[12:09] <sabdfl> night guys
[12:11] <sivang> night sabdfl
[12:16] <sivang> jdub : are we using scrollkeepr for yelp in hoary? any plans to use gnome-doc-utils ?
[12:16] <jdub> sivang: gnome-doc-utils doesn't replace scrollkeeper
[12:19] <Mithrandir> jdub: can we either get rid of scrollkeeper or get it to run faster?  it's _dog_ slow.
[12:19] <jdub> Mithrandir: everyone wants to get rid of it :)
[12:19] <jdub> Mithrandir: it'll happen at some stage
[12:20] <Mithrandir> jdub: well, we're at the front and doing crazy things, so I think we should do it. :)
[12:20] <jdub> heh
[12:20] <jdub> it's not on our feature goal list :)
[12:20] <Mithrandir> that's easily fixed ;)
[12:21] <tim1> i already proposed it in the wiki, but just in case: http://linux.thestreet.dk/
[12:21] <tim1> debinstaller written in python
[12:22] <tim1> i didn't fully test it but i think ubuntu needs sth. like that
[12:23] <Mithrandir> why?
[12:24] <Mithrandir> oh, debinstaller, should be a synaptic module, I think.
[12:24] <tim1> because all basic tasks should be available in a simple straight-forward gui
[12:25] <Mithrandir> you shouldn't go around downloading random .debs, though.
[12:26] <tim1> yeah it doesn't have to be this specific implementation but their should be sth that pops up when clicking on and .deb saying what you can/should .. do/not do
[12:26] <tim1> a synaptic module would be better, i agree with that
[12:28] <tim1> hm so you don't think one should install a .deb outside apt/synaptic at all
[12:28] <tim1> speaking for the average user
[12:28] <Mithrandir> I'm not sure we want people to do that, no.
[12:28] <Mithrandir> (I'm talking about what I think, this isn't the views of canonical or anything, though I'm a developer)
[12:30] <tim1> mh ok
[12:30] <tim1> just my proposal
[12:31] <sivang> jdub : doc team is interested in bringing some worked docs to be available under yelp for hoary, scrollkeeper is our freind ? :)
[12:36] <jdub> sivang: see how the gnome documents are structured, ie. omf files, build layout, etc.
[12:36] <jdub> sivang: i imagine you guys are working on the gnome-user-guide a bit?
[12:40] <sivang> jdub : that document yes, and I would also like to have docs that we agree on, to be available under yelp. the FAQ is already being done in DocBook..
[12:49] <sivang> jdub : this is already material which is adopted from the main documentation section on the website, and wiki.
[12:53] <tim1> good night everybody
[12:56] <jdub> elmo: around?
[12:59] <jdub> d'oh
[12:59] <jdub> there is TOTAL RADNESS stuck in NEW
[12:59] <jdub> ;)
[01:01] <pitti> jdub: what in particular?
[01:01] <pitti> jdub: inotify? :-)
[01:01] <jdub> gnome-app-install
[01:01] <jdub> heh, i'm totally not going to be uploading kernels, dude ;)
[01:04] <pitti> jdub: but if we are at it, now that gamin is in hoary, it still uses dnotify, right?
[01:04] <pitti> jdub: or do we already have a patched kernel for inotify?
[01:05] <jdub> still uses dnotify
[01:05] <jdub> inotify is on the list for hoary kernels
[01:05] <jdub> the only update so far has been a bugfix
[01:06] <jdub> so
[01:06] <jdub> http://people.ubuntulinux.org/~jdub/gnome-app-install_0+20041111-1ubuntu1_all.deb
[01:06] <jdub> Keybuk: ^
[01:06] <wasabi> where is ubuntu NEW at?
[01:06] <wasabi> well that works.
[01:07] <pitti> good night, guys!
[01:07] <jdub> night pitti 
[01:07] <wasabi> oh that's slick.
[01:08] <Keybuk> jdub: fonts are a little *HUGE* ? :p
[01:08] <wasabi> That's a neat idea
[01:08] <jdub> they ought to be big, but not stupidly so
[01:09] <Keybuk> heh
[01:09] <jdub> if you think they're buggily big, sshot it for me :)
[01:09] <wasabi> I guess it uses it's own package list?
[01:09] <jdub> wasabi: see /usr/share/gnome-app-install/
[01:09] <wasabi> Since I don't see any X-Ubuntu-Package: in the .desktop files. ;0
[01:09] <wasabi> oh.
[01:09] <jdub> those .desktop files will be generated/pulled from the real ones later
[01:09] <wasabi> There they are.
[01:09] <Keybuk> http://descent.netsplit.com/~scott/big-fonts.png
[01:09] <jdub> for now, it's just some exampls
[01:10] <wasabi> Yeah. That's about the coolest thing I've seen in awhile. 
[01:10] <jdub> Keybuk: you reckon they're buggily big?
[01:10] <wasabi> Cool because it's usable by my mom.
[01:10] <Keybuk> jdub: yeah
[01:10] <jdub> Keybuk: the other option is bold, slightly larger, and an explanation of the category
[01:10] <jdub> but categories are stupid to explain
[01:11] <Keybuk> looks silly compared to the little fonts when you open the category
[01:11] <Keybuk> oh, they're big to make the icons big? :p
[01:11] <jdub> ok, i will wait for more feedback on that one ;)
[01:11] <Keybuk> The grafical user interface (xlib) does not support your chosen language environment
[01:11] <wasabi> Makes me wonder what kind of server stuff Ubuntu is going to go for.
[01:11] <Keybuk> heh *boom*
[01:11] <Keybuk> wasabi: "server stuff" ?
[01:11] <jdub> Keybuk: isn't that bong... so annoying
[01:11] <wasabi> Weither you'll be able to add/remove appache from that proggram... like you can do in OS X Server and Windows.
[01:11] <wasabi> FOr instance.
[01:12] <wasabi> (since all those servers are gui managed)
[01:12] <jdub> wasabi: there will be some stuff like that
[01:12] <jdub> wasabi: for instance, "LAMP Web Development"
[01:13] <jdub> etc.
[01:13] <jdub> there will also be developer-oriented selections
[01:13] <wasabi> http://www.ubuntulinux.org/wiki/DomainAuthenticationUtility/view?searchterm=authentication
[01:14] <wasabi> Someday i'll figure out how to format that right.
[01:14] <jdub> wasabi: seen the libuser proposal on u-d?
[01:15] <wasabi> haven't signed up yet
[01:15] <Mithrandir> jdub: has anybody ITP-ed libuser?
[01:15] <wasabi> searching
[01:15] <jdub> wasabi: i'm actively thinking about these problems - glad you are too :)
[01:15] <jdub> Mithrandir: no, but considering it.
[01:15] <wasabi> jdub, centralized auth systems are my speciality right now. :0
[01:15] <Mithrandir> jdub: me too.. basically hanging on available time.
[01:15] <wasabi> I'm tired of editing pam and nss.
[01:15] <wasabi> and creating kerberos keys. *scream*
[01:16] <jdub> Mithrandir: do you think it would be acceptable in debian by default?
[01:16] <Mithrandir> jdub: not for sarge, possibly for etch.
[01:16] <jdub> Mithrandir: (i think you'd be more up to maintaining it than i, btw)
[01:16] <wasabi> jdub, what month was it in?
[01:16] <jdub> Mithrandir: (although i could do the legwork, and you could send patches :)
[01:16] <Mithrandir> jdub: what's the upstream URL?  I only have your repo for it.
[01:16] <jdub> wasabi: last, maybe?
[01:16] <wasabi> hmm. no search on the archives.
[01:16] <jdub> Mithrandir: in src.rpms...
[01:16] <Mithrandir> ew.
[01:17] <jdub> :-)
[01:17] <Mithrandir> with new rpm being non-free and all.. phun!
[01:17] <Mithrandir> (it uses some non-free lib, iirc?
[01:17] <wasabi> http://lists.ubuntu.com/archives/ubuntu-devel/2004-November/000906.html
[01:17] <wasabi> That one?
[01:17] <Keybuk> Mithrandir: the lib might be free
[01:17] <wasabi> Oh Rock.
[01:17] <Keybuk> nobody Debianish really bothered to check that thoroughly
[01:17] <jdub> wasabi: yeah, message above
[01:18] <wasabi> Yeah that's exactly what im thinking about.
[01:18] <Mithrandir> Keybuk: ok, I just remember that joeyh orphaned rpm because of it.
[01:18] <wasabi> that wiki thing has a walk through explaining what I think of hte process.
[01:18] <wasabi> no UI mockups... those will come later.
[01:19] <jdub> wasabi: mmm, i think the infrastructure is more important for the moment anyway
[01:19] <wasabi> The infrastructure is already done really.
[01:19] <jdub> Mithrandir: want to co-maintain libuser?
[01:19] <wasabi> It's just a matter of glueing it all together.
[01:19] <jdub> wasabi: we talked about this -> "it's the integration, stupid" :-)
[01:19] <Mithrandir> jdub: sure; you ITP it or I?
[01:19] <wasabi> Yeah.
[01:19] <jdub> Mithrandir: i can
[01:19] <Mithrandir> jdub: go ahead, then.
[01:20] <wasabi> THat's what I mean jdub. We have to combine all these disparet services: Kerberos, LDAP, DNS, NSS, PAM, together into one entity. GIve it one name.
[01:21] <wasabi> And then it needs to work, with no fuss.
[01:22] <Mithrandir> where in the stack does libuser really fit in?
[01:22] <wasabi> Might be interested in the Lorikeet project.
[01:23] <wasabi> It's an effort by a number of the Samba guys to integrate all those things into one package.
[01:23] <wasabi> Clean them all up so they work perfectly together: openldap, heimdal, samba.
[01:23] <jdub> wasabi: oh?
[01:23] <jdub> which samba guys?
[01:23] <wasabi> abartlet told me about it
[01:23] <wasabi> think vorlon is working on it too, not sure.
[01:24] <wasabi> http://64.233.161.104/search?q=cache:bCG0alB-rPkJ:www.openldap.org/conf/odd-sandiego-2004/Andrew.pdf+lorikeet+samba&hl=en
[01:24] <jdub> oh
[01:24] <jdub> hrm, i think i've read this
[01:25] <wasabi> It's less a single project and more an effort to smooth all the bumbs from all the respective services so they integrate seamlessly
[01:25] <wasabi> They've altered Heimdal to store keys in LDAP, for instance.
[01:25] <Mithrandir> jdub: drop me an URL when you have an arch archive with libuser up and I'll tag off you?
[01:25] <jdub> mmm, and now there's the password sync module in openldap, which makes life easier
[01:25] <jdub> Mithrandir: heh, ok
[01:26] <wasabi> Well the end mission is to have no passwords in LDAP, to have one source: kerberos.
[01:26] <wasabi> but yeah, the password sync module is a good start. ;0
[01:28] <wasabi> So, my client gui idea is geared towards being dirt simple: in the best case it would only ask you for a user name and password.
[01:33] <sivang> jdub : so is user-conf from the gst backend going to be dropped altogether in favor of libuser and fedora's userconfig ?
[01:34] <sivang> jdub : sorry, this includes the gui also ofcourse.
[01:37] <wasabi> hmm. does libuser have any docs yet?
[01:38] <Mithrandir> wasabi: there's a docs/ directory in the tarball, at least.
[01:38] <Mithrandir> : tfheen@yiwaz ../libuser-0.52.5/docs/sgml > ls -l libuser.sgml 
[01:38] <Mithrandir> -rw-r--r--  1 tfheen tfheen 8629 2004-09-28 22:48 libuser.sgml
[01:38] <Mithrandir> a little bit about motivation and so on.
[01:40] <sivang> jdub : I have started some work with garanacho about adding the default privilieges groups , so it might be of interest to me :)
[01:43] <jdub> sivang: no decision so far
[01:46] <jdub> mdz: aorund?
[01:47] <robertj> wasabi: Whoever did OS X's directory stuff knew their stuff
[01:47] <wasabi> robertj, amen.
[01:47] <robertj> Except for the quota business
[01:47] <wasabi> robertj, have you used it?
[01:47] <wasabi> Even their ACL system (NT based) is awesome.
[01:47] <robertj> Im an admin at a mostly mac school
[01:47] <wasabi> So very well thought out.
[01:47] <wasabi> (in tiger)
[01:48] <robertj> wasabi: haven't looked at it
[01:48] <wasabi> Beats the pants off of POSIX ACLs and then some.
[01:48] <robertj> I installed tiger, checked to see if webdav workers with Nove's NetStorage yet (no), checked whether to see secure dav worked (no), and that was it.
[01:48] <wasabi> Really though, their directory stuff is just a combination of LDAP and Kerberos, hooked together with a great UI.
[01:48] <wasabi> Just like Windows.
[01:48] <robertj> wasabi: yeah. Made me very happy
[01:48] <wasabi> I think they're even using OPenLDAP as a base aren't they? I don't know that.
[01:49] <robertj> They do use OpenLDAP with a custom schema
[01:49] <wasabi> Yeah.
[01:49] <jdub> it would be nice if we had something like netinfo
[01:49] <wasabi> netinfo is deprecated no?
[01:49] <Mithrandir> jdub: netinfo is eeeeevil
[01:49] <robertj> netinfo is just a directory service plugin now
[01:49] <robertj> and it is evil
[01:49] <wasabi> Yeah. Replaced with OpenDirectory
[01:49] <wasabi> netinfo is just a front end API now.
[01:49] <jdub> how do they do disconnected clients? ldap proxy?
[01:49] <wasabi> That's how I read it anyways. ;)
[01:49] <robertj> jdub: they dont
[01:50] <robertj> OlD is just a frontend as well
[01:50] <wasabi> jdub, probably don't. MS doesn't.
[01:50] <wasabi> ANd MS has never needed to.
[01:50] <wasabi> But, it would be neat. ;)
[01:50] <jdub> erm
[01:50] <jdub> you *can* do disconnected clients with windows
[01:50] <robertj> Openldap, dotfiles nad netinfo all plug into the directory services architecture to actually do auth
[01:50] <wasabi> no domain clients.
[01:50] <wasabi> You mean disconnected file shares.
[01:50] <Mithrandir> jdub: I think we need something like planet, but for calendars.  Preferably with two-way synchronization.
[01:50] <jdub> no, authentication
[01:50] <robertj> jdub: that
[01:50] <robertj> er that's just cacheing
[01:51] <jdub> Mithrandir: yes, have been looking at it, oddly enough :)
[01:51] <jdub> robertj: exactly
[01:51] <robertj> pull your hard drive and put it in another machine and have it auth the machine account
[01:51] <jdub> i was under the impression that netinfo did it
[01:51] <robertj> That will be offline directory services ;)
[01:51] <robertj> jdub: I never saw netinfo used as anything but a dotfile replacement
[01:51] <Mithrandir> jdub: gotten anywhere yet?  I've just thought about it for a week or so, after I saw schoolbell hit unstable.
[01:51] <robertj> it's badly documented, poorly understood, and has a bad rep
[01:52] <jdub> Mithrandir: turns out the schooltool guys have written a fairly nice ical lib, which they're considering separating out
[01:52] <mjg59> The good news so far is that it looks like we can get away with including suspend to disk in hoary without any trouble
[01:52] <Mithrandir> jdub: ooh, that'd be great.
[01:52] <jdub> robertj: you could flick a switch to make it auth against other services
[01:52] <jdub> mjg59: yay!
[01:52] <Mithrandir> mjg59: yay!
[01:52] <robertj> jdub: on 10.2?
[01:52] <jdub> robertj: believe so
[01:52] <Mithrandir> mjg59: coming to BCN?
[01:52] <robertj> I dont think it had any gui utils for doing such
[01:52] <mjg59> Mithrandir: Doubt it - too much work to do
[01:52] <jdub> the server did
[01:52] <Mithrandir> mjg59: shame :/
[01:52] <robertj> on Mac you take your life into yoru own hands when you stray from the gui
[01:52] <mjg59> Mithrandir: I'll be in Australia, with luc
[01:53] <jdub> mjg59: oh, when to when?
[01:53] <robertj> not doing stuff through the gui causes things to get ugly
[01:53] <Mithrandir> mjg59: oh well, I guess that'll be fun. :)
[01:53] <mjg59> jdub: Oh, next year :)
[01:53] <jdub> oh
[01:53] <jdub> mjg59: for lca?
[01:53] <mjg59> Yeah
[01:53] <jdub> tops
[01:53] <jdub> we'll all be there
[01:53] <mjg59> Still waiting to hear back about papers
[01:53] <Kamion> mjg59: which reminds me; are you planning to sign up to join the Ubuntu maintainer list?
[01:53] <mjg59> Kamion: Hrngh.
[01:54] <jdub> haha
[01:54] <mjg59> Kamion: That implies a level of committment I can't current offer :)
[01:54] <jdub> Kamion: tbm will take away his hair.
[01:54] <Kamion> ah well, worth a try :)
[01:54] <mjg59> Kamion: For the moment I'd rather try to coordinate some stuff and make patches someone else's problem
[01:54] <robertj> like the thing that is currently souring my day is how quotas mysteriously dissapear and quotacheck crashes
[01:54] <Kamion> jdub: to replenish his own diminishing supply?
[01:55] <wasabi> we have no need for anything netinfo like
[01:55] <wasabi> since we have NSS
[01:55] <mjg59> Uh
[01:55] <mjg59> Dudes
[01:55] <mjg59> Why is acpid suddenly packaged as a native package?
[01:55] <mjg59> jdub: Oh, and why does polypaudio-x11 have no binaries in it?
[01:56] <jdub> mjg59: on?
[01:56] <robertj> OS X's flexible directory architecture is completely defeated by our EITS here
[01:56] <robertj> they run eDirectory, which is fine, except they _will_not_ add numeric uids
[01:57] <sivang> robertj : EITS ?(this one I am not familiar with)
[01:57] <mjg59> jdub: Hoary
[01:57] <robertj> enterprise information technology services
[01:57] <mjg59> i386
[01:58] <jdub> doubt it has anything to do with the xorg upload
[01:58] <jdub> but might
[01:58] <jdub> i'll see, needs some love anyway
[01:58] <Kamion> is polypaudio in universe or something? I don't see it on my mirror?
[01:58] <mjg59> Haha
[01:58] <jdub> Kamion: universe
[01:58] <mjg59> Kamion: yeah
[01:58] <Kamion> bleh
[01:58] <jdub> Kamion: not committing to it yet :)
[01:58] <mjg59> jdub: It desperately needs an esd-style Let go of the sound device after this much inactivity option
[01:59] <mjg59> It's got an option to spawn on demand, but that only works for polypaudio clients, not legacy esd ones
[01:59] <jdub> http://primates.ximian.com/~jimmac/blog/Artwork/Eeek-a-gaim
[01:59] <jdub> ha ha ha ha ha
[02:05] <mjg59> Suspend to RAM still looks a bit dodgy
[02:05] <mjg59> Which is a pain
[02:06] <jdub> how can we make booting with suspend-to-disk 'just work'?
[02:08] <mjg59> Add a kernel that supports it. Add resume= to the default kernel options. Add support scripts.
[02:08] <mjg59> Done.
[02:09] <mjg59> We probably need a few patches for some bits of hardware, but that's something that'll come up in testing
[02:11] <jdub> resume is fairly safe? :)
[02:12] <jdub> but i guess you can mkswap early on in boot
[02:12] <jdub> and stuff
[02:13] <mjg59> The resume process redoes the swap signature even if resume fails
[02:13] <mjg59> The only issue is if someone boots with noresume 
[02:13] <jdub> ah, right
[02:14] <mjg59> Which is DANGEROUS and BAD and they SOULDN'T BE DOING IT
[02:14] <mjg59> (ahem)
[02:14] <mjg59> But yeah, it works with the craptop
[02:14] <mjg59> So it ought to work anywhere
[02:15] <mjg59> (Now that I've fixed the couple of obvious bugs)
[02:15] <Kamion> I must try it with the craptop's older brother
[02:16] <mjg59> http://store.myaopen.com/proso2000.html - coooooooooooooooool
[02:16] <mjg59> Kamion: Which one is that? Your old Sony?
[02:17] <Kamion> "default kernel options" is a bit of an interesting concept BTW; we can change {grub,lilo,yaboot}-installer, but how does it work on upgrades?
[02:17] <Kamion> no, another Averatec box but rather more powerful
[02:18] <Kamion> coooooooooooooooool> I think you misspelled "freeeeeeeeeeeeeeeeeeeeeeeeak"
[02:18] <mjg59> Kamion: Oh, I know the one
[02:18] <mjg59> This isn't an averatec - it's even cheaper
[02:18] <Kamion> oh, so it is
[02:18] <Kamion> right, I meant "another Via box" I think
[02:19] <mjg59> Heh
[02:19] <mjg59> Kamion: Re: upgrades - in theory, grub can add it itself in the postinst
[02:20] <mjg59> On the other hand, you probably don't want to do that in general
[02:20] <Kamion> yeah, semantics get a bit hairy though 'cos you're fiddling with a config file
[02:20] <mjg59> There's no real way we can autodetect which partition to resume off
[02:20] <mjg59> Swap hasn't been mounted at that point
[02:21] <mjg59> I guess it /could/ do something like RAID autostart and look for anything with the right header, but that's going to be pain
[02:21] <jdub> mjg59: i *almost* got my hands on a voyager.
[02:22] <mjg59> Ooh
[02:22] <jdub> but i stupidly installed linux on it, and the customer wanted to keep it
[02:22] <jdub> *d'oh*
[02:22] <jdub> fucking sweet machine,though
[02:24] <robertj> is disk manager an ubuntu specific tool>?
[02:24] <jdub> no, it's a gnome-system-tool
[02:24] <robertj> ahh
[02:25] <robertj> It would be really great if the hal device manager ganked those tabs
[02:26] <jdub> hal device manager is really just a hal inspector
[02:26] <jdub> we shouldn't have put it in the menu :)
[02:27] <mjg59> But seriously, guys - we need more x86 laptop testing before I try to get this stuff pushed into Hoary
[02:27] <jdub> (my fault)
[02:27] <jdub> mjg59: dude, hoary is the devel branch for testing! 
[02:27] <robertj> yeah, I used to thhink that kind of thing was needed, but I believe it's really kinda useless
[02:27] <mjg59> jdub: Yeah, but I'm not going to convince Herbert to do anything with this stuff unless I've seen it work on a decent number of machines
[02:28] <jdub> heh
[02:28] <mjg59> And you need someone to do you production-ready 2.6.9 images, because 2.6.8.1 isn't going to be good enough
[02:28] <mjg59> And I am *not* going to sort that shit out on PPC and amd64
[02:28] <Kamion> Herbert's doing that, isn't he?
[02:28] <Kamion> is he still on contract?
[02:29] <jdub> he is
[02:29] <mjg59> Kamion: As far as I know
[02:29] <jdub> dunno if he's doing that htough
[02:29] <mjg59> Kamion: My point was more that this stuff isn't going into 2.6.8.1, and I'm not doing you proper 2.6.9 images :)
[02:29] <Kamion> indeed
[02:29] <jdub> mjg59: did you dump patches in bugzilla for herbert?
