[12:43] <Kamion> lamont: ppc-netboot> yes?
[12:44] <lamont> yes
[12:44] <lamont> how?
[12:44] <lamont> on my poor iMac G3.
[12:44] <Kamion> first, have a network card supported by Open Firmware
[12:44] <lamont> most notably, overriding the boot from disk...
[12:44] <lamont> grumble
[12:44] <lamont> is builtin lan...
[12:44] <Kamion> should be fine
[12:44] <[Cliff] > hi all
[12:45] <Kamion> lamont: when you boot, hold down cmd+opt+o+f simultaneously, until the Open Firmware prompt appears
[12:45] <lamont> then what do I need where on the dhcp/tftp server?
[12:45] <[Cliff] > i'd like to tweak some stuff on metacity. can any one here lend me a hand?
[12:45] <Kamion> lamont: usual netboot stuff on the DHCP server, 'filename' should point to yaboot
[12:45] <Kamion> copy yaboot from the archive] 
[12:45] <lamont> right - copied all of powerpc/netboot
[12:46] <[Cliff] > what i'm trying to do is to enable the mousescroll on the titlebar to enable shading (kind of what xfwm does)
[12:46] <Kamion> lamont: so you have initrd.gz, vmlinux, yaboot, yaboot.conf; correct?
[12:47] <lamont> and boot.img.gz
[12:47] <Kamion> lamont: make sure yaboot.conf has vmlinux and initrd.gz paths that match what's needed over TFTP
[12:47] <Kamion> lamont: then 'boot enet:' at the OF prompt should do it
[12:47] <Kamion> lamont: failing that, try 'boot enet:0'; failing that, try 'boot enet:0,/path/to/yaboot'
[12:49] <thom> meh, how did you guys break torrent?
[12:50] <dredg> just looking at modconf (not currently building). it build-deps on kernel-source, which is not a package itself. any ideas on how to proceed? should modconf depend on a particular version of kernel-source? suggestions greatly appreciated
[12:51] <diamond> dredg: the buildd handles that situation alright, picking the first candidate for the kernel-source name,
[12:52] <diamond> dredg: the issue is that it expects the kernel source to be expanded in /usr/src
[12:52] <dredg> ah, you're right
[12:52] <dredg> my bad
[12:52] <diamond> dredg: was going to poke it earlier, but my dsl connection didn't like the look of the 40M download just to play ,-)
[12:53] <Kamion> thom: we'd all like to know ...
[12:53] <dredg> diamond: which is why i'm tempted to put an ubuntu uml together with 10M dedicated :)
[12:53] <diamond> dredg: yeah well, i hate you. -)
[12:54] <Kamion> dredg: surely that'd want to be "linux-source" for Ubuntu kernels
[12:54] <Kamion> and you'd need to cope with other related renamings ...
[12:54] <Kamion> I looked at modconf ages ago, it's non-trivial, so good luck :)
[12:55] <dredg> Kamion: no, the source packages seem to be called kernel-source
[12:55] <dredg> ie there's no linux-source-2.6.9
[12:55] <dredg> (and thanks. now i don't wanna do it ;) )
[12:55] <Kamion> dredg: I think you'll find that there *is* most certainly a linux-source-2.6.10
[12:56] <dredg> interesting
[12:56] <Kamion> the kernels we imported from Debian are kernel-*; that's the Debian naming scheme
[12:56] <dredg> apt-get build-dep modconf
[12:56] <Kamion> when we started doing our own kernel packaging, we also changed the name
[12:56] <dredg> Package kernel-source is a virtual package provided by:
[12:56] <dredg>   kernel-source-2.6.9 2.6.9-3
[12:56] <dredg> and so on
[12:56] <Kamion> yes, linux-source-* does not provide kernel-source, it provides linux-source
[12:56] <diamond> Kamion: ah. yes. i too was thinking it was weird that both the package name was kernel-source, and that there was no 2.6.10 version
[12:57] <diamond> that would certainly explain it
[12:57] <Kamion> unfortunately I don't really remember what the issues I ran into were - I just remember thinking "hmm, this is more effort than s/kernel/linux/g"
[12:57] <Kamion> it's possible I was trying to make it work with either Debian or Ubuntu kernels, although I sort of doubt it
[12:58] <dredg> ok
[12:58] <Kamion> ah, I know, I think nobody ever actually updated modconf to work properly with 2.6 kernels either
[12:58] <Kamion> even in Debian
[12:58] <dredg> i'll poke it til it either works or i cry
[12:58] <diamond> Kamion: *cringe*
[12:58] <Kamion> so IIRC it needed a fair bit of hacking for that
[12:58] <dredg> ok, i'm bailing out now then
[12:59] <Kamion> don't let me discourage you though, it would be great if somebody did it - enlist somebody who's familiar with Ubuntu kernels
[01:00] <lamont> Kamion: btw, enet:0 worked 
[01:00] <lamont> and local mirrors work best for installing when you have .udeb's. :-)
[01:02] <Kamion> lamont: yep, think yourself lucky that choose-mirror now actually asks for the local mirror ;)
[01:02] <lamont> Kamion: nah - firewall blocks the house machines from hitting archive.u.c :-)
[01:02] <lamont> so it errors out
[01:02] <seb128> thom: around ?
[01:03] <lamont> Kamion: btw, would be wonderful if the cdimage happened to include the netboot files... :0-)
[01:03] <Kamion> lamont: couldn't remember whether the 0 was needed, I haven't kept notes of how to do powermac netboot because it's more or less fresh enough that I can remember it with a little bit of trial and error as needed
[01:03] <lamont> Kamion: way beyond my memory though.  thanks.
[01:03] <Kamion> lamont: problem with that is that it makes the *3 kernel/initrd issue even worse
[01:04] <lamont> son of bitch... where does debootstrap get it's idea of the keyring?\
[01:05] <ogra> jdub, ping
[01:09] <Kamion> lamont: ubuntu-keyring-udeb
[01:10] <Kamion> lamont: the filename is passed to it by base-installer.postinst
[01:11] <lamont> yeah, fixoring
[01:12] <dredg> Kamion: on further inspection, modconf appears to have some knowledge of 2.6
[01:12] <dredg> it just expects the source to be untarred in /usr/src/linux which would currently require manual intervention
[01:14] <diamond> dredg: that's bad and wrong
[01:14] <lamont> what's the d-i magic for installing a udeb?
[01:14] <diamond> dredg: the kernel source should not be untarred into /usr/src/linux
[01:14] <dredg> many things are bad and wrong. let's see if i can make it handle it sanely
[01:14] <lamont> (what do I type in vc2, again?)
[01:14] <lamont> Kamion: ^^
[01:14] <diamond> dredg: hehe, k
[01:15] <Kamion> lamont: anna-install <udeb>
[01:15] <lamont> ciik
[01:15] <lamont> cool, even
[01:16] <dredg> diamond: well, if you a linux-source package it drops a tarball in /usr/src. i could untar it to a temp directory
[01:16] <diamond> dredg: yeah, that's much better
[01:16] <thom> seb128: yo
[01:16] <thom> mjg59: glad you made it out alive
[01:17] <lamont> Kamion: that's gotta be the most bizzare error I've seen in a while...
[01:17] <seb128> thom: update gtk and let me know if the booog is fixed for you too :)
[01:17] <lamont> 'failed to fetch adduser'  -->  Missing dists/hoary/main/binary-powerpc/Packages (although .gz existed)
[01:18] <lamont> Kamion: for breezy, d-i should understand that there might not be a Packages file...
[01:18] <thom> seb128: rock, you found it?
[01:18] <mjg59> These HP Bioses are approximately the worst thing ever
[01:18] <Kamion> lamont: er, I thought debootstrap already did
[01:18] <Kamion> lamont: it certainly looks for .bz2 and .gz
[01:18] <diamond> mjg59: on proliants?
[01:19] <mjg59> No, on HP Compaq things
[01:19] <seb128> thom: yep, I've tracked it down and a gtk guy has made a ximian bug with a patch public when I've pointed it ...
[01:19] <Kamion> lamont: might be buggy though, I guess
[01:20] <seb128> thom: bad suse guys hide patches in their bugzilla :p
[01:20] <lamont> hrm.. could be because there is an entry in Release, but no file...... 
[01:20] <thom> seb128: hahah
[01:20] <diamond> mjg59: aye, the only experience i've had with them is on some of the hp proliant servers
[01:20] <Kamion> lamont: heh. DDTT
[01:20] <lamont> DDTT?
[01:20] <mjg59> Argh now epiphany has blown up on me
[01:20] <lamont> don't do that then
[01:20] <lamont> or don't do that turdball, not sure which.
[01:20] <Kamion> although it would certainly be nice if debootstrap behaved more permissively there
[01:20] <Kamion> don't do that then :-)
[01:22] <sabdfl> night all
[01:27] <lamont> Kamion: so it gives me the choice of which kernel to install.  linux-powerpc being the only choice...  hrmpf
[01:28] <lamont> hrmm.. and it fails to install... bet that could have something to do with it...
[01:28] <Kamion> lamont: yeah, sounds like you've dropped down to expert mode somehow due to errors
[01:29] <Kamion> lamont: check /var/log/messages
[01:29] <lamont> does apt use md5 or sha1 sums?
[01:30] <diamond> lamont: md5, afaik
[01:30] <seb128> thom: bog ?
[01:31] <diamond> lamont: apt-cache show bc | grep MD5sum
[01:31] <diamond> (for example)
[01:31] <lamont> diamond: for Release files, that is
[01:31] <diamond> lamont: as in, the file list in Release files? 
[01:32] <Kamion> Release has both md5 and sha1
[01:32] <thom> seb128: can't see it
[01:32] <Kamion> I think apt knows how to check either; dunno if it can check both
[01:32] <mdz> apt only pays attention to the md5s in Release
[01:32] <mdz> (currently)
[01:32] <Kamion> ah, I just noticed apt-pkg/contrib/sha1.*
[01:33] <seb128> thom: rock
[01:33] <Kamion> I've been meaning to make a bunch of d-i stuff check sha1 instead of md5
[01:33] <mdz> yes, it has SHA1 code in it (and apt-ftparchive release will generate the SHA1 hashes), but apt doesn't check them when downloading
[01:33] <mdz> it needs some refactoring
[01:34] <Kamion> probably not worth it if apt doesn't, though
[01:34] <Kamion> imo it isn't worth the computing time to check both
[01:34] <thom> seb128: yeah, looks good to me
[01:34] <diamond> also sha1 now has known brute-force weaknesses
[01:34] <seb128> nice
[01:35] <tseng> diamond: so does every crypto algorithm
[01:35] <Kamion> diamond: only collision, not second-preimage
[01:35] <tseng> its just a matter of time.
[01:36] <Kamion> tseng: sha1 has known collisions (which is worse than the plain stupid brute-force attack), but it doesn't have a known second-preimage attack yet AFAIK which is what you'd need for significant attacks
[01:36] <Kamion> on the Release/Packages scheme
[01:36] <diamond> Kamion: aye, but it's a good enough reason to be moving away from it
[01:36] <diamond> any such migration is going to take quite some time anyway
[01:36] <tseng> as I understand it its inevitable for a known colision to happen anywhere
[01:37] <Kamion> diamond: sure, sha-256 or sha-512 - but those are just pushing back the barriers a little, and there's some reason to believe they'll fall soon. so far there exists no hash algorithm that isn't vulnerable to the basic attack
[01:37] <diamond> Kamion: *nod*
[01:37] <lamont> and then restarts the install
[01:37] <lamont> sigh
[01:37] <Kamion> tseng: when cryptanalysts speak of collision attacks, they are talking about something more interesting than that
[01:37] <Kamion> anyway, bedtime
[01:37] <diamond> i'll leave ye folks to it anyway, i have decided on sleep. nite
[01:37] <Kamion> lamont: boot with archive-copier/copy=false
[01:37] <lamont> server
[01:37] <Kamion> heh
[01:38] <thom> Kamion: i've been meaning to ask, what the heck is a "second pre-image"?
[01:38] <lamont> just give me a working install dammit!
[01:38] <thom> (appropriate pointers to RTFM happily accepted)
[01:39] <Kamion> thom: second-preimage> given M and H(M) = X, find M' such that H(M') = X
[01:40] <Kamion> i.e. find second message (preimage) with same hash (image)
[01:40] <Kamion> tseng: an interesting collision attack is one where you can find a collision more cheaply than could be done by plain guesswork
[01:41] <Kamion> and since people generally rely on the expense of finding collisions, that is not something to shrug off with "oh, but it would have happened anyway"
[01:41] <Kamion> but still, a collision attack is only: find M and M' such that H(M) = H(M')
[01:41] <Kamion> for a collision attack, you aren't aiming at any particular hash
[01:42] <mdz> Kamion: night, great work on RC
[01:42] <Kamion> in order to attack a Release file, you're trying to find another Packages file with the same hash, which is a second-preimage attack problem
[01:43] <Kamion> anyway, really bed
[01:43] <Kamion> mdz: thanks, night
[01:45] <lamont> File does not exist: /org/ubuntu/tree/dists/hoary/main/binary-powerpc/Packages.bz2
[01:45] <lamont> now what did I do to piss it off?
[01:47] <thom> Kamion: right. thanks :-)
[02:29] <lamont> "Please wait, loading kernel...\n"
[02:29] <lamont> and then all the time in the world...
[02:29] <lamont> I don't think it likes my G3...
[02:30] <mdz> we get a regular stream of bug reports about "B&W G3" machines
[02:30] <lamont> this one is color!
[02:30] <lamont> B&W is unsup, colour is officially supported, last I heard...
[02:30] <lamont> (that's why I bought it...)
[02:31] <lamont> hung with just return (or timeout), and with "/boot/vmlinux video=ofonly"
[02:45] <ogra> hmm, in total 72 from 892 hwdb submissions have a dectivated swap partition, interesting...
[02:46] <Burgundavia> ogra: I only bothered making one a little while a go, and I don't see a performance increase
[02:47] <ogra> Burgundavia, depends.... this guy wont be happy without swap: http://hwdb.ubuntu.com/?xml=3b301d3f8bc58a58edc0bafc6af59c45
[02:47] <ogra> only 240MB mem are not very good....
[02:47] <dredg> i always have one. just because you have oodles of ram doesn't mean that any apps that are sitting idle should be using it
[02:48] <Burgundavia> ogra: ouch. I have a gig, so I am in a little different boat
[02:48] <Burgundavia> dredg: that is waht sold me on getting a swap partition too
[02:49] <ogra> heh, while we speak the next one rushes in http://hwdb.ubuntu.com/?xml=b09ef584490854dff848b4952485be13
[02:49] <Burgundavia> ogra: is it mailing now?
[02:49] <dredg> or at least, i cannot hear it
[02:49] <ogra> Burgundavia, nope, HTTP-POSTing
[02:49] <Burgundavia> ogra: ok
[02:49] <Burgundavia> ogra: how do I tell what my id is?
[02:50] <ogra> Burgundavia, read whats written under the id search field ;)
[02:50] <ogra> does only work if you submitted online...
[02:50] <Burgundavia> ogra: hmm, that reading thing is really getting to me
[02:53] <zenwhen> so is the current image the RC
[02:53] <zenwhen> ?
[02:54] <ogra> zenwhen, the image with -rc- in its name... 
[02:54] <Burgundavia> ogra: usability suggestion with your send to server thing
[02:55] <Burgundavia> ogra: make it part of the same window and provide some better feedback
[02:56] <CarlK_afk> is the RC the same as rsync cdimage.ubuntulinux.org::cdimage/daily/current/hoary-install-i386.iso ?
[02:58] <ogra> Burgundavia, i'm having plans with the separation of the scripts in mind for breezy, so i cant integrate it...
[02:59] <Burgundavia> ogra: ok
[02:59] <ogra> but i'm aware of the feedback problem....
[03:00] <Burgundavia> ogra: if it timesout, it doesn't give any feedback, that is what I just discovered
[03:00] <ogra> oh, please file a bug, i havent had this case yet
[03:01] <Burgundavia> I ran it from a terminal
[03:01] <Burgundavia> then I closed the terminal after the main window went away
[03:01] <Burgundavia> and then it timedout
[03:01] <Burgundavia> I am going to try and replicate it first
[03:01] <ogra> ah, ok....i didnt think anyone would run it from a terminal, hehe
[03:02] <Burgundavia> yep
[03:02] <Burgundavia> I can replicate it
[03:02] <Burgundavia> will file a bug for you
[03:02] <Burgundavia> should I file 2? one for feedback issue and 1 for terminal issue?
[03:03] <ogra> but i guess it doesnt happen if you run it from your menu
[03:03] <Burgundavia> no, if you run from menu it works fine
[03:03] <ogra> ok
[03:08] <Burgundavia> ogra: 8424 and 8425 are yours
[03:08] <ogra> thanks :)
[03:09] <ogra> oh, that one is interesting...http://hwdb.ubuntu.com/?xml=bc62bb4efe25dc7c19845369a9453bc9
[03:10] <Burgundavia> ogra: that is me
[03:10] <Burgundavia> my bios is crap
[03:10] <Burgundavia> doesn't detect my chip
[03:10] <ogra> Unknown CPU Typ is funny
[03:11] <Burgundavia> it is a mobile 2500+
[03:11] <Burgundavia> overclocked to 2.2 GHZ
[03:11] <schweeb> I sent a record in earlier
[03:12] <ogra> schweeb, you can look up your id and look if your cpu and mem data is correct...
[03:12] <schweeb> http://hwdb.ubuntu.com/?xml=c5e70fd84c43df20f8e4e091988721f3
[03:12] <schweeb> looks good
[03:12] <Burgundavia> ogra: the hwdb thing is very very cool. It is also completely unique in the linux world to the best of my knowledge
[03:12] <ogra> yeah
[03:13] <Burgundavia> ogra: how many PPC and amd64 overall?
[03:13] <ogra> Burgundavia, lets see how big we can grow it with a real DB :)
[03:13] <Burgundavia> ogra:  it is just flat data right now?
[03:13] <ogra> Burgundavia, havent got a DB yet, its just the interim solution to win a little time...
[03:13] <Burgundavia> ok
[03:13] <ogra> its all grep and ls output ;.P
[03:14] <Burgundavia> post hoary is where the major hacking is going to happen?
