drop it in there, would it be good enough?
[07:43] <mjg59> jbailey: Sure, that would do
[07:43] <mjg59> jbailey: Any preferences about the name?
[07:43] <jbailey> 'kay.  I'll backport the patch from Debian's udev o cope with that, then.
[07:43] <Treenaks> HiddenWolf: Yes, but don't forget that the mozilla guys are being dicks too
[07:44] <mjg59> jbailey: Ah - that's a point. How early will /dev appear, and will it be writable?
[07:44] <jbailey> mjg59: As long as it's low-conflict, I don't care....  usplash-fifo ? =)
[07:44] <Keybuk> mdz: #ubuntu-meeting
[07:44] <HiddenWolf> Treenaks: I'm not familiar with the details, but I got a distinct impression to that extent, yes.
[07:45] <mjg59> jbailey: If /dev is dynamic, we'll need to create the fifo on boot
[07:45] <mjg59> So we can't really start usplash until it's appeared
[07:46] <jbailey> I can mount it right before init-top is run, I guess, to make sure it's available for you.
[07:46] <mjg59> jbailey: Ok, that would work
[07:47] <jbailey> mjg59: Are you a C application?  Can you just call mkfifo yourself, or do you need /bin/mkfifo ?
[07:47] <jbailey> mkfifo(3) rather I mean
[07:47] <mjg59> jbailey: C, so that's not a problem
[07:48] <jbailey> Hmm, klibc provides mkfifo  anyway, so it's no big deal either way.
[07:48] <jbailey> I mean klibc-utils provides /bin/mkfifo
[07:48] <mjg59> jbailey: Ok. So basically all I need at the moment is the /dev and for vga16fb to be loaded. What's the right way to hook in to start usplash?
[07:49] <Keybuk> mjg59: #ubuntu-meeting :)
[07:49] <jbailey> To get vga16b loaded, the best thing is to drop a file called "usplash" that contains the word "vga16b" into /usr/share/initramfs-tools/modules.d
[07:50] <jbailey> To start it, drop the shell script that does what you need into /usr/share/initramfs-tools/scripts/init-top
[07:50] <jbailey> It needs to optionally accept one argument
[07:50] <jbailey> mjg59: Mind if I /msg you with a flood?
[07:50] <mjg59> jbailey: Cool. Go ahead
[07:51] <sabdfl> mdz: #ubuntu-meeting for mjg59?
[07:51] <Keybuk> quickly!  before our net connection drops again! :p
[07:51] <sabdfl> *cough* *splutter* did i just see usplash?
[07:51] <Kamion> jasoncohen: dude, we didn't package the upstream version until a couple of days ago either. :-)
[07:52] <Kamion> previous performance no indicator of future performance, etc.
[07:52] <jasoncohen> Kamion, yes, i know
[07:52] <jbailey> mjg59: That emulates a dependancy-based init setup
[07:52] <mjg59> sabdfl: Yeah, package ought to be ready soon
[07:52] <jasoncohen> Kamion, but ubuntu learned the error of their ways. It's simply too much of a burden to backport security fixes for mozilla/firefox on multiple distros. why do you think warty was never patched? pitti said he spent 40 hours backporting fixes
[07:53] <jbailey> mjg59: At the moment it's a hard dependancy.  It will soon be a soft dependancy so that if the thing you depend on doesn't exist, it won't worry about it.
[07:53] <jasoncohen> that's why debian gave up on woody
[07:53] <mjg59> jbailey: Ok. In this case, do I need to check for anything?
[07:53] <mjg59> Modules will be loaded before I run?
[07:53] <jbailey> mjg59: Nope.  The snippet that I pasted you will basically just do what you need.
[07:54] <jbailey> oo, good question.
[07:54] <sabdfl> mjg59: that's awesome news. how does it look?
[07:54] <jbailey> This might be before module load. =)
[07:54] <ogra> mjg59, congrats for main uploader status :)
[07:54] <sabdfl> ogra: fastest confirmation *ever*
[07:54] <jbailey> Whups, that's before the modules are loaded.
[07:54] <ogra> sabdfl, yep, i followed it... *g*
[07:55] <jbailey> mjg59: Will you need anything more in /dev than /dev/console?
[07:55] <mjg59> sabdfl: At the moment? Ugly :)
[07:55] <sabdfl> holler if you have specs for the artwork team
[07:55] <mjg59> jbailey: /dev/fb0 - is udev running at that point? If not, then it'll need to be created
[07:55] <mjg59> sabdfl: I've passed stuff on to Jane
[07:55] <sabdfl> ok cool, can't wait to see it
[07:56] <jbailey> mjg59: It's not.  udevstart isn't called until all the modules are loaded.
[07:56] <sabdfl> of course, we're all a little scared of breezy over here in chickenpox land, since daniels b0rked keybuk's X :-)
[07:56] <jbailey> mjg59: I'm' trying to figure out where to add the hook for you.
[07:56] <mjg59> jbailey: Ok. In that case can you make sure that it exists?
[07:56] <mjg59> Thanks!
[07:56] <jbailey> mjg59: Do you want this running before the root file system is mounted?
[07:56] <mjg59> jbailey: Ideally
[07:57] <ogra> whee, http://lists.ubuntu.com/archives/ubuntu-users/2005-July/044259.html
[07:58] <jbailey> Okay.  I have specific premount hooks for when booting off of a local harddrive and when booting off of nfs.   I just need to make a generic one.
[07:58] <\sh> *yawn* g'evening 
[07:58] <mjg59> jbailey: Oh, hrm. I could do with the "open" command.
[07:58] <mjg59> Is that practical?
[08:00] <jbailey> that's another 115k.  A lighter solution would be nicer if possible.
[08:01] <mjg59> Arh.
[08:01] <jbailey> Otherwise, there's no hard limit on i386 for the initramfs.
[08:01] <mjg59> I can probably implement the VT ioctl stuff myself, I guess.
[08:02] <jbailey> mjg59: Depends how much grief it is.  Compared to a stack of kernel modules, it's not big deal, but if it's a trivial convenience it's a bit heavy.
[08:02] <\sh> hmmm..
[08:04] <mjg59> jbailey: Nah, I can probably do it myself. No idea why "open" is that big, though.
[08:04] <jbailey> It's the two console-tools libraries that it depends on.
[08:04] <jbailey> open itself is 14k
[08:09] <mjg59> Ah, ok
[08:09] <mjg59> Building it statically would probably be smaller, but still...
[08:14] <jbailey> mjg59: To confirm, I'll give you a generic mount hook.  The means that you'll be able to hop in after all the modules are loaded, and so after /dev is populated with base system devices (So, no lvm, md, or evms).  After that happens, md, lvm, evms, any cryptroot stuff, nfsroot's fetching of an ip address, nfsmount, local mount, etc. happens.
[08:14] <\sh> infinito: ping
[08:14] <infinito> \sh: hi!
[08:14] <\sh> infinito: I just had a quick look on gcfilms
[08:14] <mjg59> jbailey: Ok. So at that point, /dev/fb0 exists. I create a /dev/usplash_fifo, change to a new VT, splash stuff on the screen
[08:15] <jbailey> Yup
[08:15] <\sh> infinito: and I don't think it's ready to be synced from debian to ubuntu
[08:15] <mjg59> jbailey: After that, things call usplash_write to write stuff to usplash
[08:15] <infinito> \sh: why?
[08:15] <mjg59> When we mount the real root, the same fifo exists in the same place. 
[08:15] <\sh> infinito: it starts out with missing dpatch build-deps
[08:15] <jbailey> mjg59: 'kay.  Does usplash_write fall back to displaying on the console if usplash isn't running?
[08:15] <mjg59> jbailey: So we don't do any chrooting in this model?
[08:15] <mjg59> jbailey: No, but the model I've assumed is that usplash_write is called in addition to echo
[08:15] <infinito> \sh: i know... buff... our version is patched, but the debian dev seems to be missing
[08:15] <\sh> infinito: and if I have to change it, then it will get an ubuntu package version stamp and it's not be synced automatically anymore
[08:16] <jbailey> mjg59: 'kay.  Does usplash_write fail silently if there's no usplash to write to?
[08:16] <\sh> infinito: so it has to go to our MOTU Review Process
[08:16] <mjg59> jbailey: Yup, without blocking
[08:16] <\sh> infinito: and for that, please join #ubuntu-motu to discuss this
[08:16] <infinito> \sh: ok
[08:16] <jbailey> mjg59: With an error code?  I think I do set -e usually now, so I'd hate to suddenly have the system die because someone didn't have usplash running.
[08:17] <mjg59> jbailey: No, with 0 in that case
[08:17] <jbailey> Cool
[08:17] <jbailey> mjg59: What chroot stuff are you thinking of?
[08:17] <mjg59> It'll error if something goes bizarro-wrong
[08:17] <mjg59> jbailey: The usplash process started from initramfs ought to be running until gdm starts. Is it going to have the same filesystem namespace once / is mounted?
[08:18] <mjg59> Hmm. This may be fun :)
[08:18] <jbailey> mjg59: We'll have to test it.  I don't know what a move mount looks like to a process that's already running.
[08:18] <jbailey> So far nothing had been persistant. =)
[08:18] <mjg59> jbailey: It reopens the fifo after every command is written (because it gets closed after the writer disconnects)
[08:20] <mjg59> So, uh. I have no real idea what'll happen.
[08:20] <jbailey> Excellent!
[08:20] <mjg59> Possibly it should communicate over unix domain sockets instead, or something.
[08:20] <jbailey> Umm, given that you'll have the same /dev in the new place, I suspect it'll be fine.
[08:20] <jbailey> Like I see it as two possibilities.
[08:20] <mjg59> But is it the same place as far as an already running process is concerned?
[08:21] <jbailey> Either move mounts don't affect currently running processes, in which case you'll have the same /dev you always did.
[08:21] <mjg59> We could do with Al Viro, really...
[08:21] <jbailey> Or it'll get move and you see the new filesystem layout, in which case we just need to make sure that nothing writes to the fifo between moving dev and moving /
[08:22] <mjg59> Well, let's go with it as is for the moment and see what breaks
[08:24] <mjg59> jbailey: Ok, looks like I need /dev/tty%d as well. Is that a problem?
[08:25] <jbailey> tty's and pty's are there.
[08:26] <mjg59> jbailey: Ok, sorted.
[08:26] <mjg59> 6 lines of extra code
[08:26] <jbailey> Lovely. =)
[08:29] <mdz> ogra: have you tested ltsp with the new unionfs in -6?
[08:31] <ogra> mdz, -6 ? 
[08:32] <mdz> ogra: kernel -6
[08:33] <ogra> ogra@edubuntu:~$ dpkg -l|grep linux-image
[08:33] <ogra> ii  linux-image-2.6.12-4-386          2.6.12-4.4 
[08:33] <ogra> installed this afternoon from yesterdays daily
[08:34] <Kamion> -6 is still in NEW
[08:34] <ogra> ahh, thats the prob
[08:34] <ogra> :)
[08:35] <Amaranth> o_O
[08:35] <mdz> ogra,Kamion: processing it now
[08:38] <Kamion> mdz: apt seems to need to build-depend on xmlto now
[08:38] <mdz> Kamion: that stuff is only necessary for building the source package
[08:38] <Kamion> mdz: at least, that's what's missing when building mvo's progress-reporting branch
[08:38] <mdz> the source package ships with pregenerated docs
[08:38] <Kamion> oh, ok
[08:38] <mdz> it's sort of a pre-build-dep
[08:38] <mdz> I think it's documented somewhere-or-other
[08:39] <\sh> Mithrandir: maswan: can u please update on ravel breezy chroot with: xmms-dev (>= 1.2.10+cvs20050209) libsqlite3-dev libpcre3-dev (>=4.3) libtag1-dev libvorbis-dev (>=1.0) fftw3-dev libglib2.0-dev dpatch automake zlib1g-dev libx11-dev libxss-dev
[08:39] <Kamion> we're going to have to think about standardising machine-readable representations of that if we're pushing buildability from arch
[08:39] <Kamion> apart from anything else grumpy will have some difficulty otherwis
[08:39] <Kamion> e
[08:39] <mvo> Kamion: it's only needed when doing arch-builds IIRC
[08:40] <mdz> mvo: I want to put out 0.6.40 this weekend
[08:41] <mdz> mvo: since I forgot to update my chroot and so didn't do the C++ transition on i386
[08:41] <maswan> \sh: done?
[08:41] <mvo> mdz: apt--fixes--0 may be interessting 
[08:41] <\sh> maswan: nope :)
[08:41] <mdz> mvo: what's there, besides the file auth stuff?
[08:42] <\sh> maswan: and fakeroot is missing as well :(
[08:42] <maswan> \sh: what's missing?
[08:42] <mdz> mvo: and does it change the abi? ;-)
[08:42] <\sh> maswan: xmms-dev (>= 1.2.10+cvs20050209) libsqlite3-dev libpcre3-dev (>=4.3) libtag1-dev libvorbis-dev (>=1.0) fftw3-dev libglib2.0-dev dpatch automake zlib1g-dev libx11-dev libxss-dev
[08:42] <maswan> \sh: the breezy chroot on ravel?
[08:42] <\sh> maswan: yes
[08:42] <\sh> argl...forget it
[08:42] <maswan> fakeroot is already the newest version.
[08:42] <mvo> mdz: I hope not :p
[08:42] <\sh> dchroot missing in \sh fingers
[08:43] <mvo> mdz: very small fixes, should all be very harmless
[08:43] <mvo> mdz: what do you thing about adding hashes to the copy method? it seems to be overhead for little gain
[08:43] <mvo> gowins copy example didn't convince me 
[08:43] <Kamion> mdz: do you happen to have a complete list of the pre-build-deps? I can't find it
[08:44] <mvo> Kamion: perlsgml, xmlto, sgml2x, sgmlspl IIRC
[08:46] <mjg59> jbailey: Can you give me a directory that I need to put my init script in, and a minimum version number for the right hook to be present?
[08:47] <shaya> anyone having problems w/ firefox?
[08:48] <mdz> Kamion: it's documented in COMPILING, but it's out of date (only lists the sgml tools, not the XML ones)
[08:48] <Kamion> ok, I was missing perlsgml and sgml2x, thanks
[08:48] <mdz> mvo: if you could update that in your fixes branch, that'd be great
[08:50] <mvo> mdz: I will do that now (will be patch-15)
[08:51] <mdz> mvo: I'm not particularly interested in goswin's argument :-P
[08:52] <Kamion> apt-get.ja.8.sgml fails to build even with those pre-build-deps
[08:52] <Keybuk> but Goswin makes such convincing, eloquent and well-considered arguments
[08:53] <Kamion> nsgmls:/home/cjwatson/apt/progress-reporting/doc/ja/apt-get.ja.8.sgml:3:59:W: cannot generate system identifier for public text "-//OASIS//DTD DocBook V3.1//EN"
[08:53] <Kamion> and a slew more
[08:55] <mvo> Kamion: debiandoc-sgml
[08:56] <mvo> ?
[08:56] <Kamion> got that
[08:56] <HiddenWolf> OMG, guys, I'm booting from an old hoaryRC on an old laptop, installing from cdrom, and it's telling me it can't find the cdrom drivers :D
[08:56] <mvo> Kamion: libsgml-perl?
[08:56] <Kamion> no such package
[08:56] <mvo> Kamion: libsgmls-perl?
[08:56] <mvo> (sorry)
[08:56] <Kamion> got it already
[08:57] <Kamion> ah, might be docbook
[08:57] <tseng> is there a way to save a backtrace from gdb
[08:57] <tseng> or ill have to paste it out?
[08:58] <Keybuk> "save" ?
[08:58] <Keybuk> you can dump core
[08:58] <tseng> i did dump core
[08:59] <Keybuk> that's effectively a saved backtrace
[08:59] <Kamion> 'set logging on', 'set logging file <file>'
[08:59] <tseng> Kamion: ah, cheers
[09:00] <Kamion> mvo: ah, yes, 'docbook' did it
[09:00] <Keybuk> ah, I see what you mean now, duh
[09:00] <Keybuk> sorry ;)
[09:01] <mvo> Kamion: thanks, I added that to the coming README
[09:02] <tseng> seb128: ff backtrace @ tseng.ath.cx/firefox-backtrace.txt
[09:03] <tseng> seb128: hm its missing "Previous frame inner to this frame (corrupt stack?)
[09:07] <Kamion> mvo: that branch makes aptitude fail to build, though
[09:07] <Kamion> (progress-reporting)
[09:07] <Kamion> cmdline_simulate.cc:41: error: call of overloaded `DoInstall()' is ambiguous
[09:08] <mvo> Kamion: hm, let me check
[09:08] <Kamion> oh, it's because you have a second DoInstall() with a default first arg, which can't be distinguished from the zero-arg case
[09:09] <Kamion> just dropping "=-1" from the second DoInstall() in packagemanager.h ought to do it ...
[09:10] <mjg59> jbailey: Ok, I have a package - just need to know where to put the init script :)
[09:11] <jbailey> mjg59: /usr/share/initramfs-tools/scripts/init-premount
[09:12] <jbailey> mjg59: Do you want the patch now for testing?
[09:12] <jbailey> mjg59: It'll be version 0.15
[09:12] <mjg59> jbailey: I don't have stuff set up to test initramfs yet, I'm afraid
[09:13] <mjg59> jbailey: If I give you a package, can you be a suck^wtester? :)
[09:14] <mvo> Kamion: yes, thanks. they will later be merged, but this way the ABI does not break
[09:14] <jbailey> mjg59: Sort of.  My laptop is text-only, pending the return of mkfontdir in breezy.
[09:14] <jbailey> So I won't be able to test the switch to gdm for you/.
[09:14] <jbailey> Hmm.
[09:14] <jbailey> Unless you can do ppc
[09:14] <mjg59> jbailey: That's ok, without adding stuff to the init scripts it'll just exit after 15 seconds
[09:15] <mjg59> ppc isn't ready yet
[09:15] <jbailey> 'k.  I can guinea pig that when you're ready too.  My other boxes are all headless.
[09:16] <Kamion> mvo: right
[09:16] <mjg59> Needs someone to write support for scrolling stuff in bogl
[09:16] <Kamion> mvo: aptitude builds after I tweak apt as above
[09:16] <mjg59> jbailey: www.srcf.ucam.org/~mjg59/tmp/usplash_0.1-1_i386.deb
[09:17] <Kamion> mvo: hmm - I think you did break the ABI, though
[09:18] <Kamion> mvo: aptitude: symbol lookup error: aptitude: undefined symbol: _ZN17pkgPackageManager9DoInstallEv
[09:18] <Kamion> mvo: (which c++filts to pkgPackageManager::DoInstall(), so seems related
[09:18] <Kamion> )
[09:18] <jbailey> mjg59: Got, lemme finish getting this glibc build started and I'll do this.
[09:18] <mjg59> jbailey: Once I get connectivity next week, I'll be able to test this locally :)
[09:18] <mjg59> jbailey: Cool, thanks!
[09:20] <mvo> Kamion: hrm, adding non-virtual functions shouldn't break the abi ... oh well (/me checks again)
[09:21] <Kamion> mvo: it sounds more like the zero-arg DoInstall() is being optimised into an inline function, or something ...
[09:21] <shaya> hmm
[09:22] <Kamion> mvo: it's certainly not present in the compiled libapt-pkg.so - only pkgPackageManager::DoInstall(int)
[09:24] <seb128> tseng: thanks. I get the crash too, it's not on startup
[09:26] <Kamion> mvo: shifting the stub DoInstall() function to the .cc file seems to work fine
[09:28] <shaya> is 2.6.12-6 good?
[09:29] <Kamion> mvo: do you expect this to work without source changes in aptitude? 'sudo aptitude -o APT::Status-Fd=3 install groff 3>~/foo' leaves ~/foo empty
[09:30] <tseng> seb128: i know
[09:30] <seb128> you didn't say so
[09:30] <tseng> sorry :/
[09:30] <seb128> np
[09:31] <seb128> use epiphany as a workaround :p
[09:31] <Kamion> mvo: boggle - it appears to work with Status-Fd=2 but not =3
[09:31] <\sh> gentlemen...who is xmms pro?
[09:32] <\sh> does xmms on amd64 has QUEUE_CONTROL enabled?
[09:32] <tseng> seb128: good idea
[09:33] <tseng> i used to use ephy all the time
[09:34] <Kamion> mvo: oh, never mind, I'm being bitten by a misfeature of sudo with apt-get, but aptitude still doesn't work
[09:34] <mvo> Kamion: I /msged you a small patch, could you please try that?
[09:34] <mvo> (for aptitutde)
[09:35] <Kamion> building now
[09:40] <Kamion> mvo: that works, thanks
[09:40] <Kamion> mvo: is the progress number intended to be floating-point?
[09:41] <Kamion> I guess that's OK, I'll int() it for debconf
[09:41] <Kamion> or s/\..*// anyway :)
[09:42] <mvo> Kamion: I haven't really thought about it :)
[09:42] <mvo> Kamion: does it work now?
[09:43] <Kamion> mvo: yep, seems to work fine - I'm pondering the base-config side now
[09:43] <Kamion> will have to make sure debconf works sanely if a package wants to ask a question while the progress bar is up
[09:43] <Kamion> which will be a little tricky, since unless I'm very careful the debconf database will be locked
[09:44] <Nafallo> hmm, is it just me or is those amd64 buildds down again?
[09:44] <mvo> Kamion: right, that sounds tricky
[09:45] <Kamion> I have a big hack held in reserve in case I can't think of a better solution (run the progress bar inside a debconf frontend that opens everything read-only)
[09:46] <Kamion> but I can probably make it work with passthrough and a bit of luck
[09:46] <mvo>  that sounds not too bad to me (also I have virtually no idea about debconf)
[09:47] <\sh> debconf is nice :)
[09:47] <\sh> and it's nasty
[09:48] <Kamion> having commit access to it upstream helps
[09:48] <\sh> the debugging mode is totally crap (from a fast debug person point of view) ;)
[09:49] <Kamion> hm? I find DEBCONF_DEBUG=developer incredibly useful actually
[09:49] <Kamion> but then I speak the protocol ...
[09:49] <Treenaks> HARD
[09:50] <Kamion> of course my current project includes a debconf protocol filter and has its own DEBCONF_DEBUG=filter debugging output - that's kind of fun
[09:51] <Kamion> work on the installer and you become very familiar with debconf very quickly
[09:52] <wasabi_> So is there an architecture doc on the new LiveCD infrastructure?
[09:53] <wasabi_> I think I got a grasp on most of it, but there are some changes I want to make.
[09:53] <Nafallo> s/his/he\'s/
[09:53] <wasabi_> Like, disable network config during the dpkg part and move it back to /etc/network/interfaces.
[09:55] <\sh> Kamion: yes...it's useful...
[09:56] <\sh> Kamion: but a --debug flag would be a nice2have
[09:59] <\sh> hmmm...i fixed imms
[10:01] <Kamion> \sh: thing is that debconf is invoked all over the place, so an environment variable is really easier to use
[10:02] <Kamion> mvo: hmm, tricky, apt closes all of its fds when forking
[10:02] <Kamion> or rather, sets them close-on-exec
[10:02] <HiddenWolf> I've got a very ancient laptop on my hands, and I get a very verbose error starting up alsa. How can I figure out what it is? are you interested in the bug, or not?
[10:03] <mvo> Kamion: yes, this anoyed me too a lot 
[10:04] <Keybuk> can we pull devscripts 2.9.x from unstable?
[10:05] <Kamion> maybe some configuration directive that tweaks the lower bound to mark close-on-exec?
[10:05] <Kamion> a bit hackish, but ...
[10:06] <mvo> I have the feeling that mdz will not like that. but some way to have more control over what fds will be closed and which will kept open would be really good
[10:08] <mvo> but we would need to come up with a clean way
[10:08] <mvo> (but it has my total support)
[10:09] <Kamion> it's certainly a showstopper for using passthrough from inside apt
[10:09] <Kamion> mdz: any thoughts?
[10:09] <\sh> Kamion: yes I know...
[10:11] <mdz> Kamion: could pass a bitfield saying which FDs should stay open
[10:11] <mdz> a la select(2)
[10:11] <Kamion> bitfields from shell?
[10:12] <mdz> shell?
[10:12] <Kamion> base-config is shell ...
[10:12] <mdz> I only saw from this forward: Kamion mvo: hmm, tricky, apt closes all of its fds when forking
[10:13] <mdz> a configuration option listing file descriptors to preserve seems doable
[10:13] <mvo> mdz: yes, a way to tell apt not to close it's fds would be cool
[10:13] <Kamion> oh, ok. the background is that I want to be able to use the debconf passthrough frontend while installing packages, in order to be able to have a debconf frontend displaying a progress bar while simultaneously not causing any packages that try to use debconf to fall over in a heap
[10:13] <Kamion> (because the debconf database would be locked
[10:13] <Kamion> )
[10:13] <HiddenWolf> ogra?
[10:14] <mdz> Kamion: apt configuration already has lists; I'd say a configuration item list of file descriptors would be the way to go
[10:14] <mvo> mdz: something like a list "APT::Interal::KeepFDs" ?
[10:14] <mdz> we could use that in place of the internal hack too
[10:15] <Kamion> mdz: that would work for mee
[10:15] <Kamion> er, me
[10:15] <mdz> just APT::KeepFDs I'd say
[10:15] <mdz> or it could stop closing file descriptors altogether
[10:16] <mdz> it's not necessarily its business
[10:16] <mvo> mdz: rock, I'll work on this tonight or tomorrow morning. it makes my life in synaptic easier too :) [stop closing is fine with me too] 
[10:17] <mdz> I'm not sure what the original rationale was; that code is old
[10:18] <mvo> there are quite a few spots where I really would like to hear from jgg what the rational was (Iterator::_dummy() comes to my mind)
[10:19] <hughsie> ogra: got a minute?
[10:20] <Kamion> yes, I could cope with it not closing fds; it's not on a security boundary so I don't see a good reason for that
[10:20] <Kamion> hmm
[10:21] <Kamion> I wonder if anything expects to be able to use hardcoded file descriptors
[10:21] <Kamion> thinking of maintainer scripts that do 4<foo or whatever
[10:21] <Kamion> it's sometimes hard to avoid that in shell, although you can just pick a higher number
[10:22] <Kamion> should be fine as long as they don't use debconf within such loops, though, and such code is fragile anyway ...
[10:24] <mdz> I like the idea of apt providing a fairly sanitized and sanity-enforced environment for dpkg to run in
[10:25] <mdz> but I don't have a specific case to back it up
[10:26] <HiddenWolf> daniels, ping?
[10:36] <highvoltage> mdz: we talked about testing on #edubuntu today. I had a hectic time since i got back... will be playing with your ltsp package this weekend.
[10:36] <mdz> highvoltage: I am desperate for testing and feedback; I don't think anyone but me and ogra have tried it
[10:37] <highvoltage> mdz: i'll also be installing a copy on my laptop, so i'll take it to a tuxlab, unplug a server and see how it works from my laptop.
[10:38] <highvoltage> mdz: i feel bad for not giving you feedback yet, but i'll give you lots of feedback. i just had a real tough past 3 weeks.
[10:38] <mdz> highvoltage: thanks
[11:10] (Kamion/#ubuntu-devel) mvo: doesn't build, I think you need ExecFork(int unused=-1) in fileutl.h rather than just (int unused)
[11:10] <\sh> hmmm....
[11:10] <\sh> amd64 buildd offline?
[11:11] <mvo> Kamion: *cough*, fixing that now
[11:12] <Nafallo> \sh: that's my conclusion. (with a silent yes from both #u-d and #u-m)
[11:15] <\sh> Nafallo: amen :)
[11:17] <mvo> Kamion: should work now
[11:19] <MAPD> hey
[11:19] <MAPD> isnt here anyone from ubuntu team is it?
[11:20] <mvo> Kamion: does it build now (after a baz update)? and maybe even work :p ?
[11:23] <\sh> g'night gentlemen and ladies..
[11:35] <Kamion> mvo: builds, at least ...
[11:41] <Mez> hmmm does anyone here have any experience creating FF themes?
[11:41] <Mez> thunderbird themes*
[11:42] <Kamion> mvo: seems to partly work; dpkg-preconfigure talks to passthrough, but the .config and .postinst confmodules themselves don't. hmm
[11:43] <mvo> Kamion: is this releated to the KeepFDs configuration?
[11:43] <Kamion> I've got -o APT::KeepFDs::=4 -o APT::KeepFDs::=5
[11:44] <Kamion> tried with higher fds, same
[11:45] <mvo> Kamion: it will only keep exactly this feeds
[11:47] <Kamion> mvo: yeah
[11:48] <Kamion> stracing, this is getting a bit mad
[11:52] <Kamion> mvo: debconf itself seems to be marking those fds close-on-exec, so it's not your problem
[11:55] <mvo> Kamion: ok, I'll probably be around another 1h, keep me updated :)