[12:46] <ezequiel> hi ppl
[12:46] <ezequiel> how can I create a .deb?
[12:46] <Amaranth> ezequiel: google for the debian new maintainer's guide
[12:47] <ezequiel> Amaranth, thanks
[12:50] <ezequiel> 'cause I'd like to have my own anjuta2 deb =) long life anjuta2!
[12:51] <ezequiel> (I still remember when it was just scaffold)
[01:11] <lifeless> jbailey: so, did it look ok ?
[01:12] <lamont__> what is it that calls ifrename?
[01:13] <Mithrandir> lamont__: hotplug
[01:13] <Mithrandir> /etc/hotplug/net.agent, it seems
[01:13] <lamont__> yeah - just finally tracked that down...
[01:14] <lifeless> when it misbehaves its a biatch
[01:14] <lamont__> didn't have the USB nic in /etc/iftab - hence the pain
[01:15] <tseng> bye Mithrandir 
[01:18] <allee> allee: udu.wiki: LaptopHardwareSupport:  Special Keys: Found no infos about add/help 'preconfiguring' them out-of-the-box for breezy
[01:19] <allee> allee: How can on join/help on this topic?
[01:19] <lifeless> does ubuntu watch wired ethernet for cable-insertion by default ?
[01:19] <mjg59> allee: I'm working on that at the moment
[01:20] <mjg59> With luck, there should be some support in Breezy next week
[01:20] <allee> lifeless: no
[01:20] <lifeless> allee: thanks
[01:20] <mjg59> lifeless: That's part of NetworkMagic (which doesn't actually exist yet)
[01:20] <allee> mjg59: great
[01:20] <lifeless> mjg59: yeah, I was at the bof ;) can't blame me for hoping ;(
[01:20] <Nafallo> license issues :-/
[01:20] <allee> mjg59: via which pkgs will add the support so I can check them (I know several ways to do it ;)
[01:21] <allee> mjg59: and contribute of course
[01:21] <mjg59> allee: It'll probably be by adding a new package, to begin with
[01:22] <allee> mjg59: care to sketch how it's planed to work?  Just to get the idea about the direction and used route
[01:22] <lifeless> mjg59: I can't login to ubuntulinux.org at the moment
[01:23] <lifeless> mjg59: is there somewhere else I can put the keycodes ?
[01:23] <Amaranth> hrm, i wonder where johndong is
[01:23] <allee> lifeless: oh, I did I miss a page for keycodes :(  URL?
[01:24] <lifeless> allee: I promised to make one
[01:24] <lifeless> but nologin == no keycodes
[01:24] <allee> LinuxJones: k.  And I promise to add data when done
[01:25] <allee> s/LinuxJones/lifeless/.   <Tab> got me :(
[01:25] <lifeless> heh~
[01:27] <mjg59> allee: Effectively the following:
[01:27] <mjg59> allee: Bind machine scancodes to generic keycodes (using DMI to tell which set of scancodes to use)
[01:27] <mjg59> allee: Configure those keycodes to do the right thing in X
[01:28] <mjg59> allee: And for machines which generate ACPI events on hotkeys, generate a fake X keycode
[01:29] <lifeless> mjg59: so ..
[01:29] <lifeless> mjg59: where should I list the codes as I can't wiki them 
[01:29] <lifeless> udu wiki ? 
[01:30] <jbailey> lifeless: Sorry, I got distracted.  Where is the package? 
[01:30] <mjg59> lifeless: LaptopKeycodes
[01:30] <mjg59> Oh, argh. Wiki is broken?
[01:30] <lifeless> my login is
[01:30] <mjg59> Gah
[01:31] <mjg59> Just stick them on a website somewhere, we'll add them later
[01:31] <mjg59> Dell seem to use the same scancodes as Microsoft, which makes things easier
[01:31] <lifeless> jbailey: 
[01:31] <lifeless> http://www.cundal.net/debian/
[01:31] <pitti> night
[01:31] <Nafallo> pitti: night :-)
[01:31] <Nafallo> dooh!