[03:14] <ogra> currently its in the source package and on the server....i will set up a bazaar repo...
[03:20] <HrdwrBoB> lamont: :/
[03:53] <dholbach> goooood morning
[03:53] <ogra> heh
[04:01] <jbailey> dilinger: There?
[04:20] <dilinger> jbailey: i was about to hop in the shower, i smell like sweat and chalk.  what's up?
[04:23] <mdz> ogra: how many submissions so far?
[04:23] <ogra> mdz, hwdb.ubuntu.com ;)
[04:24] <ogra> +180 on my disk that i still have to process manually
[04:24] <mdz> ogra: nice!
[04:24] <ogra> mdz, 77 of the online submissions have no swap :(
[04:25] <mdz> ogra: why is that a problem?
[04:25] <ogra> because we had a hibernate bug during the development process ... hibernate stuff
[04:26] <ogra> that broke the swapspace...
[04:26] <whiprush> oh hot, it accepting stuff from the clients and everything?
[04:26] <mdz> ogra: that can be fixed on upgrade eventually
[04:27] <ogra> so i guess a lot of these people dont know they have no swap.... especially these with 256MB =<
[04:27] <ogra> will hvae a very bad life currently
[04:27] <ogra> ergh s/</>
[04:28] <schweeb> whiprush: yes
[04:28] <mdz> jbailey was working on a fix, was he not?
[04:28] <whiprush> neat
[04:28] <mdz> I saw some relevant changes going in
[04:28] <ogra> the fix is to have 3xmem = swap on hibernate
[04:29] <ogra> which is done by the installer afaik
[04:30] <daniels> ogra: 
[04:30] <daniels>   File "/usr/bin/hwdb-xml", line 39, in __init__
[04:30] <daniels>     for line in self.xorgconf():
[04:30] <daniels>   File "/usr/bin/hwdb-xml", line 147, in xorgconf
[04:30] <daniels>     xconf.append("  <subsection name="+splitted[1] .strip()+">")
[04:30] <daniels> IndexError: list index out of range
[04:31] <ogra> ouch
[04:31] <mdz> mako: ping?
[04:32] <ogra> daniels, can you send me the xorg.conf ?
[04:32] <mdz> mako: I received what looks like an automated email about confirming my shipit order
[04:33] <mdz> mako: it doesn't explain what I should do in order to confirm, though
[04:34] <daniels> ogra: sure.  it's my own hacked one, which might explain it
[04:34] <ogra> ah, night be :)
[04:34] <ogra> might even
[04:34] <daniels> http://amnesiac.heapspace.net/~daniel/misc/xorg.conf
[04:35] <daniels> this is probably it:
[04:35] <daniels> SubSection<tab>"Display"
[04:35] <daniels> as opposed to a space
[04:36] <ogra> i'll check it .... but not yet, 4:30am is sleep time :)
[04:37] <ogra> thanks for the file
[04:37] <ogra> night all
[04:42] <daniels> g'night dude
[05:03] <jbailey> mdz: Yes, I have a patch against file (Accepted upstream), and the rest is a bit of shell script, planning to upload tomorrow.
[05:04] <mdz> jbailey: where is the shell script going to go?  it's very late in the release to put something like thi sin
[05:04] <mdz> it has to fail absolutely safe
[05:06] <jbailey> mdz: in mountall, there's a second swapon.  The bits of script looks for any swap partitions that aren't listed in /proc/swaps, and does a file -s /dev/FOO on them.  If the output string contains the string SWSUSP1 in it, it runs mkswap on it, and reruns swapon -a.
[05:07] <mdz> jbailey: will it cope with devfs pathname borkage?
[05:08] <jbailey> I'd have to test to be sure.  I think the worst that happens in that case is that it does extra 'file -s', which is a readonly test.
[05:08] <jbailey> The biggest risk is running a swapon against a live swap partition.  mkswap doesn't seem to have any sanity checks on that.
[05:09] <jbailey> Do you think that's too risky?
[05:09] <mdz> the biggest risk would be running a swapon against a partition with a filesystem on it :-)
[05:09] <mdz> er, a mkswap
[05:09] <jbailey> Erm, right. =)
[05:09] <mdz> this is to handle the case where a resume fails, right?
[05:10] <jbailey> right.
[05:10] <jbailey> Usually caused by someone doing a suspend, where the kernel guesses which swap partition to use, but there's nothing in grub/lilo/whatever to tell it to resume.
[05:11] <mdz> so it doesn't affect fresh installs, which set that up correctly by default?
[05:11] <jbailey> I think it could also happen if someone upgrades a kernel then suspends.
[05:11] <mdz> it seems like it would expose practically every user to a certain amount of risk, in order to cover a corner case
[05:11] <mdz> the worst of the corner case is that the user has no swap
[05:12] <jbailey> Right.
[05:12] <mdz> the worst of the risk is that they lose data
[05:12] <mdz> mkswap has practically no safety checks; it'll blow away whatever you ask it to
[05:13] <mdz> it's a very uncomfortable choice
[05:13] <jbailey> Yup.  Could put it in the release notes and note it as a mostfreq reported bug.
[05:14] <mdz> I haven't seen a lot of bugzilla activity around this issue; is it really that common?
[05:14] <jbailey> I don't know how to measure that.  I've helped 2 or 3 people in #ubuntu with it, and certainly hit it myself.
[05:14] <mdz> since it has a simple workaround, I'm inclined to delay the automatic fix until Breezy
[05:14] <jbailey> Fair 'nuff.  I'll put my patch in the bug and tag it 5.10
[05:14] <mdz> thanks
[05:30] <jbailey> dilinger: Sweat and chalk?  Do you do gymnastics?
[05:31] <infinity> jbailey : That's what I was thinking... Or marathon chalkbrush cleaning.
[05:32] <jbailey> *lol*
[05:32] <infinity> I'd feel silly going back to it.
[05:33] <infinity> I was on the national team when I was 10.  Kinda haven't done much since then.
[05:33] <infinity> At 27, I'm not quite as.. Bendy.
[05:33] <jbailey> There was a lovely club in Vancouver that had an adult gymnastics program that was really nice.  Wide range of folks and a few instructors.
[05:33] <Guilmon> damn. Me expected an answer, but I guess new people are active in #ubuntu 
[05:33] <jbailey> They'd let you hang out on the machines pissing around as long as you were being reasonably safe about it and a couple times an hours go to one of the stations and demonstrate/help.
[05:34] <Guilmon> anyone want to answer me a question because me blind? :P
[05:35] <dilinger> jbailey: heh.  rock climbing.  might as well be gymnastics tonight, though.  i definitely bent ways that i'm not supposed to
[05:35] <Guilmon> Me wishes to ge mountain climbing :)
[05:36] <Burgundavia> jbailey: your a canuck eh. come to #ubuntu-ca
[05:36] <Guilmon> Maybe in aslaska :)
[05:36] <jbailey> dilinger: Nice.  There's that whole "trust your belayer" bit that I have troubles with.
[05:37] <jbailey> Burgundavia: I did tell you guys that I wouldn't be there often.  I have 5 #ubuntu-SOMETHING windows open already...
[05:37] <Burgundavia> jbailey: ok
[05:37] <Burgundavia> jbailey: I was just wondering if you were going to make it down to LFNW
[05:37] <dilinger> jbailey: i've known my belayer for 5+ years
[05:37] <jbailey> Burgundavia: dict LFNW?
[05:37] <dilinger> jbailey: i have no problems trusting him
[05:38] <infinity> dilinger : All the more reason to distrust.  People who know you well are more likely to kill you.
[05:38] <Burgundavia> jbailey: Linux Fest Northwest
[05:38] <jbailey> dilinger: Cool.  I would probably learn to climb with my while.  She's cute, but occasionally she's as distractable as a golden retriever. =)
[05:38] <dilinger> infinity: i'm not sure if i give people motive enough to kill me
[05:39] <infinity> s/shut/shuts/
[05:39] <dilinger> jbailey: so anyways, what's up/
[05:39] <jbailey> Burgundavia: I see a collection of user groups involved, I don't see a city.
[05:40] <Burgundavia> jbailey: ? it is in Bellingham, WA
[05:40] <jbailey> Burgundavia: But in any event, no, I won't be attending.  I'm not available on April 30th for heading out that way.
[05:41] <jbailey> dilinger: I was reading the link you posted to.. Lemme find the bug.
[05:41] <jbailey> dilinger: 1763, ide auto detection stuff.
[05:42] <jbailey> Bartlomiej Zolnierkiewicz's patch for move the ide stuff to a driver model, and I see that it's in his sync set from March 8th.
[05:42] <jbailey> dilinger: How does it usually work from there?  I've never followed how patches get pulled from the ide folks into the main trees.
[05:43] <jbailey> dilinger: I'm wondering what sort of craziness we would have to face if suddenly people upgrade their kernels themselves to this.  Will they get breakage with ide-generic being loaded by hotplug and such?
[05:45] <Burgundavia> mako: ping
[05:52] <dilinger> linus needs to pull it; check out the netdev-2.6 posts that jgarzik does to lkml
[05:53] <dilinger> obviously, it would need to be tested.  i haven't been following bart's work, although i still intend to play w/ it once all the travel craziness and sarge stuff passes
[06:17] <infinity> lamont : Around?
[06:20] <fabbione> morning
[06:21] <lamont> infinity: yo
[06:21] <lamont> morning fabbione 
[06:21] <mako> Burgundavia: hey dude
[06:21] <fabbione> no daniels?
[06:24] <infinity> lamont : Two things.  Can you retry 'php4-universe' on all arches?
[06:24] <infinity> lamont : And once that's kicked off, can you rety php4, and see if it's still confusing the heck out of buildd/sbuild?  (And if so, help me figure out why?) :)
[06:26] <infinity> lamont : php4-universe was failing because it was in main, not universe.  mdz fixed that.  No idea why php4 is so confused, mind you, since the string "universe" doesn't show up anywhere in its source package.
[06:27] <mdz> it had to get into the archive before I could move it, so there was a largely inevitable window there where it would fail
[06:27] <infinity> Heh.
[06:27] <mdz> what's php4 doing?
[06:27] <infinity> mdz : It's not getting far enough to do anything.  sbuild's confusing it with php4-universe.
[06:27] <fabbione> hmmm it looks like people have some kind of love for "blocker" in bugzilla
[06:27] <mdz> oh dear, that's weird
[06:27] <mdz> fabbione: indeed
[06:28] <infinity> mdz : It downloads the source, then tries to open php4-universe.dsc, which obviously fails.  Very odd.
[06:29] <schweeb> fabbione: did you say you had completed those Xen kernels, or were you still workin on them?
[06:29] <infinity> I wonder if this has to do with binary/source packages of the same name coming from different sources, or something.
[06:29] <infinity> Oh, it probably does too.
[06:29] <mdz> lamont: I think you want apt-get --only-source source
[06:29] <infinity> Some broken heuristics for "apt-get source php4".
[06:30] <fabbione> schweeb: actually i do have the kernels, but they can't boot. For some reasons the final vmlinux is an ELF binary that grub can't understand, while a normal image is marked as x86 boot something
[06:30] <mdz> for values of 'broken' including 'gets it right in every case except the weird one you just created, and has a convenient override which should have been used here in the first place', sure :-P
[06:30] <infinity> mdz : Are you sure it isn't using --only-source already?
[06:31] <infinity> mdz : Without it, I'd expect it to download php4-universe's source too (since that's the source for the php4 binary)
[06:31] <fabbione> schweeb: i was hoping to fix them sometime during this week
[06:31] <mdz> infinity: I can think of no other explanation
[06:31] <fabbione> schweeb: or at least try to
[06:31] <mdz> oh, wait, this isn't apt at all
[06:31] <infinity> mdz : Still, fixing apt bugs right now seems bad, it might be easier to move the (empty) php4 meta-package to main.
[06:31] <mdz> sbuild does the unpacking
[06:31] <infinity> Oh, good point.
[06:31] <mdz> apt is getting the right package, and then sbuild is looking for something else
[06:31] <infinity> apt only downloads.
[06:33] <infinity> mdz : If lamont can't find/fix it reasonably quickly, I can just move the php4 package from universe to main.  It's only 1120 bytes.
[06:34] <mdz> lamont: a simple retry, now that I've moved php4-universe to universe, might suffice for now
[06:36] <mdz> https://www.ubuntulinux.org/wiki/CantInstallMP3Support
[06:36] <mdz> there's a new and creative way to ask for support without using the appropriate channels
[06:36] <infinity> Oh, ffs.  net-snmp FTBFS now.
[06:37] <infinity> (And not due to my change.. It never got that far)
[06:37] <infinity> Great.
[06:37] <mdz> weird; lamont test-built the entire archive recently
[06:37] <lamont> mdz: wow.  we should make somewhere where they can create things like that.
[06:37] <lamont> mdz: is test-building again
[06:38] <infinity> COuld be cosmic rays.  This machine has occasionally flaky RAM.  <tries again>
[06:38] <lamont> as of this morning ish?
[06:38] <calc> lamont: similar to pastebin?
[06:39] <mdz> ugh, it runs auto* during the build
[06:39] <mdz> no wonder it doesn't work
[06:41] <lamont> calc: is there a reason that php4-universe has a binary named 'php4'?
[06:41] <calc> no idea
[06:41] <lamont> that's probably part of the confusion somewhere
[06:42] <lamont> Package: php4-universe
[06:42] <lamont> Binary: php4-imap, php4-curl, php4-mhash, php4-gd, php4-odbc, php4-pear, libapache-mod-php4, php4-ldap, php4-domxml, php4-mysql, php4, php4-universe-common, php4-recode, php4-sybase, php4-xslt, php4-mcal, php4-snmp
[06:42] <infinity> lamont : Yes, we know that's the confusion.  I've offered a simple solution (I can move the binary), but this situation should be fixable anyway. :)
[06:42] <infinity> lamont : Also, I'm not calc. ;)
[06:43] <lamont> infinity: doh
[06:43] <infinity> lamont : THe only reason it's in that source package is cause someone put 'php4' in universe long ago, and I just divided the source packages to match the current layout.
[06:44] <infinity> I see no reason why moving the 1120 byte package to main would hurt anyone terribly.
[06:44] <infinity> (And it would allow for people to apt-get install php4, which they seem to love so much)
[06:44] <mdz> infinity: net-snmp builds just fine here (the version in hoary)
[06:45] <infinity> mdz : My second build is still going, so it much have been ghosts in the RAM.
[06:45] <infinity> s/much/must/
[06:45] <mdz> php4-in-main doesn't sound so evil, so long as it doesn't depend on apache 1.3.x
[06:45] <lamont> php4-universe is either building or uploaded now
[06:45] <mdz> but then you need to have php4 build it, not php4-universe
[06:45] <lamont> which should break the loop nicely, I expect
[06:45] <infinity> mdz : it depends on "libapache-mod-php4 | libapache2-mod-php4"
[06:46] <mdz> or does it already?
[06:46] <infinity> mdz : Uploading two new source packages to swap who builds it isn't rocket science.
[06:46] <lamont> php4-universe was looping on apache-dev
[06:46] <mdz> infinity: the old php4 package built php4, didn't it?  why move it?
[06:46] <infinity> mdz : And no, it's built by universe, specifically to keep my sanity at bay.  If I need two source packages, I'd rather have them both target specific suites.
[06:47] <lamont> sbuild should handle php4 just fine now that php4-universe is in universe
[06:47] <mdz> retrying?
[06:47] <infinity> mdz : The old one built everything it could that didn't build-dep on stuff in universe.  That was the whole point in the split, to get the extra stuff back.
[06:47] <infinity> Anyhow....
[06:47] <infinity> lamont : If it Just Works now, the point is moot.
[06:48] <infinity> lamont : If they both get past the unpack stage, let me know. :)
[06:48] <lamont> infinity: in the new world of php4, which source package delivers the php4 binary?
[06:48] <infinity> lamont : php4-universe delivers it currently.
[06:48] <mdz> infinity: but if php4 is just a metapackage, build-depends shouldn't enter into that particular choice
[06:48] <infinity> lamont : php4 provides all binaries in main, php4-universe provides all binaries in universe.  Like I said, to make me go less crosseyed when dealing with them. :)
[06:49] <infinity> mdz : Right, build-deps don't matter for it (or for many of the modules packages I moved).  It was purely a cosmetic "who builds what" thing.
[06:50] <infinity> mdz : Honestly, though, if we're going to provide the php4 metapackage at all (hey, there's idea #3, we could just drop it), it may as well be in main, where people will expect to be able to get at it.  If it's not in main, I don't see much point in having it at all.
[06:50] <lamont> Mar 31 05:50:02 buildd-mail: php4-universe must be manually dinstall-ed -- delayed
[06:50] <lamont> php4-universe binaries are NEW
[06:50] <mdz> infinity: I agree; if you'd proposed it before uploading, I most likely would have agreed
[06:50] <infinity> lamont : Known.
[06:50] <infinity> mdz : If I had known sbuild would freak out, I would have proposed. :)
[06:51] <lamont> accepted on the 3 though
[06:51] <mdz> need to let php4-universe stuff get into the archive, then move it to universe
[06:51] <lamont> that is, uploaded, almost certainly sitting in NEW/
[06:52] <infinity> Alright, well, if sbuild doesn't sort itself, I can swap who builds the php4 metapackage with no real issues.
[06:52] <infinity> And then if you want to re-seed it to match, that's your choice.
[06:52] <infinity> Deal?
[06:53] <mdz> I would like to have php4 in main regardless, to be honest
[06:53] <infinity> Oh.  In that case, I'll just do another upload to move it from one source to the other.
[06:53] <infinity> And you can fix the seeds.
[06:53] <infinity> And we'll all live happily ever after.
[06:53] <mdz> assuming germinate DTRT
[06:54] <infinity> And net-snmp finished fine here.  False alarm.
[06:54] <mdz> is that "| caudium-php4" really useful?
[06:54] <infinity> It's not there anymore.
[06:54] <mdz> considering we don't have such a package
[06:55] <infinity> Debian doesn't have such a package either anymore.
[06:55] <infinity> And caudium-php4 shouldn't be in php4's dep list at all in either dist, unless I missed it.
[06:55] <infinity> But it's harmless if I did.
