[12:03] <lamont> ogra: thanks
[12:15] <thully_> hi - I closed a bunch of my bugs (ones that were obsoleted by changes in hoary, suggestions that really can't be considered at this point, etc etc) and opened a few new ones
[12:15] <thully_> I added some info to the timezone bug - may help in fixing this
[12:33] <mdz> haggai: amd64 has failed on oo.o2, but the others are apparently still going.  a good sign, I take it :-)
[12:34] <thom> mdz: yo?
[12:35] <mdz> lamont: it looks like the locales in the cloop image are selected, but not actually generated yet.  I want them to be pre-generated
[12:35] <mdz> thom: never mind
[12:35] <lamont> mdz: grumble.  I thought I _had_ pregened them.
[12:35] <lamont> will fix.
[12:36] <mdz> lamont: well, 
[12:36] <mdz> I'm getting the usual complaints from perl, at any rate
[12:36] <mdz> but when locale-gen runs, it lists all 4 locales
[12:36] <mdz> (and after that the errors go away)
[12:36] <lamont> right.
[12:36] <lamont> will investigate tonight\
[12:36] <mdz> thanks
[12:37] <lamont> but also mustn't be late.
[12:37] <lamont> bbiab
[12:37] <mdz> thom: have you tried the live CDs yet?
[12:39] <jdub> so the livecd image building bits... do we have a timeframe for cracking those open?
[12:39] <ajmitch> haggai: ok, finally built :)
[12:39] <jdub> have some more people interested in doing customised livecds
[12:39] <ajmitch> the .diff.gz is a mere 1.9MB
[12:39] <T-Bone> jdub: i'm one of these ;)
[12:44] <jdub> mdz: ping
[12:45] <mdz> jdub: pong
[12:45] <mdz> jdub: those two things are not related
[12:45] <mdz> sivan green and brian sutherland are in the process of writing a HOWTO at this very moment
[12:45] <mdz> based on customizing an existing Ubuntu live CD image, rather than building it from scratch, which is far preferable
[12:46] <mdz> it's very simple
[12:53] <jdub> ok, urling up the dude on the in-progress document
[12:53] <thom> mdz: not yet
[12:54] <mdz> thom: are you short on bandwidth where you are now?
[12:54] <thom> no, just short on anywhere to put computers until i get a desk on saturday
[01:07] <seb128> thom: do you have the bogus webdav around ? 
[01:07] <seb128> thom: not needed now but would be nice to try some fixes tomorrow
[01:07] <thom> seb128: no. i'll talk to old job tomorrow and see if they'll reopen it
[01:08] <seb128> ok
[01:08] <seb128> thanks
[01:17] <thom> seb128: so my new firefox works and epiphany still works so i'm gonna upload
[01:17] <thom> you can apply the ephy half of that patch and we get search again
[01:17] <stockholm> huhu!
[01:17] <stockholm> does canonical have a logo, too?
[01:17] <jdub> stockholm: canonical.com
[01:17] <seb128> thom: rock, thanks !
[01:18] <stockholm> geee. (c:
[01:19] <stockholm> on that page i dont see a telephone number...
[01:20] <mako> stockholm: what sort of telephone number are you looking for?
[01:20] <stockholm> mako: optimally for marks. 
[01:21] <stockholm> i want to harass him becaues of debconf sponsorship in general and debconf5 specifically
[01:22] <daniels> stockholm: email generally works well
[01:22] <daniels> we all have internet access
[01:22] <stockholm> daniels: oh. (c:
[01:22] <stockholm> i got delegated to jane.
[01:23] <stockholm> ok, thanks
[01:27] <jdub> yeesh
[01:29] <jdub> thom: hrm, not to, um, pour salt on the would immediately after your firefox upload, but...
[01:29] <jdub> thom: are the fileselector changes remotely useful to us at this point?
[01:31] <thom> jdub: the ones that fedora ship?
[01:32] <jdub> fedora and novell; not sure if it's the same code
[01:33] <thom> i've not looked at the novell sources, probably ought to
[01:35] <daniels> jdub: ez gtk boooog
[01:35] <jdub> ;)
[01:40] <daniels> thom: 'entertaining' doesn't do that version number justice
[01:41] <thom> heh
[01:41] <daniels> something like 'ludicrously insane'
[01:42] <daniels> comic relief while we were all stressing during the warty crunch :)
[01:51] <lamont> Kamion: you around?
[01:51] <lamont> fabbione: or you?
[01:54] <mdz> lamont: oo.o2 still chugging along?
[01:56] <lamont> mdz: only 21 MB of log file so far.
[01:56] <lamont> and it's only been  building for 170 minutes..
[01:56] <lamont> given that oo.o 1 takes 4:47....
[01:57] <toresbe> is 2.0 out?
[01:57] <lamont> this is one of those packages where no log file is a good thing... :-)
[01:57] <lamont> openoffice.org2_1.9.66-0ubuntu6
[01:57] <toresbe> aha
[01:57] <lamont> looks like a pre-rel version to me...
[01:57] <toresbe> heh, yeah
[01:58] <toresbe> there was a cool report about Linux on Norwegian TV
[01:58] <toresbe> they got some parts ludicrously wrong ("In 2003, Novell bought Linux, and sought to commercialize the previously free system")
[01:59] <toresbe> but they did some really cool interviews with Skolelinux (easy-to-setup special-purpose school distro)
[02:00] <toresbe> and a lot of the students were really positive towards Linux
[02:00] <mjg59> daniels: http://www.codon.org.uk/~mjg59/tmp/xglx.png
[02:00] <mdz> daniels: you need to kick this habit of closing bugs just because they're difficult to fix
[02:00] <mdz> especially when someone else is working on fixing it
[02:01] <mdz> lamont: well, that's about 20.99MB farther than the last one got
[02:05] <daniels> mdz: i just saw the REOPENED, not the rest.  it's usually traditional to provide a comment. ;)
[02:07] <daniels> mjg59: pimp
[02:07] <mjg59> daniels: So, *so* much faster than using composite on i855 otherwise
[02:08] <daniels> mjg59: gotta love xaa
[02:08] <thom> mjg59: holy transparent terminals, batman!
[02:08] <jdub> composite on nvidia is actually pretty respectable
[02:08] <jdub> but oh man
[02:08] <jdub> i855
[02:08] <mjg59> I have seen the future, and it is xglx
[02:09] <daniels> mjg59: s/xglx/not XAA?
[02:09] <daniels> s/.$/\//
[02:09] <mjg59> (on mesa-solo, not on X)
[02:09] <jdub> i didn't realise what all the fuss was about composite being not ready until i ran it on i855
[02:09] <mjg59> daniels: Pff. Like anyone's ever going to get round to doing decent render acceleration for i810
[02:20] <lamont> mdz: heh
[02:21] <tseng> mdz: you said the nightly was bunk, which should I grab?
[02:21] <mdz> tseng: I have no idea what is wrong with it, which is why I need people other than me to be testing them
[02:22] <mdz> even better would be if people other than me had tested yesterday's, and could confirm whether this problem is new
[02:22] <mdz> as it is, USB functionality seems completely fucked and I don't know why
[02:23] <mdz> thom: still here?
[02:23] <mdz> thom: it might help me if you could diff the manifests for the past few live CD builds
[02:23] <mdz> thom: unfortunately they're only available inside the isos, so I can't get to them on little
[02:26] <thom> mdz: i'm litterally about to fall into bed. can you send me a mail and i can check first thing tomorrow, if that's any good?
[02:26] <lamont> mdz: on little... http://mcmurdo.buildd/~buildd/livecd/
[02:26] <mdz> I'm grasping at straws here
[02:27] <lamont> you want current, and some previous directory
[02:27] <lamont> that is, for i386
[02:27] <mdz> lamont: could you grep those for kernel stuffs and see when it changed?
[02:27] <sivang> mdz: iso links on the wiki page as usual? (for testing)
[02:27] <lamont> sure - "grep linux"??
[02:27] <mdz> if you could tar up the past week or so and send it to me, that'd be good
[02:28] <mdz> sivang: I think the wiki page has a link to the index
[02:28] <lamont> heh - terranova, anyway
[02:28] <sivang> mdz: ok, current is the one right?
[02:29] <lamont> mdz: people.ubuntu.com/~lamont/mdz/manifests.tar.gz
[02:30] <mdz> sivang: yes
[02:30] <mdz> lamont: thanks
[02:31] <mdz> 20050120.1 and 20050120 are identical
[02:31] <mdz> how likely is that?
[02:31] <mdz> 20050119.3 and 20050120 are also identical
[02:31] <lamont> well, the first one is from the manual run, I expect.  the second one is from 0615
[02:32] <lamont> it's whatever was on jackass when it ran - would have to compare that to changes, I guess
[02:34] <lamont> mdz: turns out that I populate /etc/locale.gen, and somewhere decided that would cause locales to generate them...
[02:35] <lamont> anything else you want done in the chroot?
[02:41] <mdz> lamont: nothing comes to mind
[02:41] <mdz> lamont: if you could make USB work again, that'd be great ;-)
[02:44] <mdz> mjg59: for some reason I see the apm module loaded twice, once "overridden by ACPI", once "disabled on user request". do you know why?
[02:44] <ogra> night
[02:47] <mjg59> Erm. Nope.
[02:47] <mjg59> It should only be being loaded from /etc/modules
[02:47] <jdub> https://www.ubuntulinux.org/wiki/JoeZicarelli
[02:47] <mjg59> Unless apmd tries to modprobe it, or something
[02:47] <mdz> it's not in /etc/modules
[02:47] <mdz> but apmd does try to modprobe it
[02:47] <mjg59> Hrm. I thought it was meant to be.
[02:50] <lamont> mdz: right.
[02:50] <lamont> mdz: btw, _does_ debian do rsync-able .debs by default?
[02:50] <mdz> lamont: no
[02:50] <lamont> I rather expect that there is a good deal of non-determinalism in the file system, which complicates things for us as well.
[02:50] <lamont> fs layout that is.
[02:51] <lamont> mdz: number of blocks allocated in a cg affects the placement of a new block, no?
[02:53] <mdz> lamont: I think the blocks are large enough that we'd be ok with --rsyncable
[02:53] <mdz> but we could always crank it up to 4k if needed
[02:53] <sladen> lamont: this is puzzling me.  AFAIK, the cloop/cramfs effectively compresses each 64kB block separately.  ...therefore there should already be a fair amount, ---but only 64kB at a time
[02:54] <lamont> sladen: hence my assertion that the placement of files on the device is varying
[02:55] <mdz> we tested that already
[02:55] <mdz> by matching the cloop block size with the filesystem block size
[02:55] <mdz> it didn't help
[02:55] <lamont> mdz: I'm going to rsycn 20050120.1/livecd-fsimg onto a copy of 20050120/livecd-fsimg and see how much diff there is - shouldn't be much...
[02:56] <lamont> or the situation is hopelesss
[02:56] <mdz> huh?
[02:56] <lamont> if the original fsimg is rsyncable, then it makes sense to try and make the cloop rsyncable
[02:56] <lamont> if not, then not.
[02:56] <mdz> oh
[02:56] <mdz> it will be
[02:57] <mdz> I have confidence
[02:57] <mdz> I didn't realize fsimg -> uncompressed
[02:57] <lamont> yeah
[02:57] <thully> I've been fiddling with autohinter - I've found that there are some statements to test font size
[02:57] <daniels> thully: patches gratefully accepted ...
[03:00] <sladen> lamont/mdz does an ISO contain all the extra 2048->2352 overhead ?
[03:00] <lamont> livecd-20050120.1-i386.fsimg-1024
[03:00] <lamont>   2147482624 100%   14.96MB/s    0:02:16  (1, 100.0% of 1)
[03:00] <lamont> wrote 2147744895 bytes  read 40 bytes  14966863.66 bytes/sec
[03:00] <lamont> total size is 2147482624  speedup is 1.00
[03:00] <lamont> that looks like a 'NO' to me.
[03:00] <sladen> lamont: that's just the cloop filesystem?
[03:01] <lamont> uncompressed
[03:01] <mdz> lamont: interesting
[03:01] <mdz> lamont: how about -B 1024 ?
[03:01] <thully> I've got some ideas, but I'm not quite sure about the syntax of fontconfig files - and all autohinter fonts look good on my machine, so It's a little hard to tell
[03:01] <sladen> ...why is so precisel half?
[03:01] <lamont> that was 1024
[03:01] <lamont> sladen: half??
[03:02] <lamont> speed? that's just the sucky disks we ahve
[03:02] <mdz> lamont: I mean rsync -B 1024
[03:02] <lamont> oh
[03:02] <lamont> s
[03:02] <mdz> sladen: 2048->2352 overhead?
[03:03] <lamont> actually, syncing the first one back over the copy of #2
[03:03] <sladen> mdz: cd frames are 2352 bytes eg;  with Audio, all but 8(?) bytes are audio data;  data (mode 1) uses 288 bytes for extreme error detection/recovery
[03:04] <sladen> == 1 x 2kB sector per frame
[03:04] <mdz> the default block size for rsync is 700 bytes
[03:04] <mdz> so 1k filesystem blocks should be just fine, really
[03:05] <lamont> wrote 2147744893 bytes  read 40 bytes  14761133.56 bytes/sec
[03:05] <lamont> total size is 2147482624  speedup is 1.00
[03:05] <sladen> lamont: can you build a fileystem compressed with deflate=0 ?
[03:06] <lamont> sladen: if the input isn't rsyncable, it doesn't matter _what_ we do in compression
[03:06] <lamont> and it's not seeing any benefit from rsync
[03:06] <sladen> lamont: hence wanting to know what happens if you create something non-compressed
[03:06] <lamont> I'm working with non-compressed images
[03:08] <lamont> mdz: I think the choices come down to (1) come up with some way to populate the filesystem in the manner _we_ want, or (2) live with no rsyncability...
[03:08] <lamont> which is really to say (1) fix gene2fs or (2) live with it.
[03:08] <lamont> s/fix/fix or rewrite/
[03:09] <mdz> I can't think of anything genext2fs could possibly be doing wrong which would cause this
[03:09] <lamont> we're not using genext2fs
[03:09] <mdz> ext2 stores file data in blocks, period
[03:09] <mdz> or the kernel
[03:09] <lamont> were lomounting a 2GB file, and rsyncing into that
[03:10] <mdz> the 2GB file starts out zeroed, right?
[03:10] <sladen> time to try with small / empty files and investigate what is actually happening
[03:10] <mdz> GMTA
[03:10] <lamont> 2GB file starts out as a sparse file with 1KB of zeros at the end
[03:10] <lamont> I suppose I could try it with it zeroed
[03:10] <mdz> that should cause it to be zeroed when read from
[03:10] <mdz> but worth a try
[03:10] <mdz> maybe loop-mount is weird
[03:11] <lamont> exactly.  it _shouldn't_ matter...
[03:11] <mdz> lamont: yeah, and USB _shouldn't_ be fucked on the live CD :-/
[03:11] <mdz> I am at my wit's end with this thing
[03:11] <mdz> it makes no sense
[03:12] <lamont> dd if=/dev/zero of=$IMGNAME count=$SZ bs=1024 
[03:12] <lamont> that should do something, anyway.
[03:12] <mdz> bs=1M ought to be quicker
[03:12] <lamont> yeah, but then $SZ would have to be in MB. :-)
[03:15] <sladen> lamont: that's fine, you can  append   k/m/g  to bs= and count=
[03:15] <lamont> sladen: but can I tell it to divide count by 1024?
[03:15] <lamont> :)
[03:15] <sladen> eg.  count=2k bs=1M
[03:15] <lamont> doesn't matter - it's calcluated anyway...
[03:16] <mdz> lamont: it would be worthwhile to do a cmp --verbose over those two images, I think
[03:16] <sladen> lamont: do you have the command the cloop for the CD is build with?
[03:16] <sladen> command(s), script, fooage
[03:20] <mdz> sladen: create_compressed_fs <input> <block size> > output.cloop
[03:20] <mdz> we use 65536 at the moment for the block size
[03:21] <jdub> mdz: have you tried dir_index and EAs-in-inode yet?
[03:21] <mdz> jdub: no, rsyncability is a bigger problem
[03:21] <mdz> jdub: because Ubuntu developers (cough) say they can't test live CDs due to bandwidth limitations
[03:22] <lamont> haggai: /build/buildd/openoffice.org2-1.9.66/ooo-build/build/src680-m66/sal/rtl/source/macro.hxx:102:2: #error "unknown ARCH -- insert your ARCH identifier above"
[03:22] <lamont> early and violent death for ooo2 on amd64
[03:22] <sladen> mdz: no, the bit before that.  the mkisofs/genext2fs ---that's where the problem is
[03:22] <jdub> mdz: 8)
[03:22] <mdz> sladen: the mke2fs?
[03:22] <jdub> mdz: understood and greatly appreciated :-)
[03:22] <mdz> sladen: I think that uses all defaults
[03:22] <mdz> lamont: any mke2fs args?
[03:22] <lamont>  mke2fs -b $FSBLOCK $INUM -Osparse_super -F $IMGNAME
[03:23] <lamont> INUM=""
[03:23] <lamont> FSBLOCK=1024
[03:25] <sladen> yeah, the $INUM was puzzling me, the man page says  mke2fs ... $device $inode_count  (in *that* order)
[03:26] <mdz> lamont: that oo.o failure is expected, v2 isn't ported either
[03:26] <lamont> INUM is either "" or "-N nnnn"
[03:26] <mdz> i386 built, though!
[03:27] <lamont> mdz: in 3:40! woot
[03:27] <lamont> although he better not be special casing builds by the user 'buildd' again......
[03:28] <mdz> that was rene
[03:30] <mdz> it quits working pretty much the instant that init is re-execed
[03:30] <mdz> which makes no sense whatsoever
[03:30] <jdub> mdz: related to usb b0rk?
[03:30] <mdz> jdub: yes
[03:31] <mdz> frustrating as hell, I've spent most of the day on it
[03:31] <sladen> lamont: so you mke2fs the image.  then loop mount it, then copy the files in, then umount it, then create_compressed_fs it?
[03:31] <mdz> which is why I'm in such a dreadful mood
[03:31] <lamont> sladen: yes
[03:31] <jdub> mdz: let's do the warty dance!
[03:32] <sladen> lamont: any chance of the script doing the copying?
[03:32] <mdz> jdub: do you realize how far that footage has spread?
[03:32] <lamont> sladen: I'm told no.
[03:32] <jdub> bob2 showed me a debian bug on gstreamer about it
[03:32] <lamont> or rather, "not at this time"
[03:32] <jdub> how the hell did it get out, man?
[03:33] <jdub> this is totally going to ruin my porn career
[03:33] <lamont> lol
[03:33] <mdz> lamont: any insight from that cmp --verbose ?
[03:34] <lamont> I was stalling for the zero'ed images...... guess I'll fire up another window.
[03:34] <sladen> lamont: does it do anything more than  cp -a ?
[03:35] <lamont> sladen: it builds a chroot with debootstrap/apt, runs a few things, and then rsync -a's the chroot into the loopback mounted file
[03:35] <mdz> I don't see how it could possibly matter
[03:35] <mdz> but if it's using rsync, I think that means the file list is even sorted
[03:36] <mdz> jdub: more people have seen it than have looked at that photo of the giggling 12-year-old spanish girls
[03:37] <sivang> mdz: heh, it was actually not that good IMHO ;-)
[03:38] <mdz> the dance? oh, it sucked
[03:38] <ajmitch> jdub: some disturbing footage?
[03:38] <mdz> ok, that does it
[03:38] <mdz> I launchped a shell from /sbin/unconfigured.sh
[03:38] <mdz> it doesn't get any earlier than that
[03:38] <lamont> mdz: 124MB into the cmp --verbose, 1.4GB of output...
[03:38] <mdz> and USB is fucked there
[03:38] <sivang> mdz: launchped ? ;-)
[03:39] <mdz> lauched
[03:39] <mdz> launched
[03:39] <mdz> lamont: what does the output look like?
[03:39] <jdub> ajmitch: three men who should know better swinging their hips.
[03:39] <mdz> the amount doesn't tell us much, there would of course be a lot
[03:39] <ajmitch> sigh, I wanted php5, but the debs at sources.dotdeb.org are crap
[03:39] <sivang> mdz: hehe I read it  like launchpaded.
[03:39] <mdz> I would have expected there to be swaths of identical bytes
[03:40] <ajmitch> nearly no build-depends,..
[03:40] <lamont> mdz: on the order of 10-30 bytes per hundred are different
[03:41] <mdz> lamont: do you save the output from the daily d-i builds?
[03:41] <mdz> lamont: somewhere in there I think there is a list of the versions of the udebs in the initrd
[03:41] <mdz> lamont: I need to know if that changed
[03:41] <lamont>      82101   0 371
[03:41] <lamont>      82102   0  41
[03:41] <lamont>      82105   0 372
[03:41] <lamont>      82106   0  41
[03:41] <lamont>      82109   0 373
[03:41] <lamont> logs for the dailybuilds are under ~lamont/buildLogs
[03:41] <lamont> in the usual place...they're binNMU's
[03:42] <mdz> apart from the logs, there is a file
[03:42] <mdz> like the manifest you generate for the cloop
[03:42] <lamont> the above is a block that was used in one image, not the other
[03:42] <mdz> I think it's in the tarball
[03:42] <ajmitch> would php5 be welcome in universe once the debs were usable?
[03:42] <lamont> tarballs tend to not get pruned, truthfully
[03:42] <mdz> lamont: yeah, you know what's fucked up about the un-rsyncability of that ext2 filesystem?
[03:42] <lamont> ajmitch: I don't see why it wouldn't be....
[03:42] <mdz> it's like 25% empty
[03:43] <lamont> or more
[03:43] <ajmitch> lamont: it'd be a new package, so after hoary?
[03:43] <sivang> strange, nautilus-cd-burner doesn't offer  me clear the CDRW...thus doesn't let me burn..hmm
[03:43] <lamont> ajmitch: yeah
[03:43] <mdz> surely there are a hell of a lot of contiguous blocks of zeroes
[03:44] <mdz> I do mean contiguous
[03:44] <lamont> they're scattered all over the iamge
[03:44] <mdz> if 1 in 4 blocks are empty, lots of them are going to be adjacent
[03:44] <mdz> chunks of 2k, 4k, etc. all over the place
[03:44] <lamont> and we're doing 65536, since it didn't make any difference to do the 4096
[03:44] <mdz> which even if something funky is going on with file allocation, would rsync
[03:44] <mdz> I'm talking about the uncompressed image
[03:45] <mdz> it doesn't need to be in the same place, that's my point
[03:45] <lamont> or do you want 1024 with the block on either side being zero in the other?
[03:46] <mdz> if there is a block of Nk zeroes in one image
[03:46] <mdz> and also a block of Nk zeroes in the other image
[03:46] <mdz> rsync ought to find that
[03:46] <lamont> how close to each other though?
[03:46] <mdz> I didn't think it mattered
[03:46] <mdz> rsync scans the whole file before it does anything
[03:47] <lamont> in 2GB image, it's not going to cache it all forever....
[03:47] <mdz> are you sure rsync isn't being too smart?
[03:47] <mdz> noticing that you're doing a local copy, and saying fuckit on the delta algorithm?
[03:47] <lamont> that wouldn't surprise me...
[03:47] <lamont> ah, could be...
[03:47] <mdz> ARGH
[03:47] <mdz>        -W, --whole-file
[03:47] <mdz>               With this option the incremental rsync algorithm is not used and
[03:47] <mdz>               the whole file is sent  asis  instead.   The  transfer  may  be
[03:47] <mdz>               faster  if  this  option  is used when the bandwidth between the
[03:47] <mdz>               source and destination machines is higher than the bandwidth  to
[03:47] <mdz>               disk  (especially  when  the  disk  is  actually  a  networked
[03:48] <mdz>               filesystem).  This is the default when both the source and  des
[03:48] <mdz>               tination are specified as local paths.
[03:48] <sladen> I was going to suggest that
[03:48] <mdz> ok, so I get to go back to sticking to my guns and saying that the uncompressed filesystem will rsync OK
[03:48] <sladen> however, the I suspect the kernel ext2fs is still non-deterministic in its file placement;  ...and maybe the rsync search window isn't as big as I thought
[03:50] <sladen> lamont: re: mdz.  Why /are/ you using rsync if this is an empty, fresh image each time
[03:50] <sladen> lamont: and if it's /not/ a fresh image, it'll be full from crap from the last time
[03:51] <mdz> rsync makes a nice local copy command as well
[03:51] <lamont> yep
[03:51] <mdz> rsyncing against the last build would be clever, though
[03:51] <lamont> and the origin chroot is rm -rf'ed at the start of the script, and the target image szeroed
[03:51] <mdz> then the unchanged files would be left in place
[03:51] <sladen> good point.
[03:52] <lamont> cloop size grows though that way mdz
[03:52] <lamont> rather quickly, I might add
[03:52] <mdz> lamont: rsync --delete
[03:52] <lamont> (it didn't used to do that...)
[03:52] <lamont> changed files still change blocks
[03:52] <mdz> rsync --delete --inplace
[03:52] <mdz> rsync --delete --inplace --no-whole-file !
[03:52] <sladen> that's fine.  but the *unchanged* blocks within (even changed files) stay exactly where they were the last time
[03:53] <mdz> sladen: that's what --inplace would do
[03:53] <lamont> that has potential
[03:53] <sladen> mmm.  yummy.
[03:53] <mdz> though even if we completely lost on every changed file
[03:53] <mdz> we would still win big
[03:53] <mdz> because most of them don't
[03:54] <smurfix> The way this is going, you need "rsync --every-option-ever" soon. ;-)
[03:54] <lamont> smurfix: not quite, just most of them.,
[03:55] <sladen> smurfix: rsync because useful after about:  rsync -CvzapP --stats
[03:56] <lamont> mdz: then we just need a tool to rip through an ext2fs and zero any blocks that aren't being used.
[03:56] <mdz> lamont: doesn't this mean that your 4096/4096 test was invalid as well?
[03:56] <lamont> certainly
[03:56] <mdz> lamont: actually, that already exists
[03:56] <smurfix> lamont: cp /dev/zero FS/zero; sync; rm FS/zero
[03:56] <mdz> partimage
[03:57] <lamont> mdz: what package?
[03:57] <mdz> lamont: partimage
[03:57] <mdz> universe
[03:57] <lamont> ah, that'd explain that
[03:58] <sladen> you should also be able to make the image 25% bigger to avoid fragmentation and let cloop take care of the unused parts
[03:59] <mdz> hmm
[03:59] <mdz> unloading and reloading all of the usb modules fixes it
[04:00] <mdz> sladen: he does that already
[04:01] <mdz> because it's useful to have free space for runtime
[04:02] <sladen> mdz: does diffing the before and after of:  /sbin/lsmod | grep -E '(hci|usb)' | awk '{print $1,$4}'   produce the same result?
[04:02] <mdz> lamont: did --no-whole-file give us some nicer statistics?
[04:02] <mdz> sladen: dunno
[04:03] <lamont> mdz: each of these tests takes about an hour...
[04:03] <mdz> lamont: you're shitting me
[04:03] <lamont> although I suppose that if I just did it manually, I could get it down to 40 minutes of my time
[04:03] <lamont> run the script twice, look at the results
[04:03] <mdz> oh, you rebuilt the image
[04:03] <lamont> == 30 minutes * 2
[04:03] <lamont> it was already running...
[04:04] <mdz> I am just interested in taking the last two images, already built, and rsync --no-whole-file'ing them
[04:04] <mdz> which should take almost none of your time
[04:04] <lamont> so the plan is rsync --delete --inplace --no-whole-file, using the previously built fsimg as the base, correct?
[04:04] <mdz> well, not the last two
[04:04] <mdz> because they're too close
[04:04] <lamont> fsimage, or mounted tree?
[04:04] <mdz> fsimage
[04:04] <sladen> mounted
[04:05] <sladen> never mind
[04:05] <mdz> I'm taking a few steps back
[04:05] <mdz> I'd like to know what our theoretical savings would be
[04:05] <mdz> I'd also like to retry the 4096/4096 test
[04:05] <mdz> both of those should be doable with existing images, the latter recompressing them
[04:05] <lamont> wrote 324458 bytes  read 651330424 bytes  2290526.83 bytes/sec
[04:05] <lamont> total size is 2147482624  speedup is 3.30
[04:05] <lamont> *(^_^*(&+)}_*U rsync
[04:05] <lamont> that's over the network
[04:06] <lamont> yes, same to fsimages as before
[04:06] <mdz> which two were those?
[04:06] <mdz> 20050120 and .1 ?
[04:07] <lamont> yes
[04:07] <lamont> sigh..
[04:07] <lamont> not entirely
[04:07] <lamont> those are 20 and 20.1 amd64 images
[04:07] <mdz> sladen: notably, reloading only ehci_hcd (the one with the device actually connected to it) doesn't change anything
[04:10] <mdz> sladen: only reloading usbcore fixes it
[04:11] <mdz> I wonder if it's related to /proc/bus/usb somehow
[04:13] <mdz> this is insane; I have no idea what changed
[04:16] <sladen> mu.  load order, usbfs on procfs on ro, subtle timing delays.  
[04:16] <sladen> Is the broken universal for everyone and/or just ppc, just i386
[04:16] <sladen> just mdz?
[04:17] <mdz> that's what I've been trying to determine
[04:17] <mdz> and why I've been bitching about lack of testing ;-)
[04:17] <mdz> it breaks for me on amd64, powerpc and i386
[04:17] <mdz> so I highly suspect it's universal
[04:18] <sivang> mdz: done burning, testing now.
[04:26] <lamont> mdz: --no-whole-file on 0120 and 0120.1: wrote 635250483 bytes  read 324462 bytes  1905771.95 bytes/sec
[04:26] <lamont> total size is 2147482624  speedup is 3.38
[04:26] <mdz> interesting
[04:26] <mdz> not bad, but not as good as one would expect
[04:26] <lamont> of course, that's still about 500MB too much
[04:26] <mdz> sivang: does USB work?
[04:27] <sivang> mdz: lemme test, btw, sound (EMUK10) doesn't :(
[04:27] <mdz> one problem at a time
[04:28] <sivang> mdz: btw, Xorg is superb, maximum resolution, hebrew selection in d-i ingored completly
[04:31] <sivang> erggh, nothing.
[04:31] <sivang> not even appearing in hal-device-manager
[04:31] <sivang> :-/
[04:32] <sladen> sivang: okay, now  sudo modprobe -r uhci_hcd  && sudo modprobe uhci_hcd
[04:32] <sivang> sladen: ok, load them modules up :)
[04:33] <sladen> sivang: remove+reload
[04:33] <sivang> sladen: eh right , -r
[04:33] <sivang> sladen: blank password?
[04:34] <sladen> I have no idea, I'd guess so.  Though I've a feeling sudo can get unhappy about blank passwords
[04:36] <sivang> hmm, I need to search the wiki for what the root/sudo password is.
[04:36] <lamont> mdz: btw, any feedback on https://www.ubuntulinux.org/wiki/BuildDaemons ??
[04:37] <lamont> mdz: I assume that even if we have reason to use partimage, it'll stay in universe?
[04:37] <lamont> or do we want to support it?
[04:38] <sivang> mdz: any idea for what the sudo password is? I can't load the USB modules without it it seems.
[04:40] <sivang> sladen: ah! guessed it, "ubuntu" :)
[04:42] <sladen> went into the Apple Store in London last week hunting for Mac Mini's to test Ubuntu PPC on.  They didn't have any, but I did figure out all the Apple Express IP->audio units had  apple/apple  as their username/password combo
[04:47] <sivang> back in the cd,
[04:47] <sivang> (using it)
[04:48] <sivang> sladen: modprobing hung up the terminal and the whole system so I had to power reset the machine
[04:48] <sivang> sladen: however , this was when the usb disk was already in - I will try now when it's unplugged first.
[04:48] <sladen> godo point, might be prudent to add a good sleep between unloading and reloading
[04:48] <sivang> yes
[04:49] <sladen> since I think modprobe returns before it's been finished
[04:51] <mdz> lamont: when do uploads get noticed by the buildds?
[04:51] <mdz> sladen, sivang: in my case, reloading the HCI module had no effect whatsoever
[04:51] <mdz> only after unloading usbcore
[04:52] <lamont> mdz: it's in there..
[04:52] <lamont> ah, almost in there..
[04:52] <sladen> mdz: modprobe -r  should do that
[04:52] <mdz> sladen: modprobe -r ehci_hcd won't unload usbcore
[04:53] <lamont> The build daemons are frequently idle, and just waiting for packages (and therefore sleeping for up to 5 minutes), so on a good day, they'll start building your package somewhere between 0 and 5 minutes after the source enters the archive.
[04:55] <sladen> mdz: it should do if it's a dependancy and it's use-count is zero
[04:55] <lamont> mdz: updated
[04:55] <mdz> lamont:  so roughly, <=5 for the source to be accepted, <=5 for the buidds to notice, build time, <=5 for binaries to be uploaded and accepted, then the next :03 or :33 for them to appear
[04:55] <mdz> ?
[04:56] <mdz> sladen: usbcore gets loaded when /proc/bus/usb is mounted by mountvirtfs
[04:56] <lamont> np
[04:56] <lamont> no
[04:56] <sivang> mdz: would you have a modprobe line to test with? I had already tested 2 times, same behavior: terminal spitts endless CRs,
[04:57] <lamont> <5 for the source to be accepted, up to 3-32 minutes for the source to actually hit the archive, < 5 for the build to start, < 5 for the binaries to be built and uploaded, < 5 for them to be accepted, and then the next :03/33, they're in the archive.
[04:57] <mdz> sivang: perhaps you have a USB keyboard
[04:57] <sivang> system hung up other then that. (had to reset)
[04:57] <lamont> or, in other words:
[04:57] <sivang> mdz: I do, how do you know? ;-)
[04:57] <lamont> < 5 for source accepted mail, then next :03 or :33 the build starts, and 30 minutes later you have binaries
[04:57] <mdz> lamont: oh, so the source needs to be in the archive before it's built?
[04:57] <lamont> yes.
[04:57] <sivang> mdz: well, it should wotk with it also isn't it?
[04:57] <lamont> no accepted auto-building
[04:57] <mdz> why not?
[04:58] <mdz> doesn't Debian build from accepted?
[04:58] <lamont> because we run cron.daily every 30 minutes
[04:58] <lamont> yeag
[04:58] <lamont> but they only run cron.daily once a day
[04:58] <mdz> what would it cost us to build from accepted?
[04:58] <lamont> it's not instantaneous, you know...
[04:58] <lamont> as it sits, elmo has turned off building the Contents files, because that pushes us (with 4 architectures)( over 30 minutes on heavy-churn runs.
[04:59] <lamont> that and it's a pain... if you make elmo turn it on, I can certainly add the deb-src/deb-src line to the buildd's.
[05:00] <mdz> you're saying that cron.daily takes longer if we build from accepted?
[05:00] <sladen> sudo umount /proc/bus/usb && sudo modprobe -r uhci_hcd    works from here
[05:01] <lamont> no. cron.hourly takes longer then, since it has to apt-ftparchive
[05:01] <sivang> sladen: this time I didn't plug the stick at all, but mdz suggested my usb keyboard is cuasing the problems..
[05:01] <lamont> and it has a 5 minute budget
[05:01] <lamont> (since it runs every 5)
[05:01] <lamont> and the thinking was that folks who couldn't wait 30 minutes should get a life, or something like that..... dunno. :-)
[05:02] <mdz> sladen: here as well
[05:02] <sivang> sladen: using usb kbd?
[05:02] <mdz> sivang: your keyboard is USB.  When you unload the USB drivers, your keyboard will *stop working*
[05:03] <sladen> '''hi users, the first thing you'll need to do to use the LiveCD, is plug a network connection in and bring up an SSH session to fix the USB drivers
[05:03] <sivang> mdz:  woops right, I hate to be so overlookinig for so obvious details..
[05:03] <mdz> sladen: devices which were activated during stage 1 continue to work fine
[05:03] <mdz> but no new devices are noticed, and no removals are noticed
[05:04] <sivang> mdz: ok, is there a way to do mount the stick without removing ehci_hcd ?
[05:04] <sladen> is /proc/bus/usb remounted?
[05:05] <mdz> sivang: you have confirmed that the bug exists for you also, thank you
[05:05] <mdz> remounted?
[05:05] <mdz> it gets unmounted during casper/d-i shutdown, and then mounted again when init starts up, yes
[05:05] <sladen> to remove usbcore, I was having to umount /proc/bus/usb
[05:05] <mdz> I've also tried leaving the d-i mount there
[05:05] <mdz> correct, /proc/bus/usb holds a reference to usbcore
[05:06] <sladen> I don't have an image to play with;  how big is the test LiveCD to suck down (smaller is better)
[05:06] <mdz> about 550M
[05:07] <mdz> -rw-rw-r--  1 mdz mdz 525M 2005-01-20 00:02 hoary-live-i386.iso
[05:07] <mdz> 525
[05:08] <sladen> feh
[05:08] <sivang> mdz: glad to be of service, lemme know if you need further testing with that as I am learning how to customize the livecd ;-) , I am also on 1.5Mbit conneciton so downloading the cd is not biggy.
[05:08] <mdz> I just don't understand how usbcore can get into this state
[05:09] <sladen> mdz: can you drop to a shell during d-i and see what state it is at that point?
[05:10] <mdz> or what is triggering it
[05:11] <HostingGeek> does someone want to explain to me why you need windows/dos to flash bios?
[05:11] <mdz> not here
[05:11] <mdz> but sure, another time
[05:11] <HostingGeek> mdz: well if it can be asked in the devel's mail list it can be asked here
[05:12] <HostingGeek> i prefer if someone reply to it rather than explain it here....
[05:12] <mdz> don't ask it on the development mailing list either
[05:12] <sladen> HostingGeek: there are tonnes of flash chips out there.  the /dev/bios drivers and MTD drivers eat some of them.  But not most
[05:13] <HostingGeek> sladen: i am a idiot i have no idea what that ment
[05:13] <HostingGeek> are not bios a chip on the m/b with a rom loaded on to it
[05:14] <HostingGeek> if so why can not mb makers offer just the rom and the end user can use his/her fav. tool to flash
[05:14] <smurfix> sladen / HostingGeek: I think mdz said "not here" ... and I agree actually
[05:14] <HostingGeek> sry i didn't see that comment
[05:15] <HostingGeek> mdz: no its a reply to something about using freedos to flash bios
[05:15] <sladen> HostingGeek: additionally, the 'images' frequently don't come as images, but compressed files that can be patched and uncompressed for a particular machine before being hurt---the provided floppy disks have all that in one glob.  dosemu isn't really an opinion because of port and timing-critical sections that can't be achived under emulation or userspace.  Hope that answered your question.  If it didn't, pester me privately  (/query sladen).  But pref
[05:15] <mdz> I can't find any other way to trigger this behaviour, other than actually using the live CD
[05:16] <mdz> mounting /proc/bus/usb elsewhere, unmounting it, mounting it multiple times, nothing breaks it
[05:19] <lamont> mdz: what's the default finish action in partimage, I wonder...
[05:19] <mdz> finish action?
[05:19] <lamont> nm
[05:20] <thully> hi
[05:23] <mdz> unmounting /proc/bus/usb before doing the pivot doesn't make any difference, either
[05:32] <lamont> mdz: any reason I shouldn't link the BuildDaemons package in?
[05:32] <lamont> (i.e., publish it.)
[05:39] <sladen> lamont: I'm sure it would be very useful when a structure for derivities is available;  I can see if leading to temptation and confusion right at this point
[05:41] <fabbione> morning
[05:42] <mdz> fabbione: morning
[05:42] <mdz> fabbione: I have been fighting with this strange kernel problem all day
[05:42] <lamont> morning fabbione 
[05:42] <mdz> #5701
[05:42] <sladen> fabbione: how much effort was the sparc port?   (...alpha appears to be lacking at the moment :-)
[05:42] <fabbione> guys ... stop
[05:42] <fabbione> i just woke up
[05:42] <fabbione> and reading emails :-)
[05:42] <lamont> heh
[05:49] <fabbione> mdz: 5701 worksforme on normal boxes
[05:50] <fabbione> what kernel are you using right now?
[05:50] <fabbione> because i remember fixing something USB related recently
[05:50] <mdz> fabbione: can you try the live CD?
[05:51] <mdz> fabbione: it worksforme on normal installs too
[05:51] <mdz> and it _used_ to work on live CDs
[05:51] <mdz> but something has changed and I have no idea what it is
[05:51] <mdz> the live CD I'm using has 2.6.10-10
[05:51] <fabbione> mdz: yes. i will try in a bit ok?
[05:51] <mdz> fabbione: wonderful, thank you
[05:52] <fabbione> i need to see if my mirror is synced
[05:52] <fabbione> mdz: no problem
[05:52] <fabbione> mdz: when did you build the last one?
[05:52] <mdz> the most recent one is a daily build
[05:52] <mdz> 20050120.1
[05:53] <fabbione> mdz: is it the same you are using for testing?
[05:53] <mdz> correct
[05:54] <fabbione> ok it is synced
[05:57] <lamont> fabbione: all the 20050121* images are me testing stuff
[05:57] <jdub> jdz_: dude :)
[05:57] <jdz_> jdub: hey man :)
[05:58] <fabbione> lamont: i use the "current" symlink to update
[05:58] <fabbione> i don't mirror everything even if i could :-)
[05:58] <mdz> jdz sounds like a fusion of jdub and mdz
[05:58] <mdz> a frightening concept
[05:58] <fabbione> ehhehe
[05:59] <fabbione> mdz: it's burning...
[05:59] <fabbione> and i need to wake up my gf
[06:02] <jdz_> *laughs*
[06:07] <sladen> jdz_: were you thinking some KY-Jelly would solve this problem too?
[06:08] <zenrox> hahahha
[06:08] <jdz_> oh man ;P
[06:10] <fabbione> mdz: booting now...
[06:11] <fabbione> mdz: everything is working fine here
[06:12] <mdz> fabbione: you can plug and unplug USB devices?
[06:12] <fabbione> mdz: sure.. but that won't make it working properly
[06:12] <fabbione> you know that, don't you?
[06:12] <mdz> what?
[06:13] <fabbione> the usb devices, when unplugged and replugged. might get a different /dev/input/<name>
[06:13] <fabbione> it depends on how fast you do the operation
[06:13] <mdz> have you read the bug?
[06:13] <fabbione> and X will not reopen the FD
[06:13] <mdz> when I plug in a USB device, it does not even log in dmesg
[06:13] <mdz> nothing happens
[06:13] <fabbione> ok just a sec...
[06:13] <mdz> I think I have found the next level of cause
[06:13] <mdz> khubd dies
[06:14] <mdz> once khubd is dead, no more USB device events are noticed
[06:15] <mdz> yes, that is it
[06:15] <mdz> reloading usbcore restarts khubd
[06:15] <mdz> OH
[06:15] <mdz> khubd can be killed by SIGKILL
[06:16] <mdz> which init sends it, of course
[06:16] <mdz> but this never happened before
[06:16] <fabbione> hmm
[06:16] <fabbione> i unplug my mouse and replug it i get a lot of crap in dmesg
[06:16] <magnon> mdz: are we on "why doesn't my usb stuff show up in nautilus"?
[06:17] <fabbione> driver/*/usb-core: input irq status -84 received
[06:17] <magnon> my card reader just shows up the first time it's mounted
[06:17] <mdz> magnon: no, this is "why is USB broken on the live CD"
[06:17] <magnon> ah.
[06:17] <magnon> I should file a bug on this, if it isn't filed alraedy
[06:17] <mdz> fabbione: I am fairly sure I have isolated it
[06:17] <magnon> and I need sleep, to fix my typing
[06:17] <mdz> fabbione: sudo killall -9 khubd
[06:17] <fabbione> mdz: worth a try to boot with that irqpoll ?
[06:17] <mdz> fabbione: and you will see exactly what I am talking about
[06:18] <fabbione> there is no khubd runnig on that machine
[06:18] <fabbione> i already checked
[06:18] <mdz> what
[06:18] <sivang> magnon: card reader? as in smart card?
[06:18] <magnon> CF
[06:19] <mdz> fabbione: this is with the same live CD I am using?
[06:19] <fabbione> mdz: it's the last one from archive in current/
[06:20] <fabbione> if you need another version you need to tell me
[06:20] <mdz> khubd should be started as soon as a hub is detected
[06:20] <fabbione> but yes i can reproduce the problem on livecd
[06:20] <mdz> mdz@little /srv/cdimage.no-name-yet.com/www/full/daily-live/current $ md5sum hoary-live-i386.iso
[06:20] <mdz> 1d1afb2a99db7501486b6b8d7840ca0a  hoary-live-i386.iso
[06:20] <mdz> oh, you can?
[06:20] <fabbione> ok
[06:20] <mdz> you just told me you got messages in dmesg when you connected USB devices
[06:20] <fabbione> boot -> X everyuthing works
[06:20] <mdz> which would mean that you cannot reproduce the problem
[06:20] <fabbione> i remove the mouse and replug it -> the message you wsaw above
[06:21] <fabbione> 1d1afb2a99db7501486b6b8d7840ca0a  hoary-live-i386.iso
[06:21] <fabbione> looks about the same
[06:21] <mdz> no idea about that message; that is not the same as my bug
[06:21] <mdz> but I am 99% sure I have nailed it
[06:22] <mdz> on the live CD, when init re-execs, it kills all other processes
[06:22] <fabbione> mdz: ok.. let me know...
[06:22] <mdz> first SIGTERM, then SIGKILL
[06:22] <mdz> it sends a SIGKILL to khubd
[06:22] <mdz> which kills it
[06:22] <mdz> which is what causes the problem
[06:22] <fabbione> but why do you send that message around?
[06:22] <mdz> what message?
[06:23] <fabbione> SIGTERM, then SIGKILL
[06:23] <mdz> it is init, that is what it does
[06:23] <mdz> kernel threads are not supposed to die from signals
[06:23] <fabbione> mdz: that's true.. but that didn't change 
[06:23] <mdz> but khubd does this:
[06:23] <mdz>         daemonize("khubd");
[06:23] <mdz>         allow_signal(SIGKILL);
[06:24] <mdz>         /* Send me a signal to get me die (for debugging) */
[06:24] <mdz> if it is only for debugging, I say turn it off :-P
[06:24] <fabbione> i am pretty sure i didn't patch anything there of that kind :-)
[06:24] <mdz> yes, I don't understand why this didn't happen before
[06:24] <fabbione> let me check the patchsets
[06:24] <fabbione> what file is that?
[06:24] <mdz> drivers/usb/core/hub.c
[06:24] <mdz> I checked; you don't patch it
[06:25] <fabbione> yup.. i am checking if there is a fix upstream
[06:27] <fabbione> the code is clearly broken
[06:28] <mdz> I think I know why I didn't see this before
[06:28] <fabbione> the code is BAD!
[06:28] <mdz> maybe the mouse driver was loaded by d-i
[06:28] <fabbione> it sends signals to itself to exit and die
[06:31] <mdz> hmm, no, doesn't look like the mouse drivers were ever there
[06:32] <mdz> aha
[06:32] <mdz> it's in input-modules
[06:32] <mdz> I bet input-modules was on the CD before, and is now gone, and that's the problem
[06:33] <mdz> well, the trigger
[06:33] <mdz> hotplugging still would not have worked
[06:33] <mdz> but my mouse worked before, and this is why
[06:33] <mdz> confirmed: I am not insane
[06:33] <mdz> (yet)
[06:34] <fabbione> well there are no changes in the code for a long while
[06:34] <fabbione> mdz: so who is at fault and what do you want me to do?
[06:36] <fabbione> HMMM
[06:36] <fabbione> this code is still strange tho
[06:36] <mdz> yes, it makes sense that there have not been changes
[06:36] <mdz> what changed was which modules were shipped on the CD
[06:36] <fabbione> ok
[06:37] <mdz> in my opinion, it is pointless for a kernel thread to die like that
[06:37] <mdz> I see two ways to fix it:
[06:37] <mdz> 1) patch the kernel so that the thread doesn't die
[06:37] <fabbione> it is needed on cleanup
[06:37] <mdz> 2) try to unload all USB modules before pivoting on the live CD
[06:37] <fabbione> i am reading more code...
[06:38] <mdz> SIGKILL is the wrong way to clean up a kernel thread
[06:38] <mdz> in my uninformed opinion
[06:39] <mdz> but many other threads do it
[06:39] <fabbione> mdz: i am checking waht is in bk
[06:39] <mdz> so I guess this will be a bigger problem than just usb
[06:39] <fabbione> and they started using it even more
[06:39] <mdz> md_thread does it
[06:39] <fabbione> the point is that you load khubd when you need it
[06:39] <fabbione> and you need to remove when you don't
[06:39] <mdz> many filesystems do
[06:39] <fabbione> they accept to die?
[06:40] <mdz> trying to unload all the modules is not really feasible
[06:40] <fabbione> mdz: stop sending them signals :-)
[06:40] <mdz> so I guess I will need to patch init
[06:41] <fabbione> mdz: another option would be for me to provide you a boot option to diable allow_signal
[06:41] <mdz> how can I tell a kernel thread from a user process?
[06:41] <mdz> hmm actually
[06:41] <mdz> I can't fix this in init
[06:41] <mdz> I am pretty sure init does kill(-1, SIGKILL)
[06:41] <mdz> so it is a kernel problem again! ;-)
[06:41] <fabbione> mdz: nope...
[06:41] <fabbione> but i can help you with that
[06:42] <daniels> argh, the ltmodem source package is shit
[06:42] <mdz> how do you mean?
[06:42] <fabbione> the point is that either you or me needs to patch something to be used only on livecd
[06:42] <fabbione> so either i give you a boot option
[06:42] <mdz> well, let's think about this
[06:42] <fabbione> or you patch init to recognize that it is a live cd and NOT send SIG_KILL
[06:42] <mdz> won't this be a problem when rebooting?
[06:42] <fabbione> mdz: fromt he livecd?
[06:43] <mdz> fabbione: from anywhere
[06:43] <mdz> let's say your filesystem uses a thread which can be killed by SIGKILL
[06:43] <mdz> init SIGKILLs it
[06:43] <fabbione> no.. because we can enable the boot option only on livecd
[06:43] <mdz> and then runs the shutdown scripts
[06:43] <mdz> aaaah
[06:43] <mdz> ./doc/Changelog:  * killall5 now skips kernel threads
[06:43] <fabbione> tsk :)
[06:44] <mdz> well, this means that sysvinit has found a way around this problem
[06:44] <mdz> and we just need to do the same in busybox
[06:45] <mdz> man
[06:45] <mdz> this has been one of those "one bug" days
[06:46] <mdz> one bug the whole damn day
[06:46] <lamont> mdz: and a trivial fix at the end of it.
[06:46] <lamont> which fits the rule
[06:46] <mdz> not as trivial as I'd like
[06:46] <lamont> (time to fix is inversely proportional to the time to find)
[06:46] <mdz> not trivial at all, unfortunately
[06:46] <fabbione> ehheh
[06:46] <mdz> busybox uses the simplest possible method
[06:47] <mdz> kill(-1, <sig>)
[06:47] <mdz> sysvinit walks /proc
[06:47] <lamont> ouch
[06:50] <daniels> an autoconf'ed source distribution that seemingly can't be made to use srcdir!=builddir
[06:50] <daniels> whoohoo!!!
[06:52] <magnon> the beauty of autotools
[06:52] <jamesh> using automake 1.4 by any chance?
[06:52] <lamont> daniels: I'm sure keybuk can hit it over the head hard enough...
[06:53] <daniels> lamont: it uses a bastard mix of automake and kbuild
[06:53] <daniels> i've given up on srcdir != builddir for this crap
[07:45] <lamont> hrmpf.
[07:45] <lamont> it's 0645 at the data center...
[07:46] <fabbione> lamont: welcome to the concept of TZ's ;)
[07:46] <fabbione> lamont: is there anything you want me to do so you can go and crash?
[07:46] <lamont> fabbione: nah-  what I really need to do is add some mutual exclusion to the livecd build script...
[07:46] <lamont> IOW... oops.
[07:54] <lamont> | #define HAVE_STDINT_H 1
[07:54] <lamont> | #define  1
[07:54] <lamont> | #ifdef __cplusplus
[07:54] <lamont> hehe
[07:59] <lamont> no mvo yet, eh?
[08:02] <fabbione> daniels: do you have any machine where you can dri and test the new kernel modules?
[08:03] <daniels> fabbione: two
[08:03] <daniels> fabbione: (i855, which uses the i915 module, and radeon)
[08:03] <fabbione> ok.. what flavours do you need?
[08:03] <daniels> 686, k7
[08:03] <daniels> i need to start getting ready very soon though
[08:03] <fabbione> uplaoding them now
[08:03] <fabbione> no problem
[08:03] <fabbione> i am not going to release today
[08:03] <daniels> cool, ta
[08:03] <daniels> ahr sweet
[08:03] <daniels> in that case I might check them when I get back
[08:03] <daniels> lest I die for not being ready on time
[08:04] <fabbione> daniels: i am uploading them to people
[08:04] <fabbione> in my home
[08:04] <fabbione> they should be up in 30 minutes or so
[08:04] <fabbione> g'night lamont
[08:04] <lamont> night
[08:04] <fabbione> daniels: i need you to upgrade the kernels without rebooting the machine
[08:05] <fabbione> daniels: unload the old module and load the new one
[08:05] <fabbione> to see if the ABI is broken or not
[08:05] <fabbione> from my tests is not, but i can't really use dri here
[08:05] <fabbione> so my loads are fake
[08:06] <daniels> fabbione: sure
[08:06] <fabbione> cool
[08:07] <daniels> i915 should be ok
[08:07] <daniels> it's just adding more functions
[08:07] <daniels> and it bumped the version number
[08:07] <fabbione> daniels: just check both of them
[08:07] <fabbione> i did review the code again
[08:07] <daniels> yah
[08:07] <daniels> sweet
[08:07] <daniels> alright, I'm off now
[08:07] <fabbione> and there is a new module that collect all the common code
[08:07] <fabbione> drm.ko
[08:08] <fabbione> and all the others have been changed to use it
[08:08] <fabbione> ok.. later dude
[08:55] <d3vic3> doko, ping 
[09:11] <doko> d3vic3: morning d3vic3, just sent you an email
[09:13] <d3vic3> oh ok 
[09:13] <d3vic3> morning 
[10:10] <ogra> morning
[10:12] <ajmitch> morning ogra 
[10:14] <Treenaks> hey ogra
[10:15] <Treenaks> why is everyone so ignorant about the package system/apt etc?
[10:15] <Treenaks> on the ml
[10:16] <ajmitch> the -devel list?
[10:16] <Treenaks> no, users
[10:17] <ajmitch> that's no surprise then
[10:17] <Treenaks> but you can't vent steam in front of the people you're venting steam about :)
[10:17] <ajmitch> :)
[10:18] <ajmitch> I think people would like a nice easy way to install debs they've got themselves
[10:18] <Treenaks> still, there should be a huge brigt yellow-on-purple screen with "USE PACKAGES", about 10 times during installation
[10:18] <Treenaks> bright
[10:19] <ogra> i just started to create a MOTU page skeleton on the wiki....any idea what other subtopcs should be there ? https://www.ubuntulinux.org/wiki/MOTU
[10:20] <amu> moins 
[10:20] <ogra> moin
[10:20] <ajmitch> hi amu 
[10:20] <ajmitch> hmm, a little empty
[10:21] <ajmitch> ogra: want to upload a fix for a universe package for me?
[10:21] <ogra> ajmitch: skeletons mostly have less flesh ;)
[10:21] <ajmitch> this is more like a couple of knuckles than a skeleton ;)
[10:22] <ogra> ajmitch: if it was approved by haggai make him send me a "ok" and i'll do :)
[10:22] <amu> ajmitch: *waves*
[10:22] <ajmitch> ogra: ok, I'll talk to him
[10:23] <ogra> Treenaks: do you have uploader status already ? (or will we have to wait till next TB)
[10:29] <Kamion> morning
[10:29] <ogra> morning
[10:29] <fabbione> yo
[10:29] <Treenaks> hi Kamion 
[10:43] <pitti> Hi guys!
[10:43] <pitti> I just was in my uni, got my diploma certificate :-)
[10:44] <Treenaks> pitti: congratulations!
[10:44] <mvo_> pitti: congrats!
[10:44] <pitti> thanks
[10:44] <pitti> amu: ping
[10:45] <amu> pitti: moins 
[10:45] <pitti> amu: I have some CAN numbers for konversation
[10:45] <pitti> amu: I'll mail you, amu@ubuntu.com?
[10:46] <amu> pitti: congrats first, look to you incoming filedir, i fixed warty ar morning ;) 
[10:46] <amu> s/ar/at
[10:46] <pitti> amu: argh, I thought you wanted to wait until the CANs are there?
[10:46] <amu> pitti: mdz was faster ;) 
[10:47] <pitti> okay :-)
[10:48] <amu> pitti: mdz sended the CAN numbers at morning ..  
[10:50] <pitti> amu: it built fine, I'm releasing it now. Thanks for preparing!
[10:51] <ogra> pitti: yeah, prost und glueckwunsch !
[10:51] <amu> pitti: np
[10:51] <pitti> ogra: Danke :-)
[10:52] <Treenaks> "Willkommen in #Deutschland" ?
[10:54] <pitti> Treenaks: mea culpa :-)
[10:55] <Treenaks> pitti: it's just that there's a lot of Germans in here :)
[10:55] <amu> ogra: thx for your translations, you're fast like light 
[10:55] <ogra> heh :)
[10:56] <ogra> amu: somenative speaker should review it, my grammar is mostly crap ;)
[11:03] <toresbe> schni schna schnappi, schnappi schnappi schnapp :D
[11:03] <toresbe> oops
[11:04] <ogra> i added some flesh and bood to the MOTU page....feel free to extend or correct it
[11:04] <ogra> blood even
[11:04] <Kamion> hm, I think I might have nailed thully's timezone bug at last
[11:05] <ajmitch> ogra: looks better
[11:14] <Treenaks> ogra:  ...
[11:15] <ogra> huh ?
[11:15] <Astharot> good morning
[11:15] <ogra> Treenaks: do they still sell it ?
[11:15] <Treenaks> ogra: http://imdb.com/title/tt0403102/?fr=c2l0ZT1kZnxteD0yMHxzZz0xfGxtPTIwMHx0dD1vbnxwbj0wfHE9aGUgbWFufGh0bWw9MXxubT1vbg__;fc=21;ft=22;fm=1
[11:16] <Treenaks> uh
[11:16] <Treenaks> ogra: sorry, http://imdb.com/title/tt0427340/
[11:17] <ogra> Treenaks: hmm, the plot outline from there would make a good introduction :)
[11:17] <ogra> " He-Man, the most powerful man in the universe, goes against the evil"
[11:18] <Treenaks> ogra: you're NOT he-man
[11:18] <ogra> hehe
[11:18] <ogra> Treenaks: nope, haggai is ;)
[11:19] <Treenaks> ogra: nah.. wrong accent ;)
[11:19] <ogra> lol
[11:20] <ogra> sladen could you add something there that describes your universe work: https://www.ubuntulinux.org/wiki/MOTU
[11:22] <Riddell> ogra: has sladen done universe work?
[11:26] <fabbione> RIFE THE SPLIT
[11:26] <fabbione> RIDE even
[11:26] <ogra> Riddell: nope, but he has uploader status
[11:27] <Treenaks> fabbione: well.. "rife" is OK too, here on freenode ;)
[11:28] <ogra> heh
[11:28] <ajmitch> Treenaks: now, I hadn't seen a split for hours :)
[11:29] <Treenaks> ajmitch: then you've been lucky ;)
[11:29] <fabbione> ehehe
[11:36] <pitti> Hi seb128 !
[11:36] <Treenaks> seb? where? :P
[11:41] <fabbione> hey Keybuk 
[11:41] <mvo_> hi seb128, hi Keybuk 
[11:41] <Treenaks> grr... my client doesn't show joins..
[11:41] <fabbione> hey seb128 
[11:42] <Treenaks> this tends to help
[11:42] <fabbione> hi mvo_ 
[11:42] <mvo_> hey fabbione 
[11:42] <seb128> hey pitti & mvo_ & fabbione 
[11:42] <Treenaks> hey seb256
[11:42] <pitti> seb128: please ping me if you want to debug this icon issue
[11:43] <seb128> pitti: right, I've just pinged the panel guy, waiting for a pong :)
[11:48] <seb128> pitti: there is a patch in http://bugzilla.gnome.org/show_bug.cgi?id=108864, to apply to gnome-panel in gnome-panel/applets/notification-area/
[11:48] <seb128> pitti: if you could try that
[11:51] <pitti> seb128: bah, that means that I need to recompile gnome-panel?
[11:53] <seb128> pitti: yeah
[11:53] <seb128> pitti: I can make a patched package if you want
[11:53] <pitti> seb128: I don't mind recompiling it, it just will take a while :-)
[11:53] <pitti> seb128: this patch looks fairly intrusive...
[11:55] <seb128> why ?
[11:55] <seb128> looks ok imho
[11:56] <pitti> seb128: I downloaded the 18 MB of build-deps now :-)
[11:56] <seb128> ah ah
[11:56] <pitti> seb128: I compile it and report back
[11:56] <seb128> nice, thanks
[12:07] <pitti> seb128: bah, FTBFS with the patch
[12:10] <pitti> seb128: hmm, this is nontrivial
[12:12] <pitti> seb128: the patch uses some variables (tray, xev, icon) which are not declared anywhere
[12:12] <pitti> Hi carlos!
[12:13] <carlos> hi
[12:18] <seb128> pitti: 
[12:18] <seb128> vuntz: a priori, s/xev/xevent
[12:18] <seb128> vuntz: s/tray->sockets/manager->socket_table
[12:18] <seb128> icon -> socket
[12:19] <seb128> pitti: don't bother too much if that doesn't build
[12:20] <pitti> seb128: hey, it compiles now :-)
[12:20] <seb128> cool
[12:20] <pitti> thanks
[12:20] <seb128> np
[12:28] <elmo> haggai: ?
[12:28] <pitti> seb128: that's so mean! I tried the usuall killall axe to restart nautilus, panel & co, and now it works fine (still with the old version!)
[12:28] <pitti> elmo: is it okay if I upload the l-p flood?
[12:28] <elmo> pitti: no, sorry, I lost all of yesterday and didn't get a chance to implement the jennifer change.. give me 10minsw and I'll do it now
[12:28] <pitti> elmo: no problem
[12:29] <elmo> pitti: also, please get me a listing of what will be being uploaded so I can pre-add them to bypass NEW
[12:29] <elmo> preferably a list of binary packages 
[12:29] <pitti> elmo: no problem, I mail it to you
[12:29] <seb128> pitti: :(
[12:32] <pitti> elmo: I currently have a list of solely the binary package names (no version, no ".changes" suffix, etc.)
[12:32] <pitti> elmo: is that okay? or do you need the filenames of the debs?
[12:32] <seb128> pitti: perhaps that happens only in the login (depending on the order to load gaim/panel/...)
[12:32] <rburton> is the usplash source available in cvs/arch anywhere?
[12:32] <pitti> seb128: I tried to logout and back in as well
[12:33] <pitti> seb128: anyway, I try some further, and then I try the new pacakge
[12:33] <pitti> seb128: if the new pkg works as well, it's probably okay to upload :-)
[12:33] <seb128> ok :)
[12:33] <seb128> pitti: BTW why have you killer the panel with the old version still here ?
[12:34] <pitti> seb128: just testing...
[12:34] <elmo> pitti: that's great
[12:34] <pitti> seb128: probably the killall triggered a rewrite of some configuration file.
[12:34] <seb128> :(
[12:34] <pitti> elmo: sent
[12:35] <pitti> brb, new gnome-panel test...
[12:36] <pitti> seb128: I KILL this damn panel some day!!!
[12:37] <pitti> kill -9 --extra-bloody --try-harder gnome-panel
[12:37] <seb128> ah ah
[12:37] <pitti> s/kill/killall/
[12:37] <smurfix> pitti: doesn't matter, it gets restarted automatically
[12:37] <smurfix> ;-)
[12:37] <seb128> pitti: if you want to kill it just "gnome-session-remove gnome-panel"
[12:37] <pitti> find -name "*panel*" -exec rm -rf '{}' \;
[12:38] <pitti> that should do
[12:38] <seb128> :)
[12:38] <pitti> seb128: thanks for that hint
[12:38] <seb128> np
[12:38] <pitti> seb128: 
[12:38] <ogra> ....the return of the undead panel....
[12:38] <pitti> $ gnome-session remove gnome-panel
[12:38] <pitti> gnome-session: you're already running a session manager
[12:38] <seb128> gnome-session-remove
[12:38] <seb128> that's a command
[12:38] <pitti> argh, thanks
[12:38] <seb128> :)
[12:39] <smurfix> ... without a manpage, as usual :-/
[12:39] <seb128> patches are welcome :)
[12:39] <pitti> seb128: the gaim icon is still there, well and up
[12:39] <pitti> seb128: that bloody thing was broken for a week now!
[12:39] <smurfix> seb128: can I patch the upstream authors to please include manpages?
[12:39] <pitti> anyway, I test the new package now
[12:39] <seb128> bah, that's always like that, when you want to get a but that works
[12:40] <seb128> smurfix: upstreams don't care to add a manpage for that but if you send one ..
[12:40] <seb128> thom: dude, fixing a gtk bug atm :p
[12:41] <seb128> thom: have you get the webdav access btw ?
[12:41] <smurfix> seb128: I know. I want to patch the upstream people so that the *do* care. ;-)
[12:41] <seb128> smurfix: that's not an easy patch :p
[12:41] <seb128> yeah, be good guys
[12:41] <seb128> STOP USING GTK :)
[12:42] <ogra> huh ? QtGnome ?
[12:42] <seb128> no, just the command line
[12:42] <ogra> heh
[12:42] <pitti> ogra: gnome-curses :-)
[12:42] <ogra> *g*
[12:42] <ogra> pitti: the ones you suffer from ?  ;)
[12:44] <pitti> seb128: the new panel works just as well
[12:44] <thom> seb128: have emailed, not heard back
[12:45] <pitti> seb128: shall I upload it and see what happens? :-)
[12:45] <thom> elmo: please can you sync apache from unstable
[12:45] <pitti> elmo: and if you are at it: mysql-dfsg and mysql-dfsg-4.1 syncs as well please :-)
[12:46] <Kamion> pitti: the partman changes were upstream, not something I did; it'd be a lot of effort to change the UI back and I'm not sure it's worth it
[12:46] <fabbione> elmo: can you please sync libapache-mod-radius-auth and libpam-auth-radius (it's universe stuff, but since we are it ;))
[12:46] <pitti> Kamion: I think it is confusing to be forced to specify the fs type if I just want to use a partition
[12:47] <seb128> pitti: feel free :) Could you update the patch upstream ?
[12:47] <seb128> thom: ok
[12:47] <pitti> Kamion: furthermore, if I accidentially pick the wrong fs type, I can not choose to not format it...
[12:47] <Kamion> pitti: I didn't know you needed to; that sounds like a bug, it should autodetect the currently-used fs type
[12:47] <ogra> hmm, www.ubuntulinux.org is unreachable again ... ??
[12:47] <seb128> elmo: please remove "gstreamer gst-plugins gst-player kgst" from hoary
[12:47] <pitti> Kamion: the fs type is shown in the partition list
[12:48] <pitti> Kamion: but if I select it, I cannot just select "use this partition" any more
[12:48] <Kamion> yes, but the one that's currently on the partition should be the default
[12:48] <smurfix> ogra: works here
[12:48] <Kamion> that sounds like a plain bug, please report it
[12:48] <pitti> Kamion: okay, I try that out again and report it
[12:48] <ogra> smurfix: hmm, i get timeouts tracrouting from several different places.....
[12:49] <smurfix> ogra: traceroute is typically blocked by firewalls
[12:49] <ogra> 62.140.218.35 is always my last hop
[12:50] <elmo> seb128: not built from source?
[12:50] <elmo> aiee, my airo-cs dies and I get flooded with requests
[12:50] <smurfix> ogra: doesn't matter. telnet to port 80 instead.
[12:50] <ogra> probably the routing is broken anywhere at Level3.net ....(all routes i use take that path)
[12:50] <seb128> elmo: old gstreamer0.6 stuff, should be erased from the archive now :)
[12:51] <smurfix> ogra: traceroute -I works too (that uses ICMP packets instead of UDP)
[12:51] <ogra> smurfix: hmm, interesting...GTE / works....
[12:51] <ogra> GET even
[12:51] <haggai> elmo: yup
[12:55] <ogra> haggai: http://www.ubuntulinux.org/wiki/MOTU
[12:56] <thom> seb128: launch-box FTBFS :(
[12:56] <haggai> ogra: great stuff, thanks!
[12:56] <elmo> fabbione: are those the names of source packages?
[12:57] <thom> it's mod-auth-radius
[12:57] <ogra> haggai: it needs some more flesh and blood but is a start to point ppl to
[12:57] <seb128> thom: lemme see
[12:57] <fabbione> elmo: if i didn't mispell them, yes
[12:57] <elmo> haggai: what's the deal with the oo.2 crack?
[12:57] <fabbione> thom: no..
[12:57] <fabbione> thom: i didn't name them as upstream... i got them like that
[12:57] <elmo> haggai: a package called 'openoffice.org2-l10n-' ?  and what's with all the new ones?
[12:58] <thom> fabbione: ahem?
[12:58] <thom> 20:10 < willy> There is no record of the libapache-mod-radius-auth source package, and no bugs have been filed against it.
[12:58] <thom> 20:11 < fabbione> libapache-mod-auth-radius
 it's mod-auth-radius
[12:58] <haggai> elmo: bah, the l10n- should be en-us
[12:58] <fabbione> thom: ok ok.. you removed the libapache-
[12:58] <fabbione> and i am off
[12:58] <fabbione> i really need to start the weekend now
[12:59] <thom> you really ought to know the names of your own packages :P
[12:59] <fabbione> thom: <fabbione> elmo: if i didn't mispell them, yes
[12:59] <pitti> seb128: uploaded
[12:59] <haggai> thom: I know what *I* think they should be ;)
[12:59] <fabbione> thom: shushhhhh
[12:59] <fabbione> :P
[12:59] <fabbione> ah haggai
[12:59] <thom> that's not mispelling! that's out and out misremembering :-)
[12:59] <fabbione> haggai: ooo1?
[12:59] <thom> haggai: *g*
[12:59] <seb128> pitti: thanks
[12:59] <haggai> thom: that's a script going wrong :-/
[01:00] <elmo> so anyway, once you guys figure out who is who's bitch, pls let me know the names of the source packages to sync.  kthxbye
[01:00] <haggai> kthxbyelmo
[01:00] <thom> elmo: apache is the only thing that matters ;-)
[01:00] <fabbione> libapache-mod-auth-radius_1.5.7-6 and libpam-radius-auth_1.3.16-3
[01:00] <Treenaks> g
[01:01] <haggai> fabbione: well ooo2 mysteriously started working again so I guess all I have to do is reupload
[01:01] <fabbione> haggai: hold on
[01:01] <fabbione> no need to reupload to get a build
[01:01] <haggai> fabbione: oh?  what's the magic?
[01:01] <fabbione> haggai: ping lamont
[01:01] <haggai> ah,k
[01:01] <haggai> lamont: ping :)
[01:02] <elmo> fabbione: done
[01:02] <fabbione> thanks elmo
[01:03] <seb128> morning jbailey 
[01:04] <jbailey> Morning sb!
[01:08] <pitti> seb128: sent patch to the gnome bts
[01:08] <seb128> pitti: thanks
[01:21] <ogra> seb128: is it intended that the volume-control applet is hidden if the master mixer is muted on login ?
[01:24] <seb128> ogra
[01:24] <seb128> oups
[01:24] <seb128> ogra: no
[01:24] <ogra> so its a bug....
[01:25] <seb128> correct
[01:25] <seb128> at least I think so, I'm not the upstream
[01:25] <ogra> .... i will wait if your current upload still behaves like this and fle one then ;)
[01:25] <ogra> file even
[01:26] <ogra> seb128: it could also be a feature since it avoids the "no mixer found" errormessage it seems...
[01:27] <seb128> I'll ask to upstream
[01:38] <seb128> ogra: I'm opening a bug upstream
[01:39] <Treenaks> pitti: you were working on language packs right?
[01:39] <pitti> Treenaks: right
[01:39] <pitti> Treenaks: I await elmo's GO for actually uploading them
[01:40] <pitti> Treenaks: this still needs some preparations
[01:40] <pitti> Treenaks: already tried the unofficial ones?
[01:41] <Treenaks> pitti: well, I just got a mail from someone who made a list of packages to make a system more Dutch
[01:41] <Kamion> hm, just hypothetically, what would it take for me to become gdk/linux-fb maintainer?
[01:41] <pitti> Treenaks: cool, just send it to me
[01:41] <pitti> Treenaks: I can add them to language-support-nl
[01:41] <ogra> seb128: ok, thanks
[01:42] <ogra> trukulo: http://people.ubuntu.com/~lamont/buildLogs/g/graveman/0.3.0-2ubuntu1/graveman_0.3.0-2ubuntu1_20050121-1237-i386-successful
[01:42] <ogra> :)
[01:42] <thom> hrm, firefox and ephy really dn't like o2's build lg
[01:42] <Treenaks> pitti: how's your Dutch? :)
[01:43] <pitti> Treenaks: not there :-)
[01:43] <trukulo> ogra: tell me
[01:43] <trukulo> COOL !
[01:43] <ogra> :)
[01:44] <Treenaks> pitti: hmm.. http://www.lacocina.nl/artikelen/ubuntu-nederlands.html is his page.. he links to a few debs that aren't in ubuntu yet
[01:44] <decko> ogra, Are you building graveman???
[01:44] <trukulo> sp graveman is in universe for hoary?
[01:44] <pitti> Treenaks: oh, I can only depend on packages in main
[01:44] <trukulo> or warty?
[01:44] <pitti> Treenaks: I think I should be able to figure out what's written there
[01:44] <ogra> decko: i just upload it...trukulo made the package ;)
[01:44] <Treenaks> pitti: things like a Dutch openoffice dictionary file can't be in main?
[01:45] <elmo> pitti: can you do me a favour and upload two things: 1) a binary upload (of anything) signed by the language pack key, 2) a source upload of anything other than a language pack signed by the lang pack key
[01:45] <ogra> trukulo: hoary
[01:45] <trukulo> ogra: ok
[01:45] <ogra> trukulo: wartys universe wont change anymore
[01:45] <decko> ogra, Because I know a DD that made the package for the lastest version (0.3.1).
[01:45] <trukulo> that's what i tought, ok
[01:45] <pitti> Treenaks: openoffice.org-l10n-nl is in main
[01:45] <trukulo> decko: brazilian one?
[01:45] <decko> trukulo, Yes! otavio
[01:46] <trukulo> yes, sylvain told me
[01:46] <decko> trukulo, otavio is his name
[01:46] <pitti> elmo: hmm, real packages (which go into the archive), or just some mockup which you will trash?
[01:46] <Treenaks> pitti: oh wait.. it described manual installation of "myspell-nl" if the package isn't available
[01:46] <trukulo> no problem for me, if he knows to make better packages than me, then it's better
[01:46] <trukulo> i don't mind, i only want graveman avalaible
[01:46] <elmo> pitti: mockup
[01:47] <elmo> pitti: they'll be rejected, and if they aren't, I'll purge them by hand
[01:47] <decko> trukulo, I think that he upload graveman yesterday do sid
[01:47] <trukulo> decko: how do you know it?
[01:47] <decko> trukulo, he told me
[01:47] <trukulo> umm, i made an ITP in debian
[01:47] <Treenaks> pitti: and he made his own mozilla-firefox-locale-nl package, because it's not in debian or ubuntu
[01:47] <pitti> elmo: I upload a pmount version lower than hoary's current one, this should be rejected
[01:47] <trukulo> for graveman, 3 days ago
[01:48] <decko> trukulo, Wait a moment
[01:48] <elmo> pitti: nono
[01:48] <elmo> pitti: please upload something that would otherwise be accepted
[01:48] <pitti> okay
[01:49] <elmo> just trust me when I say it's not going to get into the archive proper
[01:49] <trukulo> decko: ok
[01:49] <pitti> elmo: ah, now I understand :-)
[01:50] <pitti> elmo: pmount 0.5.1-1invalid1_source uploaded (signed by langpack key)
[01:50] <trukulo> decko: could you tell ottavio my jabberid?
[01:50] <decko> trukulo, What's you jabber ID???
[01:51] <decko> trukulo, What's is your name??? Miguel???
[01:51] <trukulo> yes
[01:51] <trukulo> don't mind, i'm in query with him
[01:51] <decko> trukulo, Ohhh No problem
[01:51] <decko> trukulo, ha has send a email to you!!!
[01:52] <elmo> pitti: okay, that's 2), (1) too pls
[01:52] <pitti> I'm at it
[01:52] <pitti> I want to use another package
[01:53] <pitti> elmo pkgstriptranslations_4invalid1_i386.changes uploaded (signed by langpack key)
[01:53] <trukulo> decko: he's telling me, don't worry
[01:53] <decko> trukulo, :)
[01:54] <elmo> pitti: excellent, thanks
[01:55] <ogra> trukulo: amd64 and ia64 were successfull too ;)
[01:55] <trukulo> ogra: superb
[01:55] <ogra> yup :)
[01:56] <ogra> we got a gtk2 burning app, yay
[01:56] <elmo> pitti: I'm just creating the 'translations' section in the archive, will pre-add all the packages, then test-run lang-pack-en, and then FINALLY we'll be ready
[01:56] <pitti> cool!
[02:05] <elmo> pitti: okay, please upload the pack and pack-updates for -en, so we can test all 3 types auto-build correctly
[02:06] <pitti> elmo: done
[02:07] <elmo> ok, all 3 in accepted.. let's see how they auto-build, assuming they do, and you're happy, you're clear to upload the rest
[02:07] <pitti> nice, thanks!
[02:07] <pitti> I will follow the build progresss
[02:10] <lamont> haggai: which arch do you want ooo2 requeued on?
[02:13] <elmo> haggai: seriously, what's the point of splitting the oo.o components out by function when they all depend on a 45Mb .deb (140Mb installed) -core package?
[02:16] <haggai> lamont: just ooo1 thanks
[02:17] <haggai> elmo: modem users are very happy if they don't have to download eg 6mb of spreadsheet
[02:17] <lamont> haggai: and on which architecture?
[02:17] <haggai> elmo: and it follows upstreams' packaging
[02:18] <lamont> mvo_: you around?
[02:18] <haggai> lamont: i386 thanks
[02:18] <haggai> elmo: I did already remove 2 tiny packages which upstream had and merged them into core
[02:18] <elmo> haggai: dude, modem users are faced with the 45Mb -core, and godsknowshowmuch -common.. I'm pretty sure they've already drunk themselves unconcious rather than think about oo.o anymore
[02:19] <Treenaks> elmo: "Let's have a beer for every minute of download time" ?
[02:19] <lamont> given back
[02:19] <elmo> btw, appeals to upstream authority are particularly unimpressive when upstream are the folks who include glibc into their upstream source.. :-P
[02:20] <haggai> elmo: I've had enough people have complained about the all-one-one lump situation we've had up until now who are thankful for every saved meg
[02:21] <mvo_> lamont: yes
[02:21] <haggai> there's no glibc in the source
[02:21] <haggai> thankfully that one's a sun-onlyism
[02:21] <elmo> anymore...
[02:21] <haggai> no, it never was in the offical source
[02:21] <elmo> ok, well they still include a bucket load of crack that they ought to just link to
[02:22] <daniels> elmo: that's stupid, man
[02:22] <haggai> yeah well rene has been hard at work fixing lots of that
[02:22] <daniels> elmo: you've gotta include your own copies of zlib, freetype, fontconfig, ...
[02:22] <haggai> daniels: for us yes, but not for their broken environments
[02:22] <Mithrandir> daniels: slightly forked, preferably.
[02:22] <Treenaks> daniels: xlib 8)
[02:22] <daniels> Treenaks: x including xlib is ok, as much as i wish xlib would die
[02:23] <elmo> haggai: well, I'm not going to add the "-l10n-" package as is at least, FYI
[02:23] <haggai> elmo: the oo2 tarball has quite a few of those system lib removed already now btw
[02:24] <no0tic> hi, when will mozilla-thunderbird-enigmail be available?
[02:24] <no0tic> 1.0
[02:24] <ogra> daniels: i got a colleague with a dual opteron and a matrox four chip card ..... xorg can only intialize the first chip for him....
[02:24] <haggai> elmo: no problem it should have been l10n-en-us and I already uploaded a fix
[02:24] <elmo> haggai: it's still 155|Mb ?
[02:24] <Mithrandir> no0tic: it's available already.
[02:24] <daniels> ogra: four-head, or four-core?
[02:24] <haggai> elmo: the official upstream source is > 200mb nowadays
[02:25] <daniels> if it's a g650 (parhelia), you're sol, I'm afraid
[02:25] <daniels> matrox have refused to give us any docs at all on the parhelia
[02:25] <ogra> daniels: it has four chips and four heads...
[02:25] <no0tic> Mithrandir: I updated right now, but I found 0.90.0-1 version
[02:25] <haggai> m65 was a whopping 229 meg
[02:25] <Mithrandir> no0tic: yes, and that is the latest version.
[02:25] <ogra> daniels: he is here...., its a g450 mms
[02:25] <Mithrandir> no0tic: it works with thunderbird 1.0
[02:26] <no0tic> Mithrandir: ah, ok I thought it was for thunderbig 0.9 .. :)
[02:27] <daniels> haggai: ... that's not COMPRESSED, is it?!?
[02:27] <Mithrandir> no0tic: the depends for packages are usually correct, so you wouldn't be able to have incompatible thunderbird and enigmail installed at the same time.
[02:27] <daniels> ogra: hm, I don't think we have support for those -- certainly never seen one
[02:28] <haggai> daniels: uncompressed its over 0.8gb...
[02:28] <ogra> daniels: i could get you the data :)
[02:28] <daniels> anyway, I'm tired as hell, so need to crash -- if you want to file a bug, I'll take a look at it in the morning
[02:28] <daniels> haggai: sick sick sick sick sick sick sick sick sick sick sick sick sick
[02:28] <trukulo> ogra: tonight graveman-0.3.1 is going to enter in sid
[02:29] <no0tic> Mithrandir: thanks
[02:40] <rcaskey> graveman is a neat app but it should be two slightly less neat apps
[02:41] <rcaskey> or two different menu entries ;) graveman --audiocd or graveman --duplicate
[02:41] <rburton> serpentine!
[02:42] <rcaskey> ?
[02:42] <seb128> thom: gnome-launch-box in the archive :)
[02:42] <thom> sweet!
[02:42] <ogra> rcaskey: its just a modern replacement for xcdroast in my eyes
[02:43] <rcaskey> ogra: yah
[02:43] <rburton> rcaskey:  http://s1x.homelinux.net/projects/serpentine
[02:43] <rcaskey> but just with a bit of tweaking it could be much more gnomish
[02:43] <ogra> rcaskey: the two apps you look for are nautilus-cd-burner and rhythmbox ;)
[02:43] <rburton> ncb and rhythmbox|serpentine will rule
[02:43] <rburton> i've got half-done serpentine debs hanging around somewhere
[02:43] <rcaskey> yeah rhythmbox will take that over
[02:44] <rcaskey> but there is no point and click way to duplicate an optical disk is there?
[02:44] <ogra> rcaskey: i just urged to get graveman in universe to have _anythin_ at all for audio in gnome....to prevent tha ppl to be forced to use k3b
[02:45] <rcaskey> is RB going to get burning in in time for hoary?
[02:45] <rcaskey> I compiled the other day and it did okay but crasehd out burning a disk
[02:45] <chrisa> rcaskey: Would likely make more sense to ask in #rhythmbox on gimpnet
[02:45] <trukulo> ogra: graveman in sid tonight or tomorrow
[02:46] <rburton> but but but graveman is ugly
[02:46] <ogra> trukulo: ypu, i saw it
[02:46] <rcaskey> because if it is, maybe it's better to get experimental rb debs going instead
[02:46] <ogra> rburton: its at least _something_
[02:46] <rburton> ogra: serpentine is a better audio cd burner
[02:46] <trukulo> ogra: ok
[02:47] <rcaskey> Serpentine does look nice in the screenshots
[02:47] <ogra> rburton: btter then telling users pleas use k3b or xcdroast if you wanna burn audio
[02:47] <rburton> ogra: i understand, but i'm not saying use xcdroast or k3b
[02:47] <ogra> rburton: so when will you become a MOTU and bring it to universe ?
[02:47] <rcaskey> ogra: do you know anything about point & click copying of optical media?
[02:48] <rcaskey> (btw, I put out an SOS last night for a debian-trusted signatory for my key)
[02:48] <ogra> rburton: since you are already writing ubuntu stuff :)
[02:48] <rburton> ogra: amusingly i'm still running warty on my desktop :)
[02:48] <rburton> i must find more disk space and get a warty install
[02:48] <haggai> elmo: please can you sync dmake from Debian?
[02:49] <ogra> rcaskey: what exactly do you mean if i know anything about point n click copying ?
[02:49] <rcaskey> is there a nautilus extension for that I am missingh
[02:49] <elmo> haggai: that's in main and would violate the UVF.. please mail me, cc mdz and jeff
[02:49] <rcaskey> some little gtk python frontend to dd
[02:49] <elmo> [Updating]  dmake (4.2+cvs20031009-5 [ubuntu]  < 4.3-1 [debian] )
[02:49] <haggai> elmo: ok
[02:50] <rcaskey> ogra: like open up Computer, right click on CD drive and select "Make an Illegal Copy of this Disk" from the menu
[02:50] <elmo> haggai: (oh, btw, when you send mail, you need to include rationale for violating UVF)
[02:50] <rburton> rcaskey: iirc that's planned for ncb 2.12... hadess keeps on saying he'll write it
[02:51] <thom> Failed to fetch http://archive.ubuntu.com/ubuntu/pool/universe/g/gnome-launch-box/gnome-launch-box_0.1-0ubuntu2_i386.deb  404 Not Found
[02:51] <thom> grump
[02:51] <ogra> rcaskey: should be trivial to implement (but i wont do it yet)
[02:51] <rcaskey> ncb?
[02:51] <rburton> nautilus-cd-burner
[02:51] <rcaskey> ahh
[02:51] <rcaskey> does n-c-b work fine for data dvds as well?
[02:52] <ogra> rcaskey: yup
[02:53] <ogra> rburton: join in ;) https://www.ubuntulinux.org/wiki/MOTU
[02:53] <rburton> how the hell did that page change the bg colour of my URL entry field
[02:54] <ogra> https
[02:54] <ogra> :)
[02:54] <ogra> its firefox that does it, not the page
[02:54] <rcaskey> I've been wondering how many man-hours it would take me to get up to speed enough on c and gtk to hack on nautilus, I'm thinking maybe 100 man hours
[02:54] <opi> and that's nice feature
[02:55] <rburton> ogra: aah, new galeon feature.
[02:55] <rcaskey> some things about it are bugging me to death
[02:55] <rcaskey> #1 is the fact that items on the history bar are not gtk drop targets
[02:55] <rburton> rcaskey: nautilus team always want more hackers
[02:56] <rcaskey> yeah, I need to learn how to suck less though
[02:59] <ogra> rburton: so will we see serpentine integration in soundjuicer ?
[02:59] <rcaskey> At work I have a G4 but I'm getting a Dualie G5 next year
[02:59] <rburton> ogra: how?
[02:59] <rburton> that's the 3rd time i've been asked about serpentine integration into SJ but i can't see a single good reason
[03:00] <ogra> rburton: audio CD copy
[03:00] <rburton> ogra: that should be a right-click action on the CD icon in Computer
[03:01] <rburton> thus in ncb
[03:01] <ogra> hmm...true
[03:01] <rburton> SJ audio CD copying would encode and then decode an entire album
[03:01] <rburton> waste of time and effort
[03:01] <ogra> rburton: but currently th only way to do it with a default gnome
[03:01] <Treenaks> ogra: coaster..
[03:02] <ogra> Treenaks: audio ?
[03:02] <chrisa> People actually use coaster?
[03:02] <Treenaks> chrisa: don't know
[03:02] <chrisa> I thought it was just something harshy spoke of that no one really paid attention to
[03:03] <tseng> ogra: do multiverse bugs apply to you?
[03:03] <ogra> tseng: hmm, good question .....
[03:03] <ogra> tseng: what is it ?
[03:03] <Treenaks> MotU vs MotM ?
[03:03] <elmo> pitti: err, dude
[03:04] <ogra> Treenaks: nah
[03:04] <tseng> ogra: hm, fixed possibly anyway
[03:04] <elmo> pitti: I thought you seeded the language packs?
[03:04] <pitti> elmo: I did?
[03:04] <rcaskey> coaster is someting I see all the time on the planet and think "why"
[03:04] <pitti> $ madison language-pack-en
[03:04] <pitti> language-pack-en |   20050119 |         hoary | source, all
[03:04] <ogra> rcaskey: because libburn is the future.... ?
[03:04] <zul> hey pitti
[03:04] <pitti> Hi zul
[03:05] <rcaskey> I thought it wasn't using libburn anymore?
[03:05] <pitti> elmo: what's wrong?
[03:05] <elmo> pitti: yeah, but anastacia wants to demote them all to universe
[03:05] <elmo> pitti: anyway, if you did, don't worry for now, I'll check some more, the problem may be at my end
[03:05] <pitti> elmo: I added them to "supported"
[03:06] <pitti> elmo: so shall I wait with uploading the rest?
[03:06] <elmo> nah, you can do that if you want
[03:06] <pitti> okay, fine!
[03:06] <pitti> then let's get ready for the flood :-)
[03:07] <Mithrandir> fabbione: feel like looking at 2502?  noapic doesn't seem to apply there?
[03:15] <pitti> elmo: uploaded. I'll grab something to eat now, later
[03:23] <no0tic> I get an error installing gnome-panel-data
[03:23] <no0tic> just upgraded
[03:24] <no0tic> I can paste it in Italian, sorry, it's a problem in /schemas/apps/clock_applet/prefs/hour_format, it doesn't recognize value  for the schema
[03:31] <no0tic> brb
[03:32] <smurfix> anybody know mako's schedule?
[03:33] <tseng> 6am wake up, shave, shower, 7pm breakfast (of champions), 7:15, hit irc
[03:34] <smurfix> tseng: ... am or pm?  ;-) and which timezone?
[03:34] <Mithrandir> he lives in NYC, iirc
[03:34] <lamont_r> smurfix: he's UTC-0500 right now, but dunno what his schedule is
[03:34] <lamont_r> Mithrandir: yep
[03:34] <amu> smurfix: mako: Core Hours: roughly 10:00 - 18:00 EST (15-23 UTC)
[03:35] <sivang> Morning all!
[03:35] <smurfix> thanks, that's helpful.
[03:35] <amu> sivang: moins? it's going to be dark :) 
[03:36] <sivang> amu: yeah :), I am trying to go back to UTC , but my body keeps EST :)
[03:36] <smurfix> lamont_r: mine look more like "core dump hours" lately
[03:36] <sivang> smurfix: hehe
[03:38] <amu> lamont_r: hehe you should use UTC :) 
[03:38] <lamont_r> amu: meh
[03:39] <opi> we should use Swatch Internet time
[03:39] <opi> ;-)
[03:39] <__daniel> hai
[03:40] <ogra> oh, __daniel....hi :)
[03:40] <amu> opi: I found you *g* pls join #kubuntu-devel 
[03:40] <__daniel> did anyone notice the   Arithmetic exception   on amd64's gnome-launch-box?
[03:40] <Treenaks> ogra: beats! yeah!
[03:40] <__daniel> i would so much liked to use it :-)
[03:40] <ogra> :)
[03:40] <__daniel> hi ogra
[03:41] <Treenaks> uh. opi 
[03:41] <ogra> heh
[03:41] <opi> hi Treenaks
[03:41] <ogra> Treenaks, opi: are you sure they are not patented ?
[03:41] <opi> hum..
[03:41] <opi> never heard about it
[03:42] <opi> but I liked the idea
[03:42] <Treenaks> there should be locale-modifiers to do this!
[03:42] <opi> ,,lets meet @500''
[03:42] <opi> it was clear, without all this UTC, CET and stuff
[03:42] <Treenaks> opi: there's still the dateline problem... "the NEXT @500, or the @500 after that?"
[03:43] <opi> Treenaks, true :)
[03:43] <Treenaks> opi: "stardate" would fix that
[03:43] <ogra> Treenaks: or ubuntime
[03:44] <opi> UNIX_TIMESTAMP()
[03:44] <opi> from UTC time
[03:45] <Treenaks> opi: that's too geeky for most ;)
[03:45] <opi> naa, we need a simple program and a database ;)
[03:45] <opi> SELECT UNIX_TIMESTAMP(NOW())
[03:46] <sivang> opi: that's in shell/c ?
[03:46] <opi> SELECT meeting FROM meetings WHERE UNIX_TIMESTAMP(NOW()) == meeting_date
[03:46] <opi> sivang, SQL
[03:46] <opi> sorry, I'm in a middle of finishing a financial system
[03:46] <opi> and I'm sensible when it comes to dates ;)
[03:47] <sivang> opi: ah ok :)
[03:48] <lamont_r> wrote 174063 bytes  read 127798477 bytes  2348120.00 bytes/sec
[03:48] <lamont_r> total size is 618282296  speedup is 4.83
[03:48] <lamont_r> hrm...  that's not too bad
[03:48] <Treenaks> opi: finanicial systems in SQL... I did that once :)
[03:48] <opi> sivang, but SQL query can be called from virtal any language :)
[03:49] <opi> Treenaks, SQL's just a place to store things
[03:49] <Treenaks> opi: I know.. I did it in Perl + SQL
[03:49] <opi> Treenaks, but thanks to SQL language you can do mutch stuff with query
[03:49] <Treenaks> opi: + apache
[03:49] <opi> Treenaks, instead processing data by yourself
[03:49] <lamont_r> wrote 160329 bytes  read 17721011 bytes  701229.02 bytes/sec
[03:49] <Treenaks> opi: I put most of my "intelligence" in queries.. the programs did not much more than just show the output
[03:49] <lamont_r> total size is 524116225  speedup is 29.31
[03:50] <lamont_r> and that's just _FUNNY_
[03:50] <opi> Treenaks, that's the way it should be done
[03:50] <opi> Treenaks, database is so mutch faster at processing
[03:50] <Treenaks> lamont_r: nice transfer speeds
[03:50] <Treenaks> opi: yeah
[03:50] <lamont_r> we need a little package churn in ubuntu-desktop... where's seb128 
[03:50] <lamont_r> ?
[03:51] <opi> Treenaks, some people just SELECT * FROM table ;-)
[03:51] <lamont_r> Treenaks: that's the data center lan
[03:51] <lamont_r> it's more the speedup that I care about
[03:51] <Treenaks> lamont_r: I don't even know what it means...
[03:51] <lamont_r> since that's the cloop image for the livecd
[03:51] <seb128> lamont_r: what ?
[03:51] <ogra> opi: at least thats an easy entry threshold.....didnt we all start with select * from ... once ?
[03:52] <lamont_r> seb128: how much more gnome uploadage should I expect today?
[03:52] <opi> ogra: sure we do.. but it's lame if you won't go further
[03:52] <seb128> lamont_r: I'm making bug triage, so probably not a lot
[03:52] <lamont_r> sadly, that 17MB of data rsync transferred were from officially identical cloop filesystems...
[03:52] <opi> ogra: if you see someone who do sutch stuff as a proffesional developer, then you just cry ;-)
[03:53] <lamont_r> seb128: oh well. :-)
[03:53] <seb128> lamont_r: should I hold the upload for a while time to get a CD built or something ?
[03:53] <ogra> opi: nope, i suggest gettina an SQL class ;)
[03:53] <lamont_r> seb128: no.  I actually _want_ some churn today
[03:53] <lamont_r> but no big deal
[03:54] <seb128> why do you want that ? :)
[03:54] <no0tic> seb128: Changes submitted for bug 5717, gnome-system-tools
[03:54] <lamont_r> so I can see what affect a bit of churn has on the livecd cloop fs
[03:55] <seb128> no0tic: I get mails from bugzilla, no need to ping on IRC
[03:56] <no0tic> seb128: sorry
[03:57] <seb128> no0tic: oh no, don't worry, that's was just for information :)
[03:57] <wasabi> Heh. All that database stuff that was just talkeda bout.
[03:58] <wasabi> THat's a formula for inscalibility.
[03:58] <seb128> no0tic: seems to be a dup of #5578 which should be fixed
[04:00] <seb128> no0tic: somewhat the patch has not been applied to the build, updating the package right now
[04:01] <no0tic> seb128: I searched in bugzilla before posting, I didn't noticed there already was...
[04:01] <seb128> no0tic: it's closed ... 
[04:02] <seb128> no0tic: that's probably why you didn't notice it
[04:02] <lamont_r> haggai: ??
[04:03] <lamont_r> haggai: openoffice.org2_1.9.66-0ubuntu6_i386.changes REJECTED
[04:03] <lamont_r> Broken l10n package name.
[04:04] <lamont_r> haggai: of course, what that actually means is an elmo question
[04:08] <shlomil> hi , where can i find source for Rosetta? what is the license for it ?  
[04:11] <thom> shlomil: try #rosetta
[04:12] <shlomil> oh ok , thanks 
[04:16] <haggai> lamont_r: don't worry I asked elmo to reject it
[04:16] <haggai> lamont_r: there should be a fixed pkg building
[04:17] <lamont_r> ah, okl
[04:18] <lamont_r> haggai: so the 3.4 trampolines in ccache mean I can start hijacking it and routing it through ccache?
[04:19] <haggai> lamont_r: should do, if you prepend /usr/lib/ccache to your path
[04:20] <lamont_r> I wind up invoking it as 'ccache gcc-3.4.real ...', actually
[04:20] <lamont_r> so it probably doesn't even care about the trampoline, no?
[04:20] <haggai> oh, no
[04:21] <haggai> the trampoline is only for if you don't include ccache explicitly on the command line.
[04:21] <LetterRip> Hello all, I read the meeting minutes and saw a mention of a site blocking tool to be developed for ubuntu, I have a fairly nice design document for a tool I call safeTNet that I'd be interested in having the coders consider
[04:21] <haggai> if you do, ccache should ignore all gccs that are actually symlinks to itself like in /usr/lib/ccache
[04:23] <decko> Hi people!!! What's happen to the ubuntu mirrors??? I can't download anything from them
[04:24] <lamont_r> haggai: ok.
[04:24] <lamont_r> decko: what exactly is the failure?
[04:25] <decko> lamont-away, 404 Object not found!] 
[04:25] <lamont_r> decko: and you have a current Packages file?
[04:25] <lamont_r> that is, if you re-update and try again, does it work, or still fail?
[04:25] <lamont_r> and what exactly are you trying to download?
[04:31] <lamont_r> hrm.. no -onomtime,noctime available in mount... :0)
[04:31] <lamont_r> I guess that's a good thing
[04:32] <decko> lamont-away, I'm not using apt to download this files, I'm trying to download from the ubuntu's packages page 
[04:32] <lamont_r> ah, ok
[04:32] <lamont_r> url?
[04:32] <lamont_r> both the 404 and the page with the link, that is.
[04:34] <decko> Nope! I know, when you go to the packages page, select a package, and after this, select the archicteture
[04:35] <decko> lamont_r, And then, you select from where you want to download the file
[04:36] <lamont_r> decko: of course, this really belongs in #ubuntu...
[04:37] <lamont_r> decko: I need a url before I can see what's happening..
[04:40] <lamont_r> ironwolf: shouldn't you be asleep right now?
[04:40] <decko> lamont_r, http://higgs.djpig.de/ubuntu/www/hoary/gnome/update-manager
[04:40] <decko> lamont_r, Select the i386 architeture and try to download something
[04:41] <__daniel> decko: i had problems on archive.ubuntu.com too some minutes ago - i guess you just have to be patient for 20 minutes or something
[04:41] <Mitario> heyhey
[04:42] <decko> __daniel, But I can't download anything from any mirror!
[04:43] <__daniel> decko: i had problems too, some minutes ago
[04:43] <lamont_r> decko: that mirror looks out of sync
[04:43] <lamont_r> 0.36 is current, although that was uploaded on Jan 19
[04:44] <Mitario> hey guys, is there any active testing/development/debugging project of bluetooth stuff in hoary?
[04:45] <Mitario> or for the future :)
[04:45] <decko> lamont_r, ahauahuahua Then all the mirrors are out of sync!
[04:45] <decko> __daniel, Ok guy! Thanks so much!
[04:45] <Kamion> update-manager |       0.36 |    hoary/main | amd64, i386, ia64, powerpc, source
[04:45] <Kamion> that's on rookery ...
[04:46] <mvo_> decko: maybe it's just this http://higgs.djpig.de/ubuntu/www/ site that's not up-to-date?
[04:47] <Kamion> decko: ftp.mirrorservice.org is fine, for instance; definitely not all the mirrors
[04:47] <mvo_> seb128: is a new version of libnautilus-extensions available? it looks like g-s-t needs it
[04:49] <lamont_r> mdz: wrote 160329 bytes  read 17721011 bytes  701229.02 bytes/sec
[04:49] <lamont_r> total size is 524116225  speedup is 29.31
[04:49] <lamont_r> mdz: 4k/4k is _worse_ by loads... (127MB)
[04:49] <lamont_r> although 17MB of diff for effectively no change is kinda sad...
[04:50] <__daniel> decko: de rien :-)
[04:51] <lamont_r> mdz: doh.. comparing manifests, 10 packages are different, so I'm less concerned.
[04:51] <seb128> mvo_: no
[04:51] <seb128> mvo_: which version is required ?
[04:52] <pitti> lamont_r: the langpacks are flowing into the archive. strip time! :-)
[04:52] <pitti> lamont_r: this time for real, I hope :-)
[04:53] <lamont_r> pitti: current plan was to drop new livecd roots into things, then go to the gym, then go home...  Mind if stripping waits a couple more hours?
[04:55] <mvo_> seb128: it claims that it needs 2.9.3
[04:56] <seb128> mvo_: CVS or a release ?
[04:56] <seb128> mvo_: that's the CVS version (bumped after the 2.9.2 release)
[04:56] <mvo_> seb128: ok
[04:56] <Kamion> lamont_r: sure
[04:57] <mvo_> seb128: I'm sending garnacho a patch that makes the use of sudo a gconf option
[04:57] <Kamion> lamont_r: is this the rsyncable one?
[04:57] <seb128> mvo_: rock
[05:00] <lamont_r> Kamion: no, the next one is :-)
[05:00] <lamont_r> yes
[05:00] <lamont_r> 3/4 building... (ia64 needs more help.)
[05:00] <lamont_r> %partimage: !ia64 !alpha                                              # 64-bit is br0ken
[05:01] <lamont_r> hrm.. that bodes ill for amd64 as well...
[05:01] <lamont_r> or is the filesystem 32-bit?
[05:02] <lamont_r> Kamion: so anyway, this'll give us much more rsyncable i386/ppc images, and rsyncable amd64 images that may or may not be totally and thuroughly broken.
[05:02] <lamont_r> please advise. :-)
[05:02] <amu> mvo_: any plans for sun? 
[05:02] <pitti> lamont_r: I don't mind at all. This was just intended as a notification :-)
[05:02] <mvo_> amu: still uncertain :)
[05:02] <lamont_r> pitti: cool
[05:02] <Kamion> lamont_r: awesome
[05:03] <lamont_r> Kamion: after the gym, I'll ping you and download the image that I'll rsync against after this...
[05:03] <Kamion> lamont_r: hm, is ia64 broken? I thought it had started working
[05:03] <lamont_r> and then I'll go home and change things so that partimage is optional
[05:03] <lamont_r> the new script uses partimage for some of the rsync magic
[05:03] <Kamion> it's got a livecd-current.cloop
[05:03] <lamont_r> and partimage is PaS
[05:03] <Kamion> lamont_r: ia64 CDs should be building anyway as of tonight
[05:03] <lamont_r> yes - not the rsyncable change though
[05:04] <lamont_r> once you see ~buildd/livecd/latest populated with a cloop, then current is the same thing and ready for the CD
[05:04] <lamont_r> build
[05:04] <Kamion> I'd enabled it but forgotten to sync to little, and then I managed to run into a baz bug ... but the change is there now
[05:05] <lamont_r> partimage and 64-bit: either ia64 could build and be fine, or amd64 may be toast.
[05:05] <lamont_r> ah, ok
[05:05] <lamont_r> so someone should really test the new amd64 livecd once you burn it... :-)
[05:07] <lamont_r> anyway, off to the gym, and then the bandwidth-rich coffee shop.
[05:07] <lamont_r> back in about 90-120 minutes, give or take
[05:08] <amu> lamont_r: 20050121/amd64 ? 
[05:09] <lamont_r> amu: no, the one Kamion is going to build in about 40 minutes or so
[05:09] <lamont_r> you'll want to grab that one in any case, since future ones should rsync much nicer....
[05:10] <amu> lamont_r: no prob, i can test it 
[05:10] <fabbione> hey gys
[05:10] <fabbione> guys
[05:10] <lamont_r> amu: thanks - if it's thuroughly b0rked, then we need to fix partimage...
[05:10] <amu> Kamion: just ping me if it's ready
[05:13] <amu> *pong*
[05:17] <Kamion> what's up with partman?
[05:17] <lamont_r> it really doesn't like to run without a valid $TERM
[05:17] <Kamion> how did you manage not to have a valid $TERM?
[05:17] <lamont_r> Failed to open terminal.Unknown terminal: unknown
[05:17] <lamont_r> Check the TERM environment variable.
[05:17] <lamont_r> Also make sure that the terminal is defined in the terminfo database.
[05:17] <lamont_r> Alternatively, set the TERMCAP environment variable to the desired
[05:17] <lamont_r> termcap entry.
[05:17] <lamont_r> easy - it's inside a chroot in an at job.
[05:17] <Kamion> uh, you sure you mean partman?
[05:18] <Kamion> that's part of d-i
[05:18] <lamont_r> partimage
[05:18] <lamont_r> sorory
[05:18] <Kamion> aha
[05:22] <fabbione> jdub, thom, Mithrandir: ping
[05:22] <thom> fabbione: yo?
[05:23] <fabbione> thom: you should be able to install hoary from sparc.u.c now :-)
[05:23] <fabbione> elmo: can you bless the latest d-i please? (when you have time)
[05:23] <thom> fabbione: heh
[05:24] <stockholm> Werner_: it is in pere`s homedir on /wc
[05:24] <stockholm> heh
[05:24] <thom> stockholm: uh, wrong window one suspects
[05:24] <stockholm> thom: (c:
[05:24] <fabbione> pool/main/l/language-pack-bg/language-pack-bg_20050119_all.deb
[05:25] <fabbione> YEPPA
[05:25] <stockholm> thom: but this time it was not even embarrassing.
[05:25] <stockholm> content at least.
[05:25] <seb128> wasabi, wasabi_: no need to Cc: me I read ubuntu-devel
[05:25] <wasabi_> did i?
[05:26] <wasabi_> oh, oops. ;)
[05:26] <wasabi_> actually this one menu editing thing has incited me. i work in an office of people using ubuntu now, and one of the office guys wanted to try hoary... and was really miffed that he couldnt' edit the menus, and that he was forced to use weird text files. ;)
[05:27] <wasabi_> and i have to agree with his feelings.
[05:27] <seb128> why do you need to edit the menu ?
[05:27] <seb128> it should "just work"
[05:28] <wasabi_> They want to add their own items to it.
[05:28] <wasabi_> Stuff that is not properly packaged.
[05:28] <wasabi_> And probably never will be.
[05:28] <seb128> ie: you get a menu entry with all the user apps
[05:28] <wasabi_> third party apps
[05:28] <Kamion> if you're adding your own applications then you *obviously* need to edit the menu ...
[05:28] <lamont_r> seb128: and my list of machines menu
[05:28] <wasabi_> if it was just temporary and attributed to "hoary not being done yet", I wouldn't have a problem... but that the feature is being removed because the replacement "is not done yet"... seems odd.
[05:28] <seb128> applications:// was crap
[05:28] <wasabi_> just leave the old peice of crap in until a proper replacement is complete.
[05:28] <wasabi_> i agree.
[05:29] <seb128> you can't argue than the old menu with applications:// was better than the new one
[05:29] <wasabi_> i'm not trying to
[05:29] <seb128> the old piece of crap edit a vfolder
[05:29] <seb128> which is not used for gnome-panel
[05:29] <seb128> that doesn't work
[05:29] <wasabi_> but a feature was just ripped out from under many people who used it, and they were told to "edit some arcane text files instead".
[05:29] <wasabi_> that's just not right.
[05:29] <seb128> you prefer to do nothing so ?
[05:30] <wasabi_> I prefer to leave the menus as they are until the proper way is complete.
[05:30] <seb128> I agree than having an editor would be better
[05:30] <seb128> ok, so you would have kept the old vfolder crap with applications:/// rather than the new menu
[05:30] <lamont_r> seb128: are there plans for one before 2.10?
[05:30] <seb128> lamont_r: no
[05:30] <wasabi_> seb128: that's not the question
[05:30] <wasabi_> seb128: the question is "would you prefer to be able to edit the menus in hoary or not"
[05:30] <seb128> lamont_r: "plan" could be, people with the willing and the time ...
[05:30] <lamont_r> seb128: right
[05:31] <seb128> wasabi_: no, that's not the question
[05:31] <wasabi_> it's the question I was just asked by an ubuntu user here in the office
[05:31] <seb128> wasabi_: you are saying "keep GNOME 2.8 for hoary"
[05:31] <wasabi_> basically
[05:31] <seb128> just install menu and menu-xdg
[05:31] <lamont_r> seb128: I think the real question he's trying to ask is "is the lack of an editor for menus in 2.10 considered a release critical bug or not?  ditto for hoary?"
[05:31] <seb128> you'll get the Debian menu
[05:31] <wasabi_> My users don't give a damned how you make it happen.
[05:31] <seb128> and edit that :p
[05:32] <seb128> lamont_r: if you ask if I'm going to write a menu editor for hoary, I'm not, too busy for that :)
[05:32] <lamont_r> because wasabi_ appears to consider it release critical.
[05:32] <seb128> but patches are welcome
[05:32] <wasabi_> This is just one of those things which really pisses peple off about linux and open source in general
[05:32] <lamont_r> seb128: I know that
[05:32] <seb128> we should bounty that if 
[05:32] <wasabi_> And I'm just hilighting that.
[05:32] <seb128> jdub: ping ?
[05:32] <Kamion> hey, let's not turn this into something more general than it needs to be
[05:33] <wasabi_> heh.
[05:33] <lamont_r> wasabi_: as opposed to proprietary source,where they just don't implement the feature and don't give you the source??
[05:33] <Kamion> I've seen this kind of thing happen with software from all sources
[05:33] <lamont_r> Kamion: ditto
[05:33] <wasabi_> listen, all im saying is when I deploy hoary, my users will come to me for support, and I won't have a good answer (unless menu-xdg is a good answer, perhaps it is, trying it now)
[05:33] <seb128> wasabi_: you should consider that applications:/// was not really obvious in warty too and many people were already complaining about the menu edition, that's not new
[05:34] <seb128> wasabi_: menu-xdg install a Debian menu in applications ... have you already used debian/the debian menu ? :)
[05:34] <lamont_r> seb128: IOW, "lack of a good menu editing solution continues to be an issue looking for someone to solve it."?
[05:34] <wasabi_> seb128: you could right click on items In The Menu and edit them. add remove, delete.
[05:34] <wasabi_> You did not need to browse to applications:///
[05:34] <wasabi_> So, it was VERY obvious.
[05:34] <seb128> hum
[05:35] <seb128> so why people keep complaining about it for months :p
[05:35] <wasabi_> I just confirmed this on Warty across the office.
[05:35] <wasabi_> I dunno.
[05:35] <seb128> lamont_r: right
[05:36] <wasabi_> I mean, I don't have a good answer for you, from the developer side.
[05:36] <jdz_work> This might make a good love-fest afternoon or a good bounty.
[05:36] <wasabi_> But, from the user side, I can say, it's one of those things which will come back to haunt ya.
[05:36] <seb128> wasabi_: I've a 2.8 box here and adding a menu in the menu doesn't seems to be obvious
[05:36] <lamont_r> seb128: and 2.8->2.10 goes from a really crappy menu editor to no menu editor?
[05:36] <wasabi_> seb128: right click in the sub menu
[05:36] <seb128> lamont_r: correct
[05:36] <wasabi_> use the ... "Entire menu" or something, box.
[05:36] <wasabi_> You can add new items.
[05:36] <seb128> wasabi_: I've an option to add the menu to the panel
[05:36] <lamont_r> seb128: so if I install menu and menu-xdg I get a 'debian' menu entry that I can then play with?  that could be a start for me...
[05:36] <wasabi_> Yeah, it does suck. It is bad UI.
[05:37] <wasabi_> But, a right click can figure it out.
[05:37] <seb128> wasabi_: but not to create an entry
[05:37] <wasabi_> seb128: right click on an item in a submenu, use the Entire Menu sub pop up menu
[05:37] <seb128> lamont_r: you get a debian menu ... do we have an editor in debian for this crap ? :)
[05:38] <seb128> wasabi_: bah, that sucks, most of people don't find it and I know why :p
[05:38] <lamont_r> Kamion: eta 35 minutes or so
[05:39] <Kamion> wasabi_: if you're doing the deployment, do you have the facility to add .desktop files?
[05:39] <lamont_r> Kamion: actually, i386 is about 5-10 away from done... :)
[05:39] <Kamion> [note I don't know enough about GNOME to know whether that's sufficient to add menu items] 
[05:39] <seb128> Kamion: you just need to add a .desktop to add a menu entry (in /usr/share/applications or ~/.local/share/applications)
[05:40] <seb128> BTW perhaps KDE has an editor for the menu, since the menu are freedesktop ones now ?
[05:40] <seb128> haggai: ping ?
[05:40] <jdz_work> seb128: So what we need is a graphical editor for .desktop menu entries?
[05:40] <Riddell> seb128: kmenuedit has always been there
[05:40] <seb128> jdz_work: a way to edit menu since some people want this
[05:40] <seb128> Riddell: does it use the freedesktop spec ?
[05:41] <wasabi_> and Add new item to this menu
[05:41] <wasabi_> It pops up the launcher creator box
[05:41] <wasabi_> Let me just list some apps my users need in these menus... some of the ones I *know* about anyways.
[05:41] <wasabi_> vmware we use. it doesn't make .desktop items.
[05:41] <wasabi_> that stupid ass adobe acrobat reader
[05:41] <wasabi_> most of our developers put Eclipse in there
[05:41] <wasabi_> some install Wine applications which we need and put an item in there for each thing.
[05:41] <wasabi_> etc etc etc
[05:41] <wasabi_> anyways, i dont have a good answer for you. i realize the old system sucked and the new one is better...
[05:41] <Riddell> seb128: yes I think so
[05:41] <wasabi_> woh
[05:41] <wasabi_> lag
[05:41] <seb128> wasabi_: could you try this kmenuedit ?
[05:42] <haggai> seb128: pong
[05:42] <wasabi_> Kamion: I dont do the deployment of a lot of things. We're a very ad hoc company. We deploy the basic stuff, but really each department fends for themselves... users are given artistic license on their PCs
[05:42] <seb128> haggai: anything in KDE to edit GNOME menus ? :p
[05:42] <wasabi_> seb128: checking
[05:43] <wasabi_> Kamion: and to say that that's not the case for every home user would be foolish.
[05:43] <amu> seb128: kmenuedit ? 
[05:43] <lamont_r> Kamion: i386 is compressing, other two are chunking along in apt - I'm finally gonna go disappear for that 2 hours or so
[05:43] <seb128> I'll try kmenuedit
[05:44] <seb128> dunno if it works for the new GNOME fd menu
[05:44] <Kamion> wasabi_: I'm not trying to say that; I'm trying to help you by suggesting workarounds
[05:44] <haggai> seb128: yeah, kmenuedit.  I don't know if it works for the Gnome menu nowadays
[05:45] <Kamion> wasabi_: (and yes, every company I've ever worked for as a permanent employee has worked the way you say, and it's great)
[05:45] <seb128> bah, still, I hate pointing a KDE stuff to GNOME users :p
[05:45] <amu> haggai: should works also for gnome, both use fd.o
[05:45] <wasabi_> kmenuedit looks like it works
[05:46] <seb128> wasabi_: ok, you have a workaround so if you don't mind installing the deps ...
[05:46] <wasabi_> or not.
[05:46] <Riddell> seb128: can't they use applications:///
[05:46] <wasabi_> it reads... doesn't look like it saves
[05:46] <seb128> Riddell: that has been dropped in 2.9
[05:46] <Riddell> seb128: why is that?
[05:46] <wasabi_> seb128: kmenuedit does not work
[05:47] <wasabi_> seb128: it reads the config, but does not save changes.
[05:47] <haggai> looks like gnome menus don't get updated
[05:48] <seb128> Riddell: you should read ubuntu-devel :/
[05:48] <seb128> Riddell: gnomevfs has dropped the vfolder support which was the stuff used for the old menus, now it uses the freedesktop specifications
[05:49] <seb128> Kamion: about #4280 ?
[05:50] <Kamion> hmm
[05:50] <Kamion> I think germinate will be OK with that, let me see
[05:52] <haggai> bah it won't log back into my Gnome session
[05:52] <mdz> morning
[05:52] <haggai> afternoon
[05:53] <seb128> hey mdz 
[05:53] <fabbione> hey mdz
[05:54] <fabbione> haggai: did you talk with lamont?
[05:54] <haggai> fabbione: yes thank
[05:54] <haggai> s
[05:54] <fabbione> did he give it back?
[05:54] <Kamion> mdz: can I add a backdrop title to casper?
[05:55] <Kamion> something like "Ubuntu Live CD"
[05:55] <haggai> fabbione: hmm seems he did and it failed again
[05:55] <mdz> Kamion: yes, certainly
[05:56] <fabbione> ok
[05:56] <fabbione> no.. it's not ok
[05:56] <mdz> Kamion: I had planned to do that, and some other live CD improvements yesterday, but instead I spent all day chasing that init/khubd bug
[05:56] <pitti> Hi mdz!
[05:57] <Riddell> why are ubuntu milestones called arrays?
[05:57] <Kamion> mdz: note I'm going away for the weekend starting in about an hour; if I'm to do the busybox modifications I had better start now
[05:57] <mdz> not even so much as a log message to say "usb is useless now because khubd died, sorry"
[05:57] <fabbione> mdz: RUN! when pitti add a "!" at the end it means troubles!
[05:57] <Kamion> Riddell: collective noun: an array of hedgehogs
[05:57] <sivang> fabbione: hehe
[05:57] <sivang> pitti: Hi Martin!
[05:57] <mdz> Kamion: you already explained what needed to be done for the backtitle; I can do that if you take care of busybox
[05:58] <mdz> Kamion: are you back monday morning?
[05:58] <Kamion> mdz: ok. yes, probably Sunday evening although I don't know if I'll be around much then
[05:58] <pitti> fabbione: no worries, just sayin' hello
[05:58] <fabbione> pitti: :P
[05:58] <pitti> fabbione: I don't trash anything any more at Friday evenings :-)
[05:58] <fabbione> ehehh neither do i
[05:58] <haggai> fabbione: same problem, I'll start a local build
[05:59] <fabbione> i am laying in bed with 39 of fever :/
[05:59] <Riddell> Kamion: but of course :)
[05:59] <fabbione> haggai: ok, because i trashed my build after you told me you will look at it ;)
[05:59] <haggai> fabbione: np
[05:59] <mdz> Kamion: "Ubuntu Live CD" you think?
[05:59] <mdz> Kamion: hmm, but it will also be used on DVDs and USB sticks
[05:59] <Kamion> yeah, was wondering about that
[05:59] <T-Bone> hi
[06:00] <mdz> that's one of the reasons that casper has a proper name
[06:00] <mdz> "Ubuntu Live"?
[06:00] <T-Bone> fabbione, ping?
[06:00] <fabbione> lamont-away:
[06:00] <fabbione> cat chroot-hoary/CurrentlyBuilding 
[06:00] <fabbione> Package: aca
[06:00] <fabbione> Component: universe
[06:00] <fabbione> so the fix works :-)
[06:00] <fabbione> T-Bone: pong :-)
[06:00] <Kamion> mdz: works; reads a bit oddly at first glance
[06:00] <fabbione> i read the mail and answered to it
[06:00] <pitti> fabbione: oh, bad
[06:00] <Kamion> "Live Ubuntu Environment"?
[06:00] <fabbione> pitti: the funny thing is that i don't feel it at all
[06:00] <pitti> fabbione: anyway, I already did my trash of the day, I uploaded 246 packages (language packs) :-)
[06:00] <T-Bone> fabbione, just read your mail (i'm recovering a few boxes at school after powerfailure),
[06:00] <Kamion> "Live Ubuntu"?
[06:01] <mdz> Kamion: if we can think of something better, a number of other Casper templates could use improvement too
[06:01] <sivang> Kamion: UbuntuLive
[06:01] <fabbione> pitti: my gf noticed that i had a weird (more than usual) look
[06:01] <haggai> Ubuntu Live
[06:01] <mdz> I'm not particularly satisfied with "Preparing for live session" either
[06:01] <pitti> fabbione: hmm, but not feeling anything with high fever might also be a very bad sign...
[06:01] <T-Bone> fabbione, what do you mean by 'taking the leadership for the kernel'?
[06:01] <fabbione> pitti: nah.. it has always been like this and i saw the packages :) congratulation man
[06:01] <amu> fabbione: hehe, rest and heal :) my flu is over  
[06:02] <fabbione> T-Bone: you can take over the kernel team leadership :-)
[06:02] <fabbione> amu: i have only 9 days to finish 2 rooms in the house :(
[06:02] <fabbione> it's no good to have to stay in bed
[06:02] <T-Bone> fabbione, who's on that team. What would it involve from my side? :)
[06:03] <fabbione> T-Bone: did you read on the wiki?
[06:03] <T-Bone> fabbione, i'm installing firefox atm ;)
[06:03] <T-Bone> there, having a look
[06:03] <Kamion> mdz: BTW translations of backdrop titles won't work until we get a newer cdebconf (I fixed that upstream), but that's not urgent
[06:04] <fabbione> T-Bone: as kernel leader you lead the team and it involves basically everything 
[06:04] <T-Bone> fabbione, ok got it. Now what am I expected to do, as a (potential) team leader?
[06:04] <fabbione> :)
[06:04] <fabbione> coordinate, delegate, qa, do releases
[06:04] <T-Bone> fabbione, lol, that's an answer I can't cope with :)
[06:04] <fabbione> that's why there is a team
[06:06] <fabbione> T-Bone: that has been disabled
[06:06] <amu> fabbione: place the "brush" into you GF's hand :) 
[06:06] <fabbione> if the installer can reach one of our server it doesn't ask for proxy
[06:06] <sivang> fabbione: from where I know, a team leader also has to cover for team members who do not manage to complete tasks and who do embarrasing stuff :)
[06:06] <sivang> s/where/what
[06:07] <fabbione> sivang: as well
[06:07] <sivang> fabbione: hehe :)
[06:07] <fabbione> amu: she is working
[06:08] <amu> fabbione: you got it :) 
[06:08] <fabbione> i need to reboot 
[06:08] <fabbione> brb
[06:37] <jmones> Hello! I'm not sure if this question is appropiate here, but it seems to me that something is wrong. I need to find /usr/include/X11/Xauth.h but it's not in xlibs-static-dev as in Debian. Where is it?
[06:38] <sivang> jmones: strange, when doing apt-file search for that, I do find it there, did you install the package?
[06:39] <jmones> mmm... yes
[06:39] <sivang> jmones: sorry, my mistake
[06:39] <sivang> jmones: xlibs-static-dev: usr/X11R6/include/X11/Xauth.h
[06:39] <jmones> i installed xlibs-dev
[06:39] <jmones> and xlibs-static-dev
[06:39] <sivang> jmones: so it would be under X11R6 , not under X11
[06:39] <jmones> opps... my fault then
[06:40] <ogra> jmones: ogra@honk:~ $ dpkg -S /usr/X11R6/include/X11/Xauth.h
[06:40] <ogra> libxau-dev: /usr/X11R6/include/X11/Xauth.h
[06:40] <jmones> isn't it in X11 in debian?
[06:40] <Treenaks> ogra: ah, Xorg with split packages
[06:40] <azeem> jmones: there's a compatibility symlink
[06:40] <ogra> yup :)
[06:40] <Treenaks> jmones: Debian has XFree86, Hoary has Xorg
[06:40] <jmones> ah ok
[06:41] <jmones> i've recently changed from xfree86 to xorg and there's are some glitches 
[06:41] <jmones> btw, there's a bug that avoids the user from using ctrl+alt+f1 to switch to VT
[06:41] <sivang> jmones: on a hoary?
[06:41] <jmones> i set the sources from hoary to a warty
[06:42] <jmones> and without dist-upgrading (i was in a hurry) installed xorg
[06:42] <fabbione> jmones: nah.. that's plain wrong...
[06:42] <jmones> i noticed the problem about crl+alt+f1 (there are a lot of references on the web) is because xlibs version
[06:42] <fabbione> jmones.
[06:42] <ogra> jmones: this will certainly break
[06:42] <fabbione> either you update properly or stuff won't work
[06:43] <sivang> ofcourse, I've done dist-upgrade everytime.
[06:43] <jmones> i've now dist-upgraded and everything works :)
[06:43] <fabbione> a bug like that would be noticed even by CNN
[06:43] <sivang> moquist_: Matt ! :)
[06:43] <sivang> fabbione: CNN are using Ubuntu? ;o)
[06:43] <jmones> lol
[06:44] <jmones> thanks very much
[06:44] <fabbione> sivang: no idea.. but if something like was broken, both Daniel and I would have tons of people knocking at the door
[06:44] <sivang> fabbione: true.
[06:45] <ogra> fabbione: knock knoc
[06:45] <ogra> k
[06:45] <fabbione> ogra: come in :)
[06:46] <ogra> fabbione: on the recent ppc livecd console switching doesnt work for me *g*
[06:46] <jmones> i thought that perhaps setting a conflicts in xserver-org on old xlibs could be ok
[06:46] <fabbione> ogra: that has to be a ppc problem
[06:46] <ogra> fabbione: (baut taht is caused by the special machine i got ;)
[06:46] <jmones> ogra, that's another bug i thing related to keyboard
[06:46] <sivang> ogra: doesn't work for me also in i386
[06:47] <fabbione> jmones: please can you move these topics to #ubuntu?
[06:47] <fabbione> the are not really related in here
[06:47] <ogra> jmones: it is common for that machine....x dtdection also doesnt work ;)
[06:47] <fabbione> and if there are bugs, please report them vua bugzilla
[06:47] <fabbione> via even
[06:47] <jmones> fabbione, ok. Sorry. Thank you
[06:47] <fabbione> because otherwise people will lose track of these stuf
[06:47] <fabbione> not everybody uses a huge irc scrollback
[06:47] <fabbione> or read eons of it
[06:48] <fabbione> and the X maintainer is not here
[06:49] <fabbione> abelli: you can start testing the sparc d-i from sparc.ubuntu.com
[06:50] <abelli> fabbione: mmm... sorry for the lack of help, but i had some problems collecting the machines
[06:50] <Kamion> mdz: patch sent; see you next week
[06:50] <Kamion> mdz: as I said, no time to test it I'm afraid :-(
[06:51] <mdz> Kamion: I'll give it a shot, and fix it up if necessary. thanks
[06:51] <ogra> Kamion: thanks for fixing the MOTU typo ;)
[06:51] <mdz> Kamion: enjoy the weekend
[06:51] <ogra> Kamion: have a nice WE
[06:53] <Kamion> ogra: I like details ... :-)
[06:53] <ogra> goood :)
[06:53] <Kamion> mdz: apparently I shall be wearing a snake costume to a fancy dress party
[06:53] <ogra> lol
[06:54] <ogra> Kamion: please take pictures :)
[06:54] <thom> good weekend, folks
[06:54] <sivang> Kamion: nice, I'd join if I was in a walking distance :)
[06:54] <Kamion> my other half is in the process of buying an amelanistic corn snake
[06:54] <sivang> thom: c'ya!
[06:56] <sivang> Kamion: you should put some photos maybe ;q)
[07:10] <ogra> wow, how do i rape css: http://zengarden.20megsfree.com/
[07:10] <abelli> rape?
[07:10] <ogra> mistreat
[07:11] <__daniel> abuse :-)
[07:11] <ogra> yep
[07:11] <abelli> __daniel: yeah..
[07:11] <__daniel> ogra+__daniel: Online Thesaurus 1.0 (c) 2005 :-)
[07:11] <ogra> hehe
[07:12] <ogra> my cpu is pushed from 800Mhz to 1.6 Ghz by only displaying this page...woah
[07:12] <sivang> ogra: at least you got cpu scaling working :)
[07:13] <ogra> sivang: yup, worked from the beginning....
[07:13] <sivang> ogra: worked for me also in the beggining of hoary, not now though.
[07:13] <__daniel> sivang: which kernel do you have installed?
[07:14] <ogra> sivang: you fiddle too much with your system ;)
[07:14] <abelli> what does it mean if the installer throws a segfault?
[07:14] <abelli> :)
[07:15] <sivang> __daniel: 2.6.10-1
[07:17] <mako> smurfix: 
[07:18] <ogra> sivang: why not -2 ?
[07:18] <abelli> why not -hardened?
[07:18] <sivang> ogra: -2 sefgaulted amd tried to kill init :)
[07:19] <ogra> sivang: you definately fiddle too much with your system ;)
[07:19] <__daniel> sivang: that's bad because 2.6.10-2 made cupfreq_userspace working again :-)
[07:23] <sivang> __daniel: well, I'm upgrading the dellappi now , see where that gets me, unless I would open a bug.
[07:24] <__daniel> sivang: i'll have my fingers crossed :-)
[07:27] <sivang> __daniel: ok, MOTU also are you?
[07:28] <mdz> lamont-away: today's i386 live CD image is bigger; did you change parameters on the daily build?
[07:28] <__daniel> sivang: er... whats MOTU?
[07:28] <ogra> https://www.ubuntulinux.org/wiki/MOTU
[07:29] <ogra> sivang: __daniel will be a MOTU .....
[07:29] <ogra> sivang: i will care for that ;)
[07:29] <__daniel> sivang: i'd love to be... really, but i'm busy with my thesis at the moment
[07:29] <sivang> ogra: ah you two know each other?
[07:30] <__daniel> sivang: yes... you could say that :-)
[07:30] <ogra> sivang: we learned to know each other through ubuntu :)
[07:30] <sivang> ogra: eh nice :) please to meet you __daniel 
[07:30] <wasabi_> Hmm. I wanna be a MOTU. ;)
[07:30] <wasabi_> do they get t-shirts?
[07:31] <jdz_work> t-shirts! :D
[07:31] <__daniel> sivang: so until april/may i unfortunately wouldnt have the time to do the job properly
[07:31] <sivang> __daniel: no prob, not all of us can, work, other stuff etc.
[07:31] <__daniel> sivang: mvo_ lives 10 km away from me, too :-)
[07:31] <wasabi_> Actually. In all honesty... is anybody working on Java on Ubuntu?
[07:31] <wasabi_> or is it all just copied out of debian
[07:32] <ogra> sivang: nah, there are only 15000 packages....
[07:32] <ogra> ;)
[07:32] <sivang> wasabi_: Jeff Bailly is, IIRC
[07:32] <mdz> wasabi: jbailey is working on that for the hoary release
[07:32] <__daniel> sivang: so this area is properly ubunt-ized :-)
[07:32] <mdz> wasabi: see JavaIntegration in the wiki
[07:33] <ogra> mdz: on todays livecd the desktop worked like a charm.....(amd64)
[07:34] <sivang> mdz: anything sorted out with the usb bug?
[07:34] <eruin> does the ubuntu 2.6.9-kernels have fixes for the ptrace bug that messes with cedega/wine/ copy protection=
[07:35] <wasabi_> Oh.
[07:35] <wasabi_> Hah.
[07:35] <wasabi_> odd.
[07:36] <wasabi_> jbailey: It's just come to my attention that you ARE the Java guy for Ubuntu. Want Eclipse? :)
[07:42] <__daniel> the term MOTU is just too funny :-)
[07:42] <zul> damn now im thinking he-man and she-ra
[07:43] <wasabi_> Whose the Master of the Multiverse?
[07:43] <__daniel> zul: that's what i thought too :-)
[07:44] <ogra> wasabi_: haggai is the master of the MOTUs (which would make him he-man i guess)
[07:44] <eruin> why is #5619 marked a duplicate of 5582 ?
[07:44] <eruin> fabbione? 
[07:44] <zul> __daniel: all we need is battlecat and we would be set
[07:45] <__daniel> zul: not quite sure man-at-arms would be :-)
[07:45] <__daniel> zul: not quite sure, who man-at-arms would be :-)
[07:45] <orro> heh
[07:47] <mdz> sivang: yes, I found the bug last night
[07:48] <mdz> sivang: colin wrote an initial patch this morning
[07:48] <mdz> and I am fixing it up now
[07:48] <sivang> mdz: ok, I'll resync/download at your will.
[07:53] <mdz> lamont-away: I'm going to need d-i builds a bit later
[07:54] <sivang> yay! new 2.6.10-2 doesn't segfault init anymore and works.
[07:56] <__daniel> sivang: woohoo! :-)
[08:09] <elmo> uh
[08:10] <elmo> mdz: you know how you approved the language packs - did you realise it was going to pull in about 40 random l10n packages with it?
[08:11] <elmo> mdz: people.ubuntu.com/~james/l10n.txt
[08:11] <mdz> elmo: yeah, those should all have been seeded anyway
[08:11] <mdz> they're either new, or ones we missed
[08:11] <elmo> ok
[08:12] <mdz> in fact we should probably remove the ones which are seeded, and let the language packs handle it through germinate going forward
[08:16] <__daniel> i don't get why i'd need myspeel-de-{de,at,ch}
[08:18] <__daniel> but good to have those metapackages anyway :-)
[08:30] <abelli> is openoffice2 avaible now?
[08:33] <rcaskey> mdz: how many more arrays are left until release?
[08:33] <rcaskey> is there a torrent?
[08:33] <mdz> rcaskey: http://www.ubuntulinux.org/wiki/HoaryReleaseSchedule
[08:33] <rcaskey> (anyone on the southern light rail ;)
[08:37] <rcaskey> mdz: did the arrays get started later than planned?
[08:46] <sivang> ogra: ergh, updated to 2.6.10-2 newone, boots the syste, but now no sound and cpufreq applet still b0rked
[08:47] <ogra> hmm :(
[08:47] <__daniel> sivang: what does dmesg say when you modprobe cpufreq_userspace or the appropriate module for your soundcard
[08:47] <sivang> ogra: I should put some more time into investigatin it when I am less busy..
[08:48] <sivang> __daniel: the machine is busy doing something else at the moment, I will run this when I get a hold of it agian.
[08:48] <ogra> sivang: did you make any changes to your grub kernel line ?
[08:48] <sivang> ogra: didn't touch it
[08:48] <ogra> sivang: never ?
[08:50] <sivang> ogra: hhm, sudden;y I  not sure , I may have added something for the PCI collision workaround..
[08:50] <ogra> sivang: thats what i thought when i read dell....
[08:51] <ogra> sivang: if you added it dirctly to the kernel line it is overwritten on upgrade
[08:51] <jbailey> wasabi_: *lol*
[08:51] <ogra>  /is/was 
[08:51] <sladen> sivang: hunt around in /proc/cpufreq the display should be independent to the control
[08:52] <jbailey> wasabi_: Let's talk about it. =)
[08:52] <ogra> sladen: great to see you, could you add some info here: https://www.ubuntulinux.org/wiki/MOTU
[08:52] <abelli> can i install OOo2?
[08:53] <wasabi_> jbailey:  =)
[08:53] <wasabi_> jbailey: my mission right now is to get most of the java stuff on debian that's stuck in contrib for some reason into main.
[08:54] <wasabi_> Mostly so it makes it into Ubuntu for myself.
[08:54] <wasabi_> So, I think our goals merge.
[08:54] <jbailey> wasabi_: Yeah.  Are you one of the authors of the JavaInMain on wiki.debian.org?
[08:54] <wasabi_> no.
[08:54] <wasabi_> but, im one of the people working on it. :0
[08:54] <jbailey> =)
[08:54] <wasabi_> I dont know how out of date that page is
[08:55] <wasabi_> We've got ONE package to get into main before Eclipse hits it.
[08:55] <wasabi_> Lucene.
[08:55] <mdz> init is the biggest pain in the ass to debug
[08:55] <jbailey> Reasonably, but the sense of the page is still pretty good.
[08:55] <ogra> mdz: rewrite it :)
[08:56] <mdz> ogra: it's not the implementation
[08:56] <ogra> ouch
[08:56] <mdz> but thanks for the helpful suggestion :-p
[08:56] <ogra> hehe
[08:56] <ogra> mark would kiss you for a python replacement
[08:57] <__daniel> python replacement of what?
[08:57] <ogra> init
[08:57] <ogra> *g*
[08:58] <wasabi_> jbailey: when Eclipse hits main, that will give us eclipse-javac. It's a 100% complete javac, that's free enough.
[08:58] <wasabi_> So, we can compile anything at that point.
[08:58] <mdz> you can't trace it, you can't debug it, you need to replace it and reboot in order to test a change, and if you have a bug, the system halts
[08:59] <wasabi_> http://jack.feedbackplusinc.com/~jhaltom/eclipse    I need testing of those packages. ;0
[09:09] <sivang> wasabi_: whatever you do, make sure you include a simple DOM parser that's easier to learn then Xerces ;-p
[09:09] <sivang> and subclass..
[09:20] <Mitario> jo all
[09:20] <Mitario> jdub, here?
[09:30] <amu> Kamion: your image is finished? 
[09:32] <ogra> amu: Kamion is away for the weekend
[09:34] <amu> weekend: command not found ;)
[09:35] <ogra> heh
[09:36] <ogra> for me its... weekend, finally i can do the serious work....
[09:40] <ogra> mdz: probably a vm would be helpful, i think xen has something like that
[09:40] <ogra> debugging that is
[09:40] <mdz> not really
[09:40] <sivang> ogra: I wish QEMU would be a bit faster, then I could be using it more efficiently for cross platform needs :)
[09:42] <ogra> ah, only kernel debugging it seems....http://www.cl.cam.ac.uk/Research/SRG/netos/xen/readmes/user/#SECTION03340000000000000000
[09:42] <T-Bone> mako: ping?
[09:47] <sladen> keybuk: see #ubuntu if you're around
[09:49] <wasabi_> ubuntu cd's should come with a friendship bracelet.
[09:50] <ogra> wasabi_: great idea, start wattling ;-P
[09:51] <amu> wasabi_: ;)
[09:52] <__daniel> wasabi_: in germany we make fun of the right one of these: http://www.fernando-express.com/Petri_01.jpg
[09:52] <__daniel> wasabi_: at least 2-3 years ago :-)
[09:52] <ogra> __daniel: LOL
[09:53] <__daniel> ;-)
[09:53] <amu> __daniel: heheheh
[09:53] <wasabi_> that guy, with that hair, has more friends than me. =(
[09:54] <__daniel> wasabi_: but i guess he's not on any gpg-keyrings :-)
[09:54] <seb128> thom: here ?
[09:54] <__daniel> i thought that was the modern friendship bracelet :-)
[09:56] <ogra> __daniel: in this case ^^^, the new hoary cd will obviously ship one :)
[09:57] <__daniel> i'll wear it too, i promise ;-)
[09:57] <ogra> *g*
[09:58] <jbailey> What's the magic command to tell cdebootstrap to build a hoary chroot?
[09:59] <ogra> https://www.ubuntulinux.org/wiki/DebootstrapChroot
[10:02] <jbailey> Ah, debootstrap, not cdebootstrap.  That'd be why google didn't have it. =)
[10:03] <jbailey> ogra: Thanks! =)
[10:05] <sivang> ogra: if I already have hoary, do I need to create a chroot env? only for testing breakage stuff i,e ?
[10:07] <ogra> sivang: if you dont want your system polluted with unecessary packages a chroot is very helpful
[10:07] <sivang> ogra: k, that what I though :)
[10:07] <sivang> ogra: tnx
[10:07] <ogra> :)
[10:08] <sivang> ogra: strange, I get E: No such script: http://archive.ubuntu.com/ubuntu/
[10:08] <mdz> sivang: you gave the wrong arguments
[10:10] <sivang> mdz: used riddle's args from the wiki...
[10:10] <sivang> mdz: # udo debootstrap --variant=buildd hoary chroot/ http://archive.ubuntu.com/ubuntu/
[10:11] <ogra> sivang: after making the dir ?
[10:12] <mdz> you don't need to make the dir
[10:12] <mdz> remove the trailing slash from "chroot/"
[10:12] <sivang> mdz: k, thanks alot.
[10:13] <sivang> mdz: same, I think it can't find archive.ubuntu.com or something..
[10:15] <ogra> sivang: i just tried it again, even with the / it works here
[10:16] <sivang> ogra: could you msg me the line you executed?
[10:16] <mdz> you have an extra space somewhere, or something like that
[10:16] <mdz> you're passing 4 non-option arguments to debootstrap, instead of 3
[10:16] <ogra> sivang: i copy and pasted:  sudo debootstrap --variant=buildd hoary chroot/ http://archive.ubuntu.com/ubuntu/
[10:18] <sivang> ogra:  sudo debootstrap --variant=buildd hoary chroot/
[10:18] <sivang> I: Retrieving debootstrap.invalid_dists_hoary_Release
[10:18] <sivang> E: Failed getting release file http://ftp.debian.org/debian/dists/hoary/Release
[10:18] <ogra> sivang: now you are missing the archive ;)
[10:19] <ogra> my IRC line is wrapped....
[10:21] <sivang> ogra: working.thanks
[10:22] <ogra> hmm, it might be because the wiki adds this odd little globe graphics in the line....did you copy and paste from the wikipage before ?
[10:25] <ogra> i changed the layout to a better copy and pasteable format
[10:26] <ajmitch> I should setup a hoary chroot/pbuilder setup on my sid box
[10:27] <ajmitch> there was something on the wiki about that somewhere
[10:27] <ajmitch> can't find it now :)
[10:27] <sivang> ogra: I: Base system installed successfully.
[10:27] <ogra> yup
[10:27] <ogra> ajmitch: https://www.ubuntulinux.org/wiki/DebootstrapChroot
[10:28] <ajmitch> ogra: I'm sure there were other instructions along with that
[10:28] <ogra> https://www.ubuntulinux.org/wiki/DeveloperResources
[10:29] <ajmitch> and I was wanting something for sid
[10:29] <ajmitch> which is what the wiki had, iirc
[10:29] <ajmitch> the debootstrap line requires the hoary debootstrap scripts to be installed
[10:31] <Nafallo> ajmitch: those are copyable ;-)
[10:31] <ogra> ajmitch: there is another page, youre right:http://www.dina.dk/~abraham/religion/vi-music
[10:31] <ajmitch> the use of apt-proxy or similar will have to be mandatory for me :)
[10:31] <ogra> oops, wrong url
[10:31] <ajmitch> heh
[10:31] <ogra> https://www.ubuntulinux.org/wiki/ArchitectureBootstrapping
[10:32] <sivang> ogra: gcc-3.3 ??
[10:32] <sivang> ogra: in the chroot 
[10:32] <ajmitch> now that looks nice & evil;
[10:35] <ogra> ajmitch: but i think dropping the archive url in the end of the line and changing hoary to sid sould work too
[10:35] <ogra> from the Debootstrap wiki page ....
[10:35] <ajmitch> I've already got a pbuilder chroot setup for building sid packages
[10:36] <ajmitch> so I guess I can get another one setup for warty & then upgrade that to hoary or something :)
[10:40] <moquist_> sivang: hi!  I see you said hi a long time ago.  :)
[10:41] <__daniel> brb
[10:45] <rcaskey> ogra: is warty going to ship with esd as default?
[10:47] <wasabi_> warty shipped awhile ago. ;)
[10:48] <rcaskey> err hoary
[10:48] <rcaskey> it's 4:48 here, you know how it is ;)
[10:50] <T-Bone> fabbione: ping?
[10:51] <thully> hi - how are the dailys looking today?  are they broken in any severe way?  is it advisable to use these or array 2?
[10:51] <rcaskey> array3 is out
[10:52] <thully> that's what I meant - array 3 - my bad
[10:52] <rcaskey> dunno then
[10:52] <rcaskey> btw thully, can you just substitute in array3 or array for hoary?
[10:53] <thully> ? - array 2 and array 3 are milestone snapshots of hoary
[10:54] <rcaskey> yeah, I was wondering if there was a way to track the milestones
[10:55] <thully> no - you track hoary - the milestones are just cleaned-up daily builds for installation
[10:56] <rcaskey> yeah, but they are cleaned up ;)
[10:57] <T-Bone> lamont_r: ping?
[10:58] <lamont_r> ack
[10:58] <rcaskey> usplash debs are due next week right?
[10:59] <lamont_r> Kamion: you still around?
[10:59] <lamont_r> amu?
[10:59] <amu> yep?
[10:59] <lamont_r> how was the amd64 livecd?
[10:59] <amu> lamont_r: Kamion left, i didnt got an image :( 
[11:00] <lamont_r> curses.  I'm competing with another downloader.  only 70Kbytes/sec
[11:00] <lamont_r> oh
[11:00] <lamont_r> bummer
[11:00] <lamont_r> mdz about?
[11:00] <mdz> yep
[11:00] <rcaskey> lamont: isn't that linked to in osnews' story
[11:00] <mdz> beating on #5701
[11:01] <thully> how good are today's live CDs for i386?
[11:01] <thully> I'll download today's build if it doesn't sound like it's extremely broken today
[11:02] <amu> lamont_r: there's still time for a new download :) 
[11:02] <lamont_r> mdz: wanna kick a livecd build for us?
[11:02] <lamont_r> latest cloop image rsync'ed 17MB, with 10 packages diff in the manifest...
[11:03] <mdz> no shit?
[11:03] <lamont_r> not sure we can get much better than that without dinking around with timestamps in the fsimage
[11:03] <mdz> how did you manage that?
[11:03] <mdz> did you do the zlib thing?
[11:03] <thully> there won't be any new builds in the next hour, right (if there will be, I'll hold off on burning the current one)
[11:04] <mdz> lamont_r: sure, I can do one right now
[11:04] <mdz> lamont_r: a bit later, when I'm finished with busybox, I'll need a set of d-i builds, followed by live CD builds (but no new cloops)
[11:04] <lamont_r> woot
[11:04] <wasabi_> Is anybody here named Celso Pinto? 
[11:05] <thully> lamont_r - are you talking about live or installer builds?  because if there is a new build coming I'll hold off on burning the current build
[11:05] <lamont_r> mdz: funny thing was that the 4k:4k version was 127MB of rsync,while te 1024:64k was 17MB...
[11:05] <lamont_r> livecd
[11:06] <lamont_r> thully: although mdz is going to burn a new install cd in a while as well.
[11:06] <thully> will it be ready in next 45min-1hr?
[11:07] <thully> I have to leave for home soon, and I only have dial-up there (here I can nab an ISO faster than a kernel .deb at home)
[11:07] <mdz> should be ready in <10
[11:08] <thully> I heard a timezone fix should be in - and want to test it
[11:09] <mdz> thully: a time zone fix would probably require a new installer build, which I don't think has happened yet.  did Kamion tell you that?
[11:10] <thully> the fix is in base-config - check bug 2416 for details
[11:10] <mdz> lamont_r, thully, sivang: new live CD builds are up
[11:10] <mdz> 20050121.1
[11:10] <mdz> ah, ok, if it's in base-config, it only requires a new cloop image from lamont
[11:10] <mdz> which I think he made recently
[11:11] <thully> I'm talking about installer cd for the timezone fix
[11:11] <lamont_r> Setting up base-config (2.61ubuntu13) ...
[11:11] <lamont_r> is the last one I built, some 4-5 hours ago
[11:11] <lamont_r> or rather, is what's in the last cloop fs that I built
[11:11] <T-Bone> lamont_r: has the "libunwind" bug been fixed in that one?
[11:11] <thully> yep - that's the one that fixes it - is this on the installer build?
[11:11] <mdz> hoary-install-i386.list:/pool/main/b/base-config/base-config_2.61ubuntu12_all.deb
[11:12] <lamont_r> T-Bone: should almost certainly be...
[11:12] <mdz> ^^ that's what the latest install CD has
[11:12] <T-Bone> lamont_r: gonna test then.
[11:12] <thully> no, then
[11:12] <T-Bone> Kamion: ping?
[11:12] <thully> ubuntu13 has fix
[11:12] <lamont_r> mdz: btw, there's a risk to amd64 livecd because of partimage...
[11:13] <lamont_r> partimage is PaS'ed on ia64,alpha right now (debian), with a comment about it being b0rked on 64-bit....
[11:13] <lamont_r> so either ia64 should try it again, or amd64 is b0rked.
[11:23] <Mithrandir> is there something _really_ fucked with archive.u.c atm?  I'm getting approximately 5k/sec from it.
[11:25] <mdz> Mithrandir: maybe the cd images mirroring or something?
[11:25] <sivang> mdz: what's the rsync url? (i recall it differs form the vhost of the website)
[11:25] <mdz> the I/O on some of those servers is horrific
[11:26] <Mithrandir> mdz: I don't know, it
[11:26] <mdz> sivang: rsync://cdimage.ubuntu.com/cdimage/daily-live/...
[11:26] <sivang> mdz: thank you
[11:26] <Mithrandir> 's just _dog_ slow and wondered if anybody else saw it
[11:26] <sivang> Mithrandir: the archive?
[11:26] <lamont_r> although it's maxed out on rsync clients right now...
[11:26] <lamont_r> dammit
[11:26] <sivang> Mithrandir: I updatted my chroot, was a _pain_
[11:26] <lamont_r> OTOH, I know that it won't rsync for crap this download...
[11:26] <Mithrandir> sivang: I'm running dist-upgrade and getting 8k/sec now, yes.
[11:27] <sivang> Mithrandir: got B/s at the end...
[11:27] <sivang> :-/
[11:27] <T-Bone> Mithrandir: you're sorta lucky, i just timeouted on archive.u.c
[11:27] <Mithrandir> ok, so it's not just me.
[11:28] <thully> I've got a snails-pace download also 
[11:28] <maswan> Well, good thing there are mirrors, bad thing that the mirrors are not perfectly up to date?
[11:28] <T-Bone> this is just happening. I downloaded 200M 20mn ago at 540kB/s avg
[11:28] <Mithrandir> ooh, it's 10k/sec now.
[11:29] <sivang> lamont_r: you're not rsyncing?
[11:30] <sivang> @ERROR: max connections (15) reached - try again later
[11:30] <sivang> ergh
[11:31] <lamont_r> yeah - that'd be why sivang 
[11:31] <sivang> ok, then wget it is..
[11:32] <lamont_r> mdz: I don't suppose the livecd is bt-able, eh?
[11:33] <amu> lamont_r: i think, it make no sense atm 
[11:33] <sivang> 8.57K/s erg..
[11:34] <mdz> lamont_r: ask thom
[11:36] <sivang> 1.64K/s
[11:36] <lamont_r> I was getting 60K/s for a several seconds there...  now1.8
[11:38] <wasabi_> haha i had the same problem.
[11:38] <wasabi_> i was trying to figure out if our inet was messing up
[11:40] <lamont_r> hell - I could fetch it at home this quick...
[11:42] <Mithrandir> it should probably go on another machine?
[11:42] <Mithrandir> or be torrented?
[11:42] <lamont_r> it should tget torrented
[11:43] <lamont_r> but, ENOTHOM
[11:43] <T-Bone> heh
[11:43] <sivang> 41.09K/s
[11:43] <sivang> starting to rise up..
[11:43] <lamont_r> bouncing between 15 and 70
[11:44] <lamont_r> wow! it actually said 113K/s there for an instant. :-)
[11:48] <mdz> that was yesterday
[11:49] <Mithrandir> it's faster again now
[11:49] <lamont_r> mdz: to late, dude. :-)
[11:49] <lamont_r> yeah, much faster
[11:49] <lamont_r> wot
[11:49] <lamont_r> what you hitting ctl characters for anyway??? :-)
[11:50] <T-Bone> lamont_r: get back to the beginning of the line, as it should: ctrl-a / ctrl-e :P
[11:51] <lamont_r> T-Bone: see - that's why I don't use emacs
[11:51] <T-Bone> lamont_r: neither do I. I *hate* emacs :^)
[11:51] <T-Bone> lamont_r: pretty much as I *hate* perl :)
[11:51] <T-Bone> lamont_r: gives you the picture, heh? :^)
[11:51] <elmo> gtk-key-theme-name = "Emacs"
[11:52] <elmo> T-Bone: ^-- add that to ~/.gtkrc-2.0
[11:53] <T-Bone> doesn't work
[11:53] <elmo> sure it does
[11:54] <mdz> lamont_r: what was the secret to getting the ext2 image more rsyncable?
[11:54] <T-Bone> elmo: what it's supposed to do?
[11:54] <mdz> lamont_r: was zeroing the empty blocks with partimage the ticket?
[11:54] <lamont_r> mdz: amazing foo, thanks
[11:54] <lamont_r> mdz: partimage
[11:54] <mdz> neat
[11:54] <lamont_r> well, that and rsyncing over the top of the previous image
[11:54] <mdz> and cloop took care of itself, I guess
[11:55] <lamont_r> which either (a) necessitates partimage, or (b) makes the cloop grow over time
[11:55] <lamont_r> yeah
[11:55] <mdz> are you doing the rsync-inplace-from-previous-image trick too?
[11:55] <lamont_r> had to
[11:55] <mdz> ah, yes
[11:55] <lamont_r> that was the real trick.  partimage just cleans up the trash
[11:55] <lamont_r> Mithrandir: you gonna grab one of them thar new amd64 livecd images and tell me if it even boots?
[11:56] <Mithrandir> lamont_r: sure, I can do that
[11:57] <mdz> lamont_r: locale generation looks good now
[11:57] <sivang> 185.83K/s ah back up!
[11:57] <mdz> speeds up the boot by several seconds
[11:58] <sivang> mdz: it didn't general he_IL.UTF8 when I tested it llast time..
[11:58] <sivang> generate even
[11:58] <lamont_r> mdz: that'd be because it actually _runs_ locale-gen
[11:58] <mdz> sivang: I find that difficult to believe
[11:58] <mdz> lamont and I are talking about the pre-generated locales
[11:58] <mdz> but generating them on-demand has always worked
[11:59] <lamont_r> sivang: it pre-generates en_{US,GB,ZA},UTF-8
[11:59] <sivang> mdz: just before preparing the live session, well, I only chose "hebrew" as the language in the d-i menu.
[11:59] <sivang> lamont_r: maybe something else is needed to trigger generating other locales?
[11:59] <sivang> I mean, on the user side.
[11:59] <mdz> sivang: did you actually see that the locale wasn't generated, or are you assuming?
[12:00] <mdz> perhaps something else went wrong
[12:00] <mdz> was the locale set in /etc/environment properly?
[12:00] <sivang> mdz: I watched the "generating localeles" message on the debug console or the other consoel that allows me to see what's happening while "preparing for live session".
[12:00] <sivang> mdz: I will check that with the new iso.
[12:00] <lamont_r> mdz: the pre-gen populates /etc/locale-gen (or such) before the install, and runs locale-gen after- dunno if that's affecting the postinst..
[12:00] <mdz> sivang: and it said "Generating locales..." and your locale didn't appear in the list?
[12:01] <mdz> lamont_r: d-i goes in and manhandles it anyway
[12:01] <sivang> mdz: nope
[12:02] <mdz> getting 3MB/sec on amd64 iso download
[12:02] <lamont_r> back to 100K/s here - must be someone else downloading