[02:29] <mjg59> jdub: I dumped the kernel one, yeah
[02:30] <mjg59> jdub: My current patch isn't his ideal solution, though. Someone ought to follow up to the swsusp bug and ask him whether he's happy with that or whether I should forward port his modular swsusp one
[02:32] <robertj> I'm eager to try out Ubuntu on my laptop, but the only cds I have now don't play well with my laptops' drive :(
[02:34] <sladen> okay, how do you do tables in whatever 'reStructuredText(tm)' is?
[02:35] <sladen> || Would be nice if it just || Worked ||
[02:38] <sivang> sladen : hahah
[02:39] <sivang> sladen : was my point on the markup doc discussion..:)
[02:39] <Kamion> sladen: http://docutils.sourceforge.net/docs/user/rst/quickref.html#tables
[02:40] <Kamion> www.zwiki.org links to help on editing which links to restructuredtext stuff
[02:40] <mjg59> sladen: If you pick moinformat or whatever, the || notation works
[02:40] <mjg59> Oh, it's hard for people to do certain types of editting on this wiki
[02:40] <mjg59> I know, we'll ADD ANOTHER INCOMPATIBLE FORMAT
[02:41] <mjg59> Why are all wikis more crack-addled than Euginia?
[02:44] <sladen> been reading that quicklist, but can't find what [Nice Cleartext link to http://ob.fus.cated.org/]  is
[02:47] <sladen> mjg59: T-shirt please.  ''I should not have to use Emacs picture-mode to edit you %$#@! wiki tables''
[02:48] <robertj> sladen: if only there was a markup language with a regular syntax that everyone knew!
[02:49] <sivang> robertj : html ? :)
[02:49] <mjg59> I think we should invent a new one
[02:49] <sivang> stop this dudes, you're killing me :)
[02:49] <mjg59> I haven't researched any of the existing ones, but I once heard someone say that they were crap
[02:50] <jdub> moinmoin's is nice
[02:50] <jdub> moinmoin is nice
[02:50] <jdub> *cough*
[02:52] <Keybuk> I use Textile on my blog, it's quite nice ...
[02:52] <Keybuk> Mithrandir uses a Wiki-like syntax on his I think
[02:52] <sivang> moinmoin is nice :)
[02:52] <robertj> there needs to be some way to cooerce people into reporting comptability results ;)
[02:52] <Keybuk> my complaint about WikiSyntax is it's '''bold''' not just *bold*
[02:52] <Keybuk> /italic/ _underline_ etc.
[02:53] <mjg59> My problem with WikiSyntax is that if you try to use English it tries to interpret it as something closer to txt
[02:53] <mjg59> Ah! Capital letters! MINE MINE MINE
[02:53] <Keybuk> OhYeahAndTheLinkFormatSucks
[02:55] <robertj> Ther really needs to be a way to convert between wiki and prolog
[02:55] <mjg59> There really needs to be a way to convert between wiki and MY ARSE.
[02:55] <sivang> LetsDropThoseRadioButtonForOtherFormatsThanMoinMoin
[02:56] <mjg59> OOH! RADIOBUTTONS!
[02:56] <mjg59> MORE CRACK YES PLEASE YUM YUM YUM
[02:58] <sivang> AndAlwaysUseTheSimpleWikiSkin
[02:58] <bob2> AlwaysAlwaysAlways
[02:59] <mjg59> I hate wikis
[02:59] <robertj> I love the concept
[02:59] <robertj> I dislike the execution
[03:00] <robertj> hopefully there will be some yummy firefox plugins
[03:00] <bob2> they're documented on the MakeWikisNotSuck page, iirc
[03:01] <mjg59> We're going to make wikis not suck by fixing them in firefox?
[03:02] <robertj> is the wiki a zwiki?
[03:04] <mjg59> Yes
[03:09] <mdz> jdub: here now for a bit
[03:09] <jdub> mdz: mailed you
[03:09] <mdz> jdub: UML?
[03:09] <jdub> yeah
[03:20] <robertj> aww DTML is disabled
[03:20] <mdz> in our zwiki?
[03:20] <robertj> I think so
[03:21] <bob2> yeah, we need more markup formats
[03:21] <robertj> <dtml-var ZopeTime> didn't evaluate in a reStructuredText format
[03:22] <Keybuk> christing fuck, tla is slow
[03:25] <jdub> mdz: thanks for reply - great :)
[03:25] <lamont_r> Keybuk: you sure you're not just asking it to do a metric boatload of stuff?
[03:25] <jdub> Keybuk: not using baz yet?
[03:25] <sivang_sleep> this wiki discussion is all too fun, however night everybody!
[03:25] <jdub> $ cat ~/bin/tla
[03:25] <jdub> #!/bin/sh
[03:25] <jdub> figlet -f small "USE BAZ, DICKHEAD"
[03:25] <jdub> 
[03:25] <bob2> hahahaha
[03:25] <Keybuk> lamont_r: oh, I *am* making it do lots ... deliberately :p
[03:26] <lamont_r> hehe
[03:43] <jdub> mjg59: how do you make suspend suspend without pressing the suspend button?
[03:54] <Keybuk> jdub: echo 3 > /proc/acpi/sleep
[03:54] <sladen> jdub: grab his latest packages and  sudo /etc/acpi/suspend.sh
[04:20] <Keybuk> man, I hate benchmarks
[04:20] <Keybuk> they take too long :p
[05:15] <sladen> keybuk: just fake them!  Quicker *and* more reliable...
[06:16] <bob2> anyone know the deal with pppoe dsl and ubuntu?
[06:16] <bob2> ie, do you need just "ppp" or do you need "pppoe" from universe, too?
[06:16] <wasabi_> pppoeconf
[06:16] <wasabi_> i think you just run that.
[06:17] <bob2> on a default install?
[06:17] <wasabi_> I hope so anyways. I installed ubuntu on a co-workers PC today and told her to go home and run that
[06:17] <wasabi_> Seems so.
[06:17] <wasabi_> It pops up and tries to detect PPPoE.
[06:17] <wasabi_> I couldn't run it all the way through since I wasn't at the premisis with the actual pppoe
[06:17] <bob2> ah
[06:17] <wasabi_> but it look like it was working.
[06:20] <wasabi_> so where can I order a boxed copy of ubuntu? :)
[06:20] <bob2> send me some cash and I'll box some cds for you ;)
[06:22] <chrisa> Weren't warty cds supposed to be shipped per the order form on the site?
[06:22] <bob2> early ones were
[06:22] <lamont_r> chrisa: any day now
[06:22] <bob2> you can login and see the status of your order
[06:23] <lamont_r> chrisa: the issue is that it takes time to burn that many CD's and package and ship them.  And that couldn't start until the release was out...
[06:23] <chrisa> lamont_r: indeed
[06:32] <fabbione> morning guys
[06:39] <lamont_r> fabbione: still Xsprinting?
[06:39] <fabbione> lamont_r: yes
[06:39] <fabbione> last day
[06:40] <lamont_r> fabbione: does 2153 fit on daniels's list for the sprint, or must I bribe him with candy?
[06:41] <fabbione> lamont_r: we will have to ask daniels
[06:41] <fabbione> or jdub 
[06:41] <fabbione> to reproduce it with xorg
[06:41] <fabbione> i don't have any synaptics device here to even test it
[06:41] <fabbione> ps it's 6:40 am and my best guess is that daniels is still having hard sex with his pillow
[06:41] <lamont_r> me neither.  but the nice man who's letting me sleep on his sofa does.
[06:42] <lamont_r> he's even willing to test.
[06:42] <fabbione> well if he can test with X.org that would be nice
[06:42] <fabbione> there is a detailed readme for the setup...
[06:42] <fabbione> let me find it
[06:43] <fabbione> zless /usr/share/doc/xorg-driver-synaptics/README.gz
[06:43] <fabbione> there
[06:43] <fabbione> it explains how to enable the shmconfig in xorg.conf
[06:46] <lamont_r> can one just install xorg, or is it best to just go to hoary?
[06:46] <fabbione> i think you need an extra package to install xorg in warty
[06:46] <fabbione> but definetely i suggest a dist-upgrade
[06:46] <fabbione> pulling all the libs manually is a pain
[06:47] <lamont_r> he's upgrading now
[06:49] <fabbione> lamont_r: nice
[06:51] <fabbione> lamont_r: i am stocked with the sparc port
[06:51] <fabbione> gcc doesn't build correctly
[06:51] <fabbione> gcc-3.3
[06:51] <lamont_r> fabbione: test failures?
[06:52] <fabbione> yes
[06:52] <fabbione> a lot of them
[06:52] <fabbione> i don't think it is good to ignore them
[06:53] <lamont_r> heh
[06:53] <mdz> morning, fabio
[06:53] <fabbione> Executing on host: /build/sparcbuildd/gcc-3.3-3.3.5/build/gcc/g++ -shared-libgcc -B/build/sparcbuildd/gcc-3.3-3.3.5/build/gcc/ -nostdinc++ -L/build/
[06:53] <fabbione> sparcbuildd/gcc-3.3-3.3.5/build/sparc-linux/libstdc++-v3/src -L/build/sparcbuildd/gcc-3.3-3.3.5/build/sparc-linux/libstdc++-v3/src/.libs -B/usr/spar
[06:53] <fabbione> c-linux/bin/ -B/usr/sparc-linux/lib/ -isystem /usr/sparc-linux/include -g -ffunction-sections -fdata-sections -fmessage-length=0 -DDEBUG_ASSERT -DLO
[06:53] <fabbione> CALEDIR="/build/sparcbuildd/gcc-3.3-3.3.5/build/sparc-linux/libstdc++-v3/po/share/locale" -nostdinc++ -I/build/sparcbuildd/gcc-3.3-3.3.5/build/sparc
[06:53] <fabbione> -linux/libstdc++-v3/include/sparc-linux -I/build/sparcbuildd/gcc-3.3-3.3.5/build/sparc-linux/libstdc++-v3/include -I/build/sparcbuildd/gcc-3.3-3.3.5
[06:53] <fabbione> ckward -I/build/sparcbuildd/gcc-3.3-3.3.5/src/libstdc++-v3/testsuite /build/sparcbuildd/gcc-3.3-3.3.5/src/libstdc++-v3/testsuite/17_intro/header_cer
[06:53] <fabbione> rno.cc   -g -O2 -DDEBUG_ASSERT  -L/build/sparcbuildd/gcc-3.3-3.3.5/build/sparc-linux/./libstdc++-v3/testsuite -lv3test -lm   -o ./header_cerrno.exe 
[06:53] <fabbione>    (timeout = 450)
[06:53] <fabbione> compiler exited with status -1
[06:53] <fabbione> output is:
[06:53] <lamont_r> fabbione: talk about spam....
[06:53] <fabbione> spawn failed
[06:53] <fabbione> hey mdz
[06:54] <lamont_r> fabbione: oh - that
[06:54] <fabbione> plenty of them
[06:54] <fabbione> all the way trough
[06:54] <lamont_r> fabbione: fix the perms on /dev/pts /dev/ptmx
[06:54] <fabbione> lamont_r: ah
[06:55] <fabbione> ls -las ptmx 
[06:55] <fabbione> 0 crw-rw-rw-  1 root tty 5, 2 Sep 18 15:18 ptmx
[06:55] <fabbione> there are no pts/*
[06:56] <mdz> fabbione: are you using static /dev or udev?
[06:56] <fabbione> mdz: static. it's in a chroot
[06:56] <mdz> with static /dev you need to mount devpts
[06:56] <fabbione> point
[06:56] <lamont_r> mdz: thanks - I knew it was somethign like that
[06:57] <lamont_r> fabbione: that's probably the issue
[06:57] <lamont_r> elmo can probably tell you what we had to do to fix the build daemons
[06:58] <fabbione> lamont_r: if it is only question of mounting and changing the permissions.. that's easy
[06:59] <lamont_r> yeah - I can't remember what it was, but it was a trivial fix that we finally found - I might be able to find it in my email archive
[07:00] <fabbione> because it's like.. hmmmm a week that i am fighting with it?
[07:00] <lamont_r> fabbione: if it helps, elmo and I bounced against it for a couple months or so...
[07:01] <fabbione> so i am not completely dumb, if that's what you are saying ;)
[07:02] <lamont_r> right
[07:03] <fabbione> interesting
[07:04] <fabbione> with udev they fails 100 times faster
[07:04] <fabbione> :-)
[07:14] <fabbione> lamont_r: can you check in the buildd /dev if it has been fixed with static entries or dynamic?
[07:15] <fabbione> kinda: what is the best solution?
[07:15] <lamont_r> fabbione: has /dev/pts mounted in the chroot
[07:16] <fabbione> lamont_r: thanks
[07:17] <fabbione> lamont_r: do you suggest having /proc in the chroot?
[07:18] <fabbione> or do it on a case by case base?
[07:18] <lamont_r> fabbione: certainly have /proc mounted in the chroot - things fail otherwise
[07:18] <lamont_r> (procps being the notable example...)
[07:19] <lamont_r> having said all that, I need to get /dev/pts in hoary and the other chroots on all the buildd's...
[07:19] <lamont_r> sigh.
[07:21] <fabbione> ehhe
[08:37] <lamont_r> fabbione: let me know (email preferrably) if that fixes the gcc build issue - I'm not 100% sure that's it.
[08:37] <fabbione> lamont_r: i did try to run mkcheck and they were passing fine
[08:37] <lamont_r> online off-and-on fri afternoon-sunday evening
[08:37] <lamont_r> fabbione: sigh
[08:37] <fabbione> i am fixing the chroots now
[08:37] <lamont_r> ok.  time to fix all the buildd's then
[08:38] <fabbione> and i will re-sbuild gcc as first
[08:38] <fabbione> lamont_r: just wait that i complete the tests
[08:38] <fabbione> what i did was to run udev and chmod stuff here and there
[08:38] <fabbione> i need to test the pts solution
[08:38] <fabbione> it was just a quick and dirty hack
[09:32] <ironwolf> daniels: 2153 is fixed.  SHMConfig works with synaptics and synclient and tap touching. many thanks.
[09:33] <fabbione> ironwolf: he did nothing :-)
[09:33] <fabbione> just sitting on his chair to keep is bottom warm :P
[09:33] <fabbione> Mithrandir: you around?
[09:34] <Mithrandir> yes
[09:34] <Mithrandir> for a little bit
[09:35] <fabbione> Mithrandir: 3291?
[09:36] <ironwolf> well I'm glad it works. :)
[09:37] <Mithrandir> fabbione: I've seen it, but a lot of people seems to have problems with them.. considering waiting until Debian releases a new version as well.
[09:37] <daniels> fabbione: do you want the patch I'm doing a final diff on, or not? :P
[09:38] <daniels> ironwolf: no worries dude :) glad to hear it's all good
[09:38] <fabbione> Mithrandir: ah ok.. thanks
[09:38] <fabbione> Mithrandir: i saw the debian package coming in.. but i didn't check the bts yet
[09:39] <fabbione> daniels: yes... speed up kid :P
[09:39] <Mithrandir> fabbione: oh, we can certainly package what comes from upstream, most of the people complaining were tnt2 and mx440 users.
[09:39] <fabbione> Mithrandir: no.. let's wait
[09:39] <fabbione> i saw 2 versions from upstream
[09:39] <Mithrandir> it's snowing! :))
[09:39] <fabbione> in a raw
[09:40] <fabbione> so perhaps the last on fixes the breakage
[09:40] <fabbione> but we can certainly wait a few more days
[09:40] <fabbione> Mithrandir: cool
[09:40] <fabbione> not here... yet...
[09:41] <ironwolf> daniels: now to get the battery monitor thingy working....
[10:03] <daniels> mjg59: when you wake up -- do you agree that not switching back to the previous VT on exit is sane behaviour for -novtswitch?
[10:03] <daniels> mjg59: to me, -novtswitch implies 'bugger off, I'll deal with the switching, keep your grubby hands off it'
[10:04] <bob2> hm, should gpg be setuid these days?
[10:05] <mdz> bob2: nope
[10:06] <mvo_> hey mdz 
[10:07] <mvo_> still awake ;) ?
[10:07] <bob2> hrm, it's whinging about using insecure memory tho
[10:07] <mdz> mvo_: yes, I need to fix that shortly
[10:07] <mdz> bob2: are you not running an Ubuntu kernel?
[10:07] <mvo_> bob2: AFAIK gpg does no longer need to be suid because the 2.6.9 contains a api to get small secure memory
[10:08] <mdz> mvo_: Ubuntu 2.6.8.1 has the same patch
[10:08] <bob2> mdz: hoary default
[10:08] <mvo_> mdz: must be pretty late in your part of the world already :)
[10:08] <mvo_> mdz: ah, cool!
[10:08] <bob2> mvo_: hm, I thought it was using capabilities or something
[10:09] <mdz> bob2: hmm, I use gpg about a billion times per day, and don't get that warning
[10:09] <mdz> (and it isn't setuid)
[10:09] <bob2> I only just noticed it today
[10:09] <mdz> bob2: strace it?
[10:09] <bob2> ah, good idea
[10:09] <mdz> bob2: are you really really sure you're running the right kernel?
[10:10] <bob2> 2.6.8.1-3-powerpc
[10:11] <mdz> how odd
[10:11] <mdz> I don't think I've actually tried it on powerpc, but it would be very strange if it were broken there
[10:11] <bob2> yeah, I'm sure it wasn't saying it before
[10:11] <bob2> does the patch require me to be in a magic group or something (grasping at straws)?
[10:11] <mdz> bob2: aha
[10:11] <daniels> bob2: works on my x40 ;)
[10:12] <mdz> bob2: the patch touches include/asm-<arch>/resource.h
[10:12] <mdz> bob2: for pretty much every arch except powerpc
[10:12] <bob2> mdz: hah
[10:12] <mdz> bob2: please file a bug, component 'linux'
[10:12] <daniels> -rw-r--r--  1 daniels daniels 247M 2004-11-12 10:09 013_novtswitch.diff
[10:12] <bob2> mdz: sure
[10:12] <daniels> maybe I should try that diff again
[10:12] <mdz> daniels: yeah, maybe
[10:12] <bob2> mdz: thanks!
[10:14] <daniels> mdz: it's single-handedly larger than all of xfree86's debian/patches, believe it or not ;)
[10:16] <Micksa> heh, do a patch on xfree86 that replaces it with xorg
[10:16] <mdz> daniels: it's nearly as big as all of xfree86
[10:16] <pitti> Kamion: Hi!
[10:17] <pitti> Kamion: What do you think is better? Implementing the language packs as tasks or metapackages?
[10:17] <daniels> mdz: you tend to get patches that big when you diff full source X source trees against directories that don't exist :P
[10:17] <daniels> jesus!
[10:17] <daniels> guys, 1989 called, it said something about a UNIFIED CONTEXT DIFF
[10:17] <pitti> Kamion: I thought a bit over it, and tasks almost seems a bit more elegant to me
[10:18] <mdz> pitti: the only thing about tasks is that it's much less straightforward to propagate changes
[10:18] <mdz> to installed system
[10:18] <mdz> s
[10:18] <pitti> mdz: right
[10:18] <pitti> mdz: you mean if we get a completely new software that also has language packs?
[10:18] <pitti> that's true
[10:27] <mdz> night
[10:30] <pitti> Night mdz
[10:31] <mzarza> hi
[10:33] <trukulo> fabbione, jdub: r u awake?
[10:34] <fabbione> trukulo: yes?
[10:35] <trukulo> :) ok, i wanna talk about matar
[10:35] <trukulo> come to #ubuntu-offtopic or private?
[10:36] <fabbione> -offtopic
[10:36] <trukulo> k
[10:45] <pitti> Morning seb128!
[10:46] <seb128> hello !
[10:46] <pitti> elmo: do we have a warty dchroot somewhere that I can use?
[10:46] <mvo_> hi seb128, hi pitti 
[10:46] <pitti> elmo: no, a hoary one. Sorry
[10:47] <pitti> Hi mvo_!
[10:48] <trukulo> hi seb
[10:51] <fabbione> pitti: ping
[10:52] <pitti> fabbione: pong. Hi, howdy?
[10:52] <fabbione> pitti: hey..
[10:52] <fabbione> pitti: 280233 and 280871
[10:52] <fabbione> (from debian)
[10:53] <ddaa> mjg59: hey
[10:54] <ddaa> t42p kernel testing update
[10:54] <ddaa> Using debs fresh from your page.
[10:54] <ddaa> 1. suspend-to-ram: okay
[10:54] <ddaa> 2. suspend-to-disk: mostly okay
[10:54] <bob2> how fast does to-ram eat your battery?
[10:55] <bob2> 10%/hour?
[10:56] <ddaa> mostly: on wakeup, i get this message (copied from dmesg):
[10:56] <ddaa> eth1: Coming out of suspend...
[10:56] <ddaa> ipw2200: failed to send TX_POWER command [5 times] 
[10:56] <ddaa> ipw2200: Unable to initialize device after 5 attempts.
[10:57] <bob2> does it work fine if you rmmod before suspend, then modprobe afterwards?
[10:57] <ddaa> To get the wireless to work again, I have to "rmmod ipw2200; modprobe ipw2200".
[10:57] <bob2> ah, right
[10:58] <ddaa> External video seems to work. I guess the external video resolution configuration is an X issue now.
[10:58] <pitti> fabbione: odd thing. Do you know of similar reports about the sid version?
[10:58] <fabbione> pitti: no
[10:58] <pitti> fabbione: I did not test apache in an UML, though
[10:59] <ddaa> mjg59: and, finally, the suspend-to-disk key is inneffective, I have to run the hibernate script by hand.
[10:59] <fabbione> pitti: note that both of them are in UML env
[10:59] <fabbione> pitti: i don't think it breaks in normal env
[10:59] <pitti> fabbione: I never noticed that, too
[10:59] <ddaa> bob2: last time I checked, it was similar to daniels report, 10% per 8 hours.
[10:59] <pitti> fabbione: I have a test apache running on my ancient p60 server and it works fine :-)
[11:00] <bob2> ddaa: oh, awesome
[11:00] <pitti> fabbione: however, Joey's and my patches could be compared
[11:00] <ddaa> bob2: or something of that order. I have not tested with that new kernel though.
[11:00] <fabbione> pitti: aren't they the same?
[11:00] <fabbione> or better..
[11:00] <fabbione> they should be the same
[11:01] <pitti> fabbione: not at all
[11:01] <ddaa> bob2: yeah, awesome. It's very close to fully functional, and it beats the crap out of desktop in computing power.
[11:01] <fabbione> pitti: uh?
[11:01] <fabbione> pitti: oh right.. you patched 1.3.33
[11:01] <fabbione> or .31
[11:01] <pitti> fabbione: the second report is about apache-ssl
[11:01] <pitti> fabbione: http://packages.debian.org/changelogs/pool/main/a/apache-ssl/apache-ssl_1.3.26.1+1.48-0woody3/changelog
[11:01] <fabbione> same patch to both of them?
[11:01] <pitti> fabbione: these are completely different CANs
[11:02] <pitti> fabbione: I have to look
[11:02] <fabbione> pitti: ok
[11:03] <pitti> fabbione: I fixed -0940 in 1.3.31
[11:03] <pitti> fabbione: Matt fixed -0492 in 1.3.26-0woody5
[11:04] <pitti> fabbione: Joey fixed 0839, 0840, 0843 in -ssl
[11:04] <pitti> fabbione: the latter ones are already fixed in 1.3.31
[11:04] <pitti> fabbione: but it might be that -0940 is still missing in woody (but I notified Joey...)
[11:05] <fabbione> hmmm
[11:05] <fabbione> it sounds like a big mess to me
[11:07] <pitti> indeed
[11:27] <pitti> sjoerd: Hi!
[11:27] <pitti> sjoerd: here?
[12:11] <mjg59> ddaa: Yeah, the suspend to disk key requires ibm-acpi
[12:28] <sjoerd> pitti: i am now
[12:28] <sjoerd> pitti: morning :)
[12:37] <pitti> Hi! Back from lunch...
[12:37] <ChrisH> Hey... does anyone happen to know how far the hotel will be from the tecnocampus?
[12:38] <pitti> sjoerd: I still have the problem that I must integrate the --async switch of pmount into the GUI
[12:38] <pitti> sjoerd: but AFAIK upstream has its own idea for this problem, right?
[12:39] <pitti> sjoerd: AFAIK they automatically mount async if the size is bigger than a treshold
[12:39] <pitti> sjoerd: how do you do this in Debian?
[12:39] <pitti> sjoerd: it should be configurable, preferably in the g-v-m dialog
[12:39] <jdub> hey hey hey!
[12:40] <pitti> Hey jdub
[12:47] <rburton> yo yo jdub
[12:47] <rburton> jdub: i'll do a new deb of AppInstall with the gksudo changes now if you want
[12:47] <ChrisH> Well... I'll figure it out somehow. :)
[12:48] <jdub> rburton: ok :)
[12:49] <rburton> argh
[12:50] <Kamion> this has been made a touch more painful by the fact that four packages from the CDs have moved to pool/universe/ since warty was released :)
[12:51] <rburton> jdub: gksu isn't passing pipes to child processes -- i can't do gksu synaptic and pipe into it
[12:51] <rburton> gksudu even
[12:52] <jdub> oh yeah
[12:52] <mvo_> I could add a option to read the selections from a file?
[12:52] <rburton> fixing gksu(do) would be better
[12:52] <jdub> yeah
[12:52] <mvo_> oh, certainly :)
[12:52] <jdub> making them not kill tty and stuff would be nice too
[12:53] <mvo_> rburton: is your deb available somewhere public?
[12:54] <rburton> mvo_: burtonini.com/debian/unstable
[12:54] <rburton>  /debian has APT lines
[12:54] <seb128> rburton: just ping kov to get gksu(do) fixed :)
[12:54] <rburton> ah, kov wrote it. fab
[12:55] <seb128> yeah, and usually he's pretty fast to deal with bugs reported
[12:55] <mvo_> and ask him to return 1 as exitcode if the password is wrong  :)
[12:55] <rburton> heh
[12:58] <jdub> elmo: ping
[12:58] <jdub> rburton: i uploaded the other one too
[12:58] <jdub> but it hasn't got past NEW
[12:58] <rburton> jdub: cool
[12:59] <mvo_> rburton: it looks great! 
[12:59] <jdub> it's fricken amazingly brill
[12:59] <mvo_> you may want to add "--hide-main-window" when calling synaptic :)
[01:00] <rburton> mvo_: my synaptic doesn't support it yet iirc
[01:00] <rburton> (on sid here)
[01:00] <mvo_> do you use the warty version?
[01:00] <mvo_> oh, on sid. sure :)
[01:00] <jdub> rburton: you could check the version ;)
[01:00] <rburton> mvo_: hm, according to strings my synaptic does have it. sweeeet
[01:02] <rburton> mvo_: what version did it appear in?
[01:02] <mvo_> I think in 0.55
[01:02] <mvo_> don't add a version check :)
[01:02] <rburton> no, for the Depends:
[01:04] <mvo_> 0.55
[01:04] <jdub> oh
[01:04] <jdub> elmo just let it in
[01:05] <rburton> go elmo
[01:05] <jdub> seb128: the evo launcher is in gnome-panel?!
[01:05] <jdub> seb128: couldn't we patch ours in to do the right version things?
[01:05] <jdub> seb128: in the evo package itself?
[01:05] <seb128> jdub: the panel default layout is
[01:05] <jdub> oh
[01:05] <seb128> jdub: the panel config includes some .desktop
[01:06] <jdub> that one :)
[01:06] <jdub> gotcha :)
[01:06] <seb128> ok :)
[01:23] <pitti> Hi sivan!
[01:23] <sivang> Hey Pitti!
[01:26] <robtaylor> amu: ping?
[01:30] <Kamion> it doesn't present the "choose a mirror" option unless you have "Debian:" or "Non-US:" in the jigdo's [Parts]  section
[01:30] <Kamion> so we can't do Ubuntu:
[01:52] <daniels> mjg59: ping
[01:52] <daniels> mjg59: please email libv@skynet.be the lspci -vvv output for the video hw on the craptop, ccing myself (d.s@c.c)
[01:52] <Kamion> Successfully created `warty-release-install-i386.iso'
[02:13] <mjg59> daniels: Ok, just booting it
[02:15] <mjg59> daniels: Just the video hardware?
[02:15] <daniels> mjg59: yah
[02:16] <daniels> Kamion: cool :)
[02:16] <mjg59> Sent
[02:19] <daniels> cheers dude
[02:20] <daniels> 'Elitegroup Computer Systems'
[02:20] <daniels> that instils so much confidence it hurts
[02:22] <ddaa> mjg59: is there a particular reason why ibm-acpi is not included in your kernel/modules packages?
[02:26] <mjg59> ddaa: Because I didn't add the patch :)
[02:26] <mjg59> daniels: ECS have a long and proud tradition of producing shit
[02:26] <ddaa> :) When do you expect to have the time to add it?
[02:26] <mjg59> ddaa: Possibly over the weekend
[02:26] <daniels> mjg59: they've been around for more than like four weeks?
[02:26] <mjg59> daniels: Yeah
[02:27] <daniels> oh dear.
[02:27] <mjg59> 1987
[02:27] <daniels> !
[02:27] <ddaa> Also, it seems that there is an issue with hotplug. My usb drive was not automounted. I will give you more data as soon as I can come around testing that properly.
[02:27] <Kamion> what were they selling then, Spectrums?
[02:28] <chrisa> ECS, PCChips, PCWave
[02:28] <chrisa> Same difference, all utter crap
[02:28] <chrisa> Been around years, custom build shops love them because they can sell cheap, integrated barebones systems
[02:28] <daniels> chrisa: you forgot 'shit', apparently
[02:29] <chrisa> daniels: Figured it was implied
[02:45] <stratus> daniels, did you read my mail?
[02:46] <daniels> yeah, haven't had time to fix it though
[02:46] <daniels> that was literally a three-minute job
[02:46] <daniels> if that
[02:48] <stratus> daniels, np ;)
[02:49] <stratus> daniels, i haven't looked too deep but xcompmgr seems to be ok, transset was in that state.
[02:49] <daniels> yeah, xcompmgr seems to work for most people
[02:50] <stratus> daniels, and i've tested my transset here to make fun of some co-workers, it's ok too.
[02:51] <daniels> cool
[02:51] <stratus> wfm (slowly, of course)
[02:51] <jdub> noice'n'fast with nvidia ;)
[02:52] <stratus> daniels, i've a i845g at my desktop can i speed up it anyway?
[02:52] <daniels> i'm not entirely sure if i845g accelerates anything beyond rectangle fills :P
[02:53] <stratus> haha, i checked the logs and i think that the rectangle fills are "accelerated". :(
[02:54] <jdub> daniels: http://www.gnome.org/~jdub/blog/1100266979
[02:54] <daniels> if you want solid, reasonably reliable video output, get an i8xx
[02:55] <daniels> if you want anything resembling 'fast', or 'featureful', get ati
[02:55] <stratus> daniels, i'm not a gamer but i want to use composite and damage in the near feature.
[02:56] <Mithrandir> daniels: .. and if you are on amd64, don't get ATI. :)
[02:56] <jdub> dudes
[02:56] <jdub> stuff this 'hardware support list' business
[02:57] <jdub> let's write a hardware buyer's guide for ubuntu
[02:57] <stratus> is it serious?
[03:00] <pitti> elmo: here?
[03:00] <daniels> jdub: bring your gloves to mataro, we'll sort ati vs nvidia out properly ;)
[03:00] <daniels> Mithrandir: 3d is for weenies
[03:00] <daniels> Mithrandir: but r2xx works just fine anyway
[03:00] <daniels> Mithrandir: so kwitcherbitchin
[03:01] <stratus> daniels, can you add a micro computer to the 'hardware buyer's guide' ? http://davidguy.brinkster.net/computer/001.html
[03:02] <Mithrandir> jdub: I don't like nvidia, but nobody else seems to care about 3d on amd64, so..
[03:02] <stratus> jdub, what video card are you using? ati?
[03:03] <jdub> stratus: nvidia on my desktop/test machine, used to be ati. i855 on my laptop
[03:03] <seb128> jdub: do we have an official "minimum hardware requirement" somewhere ?
[03:03] <stratus> jdub, hmm and composite works well with your nvidia at desktop? btw, what nvidia?
[03:03] <jdub> um
[03:03] <jdub> i forget
[03:04] <jdub> oh
[03:04] <stratus> seb128, hi i'm the guy that's crashing evolution! :)
[03:04] <jdub> 0000:03:00.0 VGA compatible controller: nVidia Corporation: Unknown device 0343 (rev a1)
[03:04] <jdub> composite works nicely
[03:04] <stratus> jdub, :)
[03:04] <seb128> stratus: oh ok, it doesn't crash in gdb so ?
[03:04] <stratus> i'll test at home, i was talking about my i845g at work.
[03:05] <stratus> seb128, it freezes with that gdb output that i've attached to the bug report.
[03:05] <seb128> this one is not while hitting "[", right ?
[03:05] <stratus> no, no
[03:06] <seb128> ]  I mean
[03:06] <stratus> seb128, i'm talking about #3634
[03:06] <mwh_> hi, im wondering if one can use oss in ubuntu? my sndcard doesnt seem to work with alsa
[03:06] <seb128> stratus: oups, sorry, there is a second bug about a crash
[03:06] <stratus> seb128, np
[03:06] <seb128> stratus: I don't have this problem neither
[03:07] <daniels> jdub: WOAH
[03:07] <stratus> seb128, hmm...
[03:07] <daniels> jdub: pia has blingmouth now?
[03:07] <jdub> daniels: hard core.
[03:07] <seb128> stratus: you're on i386 right ?
[03:07] <stratus> seb128, yes i'm.
[03:07] <seb128> stratus: ok, I'll follow upstream with the backtrace
[03:08] <stratus> seb128, thanks hope that helps the guys there.
[03:08] <daniels> jdub: ill
[03:08] <stratus> seb128, let me know if you need more debugging.
[03:08] <seb128> stratus: ok, thanks
[03:08] <daniels> jdub: but why is it not gold?
[03:08] <daniels> oh man
[03:08] <daniels> a tooth in the shape of a $
[03:08] <daniels> that would be so phat
[03:14] <Mitario> ello everyone
[03:19] <sjoerd> pitti: async mount is a property in hal's database
[03:20] <sjoerd> pitti: so our pmount hal wrapper to be will pass it to pmount
[03:20] <pitti> sjoerd: is there an easy way of changing the default property setting by g-v-m=
[03:20] <pitti> s/=/?/
[03:21] <sjoerd> pitti: no should there ?
[03:21] <sjoerd> pitti: i guess, that's where the persistent hal db props stuff comes into play..
[03:21] <pitti> sjoerd: well, it could be nice if users can easily configure this
[03:22] <sjoerd> pitti: currently i want to actually go back to the upstream g-v-m ui, so it's translated nicely (as far as it's translated in gnome cvs)
[03:23] <pitti> sjoerd: you mean without the two new options (Mount static drives on session start, Mount removable drives when hot-plugged)?
[03:24] <pitti> sjoerd: I already forgot again the difference between the second and the third option
[03:24] <pitti> sjoerd: it's too confusing, this must go away
[03:24] <sjoerd> pitti: only the first is patched into it
[03:24] <fabbione> pitti, sjoerd: i have a dbus/hal question
[03:24] <sjoerd> pitti: you will be able to configure in a hal fdi if you really want too
[03:25] <fabbione> do you have to minutes to dedicate to me?
[03:25] <sjoerd> pitti: i've sent a patch to gnome's utopia list for that some time ago..
[03:25] <pitti> fabbione: who? 
[03:25] <sjoerd> pitti: so the gui is default again
[03:25] <fabbione> both of you
[03:25] <sjoerd> fabbione: fire away
[03:25] <pitti> sure, for my part
[03:25] <fabbione> ok let's put in this way
[03:25] <fabbione> scenario:
[03:26] <fabbione>  /dev/input/myvolatilekeyb0 is a symlink to /dev/input/event0
[03:26] <fabbione> now if i unplug my keyboard and plug it back
[03:26] <fabbione> that might get associated to dunno.. eventX
[03:26] <pitti> Hi silbs
[03:27] <fabbione> is thare anything that hal and dbus can do to restore the link to the proper event device?
[03:27] <pitti> ugh
[03:27] <fabbione> i can see dbus/hal can see the event happening
[03:27] <pitti> what is setting the symlink in the first place?
[03:27] <fabbione> right now me manually
[03:27] <fabbione> i am just asking
[03:27] <fabbione> to understand if i need to write my own little daemon
[03:27] <sjoerd> that's a dbus/kernel issue mostly
[03:27] <fabbione> or i can use dbus/hal
[03:27] <sjoerd> argh, udev/kernel
[03:28] <pitti> AFAICS it should work to write a hal callout
[03:28] <sjoerd> pitti: not with hal running as non-root
[03:28] <pitti> if hal can tell the devices apart?
[03:28] <pitti> sjoerd: right, you need a suid-root wrapper hten
[03:28] <fabbione> it needs to run as root...
[03:28] <sjoerd> and having hal muck around in /dev/ is fugly, that's udev space imho
[03:28] <fabbione> or similar
[03:28] <pitti> fabbione: hal is actually not meant to _change_ anything in the system
[03:28] <sjoerd> you could probably make a udev script to do that
[03:29] <pitti> if udev can tell the devices apart, I'd prefer an udev rule
[03:29] <pitti> sjoerd: but does udev actually know which eventX belongs to which physical device?
[03:30] <pitti> fabbione: as long as there is a sysfs attribute that differs your keyboard from all the other stuff, it's easy
[03:30] <sjoerd> pitti: no but udev knows which eventX entry is gonna create.. though
[03:30] <pitti> fabbione: the rule can specify "%s{sysfs_attribute_name}" to match a particular device
[03:31] <fabbione> but that means that if i change keyboard type
[03:31] <sjoerd> fabbione: what's the important one btw which eventX it is or the symlink to the right eventX
[03:31] <fabbione> i need to update that entry?
[03:31] <fabbione> sjoerd: once i know what eventX is.. recrating the symlink is easy
[03:33] <sjoerd> fabbione: i ment, you don't need your keyboard to be event0 specifically ?
[03:33] <fabbione> sjoerd: not at all
[03:33] <fabbione> my keyboard is the "symlink"
[03:33] <fabbione> so i don't want to change it everytime i unplug or plug it
[03:34] <fabbione> i just want the symlink updated to the proper eventX
[03:34] <sjoerd> then it's just a way to have udev recognize what is your keyboard
[03:35] <fabbione> hmm
[03:35] <fabbione> i guess i will have to look into it
[03:37] <mirak> what is the difference between doing sudo and su -c ?
[03:40] <ddaa> authentication is handled differently
[03:40] <ddaa> sudo is more convenient and more flexible
[03:41] <mirak> oups sorry, wrong channel
[03:41] <mirak> but thanks for the answer
[03:43] <ddaa> mjg59: I'm zeroing on the "usb drive won't automont" problem.
[03:43] <ddaa> It works okay after a reboot, a ram-wakeup or a disk-wakeup.
[03:43] <pitti> sjoerd: is it just me, or did the g-v-m reaction to USB drive plugs slooooow in the last time also for you?
[03:44] <pitti> sjoerd: mounting is quite fast, but then it needs about 10 to 15 seconds to show nautilus windows
[03:44] <ddaa> But, it won't work after "ifdown eth1; rmmod ipw2200; modprobe ipw2200; ifup eth1" even if that's the only thing done after a reboot,
[03:44] <ddaa> If you want, I can try to figure out what's the precise command that's breaking.
[03:45] <sjoerd> pitti: i only have camera's to test with, so i don't really know
[03:45] <sjoerd> pitti: it used to be the wait on sysfs props
[03:45] <sjoerd> but it's a long time ago since i last profiled is
[03:46] <pitti> sjoerd: ah, I see the new property "volume.policy.mount_option.sync"
[03:46] <pitti> sjoerd: this should be read out by the pmount hal wrapper, right
[03:46] <sjoerd> yup
[03:48] <sjoerd> in storage-policy.fdi it's configured
[03:52] <pitti> sjoerd: okay, then we need to change g-v-m to pass the hal UUID instead of (or in addition to) the device name
[03:57] <sjoerd> yeah, uhm, maybe both, need to play with it a little
[03:58] <sjoerd> depends how/if pmount searches fstab :)
[04:03] <pitti> sjoerd: btw, what do you think about tarball.mk of cdbs? I like it a lot
[04:04] <sjoerd> hrm for my fixed disk ``fix'', gvm should search fstab already... seems that it needs to be worked out a little
[04:04] <pitti> sjoerd: it keeps the package dir clean, it prevents you from accidentially changing source files and you can use dbs-edit-patch
[04:04] <pitti> sjoerd: it means that the original source tarball is kept as it is
[04:04] <pitti> sjoerd: so, in hal-0.4.1/, you only have hal-0.4.1.tar.gz and debian/
[04:04] <pitti> sjoerd: and the actual building is done in debian/build-tree/hal-0.4.1/
[04:05] <pitti> sjoerd: just like dbs
[04:05] <sjoerd> so if you do apt-get source you get a tarball in it...
[04:05] <pitti> sjoerd: right
[04:05] <pitti> sjoerd: I like it and I wanted to ask you whether you know it and consider packaging hal with it
[04:06] <sjoerd> don't like that part of it
[04:06] <pitti> okay
[04:06] <sjoerd> but if the pro's are outweighing it
[04:06] <pitti> the nice thing is, you can mess around in the build-tree, try changes
[04:07] <pitti> and if you have found a solution, just wipe the build-tree and start to package it properly
[04:07] <pitti> and dbs has a nice script dbs-edit-patch which can be used with cdbs+tarball too
[04:07] <sjoerd> that sounds like how i use svn-buildpackage 
[04:07] <pitti> sjoerd: well, it probably needs a tryout
[04:08] <sjoerd> probably
[04:16] <Mitario> mvo_, here?
[04:23] <sivang> seb128 : was yelp bug fixed?
[04:24] <seb128> ELACKOFDETAILS
[04:24] <seb128> 'bug'
[04:24] <seb128> the only one on yelp ? :)
[04:24] <sivang> seb128 : could you send me again the nunber? I can't find on the b.u.c..
[04:24] <seb128> if you provide some details
[04:24] <seb128> we are several bug on yelp
[04:24] <seb128> so which one ?
[04:26] <seb128> sivang: [Bug 3557]  yelp seems unable to display some documents <- this one ?
[04:27] <seb128> (grrr people who ping you with no details on what they want, don't answer when you ask for details and run away)
[04:28] <sivang> seb128 : no , sorry! Yes, well mine was a duplicate of that, but I included the stdout output..
[04:30] <seb128> sivang: I don't have the same output
[04:30] <seb128> sivang: http://bugzilla.gnome.org/show_bug.cgi?id=157943
[04:31] <mvo_> Mitario: yep
[04:33] <sivang> sivang : Were you referring to the db2html.xml import / missing error?
[04:33] <sivang> seb128 : Were you referring to the db2html.xml import / missing error?
[04:34] <seb128> sivang: yes
[04:35] <sivang> seb128 : hmm, when I test it now, I see the some of the development docs are viewable, but most of the user stuff is not..I am able to reproduce this stdout error every time..
[04:35] <seb128> sivang: what do you call "user stuff" ?
[04:36] <seb128> sivang: in fact the bugzilla.gnome bug is from jhbuild
[04:36] <sivang> seb128 : almost everything under 'Desktop'
[04:41] <sivang> seb128 : and 'Applications'
[04:41] <mirak> I just realised that the ati drivers where removed because they depends of Xfree86
[04:42] <seb128> sivang: I get this in the console:
[04:42] <seb128> (yelp:6896): Yelp-CRITICAL **: get_doc_type: assertion `mime_type != NULL' failed
[04:43] <mirak> can this drivers work with xorg ? or is it just a dependcies problem ?
[04:44] <sivang> seb128 : hmm..mine's a bit different, (yelp:7520): GLib-GObject-CRITICAL **: g_object_unref: assertion `G_IS_OBJECT (object)' failed
[04:46] <lamont_r> seb128: apparently upgrading to hoary breaks the evo-launcher in the panel?
[04:46] <Mitario> mvo_, thanks for the apt upload :)
[04:46] <Mitario> mvo_, any word from gustav?
[04:46] <Mitario> (or whats his name again)
[04:46] <ChrisH> lamont_r: Even worse... there does not seem to be a working exchange plugin for the new version of evolution. :)
[04:46] <sivang> lamont_r : yes, I'm having this also...
[04:47] <sivang> Bunch of other things also broke..hoary is progressing faster then sid I gues..
[04:48] <seb128> lamont_r: should be fixed with gnome-panel 2.9.1-0ubuntu2
[04:49] <mvo_> Mitario: not yet
[04:49] <seb128> sivang: I've this one too but before getting an error while trying to view a doc
[04:49] <mvo_> Mitario: I'm still waiting ...
[04:49] <sivang> seb128 : ok, maybe I should talk with shaunm about this..
[04:49] <Mitario> mvo_, ok :)
[04:50] <seb128> sivang: good idea :)
[04:50] <Mitario> brb
[05:20] <Mitario> hi all :)
[05:21] <mvo_> hi Mitario 
[05:31] <Mitario> hmm, just tested gnome-app-install, works nice :)
[05:40] <mvo_> and looks nice!
[05:40] <mvo_> rburton did a great job 
[05:40] <rburton> cheers
[05:41] <Mitario> indeed :)
[05:43] <seb128> it doesn't work fine here
[05:43] <seb128> but it looks nice :)
[05:44] <Mitario> well header fonts could be a bit smaller, but i think it looks and works great! :D
[05:45] <seb128> it complains about my locales
[05:45] <seb128> and about not being root
[06:22] <mdz> morning
[06:24] <Astharot> hi mdz 
[06:24] <seb128> hello mdz 
[06:28] <mdz> GAH
[06:28] <mdz> I think procmail just ate all of my mail overnight
[06:32] <mdz> AARARRGGHH
[06:59] <stratus> mako!
[07:09] <mako> stratus: jeu tjere
[07:09] <stratus> mako, what? :|
[07:09] <mako> wait.. hey there!
[07:09] <mako> sorry.. fingers were off a row
[07:09] <mako> "_
[07:09] <mako> :)
[07:09] <mako> hah
[07:09] <stratus> mako, np. heh
[07:09] <ChrisH> Does anyone know about the status of the exchange-backend for the new evolution in Hoary?
[07:18] <mdz> seb128 or jdub probably would
[07:43] <mdz> daniels: if I could get copies of your maildir archives for the past 13 hours, that would be fabulous
[07:45] <daniels> what's the find recipe for ctime in the last 13h?
[07:47] <mdz> -ctime -13h
[07:47] <mdz> maybe not
[07:47] <mdz> I think it can only do days
[07:47] <mdz> oh, you can do fractional days I believe
[07:47] <mdz> nope
[07:48] <mdz> daniels: give me the past 24h, and I'll filter from there
[07:48] <daniels> daniel@tycho:~/Maildir% find ubuntu-users/ -ctime 0.6
[07:48] <daniels> find: invalid argument `0.6' to `-ctime'
[07:48] <daniels> ah, right
[07:48] <daniels> 'kay
[07:49] <daniels> do you want hoary-changes and warty-bugs?
[08:34] <Matt|> evening
[08:34] <Matt|> is firefox 1 coming in hoary?
[08:35] <ChrisH> Probably... as the release date will be 4/2005.
[08:38] <Matt|> in the devel tree?
[08:39] <thom> Matt|: there's an RC in the devel tree
[08:39] <Matt|> thom, hi
[08:39] <Matt|> thom, yeah it's rc1 isn't it?
[08:40] <thom> yes
[08:40] <tim1> current version in hoary is called 0.99+RC1-4ubuntu1
[08:40] <Matt|> thom, do you know if ff1 is gonna be in the tree sometime soon
[08:41] <Matt|> i'm hoping it will fix that flash problem :(
[08:41] <thom> Matt|: dude. it'll be in the tree before release
[08:41] <daniels> Matt|: it doesn't, the problem in this case is flash itself
[08:41] <Matt|> daniels, are you sure?
[08:41] <daniels> try export XLIB_SKIP_ARGB_VISUALS=1, then running firefox from a terminal
[08:41] <Matt|> ok
[08:41] <daniels> yes, xorg with composite exposes extra visuals and flash doesn't deal with that
[08:41] <daniels> because it sucks
[08:42] <Matt|> daniels, it was the same with the xfree packages
[08:42] <daniels> oh, right
[08:42] <Matt|> daniels, and i don't think i have composite
[08:42] <daniels> well, in that case, my trick won't solve it, and neither will ff1
[08:42] <Matt|> daniels, :(
[08:42] <Matt|> so you guys all have the same problem?
[08:43] <Matt|> my gentoo box is fine
[08:43] <daniels> which problem is that?
 Matt|: it doesn't, the problem in this case is flash itself
[08:43] <daniels> i made an assumption
[08:44] <Matt|> oh
[08:44] <Matt|> ;)
[08:44] <daniels> (you have no idea how many times that argb problem has been reported lately)
[08:44] <Matt|> oh ok
[08:44] <Matt|> :)
[08:44] <Matt|> bug 3302
[08:44] <Matt|> is my problem
[08:44] <Matt|> but its not assigned
[08:44] <daniels> my private mail can largely be summaried as 'composite is slow', 'FLASH DOESNT WORK HELP' and 'composite is slow'
[08:44] <Matt|> *laughs*
[08:45] <ChrisH> :)
[08:45] <Matt|> can you add them to your spam filter
[08:45] <Matt|> i'm chuffed with xorg btw
[08:45] <daniels> unfortunately mdz is right
[08:45] <daniels> we can't even begin to debug it without the source
[08:45] <daniels> the problem could be in firefox, glibc, flash, interactions thereof ...
[08:45] <Matt|> yeah
[08:46] <daniels> heh, i still at least make an *attempt* to answer all my mail :P
[08:46] <Matt|> is there no way to look at the builds on other distros?
[08:46] <daniels> private email gets answered unless I can't answer it on the spot so I flag it for follow-up and it disappears among the other 500 messages to have befallen the same fate
[08:46] <Matt|> i suppose it is guesswork :(
[08:46] <daniels> pure guesswork, since they're all so wildly different
[08:46] <Matt|> hmm
[08:46] <Matt|> damn
[08:46] <daniels> and the build environments (which versions of gcc, e.g.) are so totally different
[08:46] <Matt|> it's really irritating
[08:46] <daniels> yeah -- this is a no-source, no-debug problem
[08:46] <Matt|> yeah guess so
[08:47] <Matt|> daniels, but you can see why I hoped it might be fixed in ff1
[08:47] <Matt|> a guy in #ubuntu installed firefox using the mozilla installer and couldn't reproduce the problem
[08:47] <Matt|> i guess i'll do that
[08:48] <daniels> the official builds of mozilla are built with different compilers, different patches, different options, etc
[08:48] <daniels> so unfortunately even that doesn't narrow it down to a version-to-version difference
[08:49] <Matt|> yeah
[08:49] <Matt|> how annoying
[08:49] <Matt|> oh well
[08:49] <Matt|> life goes on
[08:50] <Matt|> great work on xorg btw
[08:50] <Matt|> it was such an easy update
[08:51] <daniels> cheers :) credit shared with fabbione
[08:51] <Matt|> i bet it was a couple of hardcore weeks huh?
[08:52] <daniels> we were both individually working on it before that, so it's been a while of xorg packages
[08:52] <Matt|> he came to visit you in the US?
[08:53] <daniels> nah, I live in .au, and I'm visiting him in .dk
[08:53] <Matt|> oh i c
[08:54] <Matt|> i must have you confused with someone else
[08:54] <Matt|> i thought you lived in california
[08:58] <daniels> no, not I.  although when X confcalls start at 3am and team meetings at 4am, it's enough to make me wish I was ;)
[08:59] <Matt|> *grins*
[08:59] <Matt|> whereabouts in australia do you live?
[09:03] <Matt|> *grins* i've installed firefox from mozilla.org and it has the same problem. Must be something really weird
[09:03] <daniels> Matt|: in Melbourne (east coast)
[09:03] <Matt|> ah nice
[09:06] <amu> someone interessed in testing the hoarylive with new the xorg ?
[09:06] <mdz> yes
[09:07] <amu> mdz: ..uploading   
[09:08] <mdz> amu: you still can't build CDs at the data center?  what is blocking that?
[09:10] <amu> mdz: right i can't, elmo didnt accepted to build in the chroot   
[09:10] <Mitario> re everyone
[09:11] <mdz> amu, elmo: this needs to be resolved.  I thought that the last time we talked about this, the solution was to provide the root-ish bits via a cron job, and allow the rest to be built unprivileged in a chroot, or something like that
[09:13] <wasabi> user model linux.
[09:13] <wasabi> mode.
[09:13] <tim1> anybody else not being able to change the wallpaper anymore ?
[09:13] <tim1> using hoary with newest packages
[09:14] <mdz> tim1: yes, see bugzilla
[09:15] <amu> mdz: or building them with a login by request ....
[09:15] <wasabi> why not user mode linux? It's pretty slick for that kind of stuff. ;)
[09:16] <tim1> ah, ok, search on wallpaper didn't return anything, now background does ..
[09:16] <mdz> wasabi: we've considered it, but it is by no means the simplest solution
[09:16] <wasabi> I thought it just ran as a program.
[09:16] <wasabi> Few commands to set up a clean fake file system, few commands to boot it with a custom script as init.
[09:17] <mdz> I have used UML a great deal
[09:18] <mdz> and for something which needs to be reliably automated, it is unfortunately not very suitable
[09:18] <wasabi> ahh.
[09:18] <mdz> I've written a tool which provides a certain level of automation via communication with a daemon which runs inside UML, but it it turned out to be extremely difficult to make it robust
[09:19] <wasabi> umlrun?
[09:19] <mdz> yes
[09:19] <wasabi> heh i was just reading about it.
[09:19] <mdz> I have a whole pile of ideas about how we could use UML for some Ubuntu projects
[09:20] <wasabi> I've always fancied the idea of using it for automated builds and such.
[09:20] <mdz> but it's mostly automation, so we need some infrastructure for that first
[09:20] <mdz> test builds, test installs, functional testing, etc.
[09:21] <LifesizeKenDoll> question - what could I do to help Ubuntu? I'm a reasonably competent programmer in quite a few languages
[09:22] <mdz> LifesizeKenDoll: take a look at /wiki/HoaryGoals and /community/bounties/
[09:22] <LifesizeKenDoll> alright
[09:22] <mdz> if you see a project that appeals to you, let me know
[09:23] <LifesizeKenDoll> can't find /community/bounties/
[09:23] <mdz> or if you have an idea of your own
[09:23] <mdz> http://www.ubuntulinux.org/community/bounties/
[09:23] <LifesizeKenDoll> whoops - I was going to ubuntu.org
[09:27] <zul> setting up a fs for uml under debian is realiativly easy
[09:28] <rcaskey> wowzers
[09:28] <rcaskey> 100k in bounties for 04
[09:28] <zul> hmmm?
[09:29] <ChrisH> Mark should set up bounties for getting rid of SCO and Microsoft, too. ;)
[09:31] <LifesizeKenDoll> that'd be funny more than anything
[09:31] <RubenV> www.gnome.org/bounties
[09:31] <RubenV> this is a nice bounties page :)
[09:31] <LifesizeKenDoll> I like that bounty page
[09:31] <mdz> RubenV: can you be a bit more specific about what you think is lacking?
[09:32] <mdz> I'm going to be editing it some more today anyway
[09:32] <RubenV> specific tasks and/or goals
[09:32] <RubenV> more like the gnome system
[09:32] <RubenV> but my opinion shouldn't be given too much value
[09:33] <RubenV> i don't have the skills to fix a bounty anyway :p
[09:33] <mdz> RubenV: apart from the python work, they are all quite specific
[09:34] <RubenV> aha, i see the page has been updated :)
[09:34] <RubenV> time for me to m
[09:34] <RubenV> learn decent C
[09:34] <mdz> I only added to it; I didn't modify the existing items
[09:34] <wasabi> so, any of you have any suggestions about what scripts I should use to run my own apt repository locally? with an incoming queue?
[09:35] <amu> mdz: 3h ETA
[09:36] <mirak> hi
[09:36] <mirak> I was toying with shfs mount and managed to autmount location with autofs
[09:36] <ChrisH> wasabi: apt-ftparchive?
[09:37] <wasabi> Chris, what and where is that?
[09:37] <mirak> it would be nice to be able to be able to have an automount behavior that can be setted up by users
[09:37] <wasabi> n/m.
[09:37] <mirak> I am not sure it's possible to put automount suid
[09:37] <ChrisH> wasabi: comes with Warty... see the man page. It helps you create all the meta files you need to get a working repository
[09:38] <wasabi> Chris will it read from an incoming queue and put .debs in teh right place?
[09:38] <wasabi> I want something easy to use. =/
[09:39] <ChrisH> wasabi: No, it just takes the packages from where they actually are.
[09:39] <ChrisH> wasabi: There was another program that might do what you want. What was the name...
[09:39] <wasabi> yeah i want an incoming queue. This is going to be for my company's office.
[09:39] <wasabi> Developers would like to upload programs, etc etc.
[09:40] <ChrisH> wasabi: you could try mini-dinstall although I find it ugly
[09:40] <wasabi> i think i've tried it previously, and also found it ugly.
[09:41] <wasabi> does archive.ul.org use the scripts from debina.org? katie and the bunch?
[09:41] <wasabi> debian.
[09:41] <ChrisH> That's what I would like to know, too. :)
[09:42] <ChrisH> I haven't met anyone in Debian who was able to install katie. So I wrote my own set of scripts for mentors.debian.net.
[09:42] <wasabi> heh.
[09:42] <wasabi> that's about where I am. I need the same functionality, but i'd like it apt-gettable.
[09:42] <wasabi> apt-get install coolaptthingy, edit a config file, and go with it
[09:43] <ChrisH> Rebuilding the meta files is rather easy. The scare part is checking if the package is valid (PGP key, linda, lintian, mandatory entries in the control file).
[09:44] <ChrisH> *scary
[09:44] <wasabi> yeah.
[09:45] <ChrisH> Do you have such a large number of packages that you really need different directories and the full checking?
[09:45] <wasabi> No. But it sounds fun.
[09:45] <wasabi> I just want it to be managable without sshing in and running stuff.
[09:45] <ChrisH> Oh, it is. :) Took a team of three people two months until it worked with all the packages we tried to upload. ;)
[09:45] <wasabi> I need it to be usable by people other than me.
[09:46] <ChrisH> Then just let the people upload the files (copy them somewhere) and run apt-ftparchive. Not much security checking but it works automatically.
[09:46] <ChrisH> Now I remember! "apt-move" could be worth a look, too.
[09:50] <ZlinxE> Does this channel support gnoppix?
[09:50] <mdz> wasabi: Ubuntu uses katie
[09:51] <mdz> isn't there a #gnoppix on this server?
[09:52] <mdz> amu would know if there is an official channel
[09:52] <ZlinxE> no there isnt
[09:55] <amu> #gnoppix it is 
[09:56] <ZlinxE> its empty
[09:56] <amu> the other possibility is on ircnet, some guys moved to there once a time :) 
[09:57] <mdz> daniels: what's the story with the "locale not supported by Xlib" stuff?
[10:02] <mdz> elmo: around?
[10:04] <ChrisH> mdz: katie? Hmmm. Looks like I will still need to learn Python, too.
[10:05] <mdz> ChrisH: that's a good idea anyway, especially in an Ubuntu context
[10:05] <LifesizeKenDoll> what is katie?
[10:05] <ChrisH> mdz: Whatever language you choose - it's the wrong one. And Perl has actually been fun for years.
[10:06] <ChrisH> LifesizeKenDoll: The program that handles package uploads to the archive.
[10:07] <Kyaneos> hi
[10:08] <ChrisH> mdz: Is Python more suited for scripting in Ubuntu? Or why was it chosen? (Besides that Mark seems to like it personally.) Large parts of Debian have been using Perl.
[10:08] <mdz> ChrisH: my point is that we do new development in Python wherever feasible
[10:08] <mdz> I'm not arguing your preferences, but that is where we are standardizing
[10:09] <mdz> ChrisH: take a little time to learn the basics, and see if you still have the same question :-)
[10:09] <ChrisH> mdz: I'm just curious what the advantages of Pythons are. That'll perhaps simplify it to switch for me. :)
[10:09] <mdz> it's a very nice language
[10:10] <mdz> you'll find it very simple to learn
[10:10] <mdz> I used to feel the same way; "perl gives me everything I need, why would I learn python?"
[10:11] <ChrisH> Just took a closer look at katie. And besides that I'm not used to the syntax it looks quite readable.
[10:11] <mdz> I no longer write perl programs except one-liners on the command line
[10:11] <LifesizeKenDoll> I heart Python
[10:11] <ChrisH> There are a couple of questions I need to answer for myself. Like... is there equally as good support in terms of loadable modules as in Perl. Looking at search.cpan.org is really overwhelming.
[10:12] <ChrisH> But there's still time until December 5th. :)
[10:12] <LifesizeKenDoll> what happens then?
[10:12] <ChrisH> http://www.ubuntulinux.org/wiki/Conference/
[10:12] <LifesizeKenDoll> ah
[10:12] <mdz> ChrisH: loadable modules meaning extensions written in C?  or pure python modules?
[10:13] <ChrisH> I better leave my Perl t-shirt at home. :)
[10:13] <mdz> in both cases, the answer is "yes, and in many ways nicer than in Perl" :-)
[10:13] <ChrisH> mdz: Hmm. Both.
[10:13] <ChrisH> mdz: Argh. I knew you were going to say that. :)
[10:14] <ChrisH> mdz: Is there a similar agreement on repository systems? In the Doc Team we have been discussing whether to use Subversion or Arch. Well... Subversion runs well so far for the new FAQ. And that topic /really/ seems to be religious.
[10:15] <mdz> perl:                     char *handle= SvPV(ST(i),len);
[10:15] <mdz> python:    PyObject *Obj = PyString_FromString(Str);
[10:15] <mdz> ChrisH: yes, we are converging on arch
[10:17] <mdz> ChrisH: we freely admit that arch isn't yet as usable as it needs to be, but that is changing
[10:18] <ChrisH> mdz: I read an interview on /. from the arch author and just thought "yeah, bash everything else allright". Didn't really make me want to try it out yet.
[10:25] <mdz> ChrisH: an idea cannot be held responsible for those who believe in it :-)
[10:26] <ChrisH> mdz: Sounds a little esoteric. :)
[10:27] <mirak> hey
[10:27] <mirak> would this be possible to unlike ubuntu desktop from xserver ?
[10:27] <mirak> unlink
[10:27] <mirak> because I must keep xfree86 instead of xorg in hoary
[10:27] <Mithrandir> oh, why?
[10:27] <mirak> however unbuntu desktopn absolutely want xorg
[10:28] <mirak> because ATI drivers
[10:28] <mirak> I don't know if this drivers can work with xorg, but the deps prevent me to use this drivers
[10:28] <mirak> so either I tweak the deps
[10:28] <mirak> or I put back xfree86
[10:29] <Matt|> mirak, your card is not working with xorg?
[10:29] <mirak> I don't know yet
[10:29] <Mithrandir> we'll fix that before hoary releases, I hope, so we have fully-working ati binary drivers.
[10:29] <mirak> all I know is that the ati deb packege depends on xfree
[10:30] <Matt|> mirak, maybe with xorg you won't need the ati binary drivers for your card to work
[10:30] <mirak> maybe adding the deps on xorg would be enough
[10:30] <mirak> my card work
[10:30] <mirak> without the drivers
[10:30] <mirak> but there is no hardware accelaration for 3D
[10:31] <Matt|> mirak, are you sure?
[10:31] <Matt|> there is on mine
[10:31] <mdz> mirak: just remove ubuntu-desktop if it doesn't match what you want
[10:31] <mdz> it's only there for convenient upgrades
[10:32] <Matt|> can i ask a support question in here? I've had no luck in #ubuntu
[10:32] <mirak> mdz: yes. but having the deps fixed on the ati driver package would be nice also
[10:33] <mdz> mirak: the dependencies on the ati driver package are correct
[10:33] <mdz> the driver does not work with the X.org server, I don't think
[10:33] <Matt|> mirak, i think the reason for those deps is that xorg includes opensource drivers for most ati cards. give it a try
[10:33] <mirak> mdz: I have it works with 6.7 but not 6.8 or something like this
[10:34] <mirak> Matt|: well I am already trying it then
[10:34] <mirak> Matt|: mmm I have not restarted X
[10:34] <mirak> yet
[10:34] <mirak> I will see if it's ok
[10:36] <sivang_away> mdz : Hi
[10:37] <mdz> sivang: hi?
[10:42] <ChrisH> mdz: Okay, now you got me. I'll stay awake tonight and learn Python. :)
[10:42] <mdz> ChrisH: the tutorial on the python site is good, diveintopython may be even better (I haven't read the latter, but hear good things)
[10:43] <sivang> mdz : I was wondering about gnome-system-tools, I am working with garanacho on extending the perl backend and the frontend to allow for our default group priviliges, will it be obsoleted by fedora's UserConfig tool ? :)
[10:44] <sivang> mdz : diveintopython is very good indeed.
[10:44] <daniels> mdz: i don't know.  been having a look at it but it's hairy.
[10:44] <mdz> sivang: hasn't Martin Pitt already done that?
[10:44] <mdz> daniels: is there a bug open in bugzilla yet?
[10:45] <sivang> mdz : he had a patch for the frontend that made it on warty, however the patch was dropped with the new gst in hoary
[10:46] <sivang> mdz : the new way to solve this is more general, and will (hopefully) be in upstream and support even other distros.
[10:47] <daniels> mdz: yah
[10:49] <mdz> sivang: ok
[10:49] <mdz> sivang: if you want to rewrite the backend in python at the same time, that would be ideal ;-)
[10:51] <sivang> mdz : will this keep gst in ubuntu? I will talk garancho into it then....:)
[10:51] <mdz> keep? we are not removing it
[10:55] <ChrisH> sivang: Come over for a beer. We'll learn Python together.
[10:57] <sivang> ChrisH : Sure, no prob. Wait up, I will use my teleporter :)
[10:57] <ChrisH> mdz: btw, thanks for replying to my "mirror problem" on -devel. We managed to get it installed. I've been mirroring the whole archive since yesterday.
[10:58] <sivang> mdz : ok, so when userconfig will be introduced, it will be offered as another tool to do the job?
[10:58] <ChrisH> mdz: If we will really exceed the traffic limit we'll just shape it down.
[10:58] <mirak> sivang: what is userconfig ?
[10:58] <jdub> sivang: no decision has been made on any of that stuff so far, and currently only libuser has been raised.
[10:58] <mdz> ChrisH: glad to hear it
[10:58] <mdz> sivang: as jdub says
[10:59] <ChrisH> mdz: Colin didn't reply yet to add the mirror. But I'll be patient.
[10:59] <mdz> ChrisH: Colin?  or James?
[10:59] <ChrisH> mdz: Err.
[10:59] <mdz> ChrisH: did you add it to the wiki already?
[10:59] <ChrisH> mdz: James. :)
[11:00] <ChrisH> mdz: I don't seem to have edit access on the mirrors page.
[11:00] <mdz> ChrisH: the wiki page is editable by anyone
[11:01] <mdz> http://www.ubuntulinux.org/wiki/Archive
[11:01] <ChrisH> mdz: I thought about https://www.ubuntulinux.org/download/
[11:02] <ChrisH> mdz: As that's the page where people will probably try to get the CD from.
[11:02] <sivang> ChrisH : don't we have write access there already?
[11:03] <ChrisH> sivang: Looks like that area is restricted to Canonical :(
[11:07] <mdz> ChrisH: it is restricted to website managers, not Canonical
[11:08] <mdz> ChrisH: add your mirror to /wiki/Archive as a first step
[11:08] <ChrisH> mdz: Yes, Sir! :)
[11:09] <mdz> ChrisH: the mirrors on /download/ are CD image mirrors (only)
[11:09] <mdz> the wiki Archive page is the only list of package archive mirrors that we have at the moment
[11:09] <mdz> it should probably be cleaned up and moved someplace more official
[11:09] <mdz> or rather, linked
[11:09] <ChrisH> mdz: source.rfc822.org is a full mirror and also mentioned on the download page.
[11:10] <mdz> ChrisH: the URL on that page is only a CD image mirror
[11:10] <ChrisH> mdz: so would be http://ubuntu-mirror.workaround.org/4.10/
[11:10] <mdz> none of the URLs on /download/ point to packages
[11:11] <mdz> ChrisH: excellent, then it can be added to both sections on the Archive page
[11:11] <ChrisH> mdz: I'm on it.
[11:12] <mdz> ChrisH: I'm not sure if there are any established criteria for adding CD image mirrors to the download page, so I'm leaving it to the website people to edit it.  If you send me an email reminder, I'll pass on your request to the appropriate people
[11:13] <ChrisH> mdz: Sure. Thanks. Let's see how much the line will burn. :)
[11:21] <wasabi> heh. mini-dinstall looks like it doesn't accept "warty" distribution name.
[11:33] <ChrisH> wasabi: Did you try "apt-move"?
[11:34] <wasabi> I got mini-dinstall working
[11:34] <wasabi> Looks fine. I just won't look at the code.
[11:34] <wasabi> Doesn't do pools though.
[11:34] <wasabi> =(
[11:34] <wasabi> doesn't have an interface of any sort for removing packages from the archive either it looks like.