[06:55] <mdz> seed updated
[06:56] <lamont> dpkg-source: extracting php4 in php4-4.3.10
[06:56] <lamont> dpkg-buildpackage: source package is php4
[06:56] <infinity> Nope, didn't miss it.  I'm not retarded.  Yay me.
[06:56] <infinity> lamont : Ahh, so it sorted itself.  Fancy.
[06:56] <infinity> mdz : I'll get new source uploaded ASAP with the swap.
[06:57] <infinity> lamont : Thanks for the fiddling.
[06:57] <lamont> of course, having a binary package foo delivered by source package bar, and a source package foo is just, well, wrong.
[06:57] <mdz> there's a lot to be said for "apt-get install php4" doing the right thing
[06:57] <infinity> lamont : That will be fixed in the next upload.
[06:57] <lamont> infinity: near certain that's what had sbuild all confused.
[06:57] <infinity> lamont : But thanks for babysitting the brokenish situation. :)
[06:58] <infinity> lamont : I'm positive that was the source of the confusion.  Still looks like an sbuild bug to me, weird corner case though it is.
[06:58] <lamont> (but with the ogre model, php4-universe (being in universe) is not in the cache for the main build of php4
[06:58] <lamont> infinity: yeah
[06:59] <lamont> given the corner-case-ness of it, and how much I loathe perl, and the fact that lunchpad obsoletes sbuild, I'm tempted to temporarily lump that in the 'doctor it hurts when I do this' category
[06:59] <infinity> ;)
[06:59] <mdz> infinity: please test an upgrade from Warty with php4 stuffs installed and make sure everything is kosher there as well
[07:00] <infinity> mdz : Will do.  They're functionally equivalent to Debian's packages, where I've been keeping an ongoing upgrade path all the way from Woody, so it SHOULD be okay.  But I'll test nonetheless.
[07:00] <infinity> (I can't wait to scrap a lot of that code..)
[07:01] <infinity> In fact, some may be upgrade from potato era...
[07:01] <infinity> Hrm.
[07:05] <pitti> Morning
[07:07] <lamont> talk about obscure options
[07:07] <lamont> if exists dhcp-parameter-request-list {
[07:07] <lamont>         # Always send the PXELINUX options (specified in hexadecimal)
[07:07] <lamont>         option dhcp-parameter-request-list = concat(option dhcp-parameter-request-list,interface-mtu);
[07:07] <lamont> }
[07:08] <lamont> gah - bad comment
[07:10] <fabbione> morning pitti
[07:10] <pitti> Hi fabbione 
[07:10] <dholbach> hey pitti 
[07:11] <fabbione> pitti: way too early
[07:11] <fabbione> i started soffering of insomnia again
[07:11] <dholbach> pitti: at 03:45 i couldnt sleep anymore
[07:11] <dholbach> but i was in bed at 22:00, maybe that's why
[07:13] <dholbach> Burgundavia: could you turn that away-script off?
[07:13] <Burgundavia> sorry
[07:13] <dholbach> thanks :-)
[07:13] <Burgundavia> done
[07:21] <infinity> Mornin', pitti.
[07:26] <pitti> Hi infinity 
[07:27] <pitti> mdz: here?
[07:27] <mdz> pitti: somewhat
[07:28] <pitti> mdz: you had some concerns wrt. #8263 (adding the gimp help files to l-support-*)
[07:28] <fabbione> good night lamont..
[07:28] <pitti> mdz: -en would land on the CDs, so they would grow by ~ 500 KB
[07:28] <pitti> Night lamont 
[07:28] <mdz> pitti: 500kb is OK
[07:29] <pitti> okay, so I can add them?
[07:31] <pitti> mdz: -zh-cn is in universe, all other langs are in main; okay for you?
[07:31] <pitti> (i. e. okay to promote to supported)
[07:37] <pitti> mdz: I followup on the bug report and CC you, there is another issue
[07:38] <fabbione> YAY
[07:38] <fabbione> finally some multiarse good news
[07:39] <aj> multiarch?
[07:39] <fabbione> multiseat
[07:40] <fabbione> renamed as multiarse
[07:40] <fabbione> ;)
[07:40] <aj> ahhh...
[07:40] <aj> so not related to multiverse either
[07:40] <fabbione> nope.. it's stuff in main
[07:41] <fabbione> mako: eheheh
[07:41] <mako> a sonnet even
[07:41] <fabbione> but i don't own the multiarse copyright
[07:41] <fabbione> i didn't forge that nick
[07:43] <mdz> pitti: yes, ok to promote
[07:45] <mdz> pitti: hmm, looking at the bug though
[07:45] <mdz> pitti: 500k is OK, but 6M may be too much
[07:45] <pitti> mdz: it would sum up to 6.6 MB
[07:45] <mdz> pitti: that is a bit more than 500k :-)
[07:45] <pitti> mdz: yeah, I didn't take -common into account at first
[07:46] <dholbach> bbl
[07:47] <Lathiat> ~>
[08:19] <fabbione> bah no wonder why #8396 hangs the world
[08:19] <fabbione> k3b is a SIGKILLorama
[08:20] <fabbione> mdz: btw did you read that link i gave to you yesterday?
[08:49] <Jeeves_> Morning all
[08:59] <fabbione> any MOTU's around?
[09:16] <infinity> mdz : Alright, new versions built, tested, and uploaded.
[09:19] <pitti> elmo: ncpfs sync, please
[09:19] <fabbione> AHHHHHHH
[09:19] <fabbione> amen
[09:20] <pitti> hi sabdfl 
[09:20] <fabbione> schweeb: i figured why xen doesn't boot :)
[09:20] <fabbione> hi sabdfl 
[09:20] <sabdfl> morning all
[09:20] <fabbione> sabdfl: i got the AGP/PCI solved!
[09:20] <fabbione> sabdfl: but it's both bad and good news
[09:21] <sabdfl> ok?
[09:21] <fabbione> the bad news is that for hoary we will have to use the nv driver = no 3d accel
[09:21] <sabdfl> erk
[09:21] <fabbione> the good news is that the problem is not the bios or the hardware
[09:21] <fabbione> but the nvidia binary driver
[09:21] <sabdfl> ok
[09:21] <sabdfl> ah
[09:21] <fabbione> the new version fixes it
[09:21] <sabdfl> is there an update on the way
[09:21] <sabdfl> ok
[09:21] <sabdfl> mdz?
[09:21] <fabbione> but we can't upload it to l-r-m
[09:21] <fabbione> it's way to late to jump so far
[09:21] <sabdfl> hmm
[09:22] <fabbione> i think using the nv for hoary is a good compromise and not intrusive
[09:22] <fabbione> just commeting one line in multiarse-configurator
[09:25] <sabdfl> ok
[09:25] <sabdfl> isthe new nvidia binary known good or bad?
[09:26] <fabbione> sabdfl: dunno.. works here on one machine.
[09:26] <sabdfl> how new is it?
[09:26] <fabbione> checking now...
[09:27] <fabbione> Release Date: March 11, 2005
[09:27] <fabbione> 20 days old...
[09:27] <fabbione> kinda risky
[09:28] <fabbione> and i am pretty sure mdz will not like the idea of a major upstream release change after RC
[09:29] <infinity> binary-only drivers are unsupportable either way, regardless of when it's added, so maybe he'd let it slide.
[09:29] <fabbione> infinity: there is more than that. remember that the nvidia stuff also adds libs that are used by xorg
[09:30] <sabdfl> fabbione: the cape town machine, was that also nvidia, or was it ATI-based?
[09:30] <fabbione> sabdfl: afaik it is nvidia too
[09:30] <fabbione> sabdfl: can you give me the email address of the guy handling it in SA?
[09:30] <infinity> fabbione : libGL stuff, you mean?.. Or does it do more these days?
[09:30] <fabbione> i would like to get some info about that machine
[09:30] <fabbione> infinity: libGL
[09:31] <infinity> FWIW, I've never had the nVidia libGL blow up, only the video driver (occasionally) and the kernel driver (more often).
[09:32] <infinity> But that's just my personal experience on a few machines.  Not much help, I suppose.
[09:32] <infinity> (Well, my personal experience as an avid Quake player, so I do USE libGL..)
[09:33] <fabbione> infinity: no relly.. i am not confortable at all in updating the driver
[09:34] <fabbione> 1) it taint the kernel
[09:34] <fabbione> 2) it interacts with xorg
[09:34] <fabbione> this is a suicide combination if the driver is borked
[09:34] <fabbione> + we are going to release in a wekk
[09:34] <fabbione> week
[09:34] <fabbione> that means that it will not get enough testing
[09:35] <infinity> Oh, I'm inclined to agree, it's just a shame if there's a showstopper bug on the current one.
[09:35] <infinity> Not that I'm familiar with the bug you're talking about.
[09:36] <fabbione> infinity: it's a very specific setup we are talking about
[09:36] <infinity> Ahh, kay.
[09:36] <fabbione> a machine with N gfx cards
[09:36] <fabbione> one agp and the others are PCI
[09:36] <fabbione> the problem is the order in which you initialize the cards
[09:36] <fabbione> on 2 machines we have 2 different behaviours
[09:36] <fabbione> machine a) init agp first, pci after = ok
[09:37] <fabbione> machine b) same as above = hard freeze
[09:37] <fabbione> if we init agp as last
[09:37] <fabbione> the behaviour is reverted
[09:37] <HiddenWolf> daniels: ping
[09:38] <HiddenWolf> Anyone working with Xorg on now?
[09:38] <infinity> Does it relate to the BIOS setting for initialising video cards, by any chance?
[09:38] <fabbione> infinity: nope.. test that too
[09:38] <fabbione> it's just a bug in the nvidia binary driver
[09:38] <fabbione> infinity: the nv driver works perfectly in both combination
[09:38] <fabbione> HiddenWolf: daniels is not online and he is working xorg
[09:39] <infinity> Find the offending bit with a hex editor and twiddle it? :)
[09:39] <HiddenWolf> fabbione: I just installed warty > hoary-rc and I do not have X
[09:40] <fabbione> infinity: yeah right :)
[09:40] <fabbione> HiddenWolf: report the bug on bugzilla
[09:40] <HiddenWolf> fabbione: I have no clue what the bug is. Using links and irssi atm
[09:42] <sabdfl> jdub: around
[09:42] <fabbione> HiddenWolf: check the usual stuff.. /etc/X11/xorg.conf /var/log/Xorg.0.log
[09:42] <fabbione> HiddenWolf: and add them to the bug
[09:43] <fabbione> HiddenWolf: daniels will help to debug
[09:43] <fabbione> (if he needs more info)
[09:44] <HiddenWolf> fabbione: thanks. I'll go back to being totally and utterly stressed out now. (my old pc died 2am yesterday-night, before I could get my data off, and now the new one is totally fucked up)
[09:46] <fabbione> HiddenWolf: usually the log has a clear message of why X didn't start
[09:46] <fabbione> at least check that
[09:47] <fabbione> schweeb: ping?
[09:51] <jdub> sabdfl: yo
[09:51] <HiddenWo1f> fabbione: do you by any chance know what busid I should give for a pci-e graphics card?
[09:52] <fabbione> HiddenWo1f: you need to check with lspci
[09:52] <fabbione> gimme a sec...
[09:53] <HiddenWo1f> fabbione: thanks. (somehow I can't switch vt's if I have irssi open, so I'll have to quit to try modding)
[09:53] <fabbione> HiddenWo1f: run this command: lspci -Xn | grep " 0300:"
[09:53] <fabbione> the BusID is the entry that looks like: PCI:1:0:0
[09:54] <HiddenWo1f> I'll brb
[09:55] <mako> so evidently, *userlinux* doesn't know where the source is to the userlinux metapackages
[09:55] <mako> at least not to the userlinux-server-base package :)
[09:55] <Burgundavia> lol
[09:55] <fabbione> mako: ehehe yeah.. i read that :)
[09:58] <HiddenWo2f> fabbione: busid is correct, but I get the error "unable to open framebuffer device / screens found, but none has a usable configuration"
[09:59] <fabbione> HiddenWo2f: can you put your xorg.conf somewhere?
[09:59] <sabmoc> mako So after talking to the LFNW people it turns out nobody has made arrangements to setup a booth, maybe that group you talked to backed out, or maybe they just planned to hand out CDs and thats it.
[09:59] <sabmoc> mako:: Anyway, so we have made arrangements to setup a booth now, and I even roped Burgundavia into giving a presentation.
[10:00] <HiddenWo2f> fabbione: if you can tell me how to ftp from the command line
[10:00] <spiv> HiddenWo2f: http://rafb.net/paste may help
[10:00] <fabbione> HiddenWo1f: use lftp
[10:01] <HiddenWo2f> brb again
[10:02] <mako> sabmoc: awesome
[10:02] <mako> sabmoc: Burgundavia already contacted me.. i pointed him to my notes/slides from a previous set of talks
[10:02] <Burgundavia> mako: have you seen this? http://lwn.net/Articles/129465/#Comments
[10:02] <sabmoc> mako:: I pointed him to the same slides :)
[10:03] <sabmoc> mako:: what is a good ratio for attendee's/cd's at a confrence? There will be about 300 people I think.
[10:03] <mako> Burgundavia: i didn't even know that it made it onto lwn :)
[10:03] <mako> sabmoc: one for everyone? :)
[10:04] <Burgundavia> mako: came almost right away, but only made the weekly roundup now
[10:04] <sabmoc> Thats what I was thinking but I didnt want to go overkill 
[10:05] <Burgundavia> from being there last year, I can say that not a lot of people were actually taking cds, but I could be wrong
[10:06] <mako> Burgundavia: that guy doesn't know what he's talking about
[10:06] <mako> Burgundavia: bruce's message is the most sane
[10:06] <mako> at the bottom of the page
[10:07] <mako> it basically says that we shouldn't say that it *is* userlinux
[10:07] <mako> which we wouldn't, since it's not :)
[10:07] <mako> well, so it got on lwn and nobody mailed me any bug reports yet.. that's a good sign
[10:08] <Burgundavia> I suspect that is simply a non-starter all around
[10:09] <mako> yeah, i wouldn't install them on my main machine :)
[10:10] <Burgundavia> sorry, but the boat has sailed Bruce
[10:10] <Burgundavia> some mad millionare came on said go
[10:10] <Burgundavia> so it did
[10:10] <mako> well bruce doesn't seem to care all that much about userlinux any more
[10:10] <Burgundavia> and said
[10:11] <mako> he used to care about it a lot more
[10:11] <Burgundavia> it is too bad
[10:11] <Burgundavia> userlinux is a good idea
[10:12] <Hiddenwolf> fabbione: lftp is being a bitch, won't work
[10:13] <Burgundavia> mako: another misguided soul --> http://ninkendo.org/view.pl?id=49
[10:13] <sabmoc> mako:: I hear they got pretty bogged down with ideas and didnt get much done.
[10:14] <mako> well.. i am on the userlinux list now.. i'm going to post a few messages and then go to sleep i think :)
[10:17] <Burgundavia> mako: that is new on the userlinux page. A mark protection link
[10:17] <Hiddenwolf> fabbione: ftp.ibasr.nl/ubuntu/xorg.conf
[10:18] <fabbione> Hiddenwolf: invalid username/passwd
[10:18] <fabbione> it doesn't allow anon access
[10:19] <Hiddenwolf> fabbione, damn
[10:20] <spiv> Hiddenwolf: Simplest is probably to submit a bug report and attach the xorg.conf and the Xorg.0.log.
[10:23] <fabbione> Hiddenwolf: ok hold on a sec...
[10:23] <Hiddenwolf> spiv: "this version of links does not include ssl/tls support
[10:23] <Hiddenwolf> fabbione: going nowhere, save for cardiac arrest
[10:23] <crimsun> Hiddenwolf: install links-ssl instead
[10:24] <Hiddenwolf> crimsun: lifesaver
[10:24] <fabbione> Hiddenwolf: check your xorg.conf for a line like this one: Option FbDev true
[10:24] <fabbione> or "yes"
[10:24] <fabbione> Hiddenwolf: comment it out if there is one
[10:24] <fabbione> and try to start X
[10:24] <fabbione> see if something changes in the logs
[10:25] <Hiddenwolf> fabbione: which section?
[10:25] <Hiddenwolf> nm, got it
[10:25] <fabbione> Section "Device"
[10:29] <HiddenWo1f> fabbione: that did change something. Instead of a new error message, I get a psychedelic pink/gray/green x screen that isn't killable
[10:30] <fabbione> HiddenWolf: i need the logs and the config. can you boot with a liveCD and put them somewhere i can actually access them?
[10:30] <HiddenWo1f> fabbione: I'll go and figure a way, then get back to you
[10:30] <fabbione> ok
[10:31] <GheRivero> res people
[10:47] <crimsun> doko_: thanks for the python update!
[10:55] <HiddenWolf> fabbione: #8438
[10:56] <HiddenWolf> Tell me if the attachments are there, doesn't seem that links likes uploading
[10:56] <fabbione> HiddenWolf: one moment...
[11:00] <fabbione> HiddenWolf: remove the VideoRam 128000 line and the UseFBDev
[11:00] <fabbione> forcing the VideoRam is not required on nv
[11:02] <HiddenWolf> Tell me, what's the most direct way to cut off X if it starts looping again?
[11:02] <HiddenWolf> cntrl-alt-backspace didn't cut it
[11:04] <fabbione> HiddenWolf: probably not starting gdm, but it really depends on the problem
[11:08] <HiddenWolf> fabbione: uncommenting fbdev led to psychedelic effects again. had to reboot to kill x
[11:10] <fabbione> HiddenWolf: did you uncomment the VideoRam too?
[11:11] <HiddenWolf> Yup
[11:11] <HiddenWolf> The default install didn't put that in there btw, dpgk-reconfigure did
[11:11] <fabbione> HiddenWolf: i can't see anything else wrong.. we will have to wait daniels
[11:12] <HiddenWolf> fabbione: thanks. :S
[11:13] <HiddenWolf> fabbione: this wouldn't be so bad if I could figure out why my main/old rig suddenly gave in and refused to boot. :S
[11:13] <HiddenWolf> anyhow, I'll wait and shut up. :)
[11:15] <fabbione> HiddenWolf: sorry, but it is really too long since i have been digging into X for debugging
[11:16] <HiddenWolf> fabbione: doesn't matter, I'm just totally stressed about the old rig not booting, giving a few beeps only. Don't have the manual, so can't check what the beep code is, but I fear it's fucked.