[01:32] <allee> mjg59: DMI?  I'll try with dmidecode.  I'm not sure I find stuff like 'XFAudioRaise' (or so) there
[01:33] <mjg59> allee: No, you won't
[01:33] <lifeless> mjg59: people.ubuntu.com/~robertc/codes.txt
[01:34] <mjg59> allee: That's not the issue. We need to know what pressing the key does. Once we know that, we can use the DMI information to know which set of scancodes to use on an arbitrary machine.
[01:34] <mjg59> Once we have the scancodes, everything else is easy
[01:34] <lifeless> mjg59: what dmi info do you need to id the dell ?
[01:34] <mjg59> lifeless: Vendor name, model name
[01:34] <lifeless> I'll put it in that file
[01:34] <mjg59> (form DMI)
[01:35] <lifeless> this foo:
[01:35] <lifeless> Handle 0x0100
[01:35] <lifeless>         DMI type 1, 25 bytes.
[01:35] <lifeless>         System Information
[01:35] <lifeless>                 Manufacturer: Dell Inc.
[01:35] <lifeless>                 Product Name: Latitude X1                     
[01:35] <lifeless>                 Version: Not Specified
[01:35] <lifeless>                 Serial Number: 37VCF1S
[01:35] <mjg59> Yeah, that stuff
[01:35] <lifeless>                 UUID: 44454C4C-3700-1056-8043-B3C04F463153
[01:35] <lifeless>                 Wake-up Type: Power Switch
[01:35] <mjg59> But not the serial number
[01:36] <lifeless> doh
[01:36] <lifeless> well thats mine!
[01:36] <lifeless> and its public record now :)
[01:36] <Nafallo> hehe
[01:36] <mjg59> (One of the things about the hardware database spec was working out what information we didn't want to send...)
[01:37] <lifeless> mjg59: ok, its all in codes.txt
[01:37] <lifeless> if someone could redact this log to strip that serial that would be nice.
[01:37] <lifeless> fabbione: ^^ ?
[01:37] <mjg59> lifeless: Ta
[01:37] <lifeless> though I'm sure there are 4000 other logs being made
[01:38] <lifeless> np
[01:38] <lifeless> pure selfinterest here
[01:38] <jbailey> lifeless: Does this guy IRC?  I would usually have the person do some cleanups before I sponsor.
[01:39] <lifeless> you just Know that when I start using breezy I'll be looking for this to JustWork ;)
[01:39] <lifeless> jbailey: dunno, though he is responsive on email
[01:39] <lifeless> Brett Cundal <bcundal@cundal.net>
[01:39] <allee> mjg59: Sorry now I'm confused. "We need to know what pressing the key does" for me means: check showkeys output for scancode and label it with 'logical description".  How is here dmi invovled?
[01:39] <lifeless> his package has been uploaded many times, just his regular sponsor has gone busy/awol
[01:39] <jbailey> lifeless: Did he ask you to sponsor this, or are we just snarfing it off his site?
[01:40] <lifeless> so I wouldn't anticipate significant issues with the packaging itself.
[01:40] <mjg59> allee: On system boot, we need to use the DMI information to work out which scancodes to use on a given machine
[01:40] <lifeless> He asked me to help him find a sponsor
[01:40] <lifeless> he asked on debian-mentors *months* ago
[01:40] <mjg59> Dells use one set of scancodes. HP use another. Acer use a third.
[01:41] <lifeless> jbailey: quoting: Sure... I just need a sponsor to upload the package. The package needs
[01:41] <lifeless> to be looked at for packaging errors and whatnot, to see whether it's
[01:41] <lifeless> suitable for inclusion in Debian... Since it's already in unstable and
[01:41] <lifeless> the changes are fairly minor since the last upload, I would expect it
[01:41] <lifeless> to be not much work. 
[01:42] <allee> mjg59: Ah so DMI is only used to find out producer/model and that get's acociated with a scancodes config-file
[01:42] <mjg59> allee: Exactly
[01:42] <jbailey> lifeless: So far not much packaging work.  He has example scripts lying around that he doesn't need.  He renames the packaging without really saying why he does it.
[01:42] <mjg59> Once we have the scancodes and the dmi information, we can ensure that on boot that stuff works properly
[01:43] <lamont__> mjg59: any ideas on how well suspend to {disk,ram} works on the HP NC6000?
[01:43] <lifeless> jbailey: ok. want me to give it a once over first ? (I am interested in getting gst updated which is why my fingers are involved)
[01:44] <jbailey> Nah, I've already gone over the packaging, verified that it's the same as upstream source, yada yada.  Now I'm just seeing if it builds. =)
[01:44] <lifeless> ;)
[01:44] <jbailey> lifeless: My Debian machine is an ia64. =)  It's currently FTBFS on that arch...
[01:44] <lifeless> cool!
[01:44] <lifeless> jbailey: 2.1.10 has updates that Paolo said should fix at the time
[01:45] <lifeless> the FTBS is on 2.1.8 which June 2004
[01:45] <lifeless> .. thus the desire to get 2.1.10 uploaded
[01:45] <lamont__> eep.  off to training.  back in a few hours
[01:45] <lifeless> jbailey: if it doesn't work, I'd love accesss to that machine to fix it ;)
[01:46] <thom> if you need it i'll kick my ia64 bgack on to ipv6 and you can abuse that...
[01:46] <lifeless> thom: thanks!
[01:47] <mjg59> lamont__: Should be fine, but you need to blacklist i2c-i801
[01:47] <mkde> ooh
[01:47] <mjg59> (not sure why)
[01:47] <mkde> hi mjg59, i was saying the other day about some of my laptop keys not working, you said to look in dmesg, i didn't find em
[01:49] <mjg59> mkde: What sort of laptop?
[01:49] <mkde> mjg59, its a compaq presario 2100
[01:49] <mjg59> mkde: Ok. It's possible that it needs some magic.
[01:50] <mkde> mjg59, i'm all ears
[01:50] <mkde> mjg59, i feel sure that once upon a time the mute button used to work with linux
[01:51] <mjg59> mkde: I'm afraid I don't have any old Compaq hardware, and it's basically been killed since the HP merger, so it's hard to get test kit
[01:51] <mkde> mjg59, ok
[01:52] <mkde> mjg59, is it likely to be a kernel or a xorg problem?
[01:52] <mjg59> mkde: Kernel
[01:52] <mjg59> mkde: Well, kernel-ish. X knows nothing about the keyboard.
[01:52] <allee> mkde: Mute:  what does not work:  sound still on or xev show nothing/no keysym or showkeys gives nothing?
[01:53] <mkde> nothing on xev
[01:53] <mjg59> If pressing it shows nothing in xev and nothing in the kernel dmesg, then something needs to be done in the kernel
[01:53] <mkde> to be fair, even the hotkeys that work don't come up in dmesg
[01:53] <mkde> its strange that some work and others not
[01:53] <allee> mkde: and on console:  gives showkeys any output?
[01:54] <mkde> allee, what should I do to test?
[01:54] <mkde> mjg59, so is it worth filing a bug upstream or just leave it you think?
[01:54] <allee> mkde: ALT-CRTL-F1  login and run showkeys, then press not-working keys
[01:55] <mjg59> allee: It's not really a bug, it's a missing feature
[01:55] <mkde> is "showkey" the same thing?
[01:55] <mjg59> Without the hardware, it's hard to tell what's going on
[01:55] <mjg59> s/allee/mdke/
[01:55] <mkde> mjg59, ok
[01:55] <mkde> mjg59, if the hardware isn't made anymore i guess its not really worth us working on it
[01:55] <allee> mkde: ops, yes showkey
[01:56] <mkde> allee, gives nothing
[01:56] <mjg59> mdke: Well, it's something I'll look at if I can find any information about it, or get hold of some of the hardware
[01:57] <mjg59> But at the moment, I'm concentrating on current stuff
[01:57] <allee> mkde: now I'm stuck :(
[01:57] <mkde> mjg59, thats fine
[01:57] <mkde> mjg59, if you ever have a old hardware fest I can bring this laptop along ;)
[01:57] <allee> mkde: did you check linux-for-laptops archive for infos about keys for your model already?
[01:58] <mkde> allee, will do now
[01:58] <mkde> allee, linux-laptop.net?
[01:59] <allee> mkde: yes and http://tuxmobil.org/mylaptops.html
[02:00] <mkde> allee, yeah, this confirms my problem http://slist.lilotux.net/laptop/
[02:01] <mkde> no biggie
[02:02] <mkde> just in case there was a quick fix or i could report it somewhere
[02:02] <mkde> the laptop works fine otherwise
[02:14] <allee> mjg59, lifeless: what's the best way to get notified about changes in this area.  So I can jump in whenever I can contribute?
[02:15] <mjg59> allee: At the moment, probably just to monitor the ubuntu-devel mailing list
[02:15] <mjg59> I'll send something there once some code exists
[02:16] <zul> mjg59: what kind of laptops are you guys going to be sending to testers?
[02:16] <zul> im just curious
[02:16] <mjg59> zul: IBM, Toshiba, Dell, possibly HP, possibly Apple, possibly Sony
[02:16] <zul> cool
[02:17] <mjg59> Mostly looking at the business end of the market at this point, then shifting down to the consumer end in future
[02:17] <zul> nice
[02:17] <allee> mjg59: uhm, well. 'nother list broad topic list ;)  Nevertheless I'll do.  Thx!
[02:18] <mdz> ogra: aw, what happened to the xscreensaver prettiness?
[02:18] <ogra> http://www.grawert.net/xss_mockup.png
[02:18] <ogra> ;)
[02:18] <ogra> working on it
[02:18] <mdz> infinity: ping
[02:18] <Amaranth> ogra: Change User?
[02:18] <ogra> yeah
[02:18] <Nafallo> ogra: wasn't you saying /users/ where to complain ;-)? not devs :-P.
[02:19] <Nafallo> heh
[02:19] <Amaranth> ogra: I'm saying it should say 'Change User' ;)
[02:19] <Nafallo> mjg59: any amd64 laptops?
[02:19] <ogra> probably "other user"....
[02:19] <Amaranth> the people in the background are spooky
[02:20] <ogra> hehe
[02:20] <mjg59> Nafallo: None of the big vendors have really gone with them yet, so that may be Breezy+1
[02:20] <ogra> that are the people behind ubuntu ;)
[02:20] <Nafallo> mjg59: hmm. how long are you going to send out laptops? ;-)
[02:20] <Amaranth> ogra: I'm saying the way they're presented are spooky.
[02:20] <ogra> heh...
[02:21] <mjg59> Nafallo: We'll be starting in a month or so, and we'll be doing it while we have money to spend on this project :)
[02:21] <mjg59> Though the budget is not that big
[02:21] <Amaranth> ogra: Either make them a bit more visible or drop them, I'd say.
[02:21] <Nafallo> mjg59: kewl :-)
[02:21] <mjg59> People will get to keep the machines if they file enough reports
[02:21] <ogra> Amaranth, nope
[02:21] <Amaranth> ogra: It totally creeps the hell out of my as is.
[02:22] <Amaranth> s/my/me/
[02:22] <allee> mjg59: no joke?  Testers can apply for a laptop?
[02:22] <Amaranth> ogra: Can you make one without the people, just for comparison?
[02:22] <ogra> nope
[02:22] <mjg59> allee: Yeah. Read the mailing list :)
[02:22] <Amaranth> ogra: What was the point of showing us then?
[02:23] <mjg59> It'll only be 10 or so people getting them, and it'll be based on the amount of community involvement
[02:23] <allee> mjg59: wasn't subcribed ;)
[02:23] <ogra> Amaranth, mdz asked ....
[02:23] <Amaranth> I wish I was a Member. :)
[02:23] <Amaranth> I'd apply for one of those, I like breaking things.
[02:23] <Amaranth> software, not hardware
[02:23] <allee> mjg59: ah.  No problem I've nough of (Dell) laptops around me ;)
[02:24] <lifeless> allee: I'm loving my X1 ;)
[02:25] <mjg59> The Dell X series is all rebadged Samsung stuff
[02:25] <mjg59> lifeless: Does it do suspend to RAM?
[02:25] <lifeless> yes
[02:25] <allee> lifeless: good too know now that the X300 is no longer produced.   How hardware support out-of-the-box?
[02:25] <lifeless> and hibernate
[02:25] <mjg59> lifeless: Hurrah. An improvement.
[02:25] <mjg59> The X300 needed DSDT hacking.
[02:25] <lifeless> doesn't bring the ipw2200 back correctly from hibernate, does from ram
[02:26] <allee> mjg59: I had no succes with X300 here (with patched DSDT) :(
[02:27] <KaiL> lifeless: unloading the module helps?
[02:27] <mjg59> allee: Should work with stock Hoary
[02:27] <mjg59> Chat to jdub about it at some point - it works for him
[02:27] <Nafallo> baah, my laptop doesn't hibernate at all :-P
[02:27] <Nafallo> it might have worked with hoary, but I'm not sure.
[02:28] <Nafallo> (before adding rt2500 wlandriver ;-))
[02:29] <lifeless> KaiL: oh yes, its trivial to fix
[02:29] <lifeless> KaiL: rmmod, modprobe, its back.,
[02:30] <KaiL> lifeless: maybe network modules should also be unloaded? They fail quite often...
[02:30] <Nafallo> do we kill hotplug? that will bring my wlan right back in ;-)
[02:30] <lifeless> KaiL: it is removed 
[02:30] <KaiL> "should always" I mean..
[02:31] <lifeless> KaiL: its not inserting properly straight after the from-hibernate resume
[02:32] <allee> mjg59: X300: fn-suspend -> dead (power led on, black screen, ping: unreachable etc)
[02:33] <mjg59> allee: Unf. Dunno.
[02:34] <KaiL> get better Laptops :)
[02:34] <mjg59> allee: Edit /etc/default/acpi-support and change USE_DPMS to false, then see if there's any text when it hangs
[02:34] <KaiL> Samsung P35 works perfect for example (except SDIO, but that's anothr story)
[02:35] <allee> KaiL: SDIO? what's that?
[02:35] <allee> mjg59: I'll try
[02:35] <KaiL> allee: Slot for SD-Cards
[02:37] <KaiL> allee: Linux currently only supports MMC-Cards (that's more or less a subset of that) and only with a Winbond chip, not the used Ricoh
[02:37] <allee> mjg59: ha, black screen instead of grub menu.  Great ;)
[02:37] <mjg59> allee: ?
[02:38] <bob2> Ricoh has support in 2.6.12, iirc
[02:38] <mjg59> bob2: No, just Winbond
[02:38] <allee> mjg59: powered off and and.  See the f2/f12 selection then screen is black no grub menu
[02:38] <mjg59> allee: ?
[02:39] <mjg59> allee: Suspend shouldn't power off
[02:39] <KaiL> the Winbond is already in 2.6.10 (wbsd module)
[02:39] <mjg59> Are you sure you're suspending and not hibernating?
[02:39] <allee> mjg59: remember: the x300 was dead. 
[02:39] <mjg59> allee: Dead how?
[02:40] <mjg59> allee: Oh, hung?
[02:40] <allee> [02:32]  <allee> mjg59: X300: fn-suspend -> dead (power led on, black screen, ping: unreachable etc)
[02:40] <mjg59> Remove the battery and hold the power button down for a while
[02:40] <Nafallo> gaah. I got a cheap_laptop(tm)
[02:41] <Nafallo> Manufacturer: ACTEBIