[12:03] <zenwhen> eas a change made to font renering between rc and final?
[12:03] <zenwhen> was*
[12:03] <zenwhen> i cant get bitstream vera sans to not look like crap since I upgraded yesterday
[12:10] <pitti> good night everybody
[12:23] <Kamion> mdz: maybe we should post a call for awful hacks :-)
[12:25] <sladen> invite me :)
[12:26] <ajmitch_> hi azeem_ 
[12:27] <azeem_> hi Andrew
[12:35] <mdz> Kamion: fine with me
[12:35] <mdz> Kamion: I think you produced awful hacks because you had a full schedule
[12:36] <mdz> in addition to any inclination to report them ;-)
[12:42] <Kamion> mdz: probably accurate ...
[12:42] <mdz> casper has its share of awful hacks; I just haven't gotten around to adding them yet
[12:43] <Kamion> as does kickseed, for that matter
[12:43] <Kamion> but they're not very urgent to undo
[12:43] <Kamion> and in fact some of them may be further propagated ;)
[01:16] <mdz> ogra/seb128: is #8339 a GTK issue or an hwdb-client issue?
[01:16] <jcole> i need help with .udebs
[01:17] <jcole> for the live cd i'm trying to remaster wit ha different kernel
[01:17] <jcole> is there a way to convert a .deb to a .udeb?
[01:17] <mdz> doko: are you sure about #145430?  it doesn't look like a gcc-3.4 bug
[01:18] <mdz> jcole: no, they are quite different things
[01:18] <mdz> jcole: to produce the udebs, build the linux-source-2.6.10 source package
[01:18] <jcole> mdz: i saw kernel-wedgefor kernels, but what about making udebs of other apps?
[01:20] <jcole> mdz: i'm guessing i have to apt-get debian-installer and build stuff from source
[01:21] <mdz> jcole: no, you don't need debian-installer
[01:21] <mdz> jcole: what is it that you want to accomplish?
[01:21] <mdz> if it is to use a different kernel on the live CD, you only need to build linux-source-2.6.10, and that will give you udebs
[01:23] <jcole> mdz: i'm trying to something unnatural
[01:23] <mdz> doko: also #6820?
[01:24] <jcole> mdz: installing an fc3 kernel on the ubuntu live cd (so a certain proprietary app will work)... works on a regular install of ubuntu just fine, believe it or not
[01:25] <mdz> jcole: that will be complicated for the live CD
[01:25] <mdz> doko: also #93991
[01:26] <mdz> doko: also #253096. how did you generate this list?
[01:28] <jcole> mdz: but is it possible... :)
[01:28] <mdz> jcole: of course
[01:35] <jcole> mdz: does the live cd actually use the kernel in /boot in to cloop? or does it get it from a udeb? ... does it boot the /boot/ cloop kernel and load the modules from the udebs? (during the probe for hardware)
[01:35] <mdz> jcole: it doesn't use the kernel in the cloop, only the modules
[01:36] <lamont> jcole: it does, however, use the modules
[01:36] <jcole> ah
[01:36] <mdz> it boots using the kernel in /install, unpacks the modules from the udebs and uses those for hardware detection, then loads any remaining modules from the cloop while it's booting
[01:37] <mdz> the modules in the cloop must generally match those in the udebs
[01:37] <mdz> for sanity's sake
[01:38] <mdz> jcole: swapping in the CD kernel is easy, swapping in the cloop modules is also easy.  the complicated bit is making udebs out of the modules; we don't have a tool to make udebs from an arbitrary kernel
[01:38] <mdz> they're built from our kernel source
[01:39] <mdz> our kernel source and kernel-wedge (together) contain the lists of which modules go in which udebs
[01:40] <mdz> if you build a .deb out of your kernel, you could possibly use kernel-wedge to slice it up into udebs
[01:40] <lifeless> mdz: is there kernel-putter too ?
[01:41] <mdz> lifeless: no, but we have many kernel drivers
[01:41] <lifeless> )
[01:45] <jcole> mdz: sounds like if i hack up install/initrd.gz just right i may get this to work
[01:47] <jcole> mdz: along with putting the right install/vmlinuz and boot/vmlinuz (cloop)
[01:51] <jcole> after this, it will prove that i am brave, crazy, and/or stupid
[01:56] <mdz> jcole: if you can fit all of the modules you need in the initrd, yes, that could work
[01:56] <mdz> there is a size limit, which varies per architecture
[01:57] <jcole> mdz: when linux boots and searches for modules to load.... how is that different than d-i searching for modules to load?
[02:00] <jcole> mdz: seems a bit redundant, doesn't it?
[02:01] <mdz> jcole: d-i loads the modules from external locations and unpacks them into the filesystem, where LInux can find them
[02:04] <jcole> when d-i boots, modules are loaded from the install/initrd.gz and the udebs?
[02:06] <jcole> after all that, when booting into the cloop, modules are then loaded from /lib/modules/2.6.10-5-386?
[02:08] <mdz> jcole: the initrd contains a minimum set of modules needed to detect the CD
[02:09] <mdz> jcole: the remaining modules are loaded from the udebs once things are up and running
[02:09] <mdz> jcole: the udebs contain the minimum set of modules needed to detect the network, disks, etc.
[02:09] <mdz> jcole: the cloop has a complete set of modules (sound drivers, etc.)
[02:10] <jcole> mdz: oh ok... this is a bit complicated
[02:10] <mdz> jcole: what did I tell you? ;-)
[02:10] <jcole> :)
[02:10] <mdz> jcole: if you want to do it quick and dirty, you can probably get what you need by adding your modules to the initrd and the cloop
[02:11] <mdz> the installer will still unpack the udebs of course, but they'll be ignored (assuming your kernel isn't version 2.6.10-5-386)
[02:15] <jcole> mdz: what would happen if i put *all* modules from /lib/modules/2.6.xxxx into the initrd.gz?
[02:15] <jcole> mdz: attempts to load them all?
[02:16] <mdz> jcole: it will probably end up loading the hotpluggable ones
[02:16] <mdz> but that should be harmless
[02:16] <mdz> I'd be more worried about the size of the initrd
[02:16] <mdz> 55M     /lib/modules/2.6.10-5-k7
[02:19] <jcole> mdz: one last question ;) what does d-i use for hardware detection?
[02:22] <mdz> jcole: in Ubuntu 5.04, it uses hotplug
[02:23] <jcole> mdz: cool, that's good
[02:23] <jcole> mdz: thanks for your help
[02:23] <jcole> mdz: i will report here my progress tomorrow
[02:23] <mdz> jcole: good luck
[02:23] <jcole> like you guys care, but still
[02:24] <mdz> jcole: if you succeed, it would be nice to have a document in the wiki explaining the procedure
[02:24] <mdz> (hint) ;-)
[02:24] <jcole> gotcha
[02:24] <zul> heh...yay! slashdot fud
[02:25] <Lathiat> haha
[02:32] <Amaranth> mdz: PING?
[02:32] <mdz> ?
[02:33] <Amaranth> you closed https://bugzilla.ubuntu.com/show_bug.cgi?id=9048 but as a dupe but i still have no idea how to fix it :)
[02:33] <Amaranth> without manually editting /etc/environment
[02:36] <mdz> Amaranth: you can reinstall, or you can edit /etc/environment
[02:36] <Amaranth> *sigh*
[02:37] <mdz> editing /etc/environment would be the better plan
[02:37] <jba> hi guys, not sure if this is a developers question but not getting many useful answers in #ubunut. I can't seem to find the Schedule (i.e. timings) for the ubuntudownunder conference
[02:37] <Amaranth> what should it be set to?
[02:37] <jba> I live in sydney and would love to attend, but I need to know if I can make it after work?
[02:37] <mdz> jba: that's because there isn't one yet
[02:37] <jba> mdz thanks dude
[02:37] <jba> i'll check in closer to the dates then?
[02:38] <mdz> jba: if your question is whether things will still be happening in the evening, the answer is yes
[02:38] <jba> i'm a mono/pptpclient hacker so don't really have much to contribute, but i hafve been using/loving ubuntu for a while now
[02:38] <mdz> though we don't yet have a schedule showing what is happening when
[02:39] <jba> well I'd still love to attend, it's just that I have a two month old son, with whom i tend to spend all of mytime
[02:40] <jba> so if it's gonna be stuff like "kernel optimisation processes" in the evening, then I don't think I'll make it :)
[02:42] <mdz> there will be parallel tracks
[02:42] <mdz> one of them will be for Ubuntu development plans, and will be very technical
[02:42] <mdz> there will also be an Ubuntu community track, which will be less technical, happening at the same time
[02:42] <mdz> so there will always be something in which to participate
[02:43] <jba> cool
[03:12] <Nigelenki> I suppose I'm not going to find disaster recovery utils on the live/install CD
[03:12] <Nigelenki> should i submit a <bug> requesting that the LiveCD have tools added to it by default to detect and recover partitions incase i.e. WinXP system managment console decides to erase my partition table when i ask it to add an extended logical drive in 37.4 gigs of unpartitioned space in my extended partition?
[03:13] <Nigelenki> *cough*
[03:13] <Nigelenki> or would this be an "enhancement"?
[03:13] <GoneBoB> there is a tool to do that
[03:13] <GoneBoB> it's an enhancement
[03:13] <Nigelenki> oh sweet, which one
[03:13] <GoneBoB> you can file a bug with priority enhancement
[03:14] <GoneBoB> I can't remember offhand, but it saved me a few times :)
[03:14] <Nigelenki> heh
[03:17] <GoneBoB> http://www.cgsecurity.org/index.html?testdisk.html that looks like it will do it but it's not what I used
[03:17] <Nigelenki> i'll try gparted and stuff
[03:18] <mdz> Nigelenki: it already contains parted
[03:18] <mdz> which is one of the best tools for that kind of work
[03:18] <Nigelenki> mdz:  I know it has parted, but I *cough* don't like the parted cli (rather, I like it but I don't yet know how to use it)
[03:19] <GoneBoB> mdz: last time I checked parted wont recover a shagged table
[03:19] <Nigelenki> damn.
[03:19] <Nigelenki> I know a mandrake install cd will do it
[03:19] <Nigelenki> "Rescue partition table"
[03:19] <mdz> GoneBoB: see the 'rescue' command in parted
[03:19] <Nigelenki> mdz:  :)
[03:19] <GoneBoB> excellent :)
[03:19] <mdz> gpart is also good in such situations; that one is not on the live CD
[03:20] <GoneBoB> yes, gpart was what I was talking about
[03:20] <Nigelenki> mdz:  30 points to you.  Gparted however would be ideal, as well. . . even sysadmins are spoiled these days (look at MS, doing away with CLI in favor of GUI utils o_o).
[03:21] <GoneBoB> Nigelenki: knowing how to use the CLI means you generally understand how it works
[03:21] <GoneBoB> which is much better
[03:21] <Nigelenki> GoneBoB:  I tend to like understanding how things work
[03:21] <Nigelenki> it's easier for me to work with something when i know what I'm doing.
[03:22] <Nigelenki> like when I code, I code thinking about what's going on with stuff in memory, how things are being moved, locking, what things are affected by what changes, and how the changes are done in terms of steps taken by the CPU (not 1:1 instructions but closer than high level code; helps to optimize C code)
[03:22] <Nigelenki> because I love easy to use, intuitive interface; but I really REALLY want to know what it is I'm doing
[03:23] <Nigelenki> anyway, too much white noise I guess, though this place was inactive when I got here :)
[03:29] <Nigelenki> ok gparted doesn't have a rescue option
[03:31] <Nigelenki> parted CAN'T find anything
[03:31] <Nigelenki> gpart is finding stuff.
[03:32] <calc> 11.44TiB in 4 days on torrents, not bad
[03:35] <lamont> calc: your mirror?
[03:35] <calc> the t.u.c
[03:36] <lamont> ah, yes
[03:36] <thom> lamont: torrent.ubuntu.com:6969
[03:36] <lamont> yea
[03:38] <calc> are there md5sums for http://torrent.ubuntu.com/releases/hoary/release/dvd/ ?
[03:49] <wasabi> We really need a protocol as simple as windows file sharing. Just without the suck.
[03:54] <Nigelenki> ok well didn't work :(
[03:54] <Nigelenki> I didn't lose anything important
[03:54] <Nigelenki> I'll jsut reinstall.
[03:56] <Lathiat> wasabi: windows file sharing is hardly simple :)
[03:56] <Lathiat> wasabi: webdav+mdns-sd = simple :)
[03:57] <Lathiat> and thats pushing it :)
[03:57] <wasabi> webdav is missing a lot really.
[03:57] <wasabi> It is simple from the user/config point of view.
[03:57] <Lathiat> in what way
[03:57] <wasabi> Any machine can access any other (given auth constraints).
[03:57] <wasabi> And have full control: ACLs, extended attributes.
[03:57] <wasabi> Etc.
[03:57] <Lathiat> sure thats why you use authentication....
[03:57] <wasabi> I mean *SET* ACLs
[03:58] <Lathiat> you could do that with webdav
[03:58] <Lathiat> possibly not with apache
[03:58] <Lathiat> don't know anything about extended attributes
[03:58] <wasabi> I suspect WebDAV isn't very good about locking.
[03:58] <wasabi> Not as Just Worky as CIFS anyways.
[04:02] <Lathiat> wasabi: well, thats an entirely different concern
[04:02] <Lathiat> something that should be looked at i guess
[04:03] <paulproteus> wasabi: It depends on the server-side implementation.
[04:43] <fabbione> morning
[04:43] <fabbione> mornin2
[04:43] <jnc> the new protocol.  mornin2mornin
[04:52] <robitaille> fabbione,  in case you haven't noticed,  bug 7078 has been re-opened today...  I think it was an old favourite of yours
[04:53] <fabbione> robitaille: none of the above
[04:53] <fabbione> gamin is buggy
[04:54] <fabbione> really badly buggy
[04:54] <zul> and on the sucketh meter its betwen 1 and 10
[05:01] <thom> do we have breezy love yet?
[05:02] <jsgotangco> oohhh
[05:02] <zul> not as far as i know
[05:08] <maswan> ah, finally some stats:
[05:08] <maswan> http://www.acc.umu.se/technical/statistics/ftp/index.html.en
[05:09] <maswan> note that those lack the redirected install and live cds for i386
[05:10] <Lathiat> 3.1 million files?
[05:10] <thom> maswan: nice
[05:10] <Lathiat> oh thats -
[05:10] <Lathiat> so like 95000 kubuntu install cds
[05:11] <Lathiat> yowzas
[05:11] <Lathiat> and 90,000 - the redirected once
[05:12] <calc> so is breezy not really setup for use yet? i see the dirs there but no Packages files
[05:12] <maswan> the redirected ones are 4.6TB by now
[05:12] <zul> calc: i would say no
[05:13] <maswan> and does not turn up in that ftp stats page
[05:13] <Lathiat> i don't expect breezy to get into swing for another 3 weeks
[05:13] <Lathiat> altho i suspect breezy will open up sometime this week
[05:18] <sladen> maswan: 9 petabytes ?
[05:21] <sladen> terabyte even
[05:25] <maswan> sladen: 14 terabytes if you include those additional 4.6 and the smaller ubuntu directories of a few hundred gigs
[05:50] <crb> Hi all. 
[05:51] <crb> I'm trying to remaster an ubuntu installer cd with specific packages/preseed.  Started with a copy off the CD, but the installer is complaining that it's not an Ubuntu CD.
[05:51] <crb> Any ideas why?
[05:51] <crb> (well, it's obvious why; any word on how d-i knows :)
[06:25] <womble> crb: Missing the .info directory (or whatever it's called) in the root of the CD?
[06:32] <crb> womble: think thats it. *slaps rsync*  Cheers for that!
[06:39] <fabbione> mdz: ping?
[07:45] <pitti> Good morning
[07:48] <thom> hey pitti
[07:52] <infinity> Hey pittithom.
[07:57] <pitti> Hi infinity! How's mozilla going?
[08:00] <infinity> Well enough.  Right now, I'm booking flights to Sydney, though.
[08:01] <thom> infinity: you definitely UDUing then?
[08:01] <mdz> fabbione: pong
[08:01] <infinity> And my visa expires in two days.  Immigration and I have another apointment tomorrow to fix that.
[08:01] <jdub> i drank a lava lamp
[08:01] <jdub> it wasn't lava
[08:01] <infinity> thom : Apparently I am, yes. :)
[08:01] <thom> infinity: ouch.
[08:01] <ajmitch> not good
[08:02] <infinity> Could be worse, I guess.
[08:02] <ajmitch> worse as in already expired?
[08:03] <infinity> Yeah. :)
[08:03] <infinity> When I go in tomorrow and give them 1800+ AUD, they should give me a bridging visa while we sort out my de facto spousal visa thing.
[08:03] <infinity> Or something.
[08:04] <infinity> I've been told not to worry.  So, I'm worrying.
[08:04] <thom> heh
[08:04] <infinity> Which seemed a resonable thing to do.
[08:04] <thom> yeah, seems fair
[08:07] <ajmitch> flatmate only got his visa renewed on the day it expired
[08:07] <fabbione> hey pitti
[08:07] <pitti> Hey pinh^Wfabbione
[08:08] <infinity> thom : What time/day do you arrive in Sydney?
[08:09] <thom> sydney? sunday i guess
[08:12] <fabbione> thom: are you still working on the readhead stuff?
[08:12] <fabbione> and boot speed optimization?
[08:13] <pitti> elmo: ping
[08:14] <thom> fabbione: oui
[08:15] <fabbione> thom: you are going to have some extra fun with 2.6.12
[08:15] <thom> oh?
[08:15] <fabbione> thom: they added support for timestamps on printk
[08:15] <fabbione> so you can see exactly how long it takes to boot the kernel
[08:15] <thom> oh lordy
[08:15] <fabbione> and how long certain operations take
[08:17] <infinity> Well, that was a necessary feature to add in a stable branch.
[08:17] <pitti> infinity: 2.6.x is not stable any more :-/
[08:17] <jdub> Annodex talk coming up soon: http://150.203.247.2:8810/ (theatre), http://150.203.247.2:8811/ (screen) -> discuss in #annodex on freenode
[08:17] <infinity> pitti : Was it ever?
[08:17] <pitti> infinity: no, 2.4 was
[08:17] <infinity> pitti : I was using the term "stable" lightly.
[08:18] <pitti> infinity: I think it's a shame that they don't have a stable branch any more ... :-(
[08:18] <infinity> Well, they do.  2.4 still is.  Heck, so is 2.2.
[08:18] <infinity> But 2.6's patch acceptance policy does seem a bit 'out there'.
[08:18] <Burgundavia> infinity, stable and useful was what I think they were looking for
[08:18] <pitti> this reminds me of stable woody vs. unstable sid...
[08:18] <thom> a _bit_?
[08:19] <pitti> certainly not on a desktop?
[08:19] <infinity> The only really desktopish desktop I have is the laptop here, which is running a stock ubuntu 2.6 kernel right now, but probably shouldn't be.
[08:20] <pitti> noooooo! a OO.o vulnerability *sigh*
[08:20] <infinity> Or, at least, it needs some talking to.  It really doesn't get a long with the ATA drivers in the current hoary kernels.  It's happy with old 2.4 stuff.
[08:20] <Mithrandir> pitti: nooooooooooo!
[08:20] <Mithrandir> :((
[08:20] <thom> pitti: aaaaiiiieee
[08:21] <pitti> http://www.securityfocus.com/archive/1/395516
[08:23] <infinity> At least it's an easy bug to fix.
[08:23] <pitti> infinity: yeah, it's trivial, but it still requires the users to download shitloads of MB
[08:24] <infinity> And there, my flights are confirmed.
[08:24] <pitti> a friend of mine (modem user) suggested "patch" debs, just like the rpm counterpart
[08:24] <pitti> infinity: cool
[08:24] <infinity> pitti : Mail then new CDs. :)
[08:24] <infinity> s/then/them/
[08:25] <infinity> Too late for patches to work, you would have had to have the openoffice deb depend on an empty "openoffice-security-patches" metapackage or something.
[08:25] <infinity> At least, to get automatic updates to work.
[08:26] <pitti> no, I mean patch debs in general
[08:26] <pitti> not oo.o specific
[08:26] <infinity> In general, it's quite a shift.
[08:33] <spo0nman> do you guys manage your own packages? or just re-package stuff you get from gnome/other places ?
[08:36] <Burgundavia> spo0nman, most of Universe is built out of debian sid
[08:36] <Burgundavia> ubuntu is about to sync to debian sid
[08:36] <Burgundavia> there is also some packages for the stuff at apt-get.org repos
[08:37] <spo0nman> hmm ok, 
[08:37] <Burgundavia> but the stuff in main is mostly Ubuntu specific stuff, though Ubuntu tries to pass stuff back upstream as much as possible
[08:37] <infinity> (And we also have many ubuntu-specific packages, and we often have newer versions of the Debian packages than Debian has...)
[08:37] <infinity> spo0nman : To answer your original question: "Yes, both".
[08:38] <spo0nman> hmm ok.
[08:38] <rubenv> infinity: how's the PHP5 work going?
[08:38] <rubenv> and goodmorning :-)
[08:38] <spo0nman> to intorduce myself, Iv been a debian user since the last 8 years and have done some level of gtk programmin to be at bug fixing level.
[08:39] <infinity> rubenv : It's currently a "spare time" thing, but someone sent me some pretty comprehensive patches, so it's not much work for me to polish it up.
[08:39] <spo0nman> In warty thereis  no software to burn cds? does nautilus burn:/// work well I havent tried it yet fund about it only today.
[08:39] <infinity> rubenv : When breezy opens up, I'll upload it at some point, and get php4 and php5 to swap seeds (putting php5 in main for breezy)
[08:40] <rubenv> infinity: well, I've sent you 2 or 3 patches, I hope they still apply to 5.0.4
[08:40] <Mithrandir> infinity: uhm, you're going to remove php4 from main for breezy?
[08:40] <infinity> spo0nman : If you have "universe" in your sources.list, you should have access to pretty much everything you would have had on a Debian system.
[08:40] <infinity> Mithrandir : Well, that or leave php4 and php5 in. Whichever seems saner.
[08:41] <Burgundavia> universe is all free software that is not supported by canonical
[08:41] <rubenv> spo0nman: universe are basically packages done by the motu's, not the main supported packages
[08:41] <infinity> Mithrandir : Dropping php4 to unsupported seems more reasonable, from a "not causing Martin too muhc unwaranted pain" perspective.
[08:41] <rubenv> universe is a polished pul in of debian sid
[08:41] <spo0nman> another question, can i make a mix match in the source-list of ubuntu and debian server .... BAD things will happen?
[08:41] <Mithrandir> infinity: well, I'm going to hate you moderately for that, since that means we can't upgrade to breezy for a while.
[08:42] <womble> rubenv: Where does the "polished" bit come into it?
[08:42] <crimsun> spo0nman: wiki/UniversePackages
[08:42] <rubenv> womble: motu
[08:42] <infinity> Mithrandir : "we"?
[08:42] <jdub> Annodex talk is on *now*: http://150.203.247.2:8810/ (theatre), http://150.203.247.2:8811/ (screen) -> discuss in #annodex on freenode
[08:42] <Mithrandir> infinity: hardware.no
[08:42] <Mithrandir> infinity: being a medium-to-large PHP shop, it's not trivial to port to php5
[08:42] <infinity> Mithrandir : Can't move to php5 at hardware.no, or don't trust the packaging to stabilise?
[08:42] <infinity> Mithrandir : The former, I see.
[08:42] <Mithrandir> infinity: former, at least yet.
[08:43] <infinity> Mithrandir : The last place I worked, I spent a large portion of last year porting their stuff to php5, so they could move "eventually".
[08:43] <infinity> Mithrandir : No one's stopping you from running universe packages at hardware.no, are they?  I'm not going to petition to drop php4 completely, just demote it.
[08:44] <rubenv> infinity: if the patches i've sent you don't apply anymore, don't mind poking me to fix them
[08:44] <Mithrandir> infinity: yeah, sure.  Eventually, after porting a few hundred thousand lines of code.
[08:44] <Mithrandir> infinity: I'm not going to run an unsupported version of php4.
[08:44] <rubenv> Mithrandir: have you tried running it on php5?
[08:44] <infinity> rubenv : Someone just sent me a mess of patches a day or two ago.  If that wasn't you, then someone else is also just as interested. :)
[08:44] <Mithrandir> infinity: there's just too many security holes.
[08:44] <rubenv> turn on compat
[08:44] <Mithrandir> rubenv: no, as there are no packages yet. :)
[08:44] <rubenv> infinity: swell :-)
[08:45] <infinity> Mithrandir : Hrm.  True dat.  Well, argue with me about it when I actually bring up changing seeds.  Maybe I'll change my mind, or come to my sense, or someone else will override me.
[08:45] <Mithrandir> infinity: I'll just bribe you with beer at UDU
[08:45] <infinity> Mithrandir : I don't drink much.
[08:45] <rubenv> Mithrandir: unfair advantage ;-)
[08:45] <infinity> Mithrandir : Bribe me with something better, and we'll talk. :)
[08:46] <Mithrandir> infinity: cocoa, then?
[08:47] <Mithrandir> infinity: chocolate, glass beads, whatever it takes. :)
[08:57] <infinity> Heh.
[08:57] <infinity> elmo :ping.
[09:03] <thom> pitti: test.xml?
[09:03] <pitti> thom: sorry, I wanted to ask before
[09:04] <pitti> thom: I'm testing a jabber vulnerability (crash), can I test this with you?
[09:04] <thom> sure
[09:04] <pitti> thom: I have no idea where gaim has an "XML console", but I'd like to play around with that
[09:04] <pitti> in any way it crashes if I stop a file transfer...
[09:05] <thom> hrm
[09:05] <pitti> but actually I wanted to crash _your_ gaim...
[09:05] <thom> my gaim was pretty uncrashed
[09:05] <thom> i need to go now, sorry dude
[09:05] <pitti> I stopped the transfer
[09:05] <pitti> cause I wanted to warn you first
[09:05] <pitti> oh, ok
[09:05] <thom> (have to get a coach back to sydney)
[09:10] <infinity> pitti : WOuld you consider this Win32-specific? http://www.mozilla.org/security/announce/mfsa2005-25.html
[09:11] <pitti> infinity: hmm, it doesn't look particularly Win-specific
[09:13] <infinity> pitti : Well, except for the part where UNIX/Linux doesn't rely on file extensions to determine if a file is executable.
[09:13] <infinity> The BUG isn't Win32-specific, but the vulnerability might be seen as such.
[09:17] <pitti> back
[09:17] <pitti> infinity: no, but still you could download an ELF file which is named like a GIF, no?
[09:18] <infinity> pitti : Erm, you can always do that.
[09:19] <Treenaks> chmod etc.
[09:19] <infinity> pitti : This bug is that I can make a file called "foo.gif.bat", which is both a valid GIF and a valid batch file, and it'll display as a GIF in the webpage, but if you DnD it to the desktop, you can run it as a batch file.
[09:19] <infinity> pitti : I don't see how this could affect UNIX.
[09:20] <pitti> phone again
[09:20] <infinity> pitti : With any web browser, I can feed you "foo.gif" which is an ELF executable.  Nothing can "protect" against that.
[09:20] <Treenaks> unforgeable mime-types!
[09:21] <infinity> pitti : But, on UNIX, nothing you download will automatically be +x
[09:21] <infinity> (Or, it shouldn't be.. if it is, that's a different bug, though, not this one)
[09:22] <infinity> s/download/drag'n'drop/
[09:32] <pitti> infinity: back, supporting my gf at the phone.. :-)
[09:32] <pitti> infinity: okay, I'm convinced about the +x thing :-)
[09:36] <infinity> pitti : Does your GF run Ubuntu?
[09:36] <pitti> infinity: no, a pre-woody
[09:37] <pitti> infinity: pentium 100, 48 MB RAM
[09:37] <pitti> infinity: I compiled bazaar for her
[09:37] <Treenaks> *insert obvious "woody" joke*
[09:37] <pitti> infinity: but she had some troubles installing it since e. g. libxml2 was not even woody-recent
[09:38] <pitti> infinity: I offered her to give here a more recent computer very often, but she just likes that old thing
[09:40] <infinity> Scary.
[09:40] <infinity> I gave away my last machine that was that slow a few years back...
[09:41] <infinity> Several years, even.
[09:41] <infinity> Though I still have two P200 systems in service at the homes of my parents and brother, acting as firewall systems.
[09:43] <Treenaks> my slowest machine is an old unused 600MHz laptop
[09:44] <Treenaks> (but I still have 80% of a P100-era PC)
[09:45] <Treenaks> (Sony 2-speed CD player, anyone? QIC-80 tapedrive?)
[09:48] <infinity> "old unused 600MHz laptop"... I hate you. :)
[09:48] <infinity> The machine I've been working on for the last year is a 550MHz laptop.
[09:48] <Treenaks> infinity: it's falling apart
[09:48] <infinity> With no battery.
[09:48] <infinity> And a missing key.
[09:49] <Treenaks> infinity: I've only just bought a new one (2 weeks ago)
[09:49] <infinity> And X and the kernel both hate it in neat and interesting ways.
[09:49] <Treenaks> oh, mine works quite well.. one loose key, but lots of damage to the plastic "shell"
[09:50] <infinity> Oh, that too.  I have a mess of glue on this thing.
[09:50] <infinity> The back sort of pries open when you close the lid.
[09:50] <Treenaks> I have 2 huge white spots where I rest my hands when typing
[09:50] <fabbione> yay
[09:50] <fabbione> inotify 0.22 works
[09:50] <Treenaks> fabbione: works or Works ?
[09:50] <infinity> Oh well, I'll be bringing this laptop to UDU, so everyone can make fun of it/me in person.
[09:51] <jsgotangco> bah my laptop is much worse
[09:51] <Treenaks> infinity: don't worry, we've seen the stuff sladen and mako haul around :)
[09:51] <jsgotangco> when you see it in UDU
[09:51] <jsgotangco> crappy Taiwan made thing *grin*
[09:51] <fabbione> Treenaks: it works.. at least my laptop doesn't crash anymore
[09:52] <fabbione> too bad i already saw at least one problem using gamin with inotify
[09:52] <fabbione> but that will be a gamin problem...
[09:52] <Treenaks> fabbione: diverted to jdub? :P
[09:52] <fabbione> Treenaks: exactly
[09:55] <doko> jbailey: /lib64/tls is missing from your packages?
[09:55] <fabbione> hey doko
[09:56] <fabbione> doko: any ETA for gcc-4 with ppc64 support?
[09:56] <pitti> carlos: why did you go away? I wanted to test seomthing else
[09:57] <doko> fabbione: chinstrap:~doko/powerpc64 and http://people.ubuntu.com/~jbailey/glibc/ , you're welcome as an alpha tester
[09:57] <carlos> pitti: I didn't go away
[09:57] <carlos> pitti: it closed automatically
[09:57] <pitti> carlos: you are offline
[09:57] <carlos> :-)
[09:57] <doko> fabbione: that's still 3.4 
[09:57] <fabbione> doko: ENOPPC, i need a chroot on davis with that stuff
[09:57] <carlos> pitti: perhaps you kill it?
[09:58] <pitti> carlos: dunno, that's what I want to find out
[09:58] <pitti> carlos: I send you a normal file, as a comparison
[09:58] <pitti> ... as soon as you logged back in
[09:59] <carlos> pitti: but as I told you, my firewall seems to be blocking the files
[09:59] <pitti> carlos: I didn't actually send you files
[09:59] <pitti> carlos: just some h4ck1sh commands 
[10:51] <doko> smurfix: ping
[10:59] <pitti> carlos: so again, you stopped the download and gaim crashed?
[11:00] <carlos> pitti: no, I didn't stop the download, it says I did it but I didn't touch the cancel button
[11:00] <pitti> carlos: I mean, you logged off after stopping the download...
[11:01] <carlos> pitti: first time, no, It crashed, second time yes, I moved to gossip again
[11:03] <pitti> carlos: ah, so it did crash
[11:03] <carlos> but I didn't get anything
[11:37] <fabbione> jdub: ping?
[11:37] <sladen> fabbione: he was two hours idle, but could do with finding him
[11:48] <fabbione> sladen: nothing urgent
[12:01] <carlos> pitti: ?
[12:01] <pitti> carlos: I switched to psi again
[12:01] <carlos> pitti: you went offline
[12:01] <carlos> pitti: :-P
[12:01] <pitti> carlos: gaim does not have an XML console
[12:01] <pitti> carlos: can you please install the updated gaim I sent you
[12:02] <pitti> and login using gaim=
[12:02] <pitti> ?
[12:02] <carlos> pitti: I'm with a ppc computer
[12:02] <pitti> D'oh
[12:02] <pitti> carlos: could you compile the package yourself?
[12:02] <carlos> pitti: I think so, where are the sources?
[12:03] <pitti> carlos: apt-get source gaim
[12:04] <pitti> carlos: cd gaim-1.1.4
[12:04] <pitti> carlos: patch -p1 < gaim-1.1.4-1ubuntu4.1.diff
[12:04] <carlos> pitti: ok
[12:04] <pitti> carlos: ^ that is the debdiff at http://people.ubuntu.com/~pitti/gaim-1.1.4-1ubuntu4.1.diff
[12:04] <pitti> carlos: that'd rock :-)
[12:07] <mvirkkil> mvo, jdub: I'm thinking of patching gnome-app-install to store settings in a .conf file. Thoughts?
[12:07] <Treenaks> mvirkkil: why?
[12:07] <carlos> pitti: it's taking a while, getting the dependencies to compile it...
[12:08] <mvirkkil> Treenaks: Because stuff like the names of the menus to be used is hardcoded.
[12:09] <Treenaks> mvirkkil: names of the menus?
[12:09] <seb128> the menu should be the same as the panel one, so the same names
[12:10] <mvirkkil> Treenaks: Hmm.. I mean the filenamess that contain the xdm menus that are used.
[12:11] <mvirkkil> Treenaks: The tree is stored in a menu file. The name of that file is hard-coded.
[12:11] <Treenaks> ah like that
[12:11] <Treenaks> why not just a command-line option to specify a different file then?
[12:11] <seb128> xdm ?
[12:12] <mvirkkil> seb128: xdg
[12:13] <seb128> the menu is built using pythonxdg
[12:13] <seb128> is there any real usecase to change the .menu to use ?
[12:14] <Treenaks> seb128: "Which ubuntu-edu packages do you want to install?"
[12:14] <mvirkkil> Treenaks, seb128: My previous patch added the functionality to have multiple menus. Each is read from a separate file.
[12:14] <Treenaks> mvirkkil: /why/
[12:14] <seb128> Treenaks: "ubuntu-edu", is that a panel submenu ?
[12:14] <Treenaks> seb128: no, but I think that is what mvirkkil is trying
[12:15] <MrNonchalant> Treenaks: sorry to intrude but here's a thought, maintainability?
[12:15] <seb128> that's not the way g-a-i works afaik
[12:15] <mvirkkil> Treenaks: General dislike of having stuff hardcoded in the source I guess. 
[12:16] <mvirkkil> Treenaks: Plus if someone would like to use it as a base for installing restricted formats, that person wouldn't need to touch the source.
[12:16] <seb128> zillion of config files is not better
[12:16] <mvirkkil> Treenaks: to add his own menus that is-
[12:17] <mvirkkil> seb128: Hmm... It woudl be one config file for g-a-i and the rest would be resource files ;-)
[12:17] <seb128> I don't get the interest of this config file
[12:17] <mvirkkil> seb128: What did you mean by "not the way g-a-i works"?
[12:17] <seb128> you give .desktop file to g-a-i
[12:17] <seb128> and it builds a menu like the app one
[12:17] <seb128> which is its job
[12:18] <mvirkkil> seb128: No, you give it a .menu fiele, currently just applications.menu
[12:18] <mvirkkil> seb128: Though there could be several menu files.
[12:19] <seb128> that's technical detail
[12:19] <seb128> it displays the same menu as the panel one
[12:19] <seb128> what do you want to change to that and why ?
[12:19] <mvirkkil> seb128: But you are saying it's best to have the names of those files (that can be any number) should remain hardcoded?
[12:19] <seb128> I don't understand what you want to do
[12:20] <seb128> the current way makes than the panel and g-a-i match
[12:20] <seb128> they both display the applications.menu
[12:20] <mvirkkil> seb128: I'm just asking for feedback on separating the names of those files in to a separate config file. 
[12:20] <seb128> which is what the user expect
[12:20] <seb128> imho that's an useless extra config file
[12:21] <mvirkkil> seb128: But it's a _different_ applications.menu, if I'm not mistaken.
[12:21] <seb128> no
[12:21] <mvirkkil> seb128: (I might be though)
[12:21] <seb128> GNOME has one .applications
[12:21] <seb128> from gnome-menus
[12:21] <seb128> s/.applications/applications.menu/
[12:22] <mvirkkil> seb128:I thought the .deb came with it's own. 
[12:22] <seb128> no
[12:22] <mvirkkil> a*its
[12:23] <seb128> hum, it has one, lemme check
[12:23] <mvirkkil> seb128: /usr/share/gnome-app-install/applications.menu /etc/xdg/menus/applications.menu
[12:23] <mvirkkil> seb128: And those are not identical (but currently extremely similar)
[12:25] <mvirkkil> So it does use a separate .menu file. And when support for multiple .menu files is used, it would use it's own .menus anyway I suppose.
[12:26] <koke> hey, how is this "thing" called http://www.grawert.net/mataro/img004.jpeg.medium.jpeg
[12:26] <koke> can I order some of these whit shipit cd's?
[12:27] <mvirkkil> koke: CD stand? And they send you one automatically I think. Depends on the number you order.
[12:28] <seb128> mvirkkil: I don't understand why you want to complicate it and get one new config file for nothing
[12:28] <martink> koke, http://www.gnome.org/~jdub/blog/2004/12/01/
[12:28] <seb128> mvirkkil: it should just use the gnome-menus one
[12:28] <mvirkkil> seb128: And when the services.menu is used? Where does that come from? 
[12:29] <seb128> there is no services.menu
[12:29] <koke> mvirkkil: thanks
[12:29] <koke> brb
[12:30] <mvirkkil> seb128: But there will (probably) be one. For installing a "web-server", "ssh server", "web based e-mail server" or similar stuff.
[12:31] <seb128> mvirkkil: not sure, an user doesn't need to install those
[12:31] <seb128> mvirkkil: and an admin can use synaptic
[12:32] <mvirkkil> seb128: I'm not the one who is suggesting to add a services menu. But there was references to that in the sourcecode and ross/mvo/jdub/seomeone confirmed.
[12:32] <seb128> and even if we do a services.menu for that, not need of a config file, you just need to have menu selector from the UI applications/webserver/...
[12:32] <mvo> having services in g-a-i was part of the original design, but it was dropped for hoary. I think we need to discuss this in UDU 
[12:32] <seb128> imho that should be an UI option
[12:33] <seb128> not one new config file
[12:33] <mvirkkil> mvo: What do you think about specifying the .menu files to be loaded in a separate config file? Cluttering or cleaning?
[12:33] <mvirkkil> *cleansing
[12:33] <remi`> mvo, did you have time to check out the bug for update-notifier?
[12:33] <mvo> mvirkkil: sorry, no opinion yet. that part of the code was mostly ross area of work
[12:34] <madduck> could someone please tell me about how many packages are in {base,desktop,supported} together, please?
[12:34] <mvo> remi`: looked over it very briefly only. will try to do it today
[12:34] <remi`> mvo, ok thx
[12:34] <mvirkkil> seb128: So you suggest using one enormous menu, instead of using smaller ones?
[12:34] <seb128> mvirkkil: ??
[12:34] <seb128> mvirkkil: how a config file would change the menu look ?
[12:35] <mvirkkil> seb128:  "applications/webserver/"
[12:35] <seb128> I suggest to have a command line option or a menu file with the different options
[12:35] <seb128> so you pick menu, app
[12:35] <seb128> menu, webserver
[12:35] <seb128> and you have the corresponding simple meny
[12:35] <seb128> menu even
[12:35] <seb128> ups
[12:36] <seb128> s/menu file/menu menu/
[12:36] <seb128> user don't expect to go to /etc to change a config file
[12:37] <mvirkkil> seb128: Even though I disagree, I'm not saying that a user would be going to any config file.
[12:37] <seb128> just make the app listing the .menus installed and do a corresponding UI options to pick the one you want
[12:38] <mvirkkil> seb128: Now that is a suggestion I like. I scan the directory where the .menu files are suppoed to be and just generate the tabs from them.
[12:38] <mvirkkil> s/suppoed/supposed
[12:38] <MrNonchalant> user has already had to change the grub menu.lst every time synaptic adds a kernel so that their machine dual-boots properly
[12:38] <MrNonchalant> I'm sure user won't mind
[12:40] <mvirkkil> seb128: BTW: Here's the way my patch changes how the UI looks: http://users.tkk.fi/~mvirkkil/old_new_appinstall.png (new to the left)
[12:40] <Burgundavia> mvirkkil, not to be rude, but I am wondering what the point of the tab is?
[12:41] <mvirkkil> Burgundavia: Well, the point is that there will be several tabs once several menus are used
[12:41] <seb128> mvirkkil: nice, can you put a bugzilla bug with that ?
[12:41] <mvirkkil> seb128: So now each .menu file would automatically be used to generate a new tab. I like the idea. A developer would just need to drop in a .menu file 
[12:41] <mvirkkil> seb128: there already is one.
[12:41] <Burgundavia> mvirkkil, ahh
[12:41] <seb128> mvirkkil: oh, k
[12:42] <Burgundavia> mvirkkil, are we talking places/system?
[12:42] <seb128> mvirkkil: right, that's better than having a config file
[12:42] <mvirkkil> Burgundavia: Not sure what you mean. I'm referring to the services menu/tab
[12:43] <mvirkkil> seb128: https://bugzilla.ubuntu.com/show_bug.cgi?id=8762
[12:43] <pitti> carlos: any luck?
[12:44] <Burgundavia> mvirkkil, you finally removed the annoying grey space thing
[12:44] <mvirkkil> seb128: I agree. I'll try to implement something like that.
[12:44] <seb128> cool
[12:44] <seb128> thanks
[12:45] <mvirkkil> seb128: Burgundavia: I hadn't checked bugzilla lately. There is some interesting discussion there: https://bugzilla.ubuntu.com/show_bug.cgi?id=8892
[12:46] <carlos> pitti: still compiling
[12:46] <pitti> carlos: uh, I compiled the warty _and_ hoary version in the meantime :-)
[12:47] <Burgundavia> mvirkkil, thanks, taking a look now
[12:47] <carlos> pitti: I'm doing other tasks at the same time and my powerbook is three years old already...
[12:47] <carlos> :-)
[12:48] <Burgundavia> mvirkkil, services? are we talking things like servers?
[12:49] <mvirkkil> Burgundavia: yes
[12:49] <Burgundavia> mvirkkil, ah ok, we are talking different types of applications
[12:49] <mvirkkil> Burgundavia: The terminology in use is that one server offers several services like e-mail, www, ssh etc.
[12:50] <Burgundavia> ok
[12:50] <Burgundavia> nice idea
[12:50] <mvirkkil> Burgundavia: Meaning the server is the physical computer(i think)
[12:53] <madduck> could someone please tell me about how many packages are in {base,desktop,supported} together, please?
[12:54] <mvirkkil> madduck: Try something like  apt-cache dump|grep Package|wc -l
[12:55] <madduck> mvirkkil: i don't have an ubuntu system. i just writing about it. :)
[12:55] <madduck> mvirkkil: could you run it quickly?
[12:56] <mvirkkil> madduck: I don't have a standard ubuntu installation, and I'm not sure it works, it was just a pointer in one direction.
[12:56] <madduck> mh. thanks...
[12:57] <Burgundavia> madduck, are you looking for a number count of the supported software?
[12:57] <mvirkkil> mvo: Do you know about this separating .desktop files in to their own package thing?
[12:57] <madduck> Burgundavia: yes.
[12:58] <Burgundavia> madduck, ok, just a sec
[12:58] <madduck> Burgundavia: just a rough estimate for http://debianbook.info
[12:58] <madduck> right now it's at 850 packages, but that was before warty
[12:59] <dredg> niall@binky:~(0)$ grep ^Package /var/lib/apt/lists/archive.ubuntu.com_ubuntu_dists_hoary_main_binary-i386_Packages|wc -l
[12:59] <dredg> 3136
[12:59] <mvirkkil> madduck: My non standard install has over 35 thousand, though I know most of them are not in main/base/whatever :-)
[12:59] <dredg> madduck: will that do?
[12:59] <Burgundavia> madduck, around 3000 for me
[01:00] <mvo> mvirkkil: splitting it out is a planed feature. I guess we need to talk with ross and jdub about the future of g-a-i :)
[01:00] <mvirkkil> madduck: Crap. It's actually just 20 thousand.
[01:00] <madduck> Burgundavia: is that base,desktop,supported ?
[01:01] <Burgundavia> madduck, that is just the supported stuff
[01:01] <mvirkkil> mvo: How does that affect the .menu files?
[01:01] <madduck> ...ts_hoary_main_bina... -- and I thought you guys don't have 'main'
[01:01] <mvo> mvirkkil: I don't think it will affect them at all, it's just easier to maintain this way
[01:01] <mjg59> madduck: main is the supported subset of Debian main
[01:01] <mvirkkil> mvo: I mean would you separate them too?
[01:01] <mvirkkil> mvo: Or would they be distributed with gai?
[01:02] <madduck> mjg59: ic thanks.
[01:02] <mvirkkil> mvo: Not sure that it really matters, though...
[01:02] <mvirkkil> mvo: Especially if the .menu files are detected on the fly.
[01:02] <mvo> mvirkkil: my gut feeling is that I would seperate them too or at least have some mean to install additional ones via a deb. it might come handy for people who want to add additonal stuff (like games :)
[01:03] <Bnonn> urgle
[01:03] <mvirkkil> Though then gai will expect them in one directory.
[01:04] <mvirkkil> mvo: Ok. But if gai is autodetecting them, it will expect them all to be in one directory.
[01:04] <mvo> mvirkkil: that shoudn't be a problem, should it?
[01:05] <mvirkkil> mvo: No. This sounds pretty good, I think.
[01:05] <Bnonn> might be the wrong channel to ask this in, but I notice on the hardware support list the Audigy ES listed twice
[01:05] <Burgundavia> madduck, any other questions I can answer?
[01:05] <Bnonn> since i have one, I have a vested interest in wondering what the difference is between emuk10k1 and snd-audigyls
[01:06] <Bnonn> and is there any way to get my currently not-working Audigy working?
[01:07] <madduck> Burgundavia: well, no. I am not sure 3000 is the answer... has ubuntu really grown from 850 supported to 3000 supported packages in half a year?
[01:08] <Burgundavia> madduck, not having done the same on a warty system, I cannot judge
[01:08] <carlos> pitti: sorry, it's ready. I didn't see it :-P
[01:08] <carlos> pitti: installing
[01:09] <mvo> madduck: I have 2870 here in main not including the "translations" section
[01:10] <madduck> ok. thanks.
[01:13] <jbailey> doko: No, they're native NPTL packages.  It's the configuration I'm hoping to have work out, so no need for /libc64/tls.  
[01:15] <jordi> J! E! F! F!
[01:15] <ajmitch> aha, jordi lives
[01:16] <jordi> ajmitch: hello!
[01:16] <ajmitch> hey jordi, what's up?
[01:16] <jordi> I'm ultra busy
[01:16] <jordi> this sucks.
[01:16] <mvo> hey jordi 
[01:17] <jordi> hey mvo 
[01:17] <pitti> carlos ?
[01:17] <jordi> I haven't checked, but if you have done string changes to synaptic I MIGHT kill
[01:18] <carlos> pitti: it crashed :-(
[01:18] <pitti> D'OH
[01:18] <jordi> carlos: oooii carlos
[01:18] <carlos> jordi !
[01:18] <fabbione> pitti: yo dude
[01:18] <carlos> jordi: sindominio spamassassin eats launchpad's mails!
[01:18] <carlos> pitti: any chance the crash is not related to that bug report?
[01:19] <pitti> carlos: I don't know; http://sourceforge.net/tracker/?func=detail&aid=1172115&group_id=235&atid=100235
[01:19] <pitti> carlos: ^ that's spanish; maybe you can tell me?
[01:19] <carlos> pitti: try sending me a normal file so we can be sure it's not another bug...
[01:21] <carlos> pitti: it talks only about the security bug, nothing related to a general bug
[01:23] <pitti> carlos: sending...
[01:26] <lifeless> anyone seen elmo today ?
[01:27] <fabbione> lifeless: nope
[01:31] <GheRivero> res
[01:59] <jdub> fabbione: pong
[01:59] <fabbione> jdub: inotify 0.22 works here
[02:00] <jdub> fabbione: aha, wonderful! :)
[02:01] <pitti> fabbione: yay
[02:01] <fabbione> jdub: you will be happy to know that the new inotify will make gamin look even more horrible than it is now
[02:01] <fabbione> i am going to push you the kernels soon
[02:01] <jdub> fabbione: ha ha :-)
[02:01] <fabbione> and you will have to fix gamin
[02:02] <pitti> fabbione: there's still no answer from DV :-(
[02:02] <fabbione> take into account that the loop on Desktop doesn't work at all
[02:02] <jdub> fabbione: there are 0.22 patches for gamin, so we'll see how they look - should be a lot nicer in general
[02:02] <thom> pitti: he's on holiday
[02:02] <pitti> aha
[02:02] <fabbione> because nautilus doesn't register it at all
[02:02] <fabbione> or gamini
[02:02] <fabbione> or whatever
[02:02] <fabbione> well the kernel works...
[02:02] <pitti> give it to us!!! :-)
[02:02] <fabbione> now it's time to fix your crap :P
[02:03] <fabbione> pitti: in a couple of days...
[02:03] <jdub> fabbione: all that ooky desktop level mess ;)
[02:03] <fabbione> i have almost done updating all the external drivers
[02:04] <fabbione> + i want to add GFS and ipv6 statefull firewall before release -1
[02:04] <fabbione> probably something more.. i haven't decided yet
[02:05] <fabbione> GFS = teh w1n5
[02:05] <Treenaks> GFS = the redhat thingy?
[02:05] <fabbione> Treenaks: yes
[02:05] <fabbione> Global File System
[02:05] <Treenaks> l33t!
[02:06] <fabbione> that on top of AOE is really n34t
[02:06] <fabbione> Treenaks: remember.. it's not a case what i asked on -motu :P
[02:06] <fabbione> everything has a meaning in this universe
[02:06] <fabbione> specially inside my universe :)
[02:06] <maswan> AOE = ATA Over Ethernet?
[02:06] <fabbione> maswan: yes
[02:07] <jdub> fabbione: mmmmmmmm, GFS. i will kiss you.
[02:07] <maswan> Ah, that makes it a bit more resonable
[02:07] <fabbione> jdub: you already own me enough beers... 
[02:07] <fabbione> jdub: start to prepare the vaseline and bend ov^W^W^W^Wthe real liquors
[02:17] <fabbione> maswan, milli
[02:17] <fabbione> afs is exactly as upstream...
[02:17] <fabbione> we don't apply any patch to it
[02:18] <milli> fabbione: I know.  What's in sid is the 2.4 kernel stuff
[02:18] <fabbione> if you are aware of patches and stuff just open an enanchement bug with the patch
[02:18] <milli> The patches for 2.6 kernel are in experimental
[02:18] <fabbione> milli: if they are in experimental, there is a reason
[02:19] <milli> nod
[02:19] <milli> Debian sid has a 2.4 kernel offering however....
[02:20] <milli> It just sucks that the AFS hooks in 2.6 kernel got left behind by Linus et al...
[02:21] <fabbione> that's probably because the upstream for AFS isn't doing a very good job
[02:22] <maswan> fabbione: which one of them, openafs or arla?
[02:22] <maswan> or both?
[02:24] <fabbione> maswan: i don't know..but if the code is clean, Linus usually have no issues in pulling in stuff
[02:29] <maswan> fabbione: To be honest, I don't know if either of them are gpl-compatible even or if they are possible to relicense in such a way.
[02:29] <milli> the prob with OpenAFS was it hooked setgroup* calls in the syscall table
[02:29] <milli> and that's a no-no in the 2.6 world
[02:30] <Treenaks> milli: sounds evil
[02:30] <fabbione> maswan: there are not only license problems in life.. you know? ;)
[02:30] <fabbione> linus wants clean code..
[02:30] <milli> It had to do that because PAGs are not implemented in the kernel
[02:30] <Treenaks> milli: then they should implement that
[02:30] <lamont> milli: you should sleep sometime, you know...
[02:30] <milli> It's getting worked out....
[02:31] <milli> lamont: you should too
[02:31] <lamont> milli: just finished
[02:31] <milli> lamont: ditto :)
[02:31] <lamont> heh
[02:31] <Treenaks> lamont: oh, you're not pre-lagging for UDU? :)
[02:31] <maswan> fabbione: sure, but then there has to first exist clean code that can be submitted. :)
[02:31] <lamont> (morning)
[02:31] <lamont> Treenaks: sadly, kids need taken to school pretty much 5 out of 7 days.
[02:32] <Treenaks> lamont: oh yeah
[02:32] <milli> lamont: They need to be there at like 7:30?
[02:33] <lamont> milli: yeah - target is to leave the house by 6:45, so that we really leave by 7
[02:33] <milli> lamont: All too familiar with time padding...
[02:34] <maswan> milli: arla kind of works, but 0.38 isn't packaged yet,and packagin is apparently a big ammount of work due to how upstreams packages stuff
[02:34] <lamont> OTOH, wifely one is driving today
[02:34] <milli> lamont: Well, AFS 1.3.81 does seem to build fine on hoary
[02:34] <lamont> milli: coolness
[02:34] <Treenaks> (so I'm an hour early, often)
[02:35] <milli> maswan: If the kernel module from 1.3.81 openafs doesn't work reliably yet, I'm going to fallback and build a 2.4 kernel for hoary.. :(
[02:35] <milli> and live without inotify and gamin and such
[02:35] <milli> AFS is a very important part of my life
[02:35] <maswan> milli: the packaged 0.36 arla works with a few small patches, at least it worked on warty
[02:35] <maswan> same here, at least at work.
[02:36] <milli> ...as in, all of my files live on my AFS server
[02:36] <maswan> oh, well, bbiab.
[02:36] <milli> er, servers
[02:42] <mvirkkil> Ross isn't going to like me...
[02:43] <mvirkkil> I made severa invasive changes in one sweep.
[02:43] <bob2> "We continue to build libqt2 and all dependant packages with g++-3.3." <- isn't the whole kde2/qt2 tree still built with gcc 2.95?
[02:44] <lamont> bob2: not on ubuntu
[02:44] <lamont> 2.95 is not in the archive
[02:44] <bob2> hrm, ok
[02:44] <bob2> so it's binary incompatible with debian anyway?
[02:44] <lamont> wouldn't surprise me
[02:45] <bob2> heh
[02:45] <lamont> bob2: it was an intentional decision to not bootstrap gcc-2.9*, nor gcc-3.2
[02:46] <bob2> ah, ok
[02:48] <fabbione> mjg59: ping?
[02:54] <mjg59> fabbione: Hi
[02:55] <fabbione> mjg59: yo...
[02:55] <fabbione> the 2 suspend/resume patches for ipw2100 and ipw2200 were stolen from head?
[02:55] <mjg59> From their respective heads, yeah
[02:58] <zul> hey
[03:00] <fabbione> mjg59: thanks
[03:00] <fabbione> hi zul
[03:01] <zul> hey fabbione 
[03:02] <torkel> maswan, milli: neither openafs, nor arla will probably be in the kernel proper anytime soon. None of the are intrested in it
[03:03] <torkel> and the afs code that are in the kernel is just plain broken, and has no connection with openafs or arla
[03:05] <torkel> milli: if you want an unofficial arla that works with hoary I have it available at /afs/hpc2n.umu.se/home/t/torkel/www/arla/Ubuntu/
[03:06] <milli> torkel: I was just perusing Arla's page...  any issues with that interoperating with openafs file servers?
[03:06] <torkel> the right thing to do, is to drop kerberos4 support in heimdal, as krb4 is unsecure at protocol level
[03:07] <torkel> then it _should_ be a piece of cake to build a newer arla
[03:08] <torkel> milli: not that I am aware of. The performance is not the best, on the other hand it seems to be a lot more stable on amd64 than 0.36
[03:08] <milli> torkel: That's why I didn't go with Arla way back when... no krb5 support
[03:09] <torkel> milli: I think arla has supported krb5 as long as openafs
[03:11] <mvirkkil> seb128: Posted the .menu-file scanning thing: https://bugzilla.ubuntu.com/show_bug.cgi?id=8762
[03:12] <mvirkkil> Burgundavia, mvo: Posted the .menu-file scanning thing: https://bugzilla.ubuntu.com/show_bug.cgi?id=8762
[03:13] <mvo> mvirkkil: thanks, that was quick!
[03:13] <mvirkkil> mvo:Well, I made some other modifications too. Ross will hate my guts when he reviews the patch.'
[03:14] <mvirkkil> mvo: Because it's large, invasive and non trivial.
[03:14] <koke> mako: I've just sent you a mail about UDU
[03:14] <koke> have to go now
[03:14] <koke> bye all!
[03:15] <mvo> mvirkkil: heh
[03:16] <mvirkkil> mvo, seb128, Burgundavia: Any other suggestions for improvements?
[03:17] <mvo> mvirkkil: have you seen ross comment about showing the gui early and having a busy-cursor then? 
[03:18] <mvo> mvirkkil: we'll progress bar support for the initialization soon I think. my python-apt tree supports progress objects that can be used to give more visual feedback while it is initializing
[03:18] <mvirkkil> mvo: Yes, it's fixed in the new patch
[03:18] <mvo> mvirkkil: great, thanks
[03:18] <mvirkkil> mvo: That'd be nice.
[03:20] <ogra> hi all
[03:20] <mvirkkil> mvo: How are the prgoress objects implemented? I poll them or they call back?
[03:21] <mvo> mvirkkil: as callback objects, I can /msg you a example if you want
[03:22] <mvirkkil> mvo: Sure
[03:29] <mjg59> Goshawk: About?
[03:29] <Goshawk> hi
[03:29] <Goshawk> mjg59, about what?
[03:29] <mjg59> Goshawk: About as in "Are you about" :)
[03:30] <mjg59> Goshawk: sorry, I was just looking at the splashy thing again. The real problem is just that directfb doesn't work with vga16fb
[03:30] <mvo> seb128: mind if I take over #6821?
[03:31] <Goshawk> mjg59, the kind of card are you talking about are only few models (second me, it can't be true). Nowadays about 200 people are using an old version of splashy caleld usplash-0.1preview that is bugged, we need splashy-0.1 soon as possible
[03:32] <mjg59> Goshawk: Ok, I can phrase this differently:
[03:32] <mjg59> Goshawk: If Breezy is going to have a bootsplash system, it must work with vga16fb as otherwise it will not (in most cases) allow suspend and resume to work
[03:33] <Goshawk> mjg59, for Breezy relase it will be ready.. and does not Breezy will have usplash? (it's a sarcastic question..)
[03:33] <mwh_> Hi, I tried to reinstall debianutils and coreutils at the same time but it failed because coreutils replaces debianutils, is that a bug or a feature? Im wondering, if its a bug I would write a bugreport about it .. but if it is not, then please explain why its a feature
[03:33] <mjg59> Goshawk: Breezy will have bootsplash if we have code that works with vga16fb
[03:33] <mjg59> If we don't, it won't
[03:35] <Goshawk> mjg59, sure
[03:36] <Goshawk> mjg59, we will do all in our possibilities to agree with ubuntu requests
[03:55] <mvo> ping lamont 
[03:55] <lamont> mvo: ack
[03:56] <mvo> lamont: do you have a ia64 machine with update-notifier runing by chance? I got a crash report 
[03:57] <lamont> mvo: no.  the only ubuntu ia64 box I have just has a server install
[03:57] <lamont> note that archive.ubuntu.com doesn't have ia64 bits in it, could that be the source of you rcrash?
[03:58] <mvo> lamont: thanks. the crash seems to be deep inside the libhal/u-n interaction, I'm not sure yet
[03:59] <mvo> might be a 64bit uncleanless even that does not show on amd64 :/
[03:59] <lamont> mvo: I suppose I could install the 115 packages that update-notifier would require...
[03:59] <mvo> lamont: no, better leave it then :)
[03:59] <pitti> btw, daniels, if you do dbus packages, could you please put them on p.u.c?
[03:59] <pitti> daniels: I'd like to play with hal, but I don't want to duplicate your packaging for dbus
[04:00] <daniels> pitti: yeah, i'm going to put them up on p.u.c/~daniels/dbus/ tonight
[04:00] <pitti> rock
[04:00] <daniels> was hoping to get it done earlier, but had a rather hellish day
[04:00] <lamont> mvo: fwiw, one of my home-office projects is to move the production stuff on the i2000 to another machine, so that I'll have both itanium and mckinley hw for testing
[04:02] <mvo> lamont: nice! what time-frame are we talking about? days? weeks? I don't really think it's a pressing issue as ia64 is not the most common architecture around. but I still don't like crashes :)
[04:09] <lamont> mvo: before UDU
[04:11] <mvo> lamont: great, thanks
[04:22] <pitti> ogra: here?
[04:25] <ogra> pitti, yop
[04:27] <lamont> mdz awake yet, I wonder?
[04:40] <daniels> fuck
[04:40] <daniels> Keybuk: advice?
[04:40] <daniels> Keybuk: dbus-glib-1 just changed major soversion from 0 to 1
[04:41] <daniels> Keybuk: do I take this opportunity to get rid of the stupid naming and call it libdbus-glib-1-1?
[04:41] <daniels> (libdbus-glib-1.so.1)
[04:42] <Keybuk> yes, if the soname has changed, the binary package name should change
[04:42] <Keybuk> and as you're NEW-bound, it's a good time to fix it
[04:42] <Keybuk> (imo)
[04:42] <Mithrandir> Keybuk: any particular reason pkg.m4 had inconsequent quoting?  (And it quotes its secondary argument, which means stuff like gtk's configure.in falls over. :P)
[04:42] <Keybuk> Mithrandir: e.g.?
[04:43] <daniels> Keybuk: is libdbus-glib-1-1 a sensible name?  i just feel totally filthy
[04:43] <Keybuk> daniels: if that's the soname of the only library in it, sure
[04:43] <daniels> ugh
[04:47] <Mithrandir> Keybuk: dnl inside the quoted string, for instance.
[04:47] <aj> jdub: around?
[04:47] <Mithrandir> Keybuk: gtk has that
[04:48] <jdub> aj: yo
[04:48] <aj> jdub: is pia changing her name?
[04:48] <jdub> aj: yeah
[04:48] <jdub> P. Waugh
[04:48] <jdub> pwaugh!
[04:48] <aj> jdub: so getting "Pia Smith" business cards on monday would be bad, right? :)
[04:48] <jdub> everyone's been trying to convince her to change her nick to pdub
[04:48] <jdub> aj: yeah
[04:48] <Keybuk> Mithrandir: which string?
[04:48] <jdub> aj: that would be kinda mistimed
[04:49] <aj> "Pia Waugh" is correct spelling etc?
[04:49] <Mithrandir> Keybuk: they have PKG_CHECK_MODULES([foo dnl\nbar dnl\nbaz] )
[04:49] <jdub> aj: yeah
[04:49] <Keybuk> Mithrandir: that's their own silly fault then, isn't it :p
[04:49] <thom> shame her middle initial isn't 'h', really
[04:49] <jdub> aj: thanks for tonight's surreal thought
[04:49] <Mithrandir> Keybuk: depends on whether it makes sense for that string to always be quoted or not -- what's the reason for you changing it?
[04:49] <Keybuk> especially given [foo\nbar\nbaz]  would work :p
[04:49] <Keybuk> Mithrandir: consistency with auto* style
[04:50] <Mithrandir> Keybuk: is this documented somewhere so I can twap people with it?
[04:50] <Keybuk> nope
[04:50] <aj> jdub: i asked if there were any problems with her card and she said no... dear oh dear...
[04:51] <raavi02> hai folks
[04:54] <jamesh> Keybuk: most autoconf macros evaluate their arguments once though, don't they?
[04:54] <raavi02> Questions regarding codes for playing different formats of video and audio
[04:56] <jdub> aj: she has *no* idea.
[04:56] <jdub> aj: that's completely not on our radar, thus surreal for me too :)
[04:56] <aj> jdub: i trust you don't mean "she has no idea that she's changing her name, but she will, she will" :)
[04:56] <jdub> ha ha
[04:56] <jdub> she has no idea she's getting married! HA HA HA HA HA!
[04:57] <ogra> oh, you didnt tell her ?
[04:57] <spiv> I don't think weddings really work as surprise parties ;)
[04:57] <thom> ogra: she's in denial
[04:57] <ogra> heh
[04:58] <jdub> aj: you coming to the partay?
[04:58] <aj> jdub: if i work out how to get there & back; the bus only leaves in the morning right?
[04:59] <mantiena> Hi all
[04:59] <jdub> aj: could you meet someone in sydney, or are you going to be in canberra already?
[04:59] <aj> nah, brisbane to canberra flight
[04:59] <mdz> lamont: here
[05:00] <mantiena> ubuntu wiki is broken :(
[05:00] <mdz> fabbione: here
[05:00] <mdz> morning
[05:00] <ogra> morning mdz 
[05:00] <daniels> morning mdz
[05:00] <jdub> aj: hmm
[05:01] <jdub> hmm, tiger's due on the 29th
[05:03] <mantiena> guys, I can't login into ubuntu wiki about 2 weeks I get an error:
[05:03] <mantiena> Site error   This site encountered an error trying to fulfill your request.  The errors were:
[05:03] <mantiena> Error Type   UnicodeDecodeError
[05:03] <mantiena> Error Value   'ascii' codec can't decode byte 0xc4 in position 12: ordinal not in range(128)
[05:04] <mantiena> should I report a bug ?
[05:04] <daniels> mdz: got a few minutes?
[05:05] <spiv> mantiena: Yes, please do.  Also, what's your username you're logging on with?
[05:05] <mantiena> spiv: mantas@akl.lt
[05:06] <mdz> daniels: yes
[05:07] <jamesh> Keybuk: http://www.gnu.org/software/autoconf/manual/autoconf-2.57/html_node/autoconf_96.html seems to say the extra quoting pkg-config-0.16's M4 did is a bad idea
[05:07] <Keybuk> possibly ... talk to the maintainer
[05:07] <spiv> mantiena: Ouch, looks like something in our plone setup is breaking because your name contains non-ascii chars.  :/
[05:08] <Keybuk> I can't immediately think of a reason for the [[$2] ] , was probably a brain-fade
[05:08] <Keybuk> but I don't maintain it or care anymore :p
[05:09] <jamesh> Keybuk: I have :)
[05:09] <mantiena> mdz: now hoary is released, maybe you can add install parameter (template) to casper (minor modifications to casper-check.templates and S60casper-check files) ?
[05:10] <mantiena> spiv: yea, I'm from Lithuania and my family name contains ccaron and umacron symbols :)
[05:10] <mantiena> mdz: I can give you patch'es
[05:11] <spiv> mantiena: Yeah, I can see it just fine in the DB... the problem appears to be in the plone site somewhere.
[05:12] <Keybuk> hmm, is breezy open yet?  </awty>
[05:12] <spiv> mantiena: If you file a bug and tell me the bug number I'll make sure it gets sorted out.
[05:13] <Kamion> Keybuk: no
[05:13] <mdz> mantiena: you can send me an arch branch to merge; you insisted that I had to provide revision control, remember?
[05:14] <mantiena> spiv: ok, I try don't forget to fill the bug :)
[05:14] <Keybuk> ok, that explains "IOError: not a gzipped file" :)
[05:14] <mantiena> mdz: hehe, it seems learning arch is not so fast as I think :)
[05:19] <mantiena> GheRivero: hi
[05:19] <GheRivero> res
[05:20] <cartman> is it true that breeze development won't start before ~2 weeks ?
[05:20] <Kamion> cartman: I don't think it's decided yet
[05:20] <jdub> jdubtv! -> http://node.waugh.id.au:8800/
[05:20] <cartman> Kamion: humpf ok
[05:21] <Kamion> cartman: http://www.ubuntulinux.org/wiki/BreezyToolchainTransition
[05:21] <cartman> Kamion: cheers
[05:21] <ogra> jdub, hey, your beard is gotten long :)
[05:21] <jdub> yeah, have to chop it a bit for sunday :)
[05:21] <cartman> wowow
[05:21] <ogra> :)
[05:22] <cartman> that looks so cool
[05:22] <cartman> toolchain updates that is
[05:22] <cartman> finally I can stop recompiling C++ libs with gcc 3.4.0 :)
[05:24] <Amaranth> it's going to be a PITA having the kernel compiled with 3.4 and the main compiler being 4
[05:24] <cartman> is KDE 4.0 realistic for breezy? If not, transition qt and KDE as well. <-- lol
[05:25] <cartman> KDE4 is not coming soon so its unrealistic
[05:25] <Amaranth> I thought Qt puked on gcc4
[05:25] <cartman> Amaranth: fixed as Qt is in their QA suite
[05:25] <Amaranth> jdub: eww....
[05:25] <cartman> Amaranth: ie you can't release a gcc that can't compile Qt
[05:25] <cartman> stable release that is
[05:25] <hno73> mantiena: Hi. Thank you for reporting the log-in error. I will look into it.
[05:26] <Riddell> cartman: no
[05:26] <cartman> Riddell?
[05:26] <Amaranth> cartman: Sure you can, if they abuse gccisms and not standard C++.
[05:26] <Riddell> cartman: oh sorry, thought you were asking not quoting
[05:26] <cartman> Amaranth: Qt never does that ;)
[05:26] <cartman> Riddell: hehe
[05:26] <Amaranth> This is why people watch jdub, he randomly freaks out. :)
[05:27] <daniels> bugger.
[05:27] <mantiena> hno73: thank you for looking into it ;)
[05:27] <Amaranth> ...and picks his nose.
[05:28] <hno73> mantiena: np, thanks for alerting us :)
[05:28] <cartman> so breeze will be a big change
[05:28] <ogra> cartman, like oary was
[05:28] <ogra> hoary even
[05:28] <cartman> ogra: Abi transition is harder imho :/
[05:29] <cartman> ogra: like hope Debian will do the same in parallel
[05:29] <Amaranth> pfft, debian will be using 3.x until next year :P
[05:29] <ogra> cartman, python and xorg transition were not easier i guess :)
[05:29] <cartman> ogra: xorg was the reason I switched from sid to hoary :)
[05:29] <ogra> :)
[05:31] <seb128> mvo: not at all :)
[05:33] <daniels> pitti, thom: this is a bit harder than I expected.  co-ordinating the soversion transition with hal will be a bitch, since we try to restart dbus in the init script ... blurgh.
[05:33] <pitti> daniels: hm, hal needs to conflict to earlier dbus'es anyway, I think
[05:34] <pitti> daniels: maybe just stop hal on upgrades
[05:34] <daniels> pitti: but hal is stopped and started with dbus-1's init script :\
[05:34] <pitti> daniels: upgrading hal when hald doesn't run will start hal again (or should be made to for this case)
[05:34] <daniels> pitti: so we need to stop dbus before the upgrade, get hal and dbus installed at the same time, and then configured later
[05:35] <daniels> ahr
[05:35] <pitti> daniels: yeah, but if hal depends on the newer dbus, dbus will be upgraded and configured first
[05:35] <hno73> mantiena: what is your login? does it have any UTF-8 characters in it?
[05:35] <daniels> i need to crash now, but i'm putting my current sources up
[05:35] <daniels> pitti: if dbus is configured first, we lose
[05:35] <daniels> pitti: because libdbus-1-1's init script will try to start hal, and bomb out because the old hal is still installed
[05:36] <pitti> daniels: right, you must not start hal if you upgrade from 0.23 -> 0.30
[05:36] <daniels> yep
[05:36] <pitti> daniels: otherwise you can
[05:39] <pitti> daniels: you put your current package up?
[05:39] <pitti> daniels: even if it's not yet perfect, its enough for playing around with the new hal and test what breaks
[05:39] <daniels> pitti: yeah, just waiting for it to upload
[05:43] <mdz> elmo: ping?
[05:43] <pitti> Morning mdz
[05:43] <mdz> morning
[05:45] <pitti> mdz: I still had no luck with contacting elmo today; maybe he is travelling?
[05:51] <daniels> bah
[05:51] <daniels> when pitti comes back, could someone please tell him dbus sources are at p.u.c/~daniels/dbus/?
[05:51] <daniels> thom: ^^
[06:09] <lamont> milli: not yet
[06:11] <CarlK> Wasn't this just brought up on the mai list?  from #xorg: ajax: if we had a way to probe DDC on a card without running all of X, we could do that pretty reliably
[06:18] <hno73> mantiena: are you logged in ok to the website now?
[06:26] <mdke> wiki rss feeds are down?
[06:34] <astharot> ciao
[06:44] <mdke> does anyone know what's wrong with the wiki rss feeds? I get a strange message: XML Parsing Error: not well-formed Location: http://ubuntulinux.org/wiki/pages_rss Line Number 38, Column 40:      <title>Web Browsing slow ( IVP6 <-> IVP4 )</title>---------------------------------------^
[06:44] <Mithrandir> looks like the < isn't properly escaped, then?
[06:44] <sm> mdke: you use those ?
[06:45] <mdke> yeah
[06:45] <mdke> hi sm
[06:45] <sm> hi there
[06:45] <mdke> Mithrandir, i don't understand that sort of thing ;)
[06:45] <sm> looks like a bug, sorry
[06:45] <mdke> oh right
[06:45] <sm> can you paste that error into a new issue at http://zwiki.org/IssueTracker
[06:46] <mdke> np
[06:46] <mdke> its been working fine until today btw
[06:46] <sm> you can work around by changing that page's title 
[06:46] <sm> Web Browsing slow ( IVP6 <-> IVP4 )
[06:46] <sm> do something with the <->
[06:47] <mdke> me?
[06:47] <sm> oops, wrong channel, my bad
[06:47] <mdke> wanna go --> ?
[06:54] <mdz> Kamion: I noticed that we never got around to disabling the addition of ide-* to /etc/modules :-/
[06:55] <mdz> Kamion: let's do that early for Breezy
[07:07] <Kamion> mdz: bug #9099 filed
[07:11] <seb128> arg
[07:11] <seb128> bug flood
[07:11] <ogra> ?
[07:11] <seb128> I've like 20 bugs on my box, FTBFS with gcc-3.4
[07:12] <ogra> throw them away, gcc4 is the way to go ;)
[07:12] <mdz> Kamion: thanks
[07:12] <seb128> if I thow them, it'll not built with gcc3
[07:12] <seb128> 3.4 even
[07:13] <seb128> ups, nm
[07:13] <seb128> mdz: when will breezy open to upload ?
[07:14] <jordi> hmm. I wonder if all ubuntu mirrors are supossed to have cd images?
[07:16] <mdz> Kamion: bug #9101 -> I seem to recall you preparing a patch so that pcmcia wouldn't be installed if it were disabled in the installer, but I can't find a bug reference.  was there a bug filed about it?
[07:17] <mdz> seb128: I don't know; I have not received a response from elmo
[07:17] <seb128> elmo: ping ?
[07:17] <mdz> seb128: we will want to make the toolchain changes before we start with general uploads anyway
[07:18] <seb128> arg
[07:18] <seb128> the bug management is going insane
[07:18] <mdz> yes, I can't keep up with ubuntu-bugs
[07:18] <Kamion> mdz: ddetect (1.11ubuntu13) hoary; urgency=low
[07:18] <Kamion>   * Don't install pcmcia-cs if hw-detect/start_pcmcia=false (closes: Ubuntu
[07:18] <Kamion>     #8678).
[07:19] <Kamion>  -- Colin Watson <cjwatson@ubuntu.com>  Tue,  5 Apr 2005 14:14:20 +0100
[07:19] <seb128> and not beeing to upload packages is not funny
[07:20] <seb128> there is like 40 new tarballs for GNOME, some of them fix the the gcc4.0 ftbfs getting filled since redhat is working on that 
[07:20] <Kamion> mdz: I've pinged the bug
[07:20] <seb128> bah, need to find a way to manage the backlog
[07:20] <Kamion> jordi: not necessarily
[07:20] <mdz> Kamion: oh, so that actually went into hoary, then? hmm
[07:20] <Kamion> mdz: yeah
[08:11] <Kamion> lamont: any chance you could fix everything apart from the _errors business in Debian #282672? That code is copied into busybox-cvs and there's a similar bug open there; it would be annoying for busybox-cvs to have to fork as well as copy.
[08:20] <\sh> any debian packaging expert online to poke some questions and peek some answers ?? :)
[08:21] <Kamion> \sh: just ask
[08:22] <\sh> Kamion: I'm trying to build a lib package...it looks like that everything for the libfoo deb is ok..but for libfoo-dev not. it looks like it losing the predefined install directories or whatever
[08:23] <seb128> how do you move files to the this package ?
[08:29] <\sh> seb128: i'm using cdbs and using install hooks...cause the lib itself is not autotools 
[08:29] <\sh> so install/libfoo:: \ install -D libfoo.so ${CURDIR}/debian/foo/usr/lib/libfoo.so <----works
[08:30] <seb128> urg
[08:30] <\sh> install/libfoo-dev:: \ install -D foo.h ${CURDIR}/debian/foo/usr/include/foo.h <--- works too
[08:30] <mdz> tech board meeting in 30 minutes on #ubuntu-meeting
[08:30] <\sh> but dh_install -plibfoo-dev : cp: cannot stat ./usr/include/foo.h 
[08:31] <ogra> mdz, huh ?
[08:31] <ogra> mdz, not 20:00 UTC ?
[08:31] <mdz> er, 90 minutes ;-)
[08:31] <ogra> heh
[08:31] <mdz> silly DST
[08:32] <\sh> so...where am I wrong?
[08:35] <pitti> thom: here? can I please have the php4 build dependencies on concordia's amd64 dchroot?
[08:37] <seb128> ogra: what hour is the current one ? 18:40 or 19:40 UTC ?
[08:37] <ogra> ogra@honk:~ $ date -u
[08:37] <ogra> Di Apr 12 18:37:04 UTC 2005
[08:38] <ogra> says my hoary....
[08:38] <seb128> right
[08:38] <pitti> and it says right
[08:38] <seb128> rock
[08:38] <seb128> I'll have time for dinner before the meeting :)
[08:38] <ogra> :)
[08:54] <pitti> mdz: can we already upload to breezy (to have our crack waiting in the queue already) or shall we queue uploads locally until packages are actually accepted?
[08:58] <zyga> what's the difference between breezy and breezy-updates?
[08:58] <HiddenWolf> zyga, breezy-updates won't be used until breezy goes stable
[08:58] <zyga> HiddenWolf: ah, okay
[08:58] <HiddenWolf> zyga, afaik
[09:00] <zyga> hmm
[09:00] <zyga> when will archive.ubuntulinux.org actually contain something in breezy?
[09:01] <pitti> zyga: uploads are not yet allowed due to the toolchain transition
[09:01] <zyga> pitti: transition?
[09:02] <pitti> zyga: doko and jbailey still prepare gcc-4.0 and a new glibc
[09:02] <zyga> pitti: I was not following lataly
[09:09] <mdz> pitti: I don't know; I haven't heard from elmo
[09:09] <pitti> mdz: okay, I queue them on chinstrap for now 
[09:10] <fabbione> re
[09:10] <fabbione> doko: ping?
[09:10] <\sh> ok...coffee time
[09:11] <\sh> it will be a long night
[09:14] <fabbione> doko: /wind goto 3
[09:14] <fabbione> ops
[09:16] <zul> hehe
[09:17] <zyga> does breezy have release shedule or codename yet?
[09:17] <pitti> zyga: breezy badger, to be released in october 05
[09:18] <zyga> pitti: thanks
[09:19] <zyga> pitti: so that makes it 10.5? 
[09:19] <pitti> 5.10, yes
[09:20] <pitti> zyga: btw, how far is your libintl-ng? :-)
[09:20] <zyga> pitti: frozen, had no time to look at it lately
[09:20] <zyga> pitti: I've been doing python gui magick 
[09:20] <zyga> pitti: but that is still on my mind once I'm more free to do what I like
[09:22] <zyga> pitti: I still kind of suck with ngettext support
[09:23] <zyga> my preliminary idea was to create list of all known n_plural strings and add generic support later
[09:23] <zyga> I wanted to reuse as little code as possible because of the incredible mess (or my reluctance to read it...) libintl presents
[09:24] <ross> thom: ping?
[09:24] <zyga> pitti: and one more thing... I most probably will need help to get glibc compatibility right
[09:25] <zyga> pitti: all those weak_alias magic
[09:26] <jbailey> zyga: What are you doing? =)
[09:27] <zyga> jbailey: rewriting libintl
[09:27] <zyga> jbailey: I want to add configuration stuff and flexibility on catalog placement
[09:28] <pitti> ... and make the code readable in the first place? :-)
[09:28] <zyga> jbailey: the first thing will be to smoothly support language-packs
[09:28] <zyga> pitti: that too :-)
[09:28] <zyga> and proper documentation built from source
[09:28] <jbailey> zyga: Are you hoping to replace the intl/ code in upstream glibc, or just have something we link against instead?
[09:29] <zyga> jbailey: not sure yet, I heard that upstream is pretty hard to convince as far as such things go
[09:30] <zyga> jbailey: first I want full compatibility when doing LD_PRELOAD in existing apps 
[09:30] <jbailey> zyga: Right.  They're usually not fond of rip-and-replace, but also you want to take care with copyright purity.
[09:30] <zyga> jbailey: then I was thinking about getting someone to review it and decide 
[09:30] <jbailey> zyga: If you can't assign the copyright over the to FSF, there's no way it can be considered.
[09:30] <zyga> jbailey: how do I do that?
[09:31] <jbailey> zyga: First thing is to make sure that you don't refer to code other than FSF copywritten code, or code to which you have the rights.
[09:31] <zyga> jbailey: that's pretty easy, what's next?
[09:31] <pitti> jbailey: you mean, refer to code he has _no_ rights for?
[09:32] <jbailey> zyga: You need to email assign@gnu.org and ask for an assignment form.  They will ask you all of the relevant questions, like what you've already modified, and what you plan to do.
[09:32] <zyga> jbailey: this sounds like lots of paperwork ;-)
[09:32] <zyga> jbailey: once it's usable I'll think about it :)
[09:32] <jbailey> pitti: English language needs ()'s: Don'y refer to code other than ( FSF copywritten code, or code to which you have the rights)
[09:33] <pitti> ah
[09:33] <zyga> jbailey: I'd be happy if I can provide libintl-foo GPLd and compatible with ubuntu
[09:33] <jbailey> zyga: After that you need your employer, school or whoever else might have a claim over your code to send in a disclaimer saying that they won't try to make a claim.
[09:33] <jbailey> It's worth postponing the paperwork until later, but you need to make sure that it's possible for you to do it after.
[09:33] <zyga> jbailey: no employer or school is involved in this -- I's all in my free time
[09:34] <jbailey> zyga: Depending on where you live, it being in your free time may not be good enough.
[09:34] <zyga> jbailey: really?
[09:34] <zyga> jbailey: explain please
[09:34] <jbailey> zyga: In the US in particular, many employers can claim that if you're a coder during the day that they own all the code you write even on your own time.
[09:35] <zyga> jbailey: geez... that's insane 
[09:35] <zyga> jbailey: I live in Poland
[09:35] <jbailey> zyga: Here in Canada, it has to do with what field you might be working in.  (Like when I worked at the Mutual Fund place, I would have needed a disclaimer for gnucash or oleo, but not for hotplug)
[09:35] <zyga> jbailey: I have no legal knowledge about this so I might need to ask a few friends
[09:35] <jbailey> hotplug is a bad example, since it's not GNU.  bUt inetutils or mailutils.
[09:36] <jbailey> zyga: If your employer is generally friendly, it's often worth it to just ask them to sign the document and be done with it.
[09:36] <zyga> jbailey: my work is totally unrelated to this, that's for sure 
[09:36] <zyga> jbailey: but thanks for the insight, I'll have to look more into it
[09:37] <jbailey> zyga: Then you may be able to simply say that you have no need for a disclaimer.  I don't pretend to know how European law works. =)
[09:37] <jbailey> I have enough trouble keeping up with my own country's.
[09:38] <zyga> jbailey: europe == lot's of different countries and lots of different laws
[09:50] <mvirkkil> Is this BOF about restricted formats? http://udu.wiki.ubuntu.com/UbuntuDownUnder/BOFs/UbuntuDevelopment/MultiMedia
[09:51] <dilinger> jbailey: i thought that was just a contractual thing?
[09:51] <dilinger> jbailey: people in the US signing contracts when they begin employment that state all code they write is owned by The Company
[09:51] <jbailey> dilinger: I might be mistaken on the US law.  It also wouldn't surprise me if it varied state to state.
[09:51] <jbailey> Or if it were implied in the abscence of anything else.
[09:51] <dilinger> have companies actually been successful in claiming that without a contract?  that'd be evil...
[09:52] <\sh> dilinger: it's the same here in germany...u write code for company, u have the copyright, but the company the right to use it...
[09:53] <jbailey> dilinger: ISTR the assignment clerk saying it was a problem.
[09:58] <dholbach> hellas
[09:58] <ogra> dholbach, holla
[09:59] <\sh> holla dholbach 
[09:59] <dholbach> hey ogra, \sh :-)
[10:05] <pitti> mdz: meeting now?
[10:25] <lamont> Kamion: will do
[10:46] <doko> jbailey: ppc biarch looks fine. currently in the testsuite. the build will finish tomorrow
[10:47] <jbailey> doko: Nice.  I'm trying to find something that definatively sais if NPTL is known to break Xen and UML. 
[10:48] <jbailey> It looks like Xen works, but faces a performance penalty, and there's no mention of nptl on user-mode-linux.sf.net
[10:48] <jbailey> (acc to google)
[11:07] <jcole> are there any 2.4 ubuntu kernels out there?
[11:07] <dholbach> jcole: i386 ones in universe
[11:07] <jcole> dholbach: really!
[11:08] <thom> pitti:  daniels> when pitti comes back, could someone please tell him dbus sources are at p.u.c/~daniels/dbus/?
[11:08] <dholbach> jcole: yeah should be
[11:08] <pitti> thom: cool, thanks
[11:09] <pitti> thom: can I please have the php4 build dependencies on concordia's amd64 dchroot?
[11:09] <jcole> dholbach: do they have all the kernel "crack" the 2.6 kernels have?
[11:09] <thom> pitti: i suppose so
[11:10] <dholbach> jcole: they are debian's 2.4 kernels, we don' actually support them, just have them around for convenience
[11:10] <pitti> thom: I need to test-build with gcc-4.0
[11:10] <thom> pitti: doooom!
[11:10] <pitti> thom: ?
[11:12] <thom> pitti: presumably you'd also like gcc4?
[11:12] <pitti> thom: that's already installed
[11:12] <pitti> thom: I already did some builds and FTBFS fixes with gcc-4.0
[11:12] <thom> oh, so it is
[11:12] <thom> cool cool
[11:12] <mdz> doko: you wanted to talk about the toolchain?
[11:12] <pitti> indeed :)
[11:13] <thom> pitti: done
[11:13] <pitti> thanksalot
[11:14] <doko> many people are in vacation next week (me as well), so I'd like to start the ABI transition not before the UDU week. that doesn't hinder us to upgrade the packages in this time (i.e. glibc, binutils and GCC, without changing the defaults)
[11:14] <jbailey> doko: The C++ one?
[11:15] <mdz> doko: I think the decision to be made is whether we delay uploads to breezy until GCC is in
[11:15] <doko> jbailey: yes, we shouldn't consider architecture specific changes (if there are, fabbione and lamont should tell us for sparc and hppa)
[11:16] <jbailey> doko: ISTR for Sparc anyway that we decided at Debconf4 that the abi changes there didn't amount to enough to care about.
[11:16] <doko> mdz: I don't see a reason why we should. IMO it's good to check packages with a new glibc first, without transitioning the C++ ABI
[11:16] <seb128> you don't want to delay the uploads for 3 weeks ?
[11:17] <mdz> seb128: no, we don't
[11:17] <seb128> there is like 40 GNOME tarball already to package, and that's tricky for the bug management
[11:17] <seb128> *pfiou*
[11:17] <mdz> doko: we would like to be sure that things still work when rebuilt with the new gcc
[11:17] <doko> seb128: why?
[11:18] <mdz> doko: what is the best way to do that?  track which packages are rebuilt since the upload, and to manual uploads to rebuild the others?
[11:18] <seb128> doko: because accumulation new package and bug fixes for weeks is not nice
[11:18] <seb128> doko: the sooner we can get the running with breezy the better imho
[11:18] <fabbione> seb128 ++
[11:18] <doko> mdz: we can test that with a test rebuild. just recompile everything in another chroot, with the gcc-defaults package pointing to 4.0
[11:18] <seb128> doko: ie: you are flooding me with gcc4 ftbfs for GNOME stuff fixed upstream
[11:18] <fabbione> doko: but do we really need to change the toolchain for breezy?
[11:18] <mdz> doko: that does not tell us whether the packages _work_, only whethere they build
[11:19] <fabbione> can't we get to test it in deep for breezy+1?
[11:19] <mdz> s/whethere/whether/
[11:19] <doko> seb128: yes, these reports are bug reports from the Debian BTS. feel free to close them, if they are fixed.
[11:19] <seb128> fabbione: is there a reason to not do it now ?
[11:19] <mdz> we can either do it now, or in 6 months.  I would prefer to do it now.
[11:19] <seb128> fabbione: FC4 already uses gcc4 now ...
[11:20] <seb128> me too
[11:20] <dholbach> the MOTUs will suffer, but stand behind a toolchain change :-)
[11:20] <doko> fabbione: good question. if you can assure me, that sid is unfrozen after the breezy freeze, then maybe not.
[11:21] <fabbione> seb128: because i don't think we can wait 3 weeks to start breezy..
[11:21] <lamont> doing it now helps us lead the charge
[11:21] <mdz> gcc-4.0 for breezy is worthwhile
[11:21] <ajmitch_> dholbach: we'll at least get to get plenty of malone testing in 
[11:21] <mdz> we do not need to wait 3 weeks
[11:21] <lamont> much like hppa wound up doing with gcc-3.0
[11:21] <seb128> fabbione: me neither, but I would just push gcc4 NOW
[11:21] <seb128> and start uploading
[11:21] <fabbione> seb128: ok..
[11:21] <dholbach> ajmitch_: ha, exactly :-)
[11:21] <seb128> let's face the transition :)
[11:21] <doko> let's assume that sid unfreezes before the breezy freeze, then we'll have much less time for the transition
[11:21] <fabbione> seb128: it's not like i really care :) i can't compile the kenrel with gcc4 anyway
[11:22] <mdz> fabbione: oh?
[11:22] <ogra> doko, is that likely to happen ?
[11:22] <seb128> fedora guys are fixing the whole GNOME for FC4 now
[11:22] <seb128> so I don't really care neither :)
[11:22] <fabbione> mdz: according to doko is still no go and zul can't get to compile all of it yet
[11:22] <doko> lamont: there are people already doing rebuilds for amd64 and ppc64, so that should give us more confidence
[11:23] <seb128> FC4 is shipping with gcc4 in 2 months, that should be possible to face it for breezy too
[11:23] <fabbione> doko: just let me know when you are going to switch default compiler
[11:24] <fabbione> seb128: FC4 doesn't have 16000 pkgs in universe.. our MOTU's might not make it at all
[11:24] <jbailey> fabbione: ARe you at least okay on gcc-3.4?  Can -3.3 be safely dropped?
[11:24] <doko> fabbione: I think it's your decision, which compiler to use for kernel builds
[11:24] <jbailey> (after the C++ transition)
[11:24] <fabbione> jbailey: nope.. i am going to stay with 3.3 if 4.0 doesn't work
[11:24] <ogra> fabbione, hey...a bit more confidence please :)
[11:24] <mdz> doko: can we use gcc-4.0 and g++-3.3, to delay the ABI transition while still testing gcc-4.0?
[11:24] <fabbione> doko: last time we talked you asked me not to build the kernel with 4.0
[11:24] <fabbione> doko: what did change since than?
[11:25] <jbailey> mdz: Current glibc still requires gcc-3.{34}
[11:25] <fabbione> ogra: i would be really happy if you can make it easily.. but it is really a lot of work
[11:25] <jbailey> The biarch ppc testing we've been doing has been with gcc-3.4
[11:25] <mdz> jbailey: current glibc, or the one you'll be uploading soon?
[11:25] <doko> mdz: I think yes, but I'll have to check. but this strategy won't work for java (dependingon C++)
[11:25] <jbailey> The one I'll be uploading soon.  Sorry.  Upstream's view of "current" =)
[11:26] <ogra> fabbione, but were also growing and we have a dholbach ;) lets see....
[11:26] <fabbione> ogra: ehehhe
[11:26] <dholbach> ogra: hahaha :-)
[11:26] <mdz> so the kernel, glibc and a few hundred other packages in main can't be built with gcc-4.0 ;-)
[11:26] <ogra> :)
[11:26] <fabbione> mdz: kinda :)
[11:26] <fabbione> mdz: that's why i don't see the need of leading the transition
[11:26] <fabbione> what gain do we really have?
[11:26] <pitti> mdz: fortunately many main packages already have patches
[11:26] <mdz> the kernel and glibc we can use gcc-3.4
[11:27] <mdz> the other packages we can fix
[11:27] <fabbione> mdz: 3.3 for the kernel
[11:27] <doko> fabbione: nothing changed. if I did say something wrong. for woody we did have some discussion if we should provide a kgcc binary, which the kernel uses for compilations. I'm not going to propose something for the kernel, which is not proposed upstream
[11:27] <mdz> fabbione: 3.4 is no good?
[11:27] <mdz> we could use -3.3 and drop 3.4 then.  I don't want to have three bracnhes
[11:27] <mdz> branches
[11:27] <fabbione> mdz: i don't feel comfortable in changing compiler right now that i am also changing major release
[11:27] <jbailey> doko: Any idea if we can do biarch ppc with 3.3?
[11:28] <fabbione> mdz: i need to be able to see regressions clearly and not messed up by gcc even more
[11:28] <seb128> mdz: if that's any consolation GNOME works fine with gcc4 :p
[11:28] <doko> jbailey: I tried for amd64 and that one didn't work well.
[11:28] <fabbione> doko: ok
[11:28] <tseng> is fc4 building with gcc4, or just packaging it?
[11:28] <doko> tseng: building
[11:28] <fabbione> doko: but talking about gcc4.. what benefits will bring to us?
[11:29] <doko> fabbione: java
[11:29] <fabbione> becuase i am not at all that convinced yet
[11:29] <mdz> doko: I need to know if there is a way to get gcc-4.0 in while delaying the ABI transition
[11:29] <fabbione> doko: and what else?
[11:29] <jbailey> fabbione: Hand written C parser is supposed to give better error information.
[11:29] <doko> fabbione: upstream maintainance
[11:29] <mdz> upstream++
[11:30] <fabbione> i agree on the last 2 points
[11:30] <doko> mdz: you mean gcc -> gcc-4.0 and g++ -> g++-3.3 ?
[11:30] <mdz> doko: by any means
[11:30] <fabbione> brb
[11:31] <doko> mdz: what's your goal with this combination?
[11:31] <mdz> doko: the idea is to make the most of 1) the time we have over the next few weeks, 2) the build testing we get 'for free' during the sid merge
[11:32] <zyga> mvo: hey
[11:32] <mvo> hey zyga 
[11:32] <zyga> mvo: when can we start commiting changes to u-m?
[11:32] <mdz> doko: since we cannot complete the transition for several weeks, we must either a) stay with 3.3 and transition later, or b) go to 4.0 but delay the transition
[11:33] <mvo> zyga: I integrated the patch today, I need to have a look over the diff again and will probably commit tomorrow morning 
[11:33] <mdz> doko: I am asking you if b) is possible; if not, we must go with a)
[11:34] <zyga> mvo: which patch?
[11:35] <mvo> zyga: the stuff at http://www.suxx.pl/update-manager
[11:36] <doko> mdz: I understand. Thinking about b), please let me delay the answer until tomorrow. I currently don't see reason, why this should not work, but usually you discover the failing bits later ...
[11:36] <zyga> mvo: okay, I've updated that just a moment ago, warty-updates was nowhere in aptsources.py.in
[11:37] <jani> fabbione, how can I get vmlinux (not vmlinuz) for current kernel?
[11:37] <zyga> mvo: I also marked some debian strings translatable (they should be IMHO)
[11:37] <fabbione> jani: zcat vmlinuz > vmlinux
[11:37] <jani> it says it;s not gzip format
[11:38] <jani> it's x86 boot sector
[11:38] <fabbione> jani: rename the file to something .gz
[11:38] <ogra> bz2cat ?
[11:38] <jani> nope that didn't work either
[11:38] <jani> isn't that compressed with zlib but not gzip?
[11:38] <mdz> doko: ok, I do not expect to be able to upload to breezy until tomorrow at the earliest anyway
[11:38] <dholbach> i think, i'll be off to bed now... have a nice evening :-)
[11:38] <mvo> zyga: right now I only looked at the patches directory
[11:38] <jani> ie it;s no gzip in makefile that I saw
[11:38] <ogra> ciao dholbach 
[11:38] <crimsun> need to either strip the boot sector and magic, or compile it and use vmlinux in the root of your kernel source
[11:38] <zyga> mvo: they are mostyl outdated
[11:39] <dholbach> bye ogra 
[11:39] <doko> btw, does somebody know about elmo?
[11:39] <crimsun> ^ jani 
[11:39] <zyga> mvo: you probably want to take a look at a big diff or wait untill I start sending them myself
[11:39] <jani> yup I asked if we have some deb package so I don;t need to recompile
[11:39] <jani> my own kernel
[11:39] <jani> night dholbach
[11:40] <jani> ok I'll try reversing what the build prcess does to vmlinux
[11:41] <jani> another Q: what about pacthing dch so it knows about ubuntu revisions while still working fin for debian 
[11:41] <jani> like : http://sourcecontrol.net/~jmonoses/dch.diff
[11:42] <fabbione> doko: in the worst scenario.. how complex would it be to revert back to gcc-3.3 ?
[11:42] <dholbach> jani: i wrote a similar patch for dhmake :-)
[11:42] <mvo> zyga: hrm, ok. would you mind sending the diff to me tonight or tomorrow? I will have a look then
[11:42] <fabbione> doko: given that some packages cannot be easily fixed for gcc4?
[11:42] <jani> :)
[11:42] <ajmitch_> dholbach: night ;)
[11:42] <jani> 
[11:43] <doko> fabbione: if they cannot be fixed for 4.0, try with 3.4 first.
[11:43] <dholbach> ajmitch_: you're right, should be off now
[11:43] <dholbach> bye :-)
[11:43] <fabbione> doko: uh? would that generate a set of packages compiled with A and another one with B?
[11:44] <doko> worst scenario: increase the epoch on gcc-defaults binaries. we keep the shlibs anyway.
[11:44] <fabbione> doko: + mdz was talking about dropping at least one version of 3.x
[11:44] <doko> A and B are ABI compatible
[11:44] <doko> yes, 2.95 ;-)
[11:44] <doko> we'll need 3.4 for g77 anyway.
[11:45] <doko> even in main, if gfortran doesn't work
[11:46] <Gagatan> oooh.. f77.. brings back memories
[11:46] <mdke> ogra, ping
[11:46] <ogra> mdke, pongedipong
[11:47] <mdke> https://www.ubuntulinux.org/wiki/ChangeMe or http://www.ubuntulinux.org/wiki/UbuntuGIS, which is newer?
[11:47] <doko> amu, Riddell, haggai: anybody knows how KDE build with GCC 4.0 ?
[11:47] <ogra> grrr....
[11:47] <mdke> ogra, ;) they have slightly different content
[11:48] <ogra> mdke, the one with a name :)
[11:48] <mdke> ok
[11:48] <mdke> cool
[11:48] <lamont> doko: we have no 2.95 in the archive
[11:48] <lamont> well, no binaries...
[11:48] <lamont> it would need bootstrapped, and I've specifically said 'when hell freezes over' (or mdz tells me I must...)
[11:49] <doko> ouch ;-) but Mithrandir wanted to put the whole GCC history back to egcs into ia32-libs ... ;-P
[11:49] <Riddell> doko: should build fine
[11:49] <lamont> heh
[11:49] <fabbione> doko: that's because he needs to :)
[11:53] <doko> Riddell: please have a look at amu's buglist, I reassigned all GCC 3.4/4.0 related bug reports to him
[11:53] <Riddell> doko: kde developers say 'it compiles fine but the resulting code may be broken'
[11:54] <doko> yes, that should be addressed in RC2 or the final release
[11:55] <doko> Riddell: I see you modified the wiki, so you are aware of the renaming policy ...
[11:55] <Riddell> renameing?
[11:56] <Riddell> oh for C++ packages
[11:56] <doko> yes, renaming of library packages depending on libstdc++
[11:57] <Riddell> sounds fun
[11:59] <Riddell> doko: is there a recommended pattern to follow for the rename?
[12:00] <doko> https://www.ubuntulinux.org/wiki/BreezyToolchainTransition