[11:17] <fabbione> HiddenWolf: the beep have a standard format. usually a few beeps = bad ram
[11:19] <HiddenWolf> had two pc's running the other day. Each on a bank of the same identical ram. Switching it didn't work either.
[11:26] <HiddenWolf> Fabbione: It seems 'nv' driver doesn't support 6600 cards yet. Changing to nvidia does the trick
[11:26] <doko> d3vic3: ping
[11:26] <d3vic3> doko, pong 
[11:27] <Mithrandir> hm http://www.ubuntulinux.org/ubuntu/releases/document_view still lists 5.10 as grumpy
[11:32] <fabbione> hey pitti
[11:32] <fabbione> i was just waiting for you
[11:33] <fabbione> pitti: if you have time, can you just check teh security history of xen?
[11:33] <pitti> fabbione: yeah, I attended the diploma presentation of a friend
[11:33] <pitti> sure
[11:33] <fabbione> thanks
[11:47] <sabdfl> Kamion: is it safe to say that fabbione's multiseat work will get into tomorrow's daily if it is uploaded today>
[11:47] <sabdfl> ?
[11:48] <sabdfl> is anybody else seeing aplay getting stuck?
[11:49] <sabdfl> try /usr/bin/aplay /usr/share/sounds/question.wav
[11:49] <sabdfl> and let me know if it works for you?
[11:49] <sabdfl> for me it just hangs
[11:50] <Mithrandir> worksforme
[11:50] <trukulo> hi
[11:50] <Mithrandir> won't aplay try to use alsa directly, so it'll hang if you don't have hardware mixing?
[11:51] <pitti> sabdfl: aplay can't possibly work
[11:51] <pitti> sabdfl: because esd already claims the alsa interface
[11:51] <Mithrandir> pitti: it can, but you need hardware mixing
[11:51] <pitti> sabdfl: try esdplay
[11:51] <sabdfl> pitti: i thought the whole point of alsa was to make that possible
[11:52] <pitti> Mithrandir: ALSA can mix, too, but its not configured to do so by default
[11:52] <pitti> sabdfl: ^
[11:52] <Treenaks> dmix?
[11:52] <pitti> sabdfl: it is possible, but we don't do it by default
[11:52] <pitti> Treenaks: exactly
[11:52] <sabdfl> pitti: does it break on some setups? is that why we don't try by default?
[11:53] <pitti> sabdfl: I don't know it overly well; I tried it the other day and it worked relatively easy, but I lost dynamic frequency adaptio
[11:53] <sabdfl> seb128: grrr, what happened to left-double-click-closes-folders-when-opening-new-folders?
[11:53] <sabdfl> k
[11:53] <pitti> sabdfl: i. e. I could not play a 22.050 Hz sample at correct speed
[11:53] <sabdfl> pitti: ok
[11:54] <pitti> it should be possible to combine mix and resampling, but esound already does it
[11:54] <pitti> seb128: is there any possibility that you upload another gdm before release?
[11:54] <trukulo> one question, boot process is as fast in fresh installs of hoary, than in upgrades from warty with aptitude?
[11:54] <pitti> seb128: the gdm translation tarball was lost; of course I can recompile it, but if you upload anyway, I don't need this
[11:55] <pitti> trukulo: hm, should
[11:55] <seb128> sabdfl: you said to wait a day, that you were going to mail the list about it and no news ...
[11:55] <trukulo> pitti, i don't see any fast boot on my computer, in fact, sid is much more faster than hoary
[11:56] <seb128> pitti: k, I'll do an upload
[11:56] <pitti> seb128: not if you don't need to
[11:56] <Treenaks> trukulo: how much memory do you have?
[11:56] <seb128> sabdfl: I assumed that you talked with other guys and changed your mind ... do you want the change to go now ?
[11:56] <trukulo> Treenaks, 768 MBs
[11:56] <Mitario> lo everyone
[11:56] <Treenaks> hey Mitario 
[11:56] <sabdfl> seb128: yes please
[11:56] <seb128> pitti: for the moment I don't need to
[11:56] <seb128> sabdfl: k
[11:57] <sabdfl> seb128: there are some subtleties, apparently that required more than a straight swapping of the keys
[11:57] <sabdfl> buttons
[11:57] <seb128> like ?
[11:57] <sabdfl> if you double click anything other than a folder, the folder should stay open
[11:57] <sabdfl> is that the case?
[11:57] <seb128> right
[11:57] <seb128> that's only a swap of the "folder opening" action
[11:57] <sabdfl> so double-clicking an image loads the image viewer, leaving the folder open
[11:58] <sabdfl> double-clicking a folder opens the new folder, and closes the existing one
[11:58] <Treenaks> I hope that's configurable then
[11:58] <seb128> hum, need to check but that should be fine
[11:58] <crimsun> pitti: err, regarding dmix, which virtual device were you using?
[11:58] <sabdfl> shift-double-clicking a folder opens the new folder, leaving the older one open too
[11:58] <crimsun> pitti: did you create an ~/.asoundrc || /etc/asound.conf, or were you using plug:dmix ?
[11:58] <seb128> sabdfl: k, understood. I'll change that today
[11:59] <sabdfl> thanks seb
[11:59] <pitti> crimsun: I created an /etc/asound.conf which set up dmix
[11:59] <seb128> np
[11:59] <pitti> crimsun: 
[11:59] <crimsun> pitti: if you have time, please test instead aplay -Dplug:dmix foo.wav
[11:59] <pitti> pcm.!default {
[11:59] <pitti>   type plug
[11:59] <pitti>   slave.pcm "dmix"
[11:59] <pitti> }
[11:59] <pitti> pcm.dsp0 pcm.default
[12:00] <crimsun> hmm
[12:00] <pitti> crimsun: that doesn't work, because esound still claims the interface
[12:00] <pitti> crimsun: you need to run esd with dmix, too
[12:00] <crimsun> pitti: right, I'm just concerned with dmix at the moment :)
[12:01] <pitti> crimsun: it works, but it doesn't do resampling with this conf
[12:01] <pitti> crimsun: but resampling is crucial
[12:02] <crimsun> the built-in dmix (if you don't have /etc/asound.conf or ~/.asoundrc) should work with resampling as plug:dmix.  If it doesn't, that's a bug.
[12:19] <pitti> fabbione: so far it seems that _every_ user on xen domain0 (i. e. the host platform) can control all domains
[12:20] <pitti> fabbione: that's not really desirable...
[12:20] <fabbione> pitti: hmm right
[12:20] <fabbione> that kinda sucks
[12:21] <pitti> fabbione: the intention obviously was to have a user-free dom0
[12:21] <pitti> fabbione: an all users work in a domain
[12:21] <fabbione> pitti: yes, that was my point too
[12:21] <fabbione> there should be no users in dom0
[12:22] <pitti> fabbione: so this should at least be documented properly as long as there is no access control (upstream plans something like this)
[12:23] <fabbione> pitti: ok. for now it is all limited to userland stuff
[12:24] <fabbione> kernel appears to be fine tho
[12:24] <pitti> yeah
[12:24] <pitti> fabbione: is the kernel part like UML? I. e. you have a normal -i386 kernel for dom0, and -xen arch kernels on top of it?
[12:25] <fabbione> something like that yes
[12:33] <smurfix> fabbione: you don't have a normal kernel for dom0 though
[12:35] <ross> mvo: ping?
[12:37] <fabbione> smurfix: yes i am aware of that
[12:37] <fabbione> smurfix: right now i manged to build some kernel images, but they don't boot yet
[12:37] <fabbione> i need to understand why
[12:38] <fabbione> but it will be another day work
[12:38] <fabbione> today i am pretty busy
[12:41] <smurfix> fabbione: I can imagine. (NB: that was a thinko, I should've typed pitti:.)
[12:42] <pitti> I read it anyway 
[12:42] <mvo> ross: pong
[12:45] <fabbione> smurfix: ehehe
[01:19] <Kamion> sabdfl: multiseat, tomorrow's daily> yes, certainly
[01:20] <Kamion> as long as it builds, at least. multiseat isn't in the installer initrd, which is where extra delays get introduced
[01:20] <sabdfl> Kamion: thx
[01:22] <fabbione> Kamion: yes, it builds :-) already checked on 5 arches
[01:22] <fabbione> i am doing some tests at the moment and some bug fixes
[01:23] <fabbione> it will need NEW love tho
[01:23] <fabbione> arch: all -> arch: any
[01:23] <Kamion> why should that need NEW love?
[01:23] <Kamion> (it doesn't, unless you've also changed package names)
[01:24] <HiddenWolf> fabbione: thanks again for the effort this morning
[01:24] <fabbione> HiddenWolf: it's ok. don't worry
[01:25] <fabbione> Kamion: hmm right...
[01:29] <mvo> ping doko 
[01:30] <doko> mvo: pong
[01:31] <mvo> doko: could you please run the testcode for #8447 on your amd64 box? when you have a moment, it's not urgent
[01:35] <doko> mvo: same behaviour
[01:38] <mvo> doko: thanks. it doesn't crash on my i386 box, I was wondering if it is amd64 specific
[01:39] <doko> seems so ...
[01:47] <ogra> morning world
[01:47] <trukulo> morning oliver
[01:47] <dholbach> hey ogra 
[01:48] <thom> morning ogra 
[01:48] <ogra> thom, http://hwdb.ubuntu.com ;)
[01:49] <ogra> just injecting your laptop-detect code in the output scripts....
[01:50] <koke> mvo: FYI, spanish translation updated on update-manager
[01:51] <thom> ogra: kick ass
[01:51] <ogra> hehe :)
[02:00] <mvo> koke: thanks :) 
[02:07] <mjg59> Is the gnome-bluetooth stuff in Universe expected to work?
[02:08] <dholbach> mjg59: i can test... it worked some weeks ago
[02:08] <mjg59> The browser thing seems a bit... fragile
[02:08] <dholbach> browser thing?
[02:09] <mjg59> Bluetooth manager
[02:09] <mjg59> And I don't seem to get a Send via Bluetooth thing in Nautilus
[02:09] <dholbach> i always used gnome-obex-send (since the nautilus-"send to bluetooth..." is broken)
[02:09] <dholbach> yes
[02:09] <Treenaks> mjg59: send via bluetooth might be an old-style nautilus extension
[02:10] <mjg59> Ah. Hrm.
[02:10] <dholbach> oh... gnome-bluetooth-manager just breaks, when starting up
[02:11] <Lathiat> bluetooth manager doesnt do much
[02:11] <Lathiat> just lists devices
[02:11] <Lathiat> it was broken cus gnomeappbar wasnt in python
[02:11] <pitti> Moin jbailey 
[02:11] <Lathiat> i patched it to use a normal statusbar
[02:11] <Lathiat> dunno if that was ever done upstream
[02:11] <Lathiat> the gnome obex stuff works but
[02:11] <Lathiat> as does gnome phone manager
[02:12] <mjg59> Ah, gnome phone manager?
[02:12] <Lathiat> yeh
[02:12] <Lathiat> lets you send/receive sms' via bluetooth
[02:13] <Lathiat> through a mobile
[02:13] <dholbach> gnome-obex-server & gnome-obex-send work fine
[02:13] <dholbach> both directions
[02:13] <Lathiat> they do
[02:13] <mjg59> Ok, cool
[02:14] <mjg59> g-p-m is just an SMS manager?
[02:14] <mjg59> gnome-bluetooth and gnome-phone-manager all seem to be missing menu icons
[02:15] <Lathiat> restart your session see if they work
[02:15] <Lathiat> i have a bug where menu icons dont show up until my session is restarted
[02:16] <mjg59> I've had gnome-bluetooth installed for a while
[02:16] <Lathiat> oh
[02:16] <Lathiat> in that case yeh i seem to recall they didnt anyway
[02:16] <Lathiat> prolly just broken .desktop files as they do have icons on the app
[02:17] <Lathiat> also im not sure 'System Tools' is the right place to have them
[02:17] <mjg59> Yeah
[02:18] <Lathiat> accessories perhaps?
[02:18] <jbailey> moin, pitti!
[02:18] <jbailey> ( et al...)
[02:19] <pitti> jbailey: I'm currently diving into the RT source code
[02:19] <dholbach> hey jbailey 
[02:19] <jbailey> pitti: Yay, thanks!
[02:19] <Lathiat> pitti: request tracker?
[02:19] <pitti> Lathiat: yes
[02:19] <Lathiat> sounds like pain :)
[02:19] <pitti> it _is_ :-)
[02:19] <Lathiat> now why am i not surprised :( heh
[02:19] <Mithrandir> pitti: RT is good for you.
[02:19] <pitti> Lathiat: 200.000 lines of code seems to be quite much for a distributed TODO list...
[02:20] <mjg59> Ok, cool, gnome-obex-send works nicely
[02:20] <Lathiat> archer:/usr/share/applications> cat gnome-obex-server.desktop|grep Icon
[02:20] <Lathiat> Icon=../gnome-bluetooth/pixmaps/blueradio-48.png
[02:20] <Lathiat> mjg59: :)
[02:20] <Lathiat> the icons arent even installed
[02:20] <Lathiat> let alone referenced right :)
[02:21] <jbailey> pitti: 200,000 lines of perl sounds like alot for anything.  I thought you could express all programs at the same time in a hologrpahic manner in about 100... ;)
[02:21] <mjg59> Grr.
[02:21] <Lathiat> mjg59: got bluez stuff installed?
[02:21] <pitti> jbailey: just state the offset into the pi decimal representation :-)
[02:21] <mjg59> Lathiat: Yeah
[02:21] <Lathiat> mjg59: like is sdpd running?
[02:21] <mjg59> It's not finding any of the machines in here
[02:21] <Lathiat> :(
[02:22] <mjg59> It did a minute ago
[02:22] <dholbach> mjg59: try setting class 0x100100;
[02:22] <mjg59> dholbach: No, the phone's even refusing to discover things
[02:22] <dholbach> hrm ok
[02:22] <dholbach> mjg59: anything in dmesg?
[02:23] <mjg59> Rebooting the phone seems to have fixed things
[02:23] <dholbach> mjg59: what phone is it?
[02:23] <Lathiat> hmm gnome-bluetooth doesnt tell you what the file is/called
[02:23] <mjg59> T68
[02:23] <mjg59> (Not a T68i)
[02:24] <mjg59> Yeah, works fine now
[02:24] <dholbach> t68 is ericsson?
[02:24] <mjg59> Yup
[02:24] <dholbach> i have one of the new sonyericsson ones... it's a nightmare... rebooting every now and then
[02:24] <Lathiat> dholbach: what model?
[02:25] <Lathiat> frozen a couple times in the 4 months i've had it tho
[02:25] <Lathiat> acutally i thinks its mor elike 9 months now, i dont keep track of time
[02:25] <dholbach> Lathiat: k700i
[02:25] <Lathiat> ah those
[02:25] <Lathiat> they look ok
[02:25] <mjg59> Woo, working IR
[02:25] <dholbach> hrm
[02:25] <Lathiat> prolly what i would have got if i didnt get this
[02:26] <dholbach> and you have to go to a shop to do a software update
[02:27] <dholbach> siemens offered downloadable updates ... for windows, but still...
[02:27] <dholbach> enough ranting :-)
[02:28] <SlackShrike> hi
[02:30] <fabbione> elmo: ping?
[02:30] <elmo> fabbione: ?
[02:30] <SlackShrike> howto create a live cd of ubuntu ? I am like contribute with the live of ubuntu. where I find the documentation of live-cd ?
[02:30] <dholbach> SlackShrike: on the wiki
[02:30] <fabbione> elmo: can you please make sure that all the Build-Dep for l-r-m are installed on concordia/davis/halley hoary chroot and that they are uptodate?
[02:31] <elmo> btw, new cdimage machine just added - if anyone notices any problems, please shout
[02:31] <elmo> fabbione: k, doing
[02:31] <fabbione> elmo: thanks a lot
[02:32] <SlackShrike> dholbach : The wiki has the documentation of as to customize the live-cd !
[02:33] <dholbach> SlackShrike: could you please elaborate on what you're exactly trying to do?
[02:35] <thom> so, i guess consensus is that we shouldn't p4-clockmod on laptops, either
[02:36] <Kamion> hm, might be a good idea to actually turn CD builds back on
[02:37] <elmo> p4-clockmod sounds like it shouldn't be compiled
[02:37] <SlackShrike> dholbach: To study casper and the process of creation of live of ubuntu to be able to help in the hardware detention that is sold in Brazil and the translation and configuration of the system for pt_BR
[02:39] <dholbach> SlackShrike: i hope someone read what you just wrote and will be able to help you... i have no real clue about the internals
[02:40] <mjg59> I can't work out whether my inability to get the machines talking is down to something being broken or IR just being cra
[02:40] <mjg59> p
[02:41] <SlackShrike> dholbach: thanks
[02:42] <dholbach> :-)
[02:43] <thom> elmo: well, that's the other option yes
[02:44] <mjg59> Oh, gah, it seems to be an Ericcson bug
[02:44] <mjg59> Why does ALL HARDWARE SUCK?
[02:44] <Treenaks> mjg59: your ericsson is bug-free compared to my nokia
[02:45] <Mithrandir> mjg59: apply hammer.
[02:45] <Mithrandir> hammers don't suck
[02:46] <jbailey> Mithrandir: No, hammers bite. ;)
[02:47] <mjg59> Hrngh. I still get nothing.
[02:47] <Mithrandir> mjg59: apply hammer, I say.  It'll solve all your worries.
[02:51] <dholbach> woohoo, i have memtest86+ now! :-))
[02:51] <Kamion> jbailey: hammers claw, surely ...
[02:51] <Mithrandir> dholbach: does it work?
[02:51] <dholbach> Mithrandir: didn't test yet... does it have update-grub in postinst?
[02:52] <dholbach> Mithrandir: at least i didnt see the msg
[02:52] <Mithrandir> dholbach: unsure, I actually haven't rebooted to see
[02:52] <dholbach> Mithrandir: i'll report later
[02:52] <Mithrandir> goodie
[02:52] <jbailey> Any evms guru's around?
[02:53] <jbailey> I've tried a few times to figure this stuff out to solve the boot bugs that are showing up with raid, and the documentation is just a bit convoluted.
[02:53] <dholbach> Mithrandir: postinst only has lilo crack
[02:53] <Mithrandir> dholbach: hm, I'm not going to change that for release.
[02:54] <dholbach> Mithrandir: hrm... why not? if the user updates/newly-installs a kernel, it will get dragged in anyway
[02:54] <Kamion> dholbach: is this a kernel package? update-grub is done in a kernel-package hook
[02:54] <Mithrandir> dholbach: because we're releasing in a week and I'd like not to touch more than needed.
[02:55] <Kamion> /etc/kernel-img.conf
[02:55] <mjg59> Oh, woo, that worked
[02:55] <ogra_> 1140 submissions on hwdb.ubuntu.com, 165 of them are amd64, 24 are ppc
[02:55] <dholbach> Kamion: ok
[02:56] <mjg59> Whee! Patches over mobile link.
[02:56] <mjg59> Uh, packets.
[02:56] <mjg59> Christ.
[02:56] <HiddenWolf> ogra: Wait for another one. Pumping over my data atm, then watch me. 
[02:56] <ogra_> 97 have no working swap !
[02:56] <ogra_> about 10%
[02:57] <jbailey> ogra_: Is that page supposed to be for finding supported hardware?
[02:58] <ogra_> jbailey, naah, thats my interim until we have a real DB.....its currently just grepping through the entry....
[02:58] <elmo> Mithrandir: did you test our version of memtest86+ on amd64? it's quite old
[02:59] <Mithrandir> elmo: it's the same version as is in debian, iirc?
[02:59] <elmo> no, debian has 1.50-1
[02:59] <elmo> we're on 1.30 or so
[03:01] <ogra_> jbailey, all these files will go in a SQL db after release, that will be searchable then and show you the percentage of support a device gets (at least thats the plan), but currently i only want something the user can be pointed to to aprove everything was detected right....
[03:01] <jbailey> ogra_: Cool, thanks.
[03:01] <elmo> fabbione: done, btw
[03:02] <fabbione> elmo: thanks.
[03:05] <Mithrandir> elmo: works on my box at least.
[03:06] <elmo> Mithrandir: ok, cool
[03:06] <maswan> Mithrandir: are you sure it actually detects errors above 4 gig? I seem to remember we had some issues like that.
[03:06] <Mithrandir> maswan: I only have 2G memory in the box.
[03:06] <maswan> Mithrandir: ah
[03:06] <Mithrandir> maswan: it probably doesn't, since it runs in 32 bit mode.
[03:07] <maswan> I wonder. Hmm.. Let me check around a bit.
[03:07] <elmo> I suspect memtest86+ is underwhelmingly used - it seems to be utterly broken on (super common) poweredge 2650's
[03:07] <Mithrandir> if you ship me a box with 6G memory, I can test, no problem.
[03:10] <elmo> Mithrandir: how?  going to poke out some of the pins on the top 2Gb? :P
[03:11] <Mithrandir> elmo: see whether it detects 6G of memory for a start. :)
[03:11] <elmo> oh, well, there is that
[03:11] <maswan> Ok, it was local patches to make it work on opteron
[03:11] <Mithrandir> I could probably also find some faulty memory, yes.
[03:11] <elmo> I could try memtest86+ on emperor for fun
[03:11] <Mithrandir> elmo: please do.
[03:11] <elmo> Mithrandir: it's not amd64, just x86 PAE with LOTS of memory
[03:12] <Mithrandir> isn't PAE like dog slow?
[03:13] <maswan> Ok, it wasn't 4gig+ issues, it was Opteron-internal memory controller problem.
[03:13] <elmo> Mithrandir: yes - I hate it passionately
[03:13] <elmo> and the stupid NX protection on newer Xeons requires it
[03:14] <maswan> So, do you guys want patches for Opteron support? :)
[03:15] <elmo> maswan: they not upstream yet?
[03:15] <elmo> I suspect it's too late to get them in hoary in any event
[03:15] <elmo> given the < 7 days thing and all
[03:16] <Mithrandir> yeah, I'm not comfortable putting them in now; breezy, sure.
[03:16] <maswan> elmo: well, the guy that did them just muttered about the problem and him having patches, not actully submitted it.
[03:16] <maswan> so no, it's no big deal to build them either yourself
[03:16] <elmo> yeah, that's the other thing about memtest86+; you can always grab the latest .deb, no matter what distro you're running
[03:17] <maswan> yeah
[03:17] <Mithrandir> as long as your system doesn't die halfway through booting linux. ;)
[03:17] <elmo> Mithrandir: then you grab the .iso from www.memtest86.org :P
[03:17] <elmo> or whatever it is
[03:17] <maswan> the part about "Do a trivial check to make certain we can see a host bridge." assumes that it is pci device id 0, where the opterons have 24, which makes it break in spectacular ways. :)
[03:18] <HiddenWolf> mithrandir: I like it worse if the system dies on post, really
[03:18] <Mithrandir> maswan: ah, sounds fun.
[03:18] <maswan> but yeah, these patches should probalby rather be pushed upstreams
[03:18] <maswan> I just suck at dealing with most upstreams. :)
[03:19] <torkel> maswan: I think ake did, but they have not been accepted yet...
[03:19] <maswan> torkel: ah, ok. not just mubmled about having a problem with a patch? :)
[03:19] <elmo> maswan: submit them to the debian bts
[03:19] <maswan> elmo: ah, good plan
[03:22] <HiddenWolf> Why is there a debian logo in the synaptic popup to configure a package that asks a question? (for the love of god forgot the name )
[03:23] <sabdfl> maswan: is that patch in hoary too?
[03:24] <maswan> sabdfl: The patch isn't anywhere but locally here, I'm going to submit it to the debian bts so other people can see it.
[03:24] <maswan> (well, parts of the local patch, I don't think we need some of the default config changes propagated :) )
[03:27] <jani> elmo, please sync sylpheed, fixes two security bugs
[03:28] <maswan> I just remembered these issues because memtest86+ was mentioned in here. :)
[03:29] <elmo> jani: done
[03:29] <jani> thanks
[03:31] <crimsun> elmo: if you have time, would you please sync xfce4.2.1 from os-works?  The packages are also in experimental if you'd rather sync from there.
[03:32] <elmo> crimsun: what's os-works?
[03:32] <crimsun> elmo: Benny's repository (deb-src  http://www.os-works.com/debian/ testing main)
[03:32] <crimsun> elmo: those packages are also in Debian experimental
[03:32] <elmo> crimsun: well, which would you prefer?  is one better than the other?
[03:32] <elmo> oh, they're the same?
[03:33] <jani> crimsun, I was just about to ask elmo the same :)
[03:33] <elmo> I'll do experimental then - have a trust path to it
[03:33] <crimsun> elmo: Benny's is slightly newer
[03:33] <crimsun> elmo: ok, that works for us!
[03:33] <crimsun> either way, it means we can clear up xfce 4.2.1 for Hoary :)
[03:33] <crimsun> elmo: thanks a bunch
[03:33] <jani> cool
[03:34] <sabdfl> maswan: cool, as long as it's in hoary :-)
[03:34] <jani> so exeprimental does not have the 4.2.1.1 fixes ?
[03:34] <crimsun> jani: no, but they're not terribly intrusive; we can use dpatches
[03:34] <jani> so we will apply those in time for hoary right ?
[03:34] <crimsun> yep
[03:35] <crimsun> as soon as the sources are in, I'll begin updating the diffs
[03:35] <jani> cool, what about the other lightwaight benny pack which are orthogolnal (libexo, terminal and mousepad)?
[03:35] <jani> those go well with xfce4 on limited ram machines 
[03:36] <jani> mousepad is the only X app that starts in <1 sec for me :)
[03:36] <crimsun> those are only in Benny's os-works repo
[03:36] <Treenaks> jani: how about xclock? :P
[03:36] <jani> treenaks, who needs that ;)
[03:36] <elmo> crimsun: if they're not the same, I can sync from this newer site, if you want?
[03:37] <jani> elmo, crimsun that would be nicer I think it's the same maintainer and xfce4 core devel
[03:37] <crimsun> elmo: merging from Benny's would be more helpful for jani and me, but the trust issue is understandable.
[03:37] <elmo> crimsun: I only brought that up, 'cos I thought you meant they were the same ..
[03:38] <crimsun> elmo: the core packages are listed here in sorted dependency order: http://sh.nu/~crimsun/list.sources
[03:38] <crimsun> (source package names)
[03:38] <jani> that's not a private site though it's a company that makes xfce4 livecd among others, but I can understand if exp is more trusted
[03:38] <elmo> i.e. all other things being equal, I'd prefer debian over random site.  if they're not, I'll go with whatever has best sw
[03:38] <elmo> s/sw/software/
[03:39] <jani> os-works has the best, I think crimsun and I  agree
[03:39] <crimsun> elmo: well, it's easiest to sync from experimental
[03:39] <elmo> crimsun: how come?
[03:39] <crimsun> jani: right
[03:39] <jani> crimsun , let elmo do the work instead of us ;)
[03:39] <crimsun> elmo: well, I'm considering the amount of effort it'd require you
[03:40] <elmo> crimsun: don't worry about that - the syncing program is generic, as long as the site has a Sources, I can sync from it
[03:40] <elmo> if it doesn't, I still can, it's just a bit more work, but it looks like os-works has a Sources
[03:40] <crimsun> elmo: ok, then os-works is jani's and my preferred source
[03:40] <jani> elmo, is that program public, it could be useful for MOTU staff
[03:42] <zul> hey
[03:42] <dholbach> hey zul
[03:42] <elmo> jani: not yet, hopefully will be eventually - not sure it'd be useful useful for MOTU folks tho as it plays various games to get katie to bypass some of her checks, and those games are tied to certain PGP keys etc.
[03:42] <zul> hey dholbach 
[03:42] <jani> elmo, I see thanks
[03:46] <jani> elmo, one other thing I've been meaning to ask, do all syncs end up in hoary-changes? I was notified for the darcs 1.0.2 sync a couple days ago but it did not show up on the list
[03:47] <elmo> jani: hoary-changes has some overly fascist spam filtering - mithrandir wrote up a better filter for it, which we can hopefully get installed there soon
[03:47] <elmo> but yes, all syncs go to hoary-changes
[03:47] <elmo> (in theory)
[03:47] <jani> I see, thanks
[03:49] <elmo> crimsun: hum, that's not a list of source packages?
[03:49] <elmo> e.g. it has xffm4, but source is xffm?
[03:49] <elmo> OTOH, do you just want me to sync everything in that repo?
[03:49] <HiddenWolf> elmo: for -changes, why not whitelist instead of filtering?
[03:50] <elmo> HiddenWolf: way more work than the filter mithrandir has already done
[03:50] <crimsun> elmo: err, odd, the source package is xffm4 in experimental and xffm in os-works.
[03:51] <jani> hmm maybe syncing all and we sorting it out later would be easier on elmo
[03:51] <crimsun> elmo: yes, please sync all
[03:51] <jani> we can get rid of xfld stuff
[03:53] <crimsun> yes, per our discussion regarding meta-xfce4
[03:55] <mjg59> thom: So, do we get a release party next week?
[03:56] <dholbach> mjg59: i set up wiki.ubuntu.com/ReleaseParty ages ago... but hrm... see for yourself
[03:56] <zul> hell yes
[03:57] <mjg59> Oh.
[03:57] <mjg59> Enabling DMA while in the middle of burning a CD makes bad things happen
[03:57] <Lathiat> i noticed that
[03:57] <Lathiat> :(
[03:57] <Lathiat> unfortuantely
[03:57] <Lathiat> hdparm init scripts seems to run before ide-cd is loaded
[03:57] <Lathiat> i shoudl file a bug about that
[03:58] <zul> dholbach: it needs to be updated you still have it april 6th :)
[03:58] <smurfix> dholbach: Is that page linked anywhere?
[03:59] <dholbach> zul: true that
[03:59] <dholbach> smurfix: dunno
[03:59] <elmo> crimsun: can't sync startup-notification, it's in main.  doing the rest
[03:59] <smurfix> dholbach: no wonder nobody subscribed if people do't even know the page exists ;-)
[03:59] <ogra_> hehe
[04:00] <dholbach> don't laugh... link it then ;-)
[04:00] <ogra_> put t in the release announcement ;)
[04:00] <crimsun> elmo: all right, thanks again
[04:01] <thom> Lathiat: kindly don't; check that /etc/dev.d/block/hdparm is 755 and then find out why it's not being run
[04:02] <thom> mjg59: dunno, i'm gonna be on a plane since we rescheduled to friday
[04:02] <seb128> pitti: here ?
[04:02] <pitti> yeah
[04:02] <seb128> any new on the gamin plan ?
[04:03] <pitti> not yet
[04:03] <seb128> we need to work with upstream now if we want to get that fixed for hoary
[04:03] <pitti> jbailey: here?
[04:03] <seb128> don't count on upstream beeing available to fix that on 1 day
[04:03] <jbailey> pitti: Yup
[04:04] <pitti> jbailey: do you know DBIx::SearchBuilder?
[04:04] <pitti> jbailey: this is what RT uses for DB access
[04:04] <jbailey> pitti: I don't, but I'm not a Perl guru at all.
[04:04] <pitti> jbailey: hmm, okay
[04:05] <jbailey> pitti: I have someone I trust that I can ask for info...
[04:05] <pitti> jbailey: I try to find out whether anything in RT/DBIx/DBI quotes strings to prevent SQL injection
[04:05] <mjg59> thom: Teh suck
[04:06] <elmo> crimsun: hmm, FYI, lots of epochs used in this repo
[04:06] <thom> mjg59: yah
[04:06] <crimsun> elmo: indeed.
[04:07] <zul> q
[04:07] <zul> doh..
[04:08] <Lathiat> thom: oh, hrm....
[04:09] <pitti> seb128: I pinged him
[04:09] <seb128> thanks
[04:10] <Kamion> once you have two uses for something ...
[04:10] <Lathiat> thom: aha :)
[04:10] <Lathiat> thom: that must have been added sine it wasnt working a while back
[04:10] <Lathiat> thom: i wont bother you now :)
[04:11] <Kamion> I suppose it should have occurred to me earlier that prebaseconfig was no longer necessarily right for various things
[04:12] <tritium> Lathiat, I've had the same problem with hdparm running before modules are loaded.  How did you fix it?
[04:12] <Lathiat> tritium: hal now runs it when devices come up
[04:12] <mjg59> Waa.
[04:12] <Lathiat> im gonna reboot and see if it works
[04:12] <Lathiat> its possible ide-cd is loaded after hdparm and before hal
[04:12] <mjg59> My write died halfway through, and now cdrecord refuses to blank my CD.
[04:13] <Lathiat> mjg59: eww :(
[04:13] <mjg59> Cock.
[04:13] <Treenaks> mjg59: even if you do a complete blank?
[04:13] <mjg59> Treenaks: Yeah
[04:13] <abelli> mjg59: the same for me ..
[04:13] <abelli> i mounted it another time and it worked
[04:14] <Treenaks> that's just plain wankage
[04:14] <mjg59> Power calibration area error
[04:14] <abelli> huh .. this may mean that the burner is gone
[04:14] <abelli> s/may(might
[04:15] <tritium> Lathiat, so you re-ordered some init scripts?
[04:15] <Lathiat> no
[04:16] <mjg59> Hmm. It's making unhappy noises with this one, too.
[04:16] <Lathiat> mjg59: :(
[04:16] <abelli> mjg59: DMA huh ?
[04:16] <elmo> mjg59: you and hardware are just the bestest of friends
[04:16] <mjg59> abelli: Problems started when I turned on dma
[04:17] <abelli> mjg59: ahh right... 
[04:17] <mjg59> Gah. No.
[04:17] <mjg59> Maybe it needs a full power cycle.
[04:17] <mjg59> Either that or I've managed to kill it.
[04:17] <mjg59> (But it's not mine and it's under warranty, so...)
[04:18] <abelli> but it's still strange.
[04:18] <elmo> crimsun: small problem
[04:18] <abelli> dma shouldnt break anything
[04:18] <elmo> crimsun: xfce4-theme-brushedchrome is under a nasty non-commercial CC license
[04:19] <abelli> i mean there's a control circuit that should avoid this kind of things . and block the lens.
[04:19] <elmo> how important is that package?
[04:19] <Kamion> abelli: large numbers of bug reports suggest otherwise
[04:19] <abelli> Kamion: yeah i know.
[04:20] <Kamion> abelli: I think you're assuming non-broken hardware :)
[04:20] <abelli> Kamion: yeah ... its a sw problem :)
[04:22] <mjg59> Ok, power cycling seems to have made it happier
[04:23] <abelli> :-D
[04:23] <Kamion> fabbione: think we can switch multiseat to priority: standard?
[04:24] <fabbione> Kamion: let
[04:24] <Kamion> fabbione: requirement for that would be that it does nothing unless the hardware you're in front of is multiseat
[04:24] <fabbione> Kamion: just a sec.. 
[04:24] <fabbione> re
[04:24] <fabbione> let me check the code Kamion
[04:25] <fabbione> i didn't write the udeb at all
[04:25] <Kamion> that was my impression of how it worked last I checked
[04:27] <fabbione> Kamion: the postinst looks ok to me and it seems to only ask for multiarse in case it finds 4 gfx
[04:27] <fabbione> Kamion: i think we can switch it safely
[04:27] <fabbione> but it would be nice if you can double check
[04:27] <Kamion> elmo: could you please override multiseat-udeb to Priority: standard?
[04:27] <fabbione> or otherwise switch it
[04:27] <fabbione> and i can do a netinsta
[04:27] <elmo> Kamion: done
[04:27] <fabbione> and see if it work
[04:27] <Kamion> thanks
[04:28] <elmo> crimsun/jani: xfce4-genmon-plugin-1.1 has an incorrect debian/copyright file too
[04:29] <Kamion> fabbione: yeah, I'll give it a try in a bit; booting with 'multiseat' on uniseat hardware is a good test, it should do nothing
[04:29] <fabbione> exactly
[04:29] <Kamion> ask no more questions, I mean
[04:29] <Kamion> if that works, I'll remove the multiseat boot option
[04:29] <fabbione> sounds good
[04:31] <fabbione> guys please: http://lists.ubuntu.com/archives/ubuntu-devel/2005-March/006353.html
[04:31] <fabbione> and let me know
[04:36] <pitti> fabbione: yay, nvidia is supposed to work again?
[04:37] <HiddenWolf> pitti; is that nv or nvidia
[04:37] <pitti> HiddenWolf: binary nvidia drivers (l-r-m)
[04:37] <HiddenWolf> pitti; pity, I installed hoary today and was forced to go to nvidia because nv wouldn't work
[04:39] <pitti> HiddenWolf: for me it's the other way round
[04:40] <HiddenWolf> pitti, is there any difference between a fresh and an upgraded hoary?
[04:40] <pitti> HiddenWolf: yes, several if you upgraded from Warty
[04:40] <HiddenWolf> pitti: damn. I was forced to upgrade from warty, indeed. 
[04:41] <pitti> HiddenWolf: shouldn't affect the graphics drivers, though
[04:41] <HiddenWolf> (my old rig died when I connected the burner that would burn my hoary-rc iso I had been begging for all day)
[04:42] <pitti> although, there is a note about it
[04:42] <pitti> http://www.ubuntulinux.org/wiki/HoaryUpgradeNotes
[04:42] <pitti> HiddenWolf: ^
[04:43] <HiddenWolf> pitti, thanks
[04:43] <elmo> crimsun/jani: and exo too
[04:44] <tritium> fabbione, I'm currently using your 1.0-7167 nvidia.  So far so good!
[04:49] <LeeJunFan> hrm, just doing a server install with 3/28/05 dvd snapshot it asked for adding more repositories. When I hit 'YES' it really screwed up the screen with out of place text.
[04:51] <pitti> fabbione: ROCK!
[04:51] <pitti> fabbione: your version finally works again for me
[04:51] <pitti> fabbione: the version currently in Hoary doesn't, it was broken for ages
[04:52] <Kamion> LeeJunFan: should be fixed since
[04:52] <Kamion> specifically base-config 2.62ubuntu20, 29 March
[04:52] <LeeJunFan> Kamion: cool.
[04:58] <lamont> morning world
[04:58] <ogra> hey lamont
[04:58] <lamont> Kamion: thoughts on why my freshly installed G3(color) hangs on loading the kernel?
[04:58] <schweeb> fabbione: pong
[04:58] <schweeb> fabbione: sorry, cable modem died right after I pressed "enter" that first time
[04:59] <pitti> Hi lamont
[05:09] <pitti> elmo, jbailey: https://www.ubuntulinux.org/wiki/RequestTrackerSecurityReview
[05:10] <jbailey> pitti: Ooo, nice!  So aside from the lack of taint mode, you'lre generally happy?
[05:11] <pitti> jbailey: it does not verify its input, but since you cannot use metacharacters to modify HTML/SQL queries, I don't really mind
[05:11] <pitti> jbailey: so at most you can modify some strings in the HTML
[05:12] <pitti> jbailey: but not insert tags
[05:12] <lamont> fabbione: but I was hoping to build a 2-seat multiseat... :-(
[05:12] <pitti> jbailey, elmo: in short: looks good
[05:13] <tritium> Lathiat, did hdparm enable dma for your cdrom on reboot?
[05:15] <Mithrandir> Kamion: if you get approval from mdz, the memtest86+ package can be seeded, put in the debootstrap list and so on for amd64 as well
[05:15] <jbailey> pitti: Great, thanks.  I appreciate it.
[05:16] <HiddenWolf> Is there any way to enable dma on an nforce 4 chipset?
[05:18] <tritium> something in recent kernels is giving me this problem: hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
[05:19] <tritium> followed by: ide: failed opcode was: unknown, and hda: drive not ready for command
[05:19] <Lathiat> tritium: yep
[05:19] <Lathiat> works now
[05:19] <Lathiat> fails when hdparm is started
[05:19] <Lathiat> but hal runs it
[05:19] <Lathiat> and works 
[05:19] <Lathiat> tritium: just you have to set it up in /etc/hdparm.conf
[05:20] <dholbach> bbl
[05:23] <ogra> fabbione, ....
[05:23] <tritium> sorry to whoever replied.  my system froze and I had to reset
[05:24] <ogra> fabbione, nvidia doesnt work right with my xorg.conf, i get only 1280x768 instead of 1280x800 the same config works fine with nv (amd64 here)
[05:25] <ogra> i generated the config brandnew new with the last xorg update
[05:25] <tritium> Who had replied to me about hda: status error ?
[05:26] <pitti> tritium: nobody so far
[05:28] <crimsun> elmo: ok.  Multiverse ok?
[05:29] <tritium> pitti, hmm...I heard the sound I setup for nick highlights, so I alt-tabbed over to xchat when my system locked up.  strange.
[05:29] <pitti> tritium: <Lathiat> tritium: just you have to set it up in /etc/hdparm.conf
[05:29] <tritium> thanks, pitti.
[05:29] <pitti> trulux: this was the last message before you disconnected
[05:29] <tritium> I appreciate it.
[05:30] <pitti> did you get it? there were some msgs before this
[05:30] <tritium> Yes, diamond /msged me the rest.  Thanks.
[05:30] <tritium> Thanks, Lathiat for the info.  Unfortunately, I have setup my hdparm.conf, to no avail.
[05:33] <Nafallo> hmm, we upgraded my girlfriends shuttle yesterday and xorg only gives 640x480. xfree gave 1024x768. and that should be no-good for a release? i845GE chipset.
[05:39] <Nafallo> daniels: hi there. got time to check an Xorg.0.log? ;-)
[05:40] <daniels> Nafallo: emailing me is better -- it's late, and i'm tired and kinda busy, sorry
[05:40] <elmo> crimsun: yes, multiverse looks okay
[05:40] <Nafallo> daniels: oki. *defending myself from angry girlfriend* ;-)
[05:41] <daniels> heh
[05:41] <Nafallo> daniels@ubuntu.com, right?
[05:41] <daniels> yeah
[05:42] <fabbione> lamont: that's easy.. i will help you to do that :)
[05:42] <fabbione> ogra: check the xorg logs please
[05:42] <lamont> fabbione: is post-UDU, btw
[05:43] <schweeb> fabbione: poke
[05:43] <fabbione> lamont: better.. so we make a better multiarse package
[05:44] <fabbione> schweeb: yes.. i figured why xen didn't boot
[05:44] <schweeb> cool
[05:44] <schweeb> fabbione: my cable modem died last night, right after I asked that first question, lol
[05:44] <fabbione> schweeb: but there are other problems.. basically i get it to boot the xen.gz microkernel, but it reboots as soon as it tries to load the xen0 kernel
[05:44] <fabbione> schweeb: no problem.. it's not urgent
[05:44] <schweeb> yea
[05:45] <schweeb> you have the package or whatever you're using to generate the kernels anywhere?
[05:45] <ogra> fabbione, hmm, funny, the detection seems to have detected a 1280x768, but only nvidia seems to respect this...nv ignores it
[05:45] <ogra> s/1280x768/1280x768 Display
[05:46] <Lathiat> ogra: probably has no modeline for it, google?
[05:46] <Lathiat> i seem to recall seeing a bug about that
[05:46] <ogra> (WW) (1280x768,Standardbildschirm) mode clock 80.14MHz exceeds DDC maximum 70MH
[05:46] <schweeb> fabbione: you sure you selected the right CPU type?  I think I had that problem when I selected K7 rather than P4
[05:46] <ogra> Lathiat, it was solved...
[05:46] <ogra> Lathiat, (or should have been)
[05:46] <Lathiat> not for you obviously :)
[05:46] <fabbione> schweeb: yes. it's on a P4 and i did set MPENTIUM4
[05:46] <fabbione> schweeb: i did compare the 2 configs 
[05:46] <schweeb> odd
[05:47] <fabbione> perhaps it is some other config that makes problems
[05:47] <fabbione> i will have to dig into it
[05:47] <ogra> Lathiat, despite the prob with nvidia (which i dont touch normally) it works fine ;)
[05:47] <schweeb> you're more of a kernel guru than I, I'll let you figure it out ;)
[05:47] <Lathiat> nv driver looks like ass on my laptop
[05:47] <Lathiat> gradients in the ubuntu images go all bandy
[05:47] <fabbione> ogra: yes, the 2 drivers have different opinions on clockmode
[05:47] <Lathiat> despite claiming its in 24bit
[05:47] <Lathiat> works fine under nvidia
[05:48] <Lathiat> its rather annoying
[05:48] <ogra> fabbione, i'll try to regenerate my xorg.conf then, lets see
[05:48] <fabbione> ogra: ok thanks
[05:48] <mdz> Mithrandir: I don't think we want to change debootstrap etc. at this point, but it would be good to have memtest86+ in the archive for amd64. is that done already?
[05:48] <mdz> fabbione: what link that you gave me yesterday?  I don't remember
[05:49] <fabbione> mdz: http://kernelplanet.org/ 
[05:49] <Nafallo> daniels: mailed :-)
[05:49] <fabbione> mdz: check the entry from the 29th of March about memset
[05:49] <mdz> fabbione: yes, I did read that
[05:49] <fabbione> mdz: he is basicaly right and we have the opportunity to fix these bugs
[05:50] <fabbione> since we are aware of them
[05:50] <zul> fabbione: already started last night
[05:50] <zul> i just got back too late from soccer last night 
[05:50] <ogra> fabbione, funny....there is no option for 1280x800@60 in the monitor list
[05:50] <fabbione> ogra: rant with daniels about that :)
[05:50] <ogra> heh
[05:50] <Mithrandir> mdz: yes, as of per some hours ago
[05:51] <mdz> fabbione: I don't understand; those memset calls look OK to me
[05:51] <mdz> Mithrandir: cool
[05:52] <Mithrandir> mdz: didn't you like, approve it, yesterday? :)
[05:52] <fabbione> mdz: void *memset(void *s, int c, size_t n);
[05:52] <pitti> Morning mdz
[05:52] <fabbione> mdz: what's the point of a size 0 ?
[05:53] <fabbione> and with that pattern?
[05:53] <fabbione> it is morelikely that they want to 0 the mem with size of that struct
[05:53] <mdz> oh, heh
[05:53] <fabbione> memset(&bpart, sizeof(struct blkpg_partition), 0); <-
[05:54] <mdz> I've probably made that mistake myself; it should be the other way :-)
[05:54] <daniels> Nafallo: thanks
[05:54] <fabbione> mdz: yeah.. it took a bit to understand it too :)
[05:54] <mdz> fabbione: I even looked at the man page
[05:54] <fabbione> ehehe
[05:57] <ogra> fabbione, 1024x768 :)
[05:58] <ogra> fabbione, with the new generated xorg.conf and selected nvidia driver
[05:59] <fabbione> ogra: weird...
[05:59] <fabbione> ogra: so let me understand all the details
[05:59] <fabbione> if you run nv it works ok
[05:59] <ogra> fabbione, but i have not the slightest idea what to choose in the monitor list :)
[05:59] <fabbione> if you run nvidia (new release) you only get 1024x767
[05:59] <fabbione> what about the old nvidia drivers?
[06:00] <fabbione> ogra: ahhhhhh
[06:00] <mjg59> Any MOTU about?
[06:00] <crimsun> mjg59: a few, yup
[06:00] <diamond> ogra: i genereally remove any monitor specs from the .conf and let X figure it out.
[06:00] <mjg59> sl-modem-daemon really shouldn't depend on sl-modem-source - the kernel support will work in a lot of cases
[06:00] <ogra> no, with the same xorg.conf i get 1280x768 (nvidia) instead of 1280x800 (nv)
[06:01] <daniels> ogra: which one is right?
[06:01] <daniels> if nv is right, and nvidia is wrong ... *shrug*
[06:01] <ogra> daniels, the latter one
[06:01] <daniels> ok
[06:01] <daniels> the fact all my machines run pure open source X drivers isn't just good luck ...
[06:01] <ogra> daniels, i generated a new conf with the update yesterday...which runs very fine (thanks) 
[06:02] <ogra> (with nv)
[06:02] <ogra> but i noted that there was no resolution for my display in the list
[06:02] <daniels> fabbione: my only comment so far is that you haven't put a changelog entry in :P
[06:02] <fabbione> daniels: right :)
[06:03] <mdz> jbailey: what happened with #5421?
[06:03] <daniels> ogra: hm -- if it's not in the list, then xserver-xorg shouldn't tell you about it
[06:03] <ogra> daniels, 1280x800@60 would be nice to have in the last screen of the configuator (monitor selection that is)
[06:04] <daniels> ahhh, I see
[06:04] <ogra> i got it in the device list, but not in the monitor list
[06:04] <daniels> ogra: yeah, just found the problem then, will fix it in the morning
[06:04] <mjg59> ogra: Would it be possible to upload an sl-modem build with the sl-modem-source dependency removed from the sl-modem-daemon binary package?
[06:04] <ogra> great
[06:04] <zul> heh if we only have slrmodem in l-r-m oh wait there is a patch in bugzilla ;)
[06:04] <ogra> mjg59, absolutely.... :)
[06:04] <mjg59> zul: intel-8x0m works in several cases
[06:05] <mjg59> ogra: Could you do it? :)
[06:05] <ogra> mjg59, during the evenin :)
[06:05] <zul> mjg59: i know but its blacklisted in hotplug i think
[06:05] <crimsun> ogra: / mjg59: I'm on it
[06:05] <mjg59> zul: Doesn't seem to be
[06:06] <mjg59> It's been autoloaded here
[06:06] <ogra> crimsun, thanks :) you just donated to hwdb deveopment *g*
[06:06] <mjg59> crimsun: Thanks!
[06:06] <crimsun> ogra: great :)
[06:06] <daniels> fabbione: ok, I've looked through it and I can't see any of the usual mistakes; it looks pretty clean
[06:06] <fabbione> daniels: ok thanks
[06:06] <daniels> actually
[06:06] <daniels> fabbione: nvidiaminor -> 0ubuntu1
[06:06] <daniels> come to think of it
[06:07] <daniels> since it has a new base version
[06:07] <daniels> which should generate you 1.0.7167-0ubuntu1
[06:07] <daniels> instead of -0ubuntu24
[06:07] <fabbione> daniels: ah right, but it's too late for that
[06:07] <fabbione> some people have it installed already
[06:07] <jbailey> mdz: I've been looking for help from someone who knows evms.  I haven't been able to get a system working that can make it work.  I'd like to just follow the least prefered suggestion for now and remove evms support for root devices until our installer supports it.
[06:07] <ogra> btw, could anybody who submitted data online through hwdb-client please check his data on hwdb.ubuntu.com (there is not much shown yet, but i'd like to know if the laptop detection works)
[06:07] <daniels> fabbione: ah, ok
[06:07] <fabbione> daniels: it's just cosmetic anyway
[06:07] <daniels> yeah
[06:08] <daniels> it shouldn't make much difference
[06:08] <fabbione> i am more interested in real packaging issues :)
[06:08] <daniels> heh :)
[06:08] <daniels> well, it all looks OK to me
[06:08] <fabbione> jbailey: hey dude..
[06:08] <jbailey> Heya Fabio.
[06:08] <fabbione> jbailey: did you ever upload mpeg2dec again?
[06:08] <fabbione> 8455 looks scary
[06:08] <jbailey> fabbione: Nope, since Sparc wasn't a release target it didn't seem to make sense to do the upload.
[06:09] <daniels> fabbione: if you don't have anything else, I think I'm going to hit the bed now
[06:09] <daniels> fabbione: was travelling for a fair while to get back home, then had to walk a few km back from the station
[06:09] <fabbione> daniels: good night
[06:09] <daniels> fabbione: ok, night dude.  thanks very much for all this, beer counter goes up once again :)
[06:09] <daniels> let me know if you need anything else from me
[06:09] <jbailey> fabbione: It needs to be fixed properly right after release anyway.  The reason it actually works on ppc now is mostly because the altivec detection fails rather than it actually fixing altivec capabilities.
[06:09] <fabbione> jbailey: you said that you had to rever to gcc-3.3 anyway for other reasons and special case ppc
[06:09] <fabbione> ah ok
[06:10] <fabbione> i guess i will do just a bin NMU for sparc
[06:10] <fabbione> since it's not a release candidate nobody will care :)
[06:10] <jbailey> fabbione: If you remove the inline from that cpu_accel function, it should work fine on gcc-3.4, I think.
[06:10] <fabbione> jbailey: i will just compile with gcc-3.3
[06:11] <jbailey> fabbione: Or revert to gcc-3.3.  Either way it'll be broken just after hoary releases and we bump up to gcc-4.  Depends how persistant you want the fix to be. =)
[06:11] <fabbione> jbailey: just for hoary basically
[06:11] <fabbione> since i have all of main there
[06:11] <fabbione> i am quite happy as it is
[06:12] <fabbione> it would really upset me to see sparc missing one package for a stupid reason, when it can be there
[06:13] <jbailey> fabbione: Well, and it's useful if you want to use totem-gstreamer to play mpegs.
[06:13] <fabbione> right :)
[06:13] <pitti> seb128: hate, hate, hate, hate, hate!!!
[06:14] <pitti> seb128: now I compiled everything, set up gamin for testing, and now I can't reproduce it any more
[06:14] <pitti> heisenbug...
[06:15] <mdz> jbailey: you haven't been able to get evms to work, you mean?
[06:16] <mdz> jbailey: I know evms (as you might imagine), and the guys on #evms are generally helpful
[06:17] <Nafallo> daniels: I belive my girlfriend just mailed you ;-)
[06:18] <jbailey> mdz: 'k, I'll check there.
[06:18] <mdz> infinity: ping
[06:18] <thom> seb128: you happy for me to resolve 7552?
[06:18] <pitti> seb128: hmm, logged out/in, now I see it again *phew*
[06:25] <mvo> quick poll: what do you think about http://people.ubuntu.com/~mvo/update-icon.png for update-notifier?
[06:25] <thom> mvo: i like that
[06:26] <elmo> is it blurry? :P
[06:26] <Mitario> mvo, it's cool :)
[06:26] <lamont> mdz: why are there binary packages in pool/main that have no source?
[06:26] <Mitario> hi everyone
[06:27] <mvo> elmo: phhh :p
[06:27] <mvo> thom, Mitario: thanks, will be part of the next upload then
[06:27] <trukulo> hi ppl
[06:27] <Mitario> nice!
[06:27] <seb128> pitti: k
[06:27] <Mitario> hi Treenaks 
[06:27] <ogra> mvo, does that work for the red/green blind ?
[06:27] <Mitario> eh trukulo 
[06:27] <trukulo> daniels, i want wo ask you something. are you there?
[06:27] <trukulo> hi Mitario :)
[06:27] <seb128> pitti: running the CVS ? for the session I don't know, gam_server is not in the session but started when needed
[06:28] <trukulo> ogra, have you seen amaranth?
[06:28] <ogra> trukulo, nope
[06:28] <pitti> seb128: it doesn't matter, I can start a second server with another id
[06:28] <mvo> ogra: should be ok (I'm partly color-blind and can make out the difference)
[06:28] <ogra> ok
[06:28] <seb128> pitti: right, that's on the website :)
[06:28] <seb128> thom: I've closed it, thanks :)
[06:29] <pitti> seb128: I can't find any errors with test_gam, but I see the bug with the desktop icons. weird
[06:29] <thom> seb128: rocking
[06:29] <ogra> mvo, that was my only concern, otherwise its awesome :)
[06:29] <seb128> now we have to fix this gamin issue
[06:29] <seb128> if we do I'm happy with the desktop for hoary :)
[06:30] <mdz> fabbione: new driver works OK on amd64 here
[06:31] <fabbione> mdz:  thanks
[06:32] <fabbione> mdz: enjoy your first non-free experience and run glxgears!
[06:32] <fabbione> :P
[06:32] <pitti> tuxracer
[06:33] <pitti> seb128: can you please try something?
[06:33] <pitti> seb128: while true; do touch Desktop/baz; sleep 1; rm Desktop/baz; sleep 1; done
[06:33] <mdz> fabbione: I did
[06:33] <fabbione> eheheh
[06:33] <mdz> if there is something better I can run to stress it, tell me
[06:33] <fabbione> mdz: tuxracer
[06:33] <fabbione> q3a
[06:34] <pitti> seb128: or, rather: while true; do echo Hello > Desktop/baz;  sleep 1; rm Desktop/baz; sleep 1; done
[06:34] <trukulo> enemy territory
[06:34] <ogra> so many erroneous hwdb reports....
[06:34] <pitti> seb128: for some reason the touch worked ad infinitum, but with echo it stopped after a few iterations
[06:34] <trukulo> more nonfree stuff for your experience :)
[06:34] <seb128> pitti: touch works fine
[06:34] <seb128> pitti: echo boooog
[06:35] <pitti> HEY!
[06:35] <pitti> seb128: probably if the file size is prime, or so
[06:36] <mdz> I can't see which menu choice is highlighted in tuxracer; is that normal?
[06:36] <mdz> I have to navigate blind
[06:37] <trukulo> mdz, umm, think it's not normal, menus are suposed to select choose, not to randomly choose
[06:38] <trukulo> when i played tux on the past (not now with ati igp) i can view menus, with nvidia binary drivers on debian
[06:38] <mjg59> Kamion: About?
[06:39] <mjg59> Anyone know how I generate a Packages file that includes all the correct Task information?
[06:40] <elmo> use apt-ftparchive and 'ExtraOverride' in a conf file
[06:40] <elmo> look at katie (av. on cvs.d.o or in 'dak' package in sid) for an example
[06:40] <mjg59> Ah
[06:40] <mjg59> Is the Ubtuntu taskfile handily available?
[06:41] <Kamion> mjg59: yep?
[06:41] <mjg59> Kamion: ^
[06:42] <Kamion> mjg59: /indices/ in the archive
[06:42] <mjg59> Kamion: Ta!
[06:43] <mjg59> So I just feed those to apt-ftparchive and point it at my pool?
[06:44] <elmo> pretty much
[06:44] <elmo> you need to be using the apt.conf style invocation of apt-ftparchive
[06:44] <elmo> not old-skool dpkg-scanpackages imitation mode
[06:45] <crimsun> mjg59: uploaded
[06:45] <mjg59> crimsun: You rock
[06:46] <mjg59> Making custom ISOs is a bit of a pain
[06:47] <sabdfl> mvo: i like it, but the exclamation mark might look better in yellow
[06:49] <_mvo_> sabdfl: yellow is a interessting idea, I'll try it out
[06:50] <ogra> crimsun, thanks! :)
[06:50] <lamont> mjg59: yes, it is.
[06:51] <trukulo> one question, in booting speed, it's the same install freshly than upgrade from warty? because it seems to me that booting is slower now than before (in warty)
[06:51] <seb128> pitti: this gamin issue will be funny to debug :/
[06:52] <koke> mako: ping?
[06:52] <pitti> seb128: we currently discuss it in #g-hackers
[06:53] <pitti> seb128: it doesn't seem a gamin-only bug after all
[06:53] <seb128> I know, I'm here :)
[06:53] <seb128> that's why I say it'll be funny to track :/
[06:53] <dholbach> re
[06:54] <ogra> hiya
[06:54] <pitti> seb128: in any case I can't reproduce it with test_gam as client
[06:54] <dholbach> Mithrandir: memtest86+ on amd64 works fine
[06:54] <mako> koke: hey dude
[06:54] <koke> do you know which is my key status atm?
[06:54] <koke> IIRC, elmo was waiting for you
[06:55] <Mithrandir> dholbach: yeah, 'cept on opteron, maswan says
[06:55] <dholbach> Mithrandir: that i couldnt test
[06:55] <mako> koke: i uploaded the approved file yesterday
[06:56] <maswan> With 1.5TB ram in an opteron cluster, you kind of end up needing a working memtest86+ :)
[06:58] <koke> mako: ok, thanks, so next step is for elmo or I can already upload?
[07:00] <bronson> I just upgraded Hoary and now all my moz-based browsers crash when visiting http://anandtech.com .  Konq works fine.  Not looking for tech support -- just letting you guys know.
[07:01] <bronson> It appears ad-related...?
[07:01] <seb128> works fine here
[07:01] <mako> koke: i emailed telling him i updated the file.. so he should handle it
[07:02] <bronson> seb128: are you using an ad blocker?
[07:02] <bronson> x386?
[07:03] <mjg59> Hm.
[07:04] <mjg59> Oops
[07:04] <seb128> bronson: no
[07:05] <mjg59> And the Task headers are missing, too.
[07:06] <trukulo> bronson, seem to work well here (x86 and firefox)
[07:06] <bronson> WEIRD
[07:06] <trukulo> bronson, hoary last updated
[07:07] <bronson> Last night.
[07:07] <jani> fabbione GeForce2 MX works fine with the new driver
[07:07] <trukulo> bronson, same here
[07:08] <trukulo> Amaranth, how about bugs on gnome-editor?
[07:08] <Amaranth> trukulo: ?
[07:08] <fabbione> jani: awesome! thanks
[07:08] <trukulo> Amaranth, kde and translation related (yesterday)
[07:08] <Amaranth> trukulo: oh, i haven't gotten to them yet
[07:08] <Amaranth> trukulo: been working on PyMusique
[07:08] <trukulo> Amaranth, ok, keep on pymusique ;)
[07:13] <mjg59> Kamion: Oh, hrm. Do I need a filelist?
[07:13] <lamont> install -m0755 post-base-installer debian/localechooser/usr/lib/post-base-installer.d/05localechooser
[07:13] <lamont> install: cannot create regular file `debian/localechooser/usr/lib/post-base-installer.d/05localechooser': No such file or directory
[07:13] <Kamion> the CD uses a filelist
[07:13] <Kamion> dunno what katie does
[07:13] <Kamion> lamont: bugger
[07:14] <lamont> ah, was Kamion.  just looking that up
[07:14] <lamont> missing dirs entry?
[07:14] <Kamion> yeah, will fix
[07:17] <Amaranth> ugh, i hate it when people write nasty code with no documentation
[07:17] <mjg59> Kamion: Is the filelist on the CD?
[07:20] <fabbione> ok new l-r-m uploaded
[07:20] <fabbione> fasten your seatbelts
[07:21] <T-Bone> heh
[07:21] <lamont> but I want 9.3.1 in breezy
[07:22] <Kamion> mjg59: no, but it's just the output of 'find' over the pool
[07:22] <Kamion> (sorted)
[07:24] <elmo> eh, in ubuntu or debian?
[07:24] <Kamion> Ubuntu
[07:24] <elmo> was it NEW or part of a sync?
[07:24] <Kamion> NEW
[07:24] <Kamion> it's, er, kinda lacking a few things ...
[07:24] <lamont> hrm... actually....
[07:26] <elmo> Kamion: eh, you added it
[07:26] <Kamion> elmo: yes, I know, that's why I'm eyeing it ;)
[07:26] <elmo> udebs by you are basically whitelisted in ubuntu NEW, TBH
[07:26] <elmo> as I assume if _you_ don't know what you're doing, we might as well all go home :p
[07:26] <Kamion> it's a .deb actually
[07:26] <elmo> oh
[07:27] <Kamion> and it's fixed now :)
[07:27] <elmo> boggle
[07:28] <elmo> ok, yeah, I was crap too for not noticing that - I think I thought it was a udeb
[07:30] <Kamion> are missing changelog/copyright not covered by an auto-reject then? I suppose they run into all the fun with /usr/share/doc symlinks to other binaries in the same source package that you depend on
[07:31] <elmo> yah
[07:35] <mjg59> Kamion: Sorry, a list of files from which standpoint? The top of the archive?
[07:35] <mjg59> Or the top of the pool?
[07:39] <HiddenWolf> Hm. powernowd is *very* agressive
[07:39] <Kamion> mjg59: it's "find pool/main | sort"
[07:40] <mjg59> Kamion: Rock, thanks
[07:40] <fabbione> mdz: if there is nothing urgent left for today i would be off until tomorrow.
[07:40] <mdz> fabbione: ok
[07:40] <fabbione> mdz: i have my mobile phone with me in case
[07:40] <fabbione> it's turned on
[07:40] <fabbione> if i don't answer, either SMS or leave a message
[07:40] <elmo> (till he gets out the door ;)
[07:40] <fabbione> elmo: nah.. i am not going out this evening
[07:40] <fabbione> my wife is out.. me party at home!
[07:41] <Kamion> mjg59: if you want what cdimage does, get colin.watson@canonical.com--2005/cdimage--mainline--0 from http://people.ubuntu.com/~cjwatson/archives/colin.watson@canonical.com--2005, baz build-config configs/devel, and look in debian-cd/tools/scanpackages
[07:42] <fabbione> rocking.. only one package missing in main for sparc
[07:42] <fabbione> elmo: do you think you can manage to move sparc.u.c?
[07:43] <fabbione> elmo: never mind.. i promised not to bitch.. sorry 
[07:44] <elmo> fabbione: yah, sorry - just need to concentrate on stuff I need to do while still at DC ATM
[07:44] <fabbione> elmo: yes.. sorry...
[07:44] <fabbione> don't worry...
[07:45] <fabbione> i can still live with the cache :)
[08:01] <mdz> Kamion: can you update the germinate-output for hoary?
[08:02] <T-Bone> fabbione: you're way behind. We're left with less than 1900 packages to build on hppa (main/restricted/universe) =}
[08:02] <T-Bone> about 30% uncompiled
[08:02] <fabbione> T-Bone: you have a cluster.. i have one machine.. oh btw.. does the installer work on hppa?
[08:02] <fabbione> because on sparc does
[08:02] <T-Bone> lol
[08:02] <Amaranth> is ubuntu going to have something like debian unstable?
[08:02] <T-Bone> damn you :)
[08:03] <T-Bone> fabbione: installer is on the todo list :)
[08:03] <Amaranth> people in #ubuntu are saying breezy is the name of the next release and that grumpy is the name of what will be like unstable
[08:03] <elmo> Next week on PORTS WARS!
[08:03] <T-Bone> elmo: hehehe ;)
[08:04] <elmo> fabbione and t-bone are shipped off to a desert island.  they're given nothing but spare m68k parts.  to survive they must collaborate to build a radio and call for help.
[08:04] <T-Bone> elmo: rotfl ;o))
[08:04] <fabbione> elmo: ahahahahahah
[08:05] <fabbione> elmo: be careful :) i still have a bunch of m68k's around..
[08:05] <fabbione> you don't want m68k.ubuntu.com, do you?
[08:05] <fabbione> :)
[08:05] <Amaranth> heh, that'd be funny
[08:05] <Amaranth> debian is talking about dropping archs while ubuntu starts adding more :P
[08:05] <fabbione> ok time to cook dinner :)
[08:21] <Kamion> Amaranth: not quite
[08:21] <Kamion> Amaranth: breezy is the name of the next release, grumpy (if ever implemented) will be something like builds from upstream CVS
[08:22] <Amaranth> wouldn't that be a lot more work for little gain? :)
[08:22] <elmo> no
[08:22] <elmo> building from upstream CVS has a lot of benefits
[08:22] <HiddenWolf> Kamion: Breezy .... what is it? :)
[08:22] <Kamion> HiddenWolf: huh?
[08:22] <elmo> HiddenWolf: badger
[08:23] <HiddenWolf> elmo, thanks
[08:25] <Kamion> ah, I thought you were asking for the meaning of breezy or something
[08:27] <mjg59> Kamion: Do I need to do something special about language packs?
[08:28] <Kamion> mjg59: what sort of prompt?
[08:28] <mjg59> Kamion: The CD does not contain full support for your chosen language
[08:29] <mjg59> Kamion: Ok, the apt-get that installs most of the system has prompted me due to lack of authentication
[08:29] <mjg59> Oh, and it wanted the CD back in - do I need to add extra packages to the list of stuff to be copied from the CD?
[08:29] <mjg59> (I've added a multiverse section, so things may be complicated)
[08:35] <mjg59> Grah. And now X has blown up, in a slightly unexpected turn of events (it wants resolution prompting)
[08:35] <mjg59> Which is odd, because I used the same package earlier today without any trouble
[08:36] <tritium> fabbione, I can't seem to use NVagp with the newer nvidia packages.  I had to switch to agpgart.
[08:37] <fabbione> tritium: the packages from archive or people?
[08:37] <Kamion> mjg59: you'll need to make sure that multiverse is in /.disk/base_components, and if you want a package to be copied to the disk, put 'Task: ubuntu-ship' on it
[08:38] <mjg59> Kamion: Ah, right
[08:38] <tritium> fabbione, people could use NVagp, but I experienced freezes about once per hour.  Just tried the one in archive.
[08:38] <Kamion> mjg59: (for a CD, every package should probably have one of ubuntu-{base,desktop,ship})
[08:38] <fabbione> tritium: yes, that is correct. there is an errata from nvidia in the pkgs in the archive.
[08:38] <mjg59> Ah, it must have been pulling in dependencies
[08:39] <tritium> fabbione, okay, thanks.
[08:39] <mjg59> Kamion: apt-get failed with an error - is that logged?
[08:39] <fabbione> tritium: http://www.nvnews.net/vbulletin/showthread.php?t=47405 <- this one
[08:39] <Kamion> mjg59: in the second stage? /var/log/base-config.log
[08:39] <Kamion> mjg59: the "full support" thing comes up if language-support-$LL isn't on the CD
[08:40] <mjg59> Setting up python2.4-ldap... dpkg-query: error writing `stdout': Broken pipe
[08:40] <mjg59> Kamion: Hrm. I copied files off a .iso, then mkisofsed it and stuffed it back on
[08:40] <fabbione> Kamion: can you please build a cd image with the new l-r-m and multiarse?
[08:40] <fabbione> Kamion: i need to give it a shot before tomorrow's daily
[08:40] <mjg59> Ah, I see why stuff broke
[08:40] <zul> multiarse?
[08:41] <tritium> thanks, fabbione.  I'm looking it over.
[08:41] <fabbione> zul: dpkg -p multiseat
[08:42] <zul> Package `multiseat' is not available.
[08:42] <zul> and yes i did an update
[08:43] <fabbione> zul: your box is on crack
[08:43] <fabbione> zul: it's not in warty....
[08:43] <zul> probably
[08:43] <zul> not on hoary
[08:43] <zul> er..warty
[08:47] <mdz> pitti: gah, I had already run the script and shut down the machine again when I saw you attached a new version
[08:47] <mdz> pitti: is it needed?
[08:47] <elmo> do we have the unpacked source of hoary anywhere?
[08:47] <pitti> mdz: yeah, it checks for the drive status as well
[08:47] <pitti> mdz: I just send some explanatory text
[08:48] <mdz> pitti: ok, I'll re-test
[08:48] <pitti> mdz: sent, even
[08:48] <Kamion> elmo: rookery:~scott/merge-o-matic/sources/?
[08:48] <Kamion> hm, doesn't have everything
[08:48] <kagou> hi
[08:49] <elmo> I mostly care about main
[08:50] <elmo> not to worry, plent of space, I'll find somewhere to unpack it to
[08:50] <Kamion> doesn't have all of main either
[08:50] <elmo> /dev/sda3             537G  490G   20G  97% /
[08:50] <elmo> just not there ;)
[08:50] <Kamion> heh
[08:50] <zul> hah
[08:50] <Kamion> what's that?
[08:50] <mdz> pitti: sent the output to bugzilla; is it safe to shut down the machine?
[08:50] <mdz> Kamion: jackass
[08:50] <elmo> kamion: oh, only jackass
[08:50] <pitti> mdz: I think so, there is not much I can do about this
[08:50] <pitti> mdz: thanks for testing
[08:51] <Kamion> elmo: bit concerning ... is that mostly the morgue?
[08:51] <elmo> kamion: no, it's my BACKUP THE WORLD policy for the DB
[08:51] <elmo> cron."""daily""" still takes a pre/post snapshot of the DB
[08:51] <Kamion> hah
[08:51] <elmo> I think it's responsible for 150Gb or so
[08:52] <mdz> elmo: rdiff-backup works well for that
[08:52] <Kamion> sounds like how cdimage used to be
[08:52] <mdz> I used to store months and months of hourly db snapshots that way
[08:53] <jani> elmo I did my first uni upload about 20 min ago and got no ack/nack mail, any idea?
[08:53] <elmo> jani: pkg name?
[08:53] <jani> python-soya-doc
[08:54] <jani> the file name was soya-doc_0.9-1ubuntu1_source.changes
[08:54] <Kamion> fabbione: new CD building, ETA an hour or so; I'll be gone well before it arrives
[08:54] <pitti> mdz: argh, status 0 means that there is no CD in the drive
[08:54] <pitti> mdz: so you tested without a CD?
[08:54] <pitti> mdz: (sorry, I got the followup just now)
[08:54] <fabbione> Kamion: that's fine thanks.
[08:55] <elmo> From: Ubuntu Installer <katie@jackass.warthogs.hbd.com>
[08:55] <elmo> To: Jani Monoses <jani@email.ro>
[08:55] <mdz> pitti: the _first_ time I ran the script I had a CD in the drive
[08:55] <elmo> Subject: soya-doc_0.9-1ubuntu1_source.changes ACCEPTED
[08:55] <elmo> Message-Id: <20050331181512.938653B9C029@jackass.warthogs.hbd.com>
[08:55] <mdz> pitti: I got identical results whether the disc was in or not, so the second time I did not re-test
[08:55] <pitti> mdz: *sigh* sorry
[08:55] <jani> hmm did not get it :(
[08:55] <mdz> this is why I asked if we were finished :-/
[08:55] <jani> what time did it send it?
[08:55] <elmo> 2005-03-31 19:36:12 1DH4CG-0001fl-Jt mx0.email.ro [193.226.99.16] : Connection refused
[08:55] <elmo> 2005-03-31 19:36:12 1DH4CG-0001fl-Jt == jani@email.ro R=lookuphost T=remote_smtp defer (111): Connecti
[08:55] <pitti> mdz: hmm, if the status is 0 even if a CD is inserted, then there is really something broken
[08:55] <elmo> on refused
[08:56] <jani> oh, crappy free mail hosting
[08:56] <mdz> pitti: the old script didn't show the status, only the capabilities
[08:56] <pitti> yeah
[08:56] <mdz> pitti: I didn't test the new script with a CD inserted
[08:56] <jani> I think the addres is why it is not seen on changes the spamfilter does not like my address :)
[08:57] <jani> ok thanks elmo, at least it means my key and all are ok I can go on uploading
[08:58] <mdz> pitti: status: 0 with a non-blank DVD+RW inserted
[08:58] <andred> seb128,  So with the new changed spatial thing in Ubuntu, folders will close automatically when you open a new folder?
[08:58] <andred> If so, why?
[08:58] <pitti> mdz: this is weird
[08:58] <pitti> mdz: the kernel blatantly lies...
[08:58] <kagou> when i save an archive downloaded with firefox on the desktop, i must to refresh the desktop to see it. Is it normal ?
[08:59] <mdz> kagou: it is normal, but it is also a bug
[08:59] <kagou> mdz, it's the gamin's goal to do that no ?!
[09:00] <mdz> kagou: yes, the bug is most likely in gamin
[09:00] <mdz> kagou: https://bugzilla.ubuntu.com/show_bug.cgi?id=7078
[09:00] <kagou> thx mdz
[09:02] <HiddenWolf> I just burned an iso with nautilus. The cd looks physicly like it's burned for about 200mb, but nautilus says it's a 600mb cd
[09:04] <mdz> pitti: I am not able to reproduce #7078 with your loop after killall gam_server
[09:04] <mdz> pitti: how many iterations does it take for you?
[09:04] <pitti> mdz: sometimes 5, sometimes it fails immediately
[09:04] <mdz> I have been through about 30
[09:04] <pitti> mdz: in any case less than a minute
[09:04] <mdz> and it is still working fine
[09:04] <mdz> I could reproduce it before, e.g. saving files from firefox
[09:04] <andred> seb128, With that change, spatial nautilus is absolutely worthless.
[09:05] <pitti> mdz: I reproduced it on ppc and i386, seb128 had it too
[09:05] <pitti> mdz: I talked with upstream about this
[09:05] <pitti> mdz: it's not a bug in gamin itself, but a race condition between nautilus and gamin
[09:05] <pitti> mdz: nautilus accesses libgamin concurrently and opens several watches on a directory
[09:05] <pitti> mdz: it seems that this clashes from time to time
[09:06] <mdz> pitti: can we have nautilus use a mutex and avoid the concurrency?
[09:06] <kagou> i'm upto date, and all tar.gz downloaded do not appear
[09:06] <jani> elmo are you making ubuntu.com addresses for members?
[09:06] <pitti> mdz: it could help to open just _one_ watch per directory
[09:06] <pitti> seb128: can this be done? ^
[09:07] <mdz> pitti: I can't reproduce the bug anymore
[09:07] <pitti> mdz: I couldn't either, then I restarted my session and it "worked" again
[09:07] <pitti> it really seems to be a race
[09:07] <pitti> mdz: I told upstream everything I found out, maybe he finds something
[09:08] <mdz> pitti: ok, I tried on my laptop and reproduced after only 2 iterations
[09:08] <mdz> Kamion: can you re-germinate before you go?
[09:08] <mdz> (if you haven't already)
[09:09] <Kamion> mdz: er, yeah, already did, forgot to say
[09:09] <Kamion> mdz: germinate is currently set to run every hour though, so it doesn't generally need me
[09:10] <Kamion> 2 * * * *               update-germinate
[09:10] <mdz> pitti: I see so many EAGAIN and EINTR in gam_server; I wonder if it is handling them correctly
[09:10] <mdz> Kamion: ok, night
[09:11] <elmo> jani: yes, eventually
[09:11] <mdz> elmo: OK to run cron.sync to verify that php4 is happy nw?
[09:11] <mdz> now?
[09:11] <jani> ok, because it looks like hoary-changes does not get any of my changes
[09:12] <elmo> mdz: already done
[09:12] <jani> with my current mail address
[09:13] <mdz> elmo: yay
[09:19] <elmo> what's a shell way to say "show me everything after REGEXP in a file" ?  csplit seems a bit heavy handed
[09:20] <diamond> elmo: i'd probably use awk, but messey
[09:21] <Treenaks> grep with an incredibly large -A
[09:21] <Treenaks> ?
[09:21] <elmo> grep -A $(expr 2 \*\* 99) ?
[09:21] <Treenaks> you could try..
[09:21] <elmo> hmm, expr doesn't do ** or ^, lame
[09:22] <Treenaks> well, 1024*1024*1024*4 should be enough
[09:22] <Treenaks> MOST of the time
[09:24] <pitti> dholbach: I saw that you added libxinerama-dev to qiv
[09:24] <pitti> dholbach: is this required for Debian as well? (I'm the maintainer)
[09:25] <dholbach> pitti: erm... must be x.org related
[09:25] <mdz> elmo: sed
[09:25] <dholbach> pitti: so for the moment... no
[09:25] <elmo> ah, right
[09:25] <pitti> dholbach: okay, thanks
[09:25] <dholbach> pitti: sorry, wanted to notify you about the change but forgot it in the hurry
[09:25] <pitti> no worries :-)
[09:27] <dholbach> good :-)
[09:28] <jbailey> elmo:  tail -n+$(grep -n bar infile | head -n1 | while IFS=: read a b; do echo ${a}; done) infile
[09:28] <jbailey> elmo: Where bar is the regexp and infile is the file to search in.
[09:29] <elmo> see, I _knew_ someone would come up with a sick shell solution
[09:29] <jbailey> Posix shell is my bitch. =)
[09:30] <mdz> sed -ne '/regexp/,$p' is just a bit more elegant
[09:31] <diamond> jbailey: hah. nice.
[09:31] <lamont> jbailey: do you _know_ there are no meta characters in a?
[09:32] <jbailey> lamont: grep returns the line number, so I think it should be safe.
[09:32] <lamont> ah, yes
[09:32] <elmo> and the inevitable critique of the sick shell
[09:32] <lamont> mdz: he said 'after regexp'...  sed '1,/regexp/d' 
[09:32] <elmo> all we need now is a shoop pimp and I'll be full on flashbacking to #d-d
[09:32] <diamond> elmo: -)
[09:33] <lamont> lol
[09:33] <lamont> elmo: depending on whether you want the regexp or not, see mdz/my version
[09:33] <lamont> s/regexp/line with the regexp/
[09:34] <mdz> pitti: does gamin upstream acknowledge the bug yet?
[09:35] <pitti> mdz: he acks that this is a problem, not that it is a gamin bug
[09:35] <pitti> mdz: he said that he needs to evaluate this
[09:35] <doko> mvo: ping
[09:36] <mvo> doko: pong
[09:48] <elmo> god it freaks me out how many people must use the non-free nvidia driver
[09:49] <ogra> gamers :P
[09:49] <zenrox> elmo, for me its casue the free one dont work
[09:50] <jbailey> Keeping my wife happy when she asks why the screensavers suck...
[09:50] <ross> quake really sucks with the nv driver
[09:50] <elmo> tho, having this loaner laptop, it has to be said.. wow, xscreensaver so pretty pretty these days
[09:50] <zenrox> lol
[09:50] <zenrox> that too
[09:50] <ogra> jbailey, electricsheep looks way cooler and doesnt need GL
[09:50] <zenrox> i like the smoke one
[09:51] <zenrox> hufo's smoke
[09:51] <zenrox> execlet screen saver
[09:52] <jbailey> Where's the bit in the installer where I'm supposed to be able to setup a partition as raid?   I thought it was in the same list as where you picked the filesystem for the partition.
[09:53] <elmo> jbailey: it is
[09:53] <elmo> it's in the "what to use this partition for" bit
[09:53] <elmo> IIRC
[09:54] <jbailey> elmo: Right, I'm in that screen, I see ext3, ext2, reiserfs, jfs, xfs, fat16, fat32, swap, newworld boot, and do not use the partition.
[09:55] <elmo> what arch?
[09:55] <jbailey> Mac G5
[09:55] <elmo> raid's not supported
[09:55] <elmo> by the partitioner
[09:55] <elmo> apparently svenl was working on it, but it didn't/hasn't gone in in time for us
[09:56] <jbailey> Ah, that would explain why I've not been able to find it then.
[09:56] <jbailey> Thanks.
[10:45] <lamont> xsltproc -o output-html/ ./../aptitude-html.xsl ./aptitude.xml
[10:45] <lamont> error : Operation in progress
[10:45] <lamont> ./aptitude.xml:65: warning: failed to load external entity "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd"
[10:45] <lamont> hrm... would that be like something we should fix?
[11:13] <dredg> elmo: ping
[11:14] <trulux> tritium: writing on the paper, it's getting long and long and I don't see the... end!
[11:14] <tritium> trulux, you'll get there :)
[11:15] <tritium> seb128, jpr just told me there will be a patch in evolution 2.2.2 on Monday to fix vcalendar breakage.  Will that make it into Hoary?
[11:16] <trulux> tritium: I hope, really, I might cry then, 'cos too much work around it creates strong feelings like: what the hell made my LaTeX package to break the pdf!?
[11:16] <trulux> "oh my dear, you broke it!"
[11:17] <trulux> it's preparing me for being married in a far away future
[11:18] <tritium> heh
[11:20] <seb128> tritium: if somebody send me the patch or if they roll a tarball yep
[11:20] <tritium> seb128, I'll try to do that for you then.
[11:20] <seb128> andred: ?
[11:21] <seb128> tritium: they will probably roll tarballs for GNOME 2.10.1, don't bother
[11:21] <tritium> seb128, okay, thanks for telling me.  :)
[11:22] <seb128> pitti: hum, I can have a look but I don't the nautilus code well, this changes are probably not trivial
[11:23] <pitti> seb128: that's what I was afraid to hear :-/
[11:23] <pitti> seb128: however, I will continue to debug gamin tomorrow
[11:23] <pitti> seb128: maybe I can find something and there is an easy workaround
[11:23] <seb128> I can look on it too if you want, let me know
[11:24] <pitti> seb128: maybe you can look why nautilus opens so many watches onto the same directory?
[11:24] <seb128> I'll have a look
[11:26] <andred> seb128, What did you not understand?
 seb128, With that change, spatial nautilus is absolutely worthless.
[11:26] <seb128> you have a gconf key to set if you don't like the changes
[11:27] <andred> seb128, Yeah, but why go against upstreams in this way?
[11:27] <seb128> sabdfl decision
[11:27] <andred> seb128, It breaks the whole spatial metaphor.
[11:27] <andred> seb128, sabdfl?
[11:28] <seb128> /whois sabdfl
[11:28] <andred> Shuttleworth?
[11:29] <seb128> andred: right
[11:29] <seb128> andred: I know that has been discussed upstream again and again
[11:29] <seb128> default is spatial is nice if you do through several directories
[11:29] <andred> Yeah, but that was mostly browsing vs. spatial.
[11:29] <seb128> if you does you have a lot of windows
[11:29] <seb128> and that sucks
[11:29] <seb128> no
[11:30] <seb128> they windows still have they properties
[11:30] <seb128> one by folder
[11:30] <seb128> size, position, content, etc
[11:30] <andred> Yeah, but then default to browsing nautilus, not thos half-done spatial.
[11:30] <seb128> that's well done spatial :)
[11:30] <dredg> not the browse-vs-spatial argument again :(
[11:30] <seb128> arguing for hours will not go anywhere
[11:30] <seb128> you don't like the change turn the gconf key
[11:30] <andred> I like spatial. I don't like broken spatial.
[11:30] <azeem_> is this about closing the parent windo automatically?
[11:31] <azeem_> window, even
[11:31] <maswan> pfft. your release candidates are week. not even twice the "normal" bandwidth used. ;) http://www.acc.umu.se/technical/statistics/ftp/monitordata/index.html.en 
[11:31] <seb128> azeem_: right
[11:31] <T-Bone> maswan: heh ;] 
[11:31] <seb128> azeem_: change left/middle click for the folders
[11:31] <seb128> andred: that's why there is a gconf key
[11:32] <azeem_> sounds sensible to me
[11:32] <andred> seb128, I know, but I'm talking about sane defaults.
[11:32] <seb128> look on the number of user complaining on the list about the default
[11:32] <andred> I mean, what's the metaphor here? I open a folder and the previous one closes? That sounds like browsing metaphor.
[11:32] <azeem_> it's a function of your folder depth I guess
[11:32] <seb128> having ton of windows open
[11:32] <seb128> correct
[11:32] <seb128> exactly
[11:32] <azeem_>  if you only have two or three levels of folders under $HOME, the old behaviour is alright
[11:33] <azeem_> if you have more, closing the parent makes more sense
[11:34] <seb128> there is a mail of alex about that in the spatial discussion saying than if users have to go through 5-6 folders to reach what they want then the spatial mode miss the target
[11:34] <seb128> that's a discussion about how you should organize your files
[11:35] <seb128> I don't have strong opinion either way
[11:35] <andred> I wouldn't have minded if they completely switched to browsing instead then, but this compromise seams weird.
[11:36] <andred> Oh well, lets hope people like that default.
[11:36] <pitti> seb128: sabdfl doesn't want browsing?
[11:37] <trulux> oops, hoary-livcd gets frozen in my laptop while loading the kernel image
[11:38] <seb128> pitti: dunno, ask him :)
[11:38] <seb128> pitti: browser kind of suck imho
[11:38] <ogra> yeah
[11:39] <schweeb> are you guys discussing what I just happened to have noticed? nautilus only keeps one window open now?
[11:39] <ogra> we dont discuss it ;) 
[11:40] <ogra> only commenting
[11:40] <diamond> don't mention the war? -)
[11:42] <ogra> hwdb submissions: yesterday 527, today 639 , Yay
[11:43] <andred> ogra, It's surprising that that many people discover that feature.
[11:43] <ogra> yeah, i've hidden it well ;P
[11:43] <dholbach> andred: there are LOTS of people
[11:44] <dholbach> :-)
[11:44] <andred> So all that data will be collected on a webpage eventually to show what hardware is supported+
[11:45] <andred> ?
[11:45] <ogra> and many many more things, yes
[11:45] <andred> Like what? I'm curious.
[11:45] <HiddenWolf> ogra; what are those possibilities?
[11:46] <ogra> it starts with simple statistics like andred menitioned....
[11:46] <pitti> ogra: nice idea, I'll send you mine, too :-)
[11:47] <zenrox> + stuff like debs installed
[11:47] <ogra> but also the xorf.conf and your bootlog data is collected, you can ues this data for a cool helpdesk system, or attach your id for the datasheet to a bug instead of collecting logfiles
[11:48] <ogra> etc etc....having the data opens a world of options...just think a bit creative about it, there are endless possibilitys
[11:48] <andred> ogra, Are you the one coding hwdb-client?
[11:48] <ogra> yup
[11:49] <torkel> phonenumber, credit card info,... :-)
[11:49] <ogra> hehe
[11:49] <pitti> ogra: nice work :-) just sent you my iBook stuff
[11:50] <ogra> pitti, i see it ;)
[11:50] <ogra> http://hwdb.ubuntu.com/?xml=3761d0be1cbf20870f4b24841806eef4
[11:50] <andred> I discovered a bug in it the other day. If you type something into the comment field, then change to "partial" the "Forward" button is greyed out, even though you already have stuff written in the text field.
[11:50] <pitti> ogra: erm, that's everything? where is the other stuff?
[11:50] <ogra> andred, yeah, i'm aware of this, havent fond a way around yet....
[11:50] <andred> ok, good that you nkow it
[11:51] <seb128> https://bugzilla.ubuntu.com/show_bug.cgi?id=8465
[11:51] <seb128> a guy like the nautilus change :)
[11:51] <ogra> pitti, this is a quickly hacked script that greps cpu, mam and laptop recognition data from the file ;) 
[11:51] <pitti> ogra: okay, so you do have the other info
[11:52] <ogra> pitti, sure, but i havet got the time to write a big server app before release, so this must be sufficient for now :)
[11:52] <pitti> ogra: oh, don't worry, I was just concerned that something got lost
[11:52] <andred> How do you find the ID so you can find your submission like that?
[11:54] <andred> aha, in ~/.hwdb
[11:54] <ogra> yeah...the next version will have a button that opens the webpage for you
[11:55] <ogra> (i had no server when i released the last)
[11:57] <andred> Exactly what fields can you search for on that webpage? I can't find a single entry when I search:-)
[12:00] <ogra> andred, only cpu and memory data of your system and a little laptop detection is done....this is plain flatfile work, no database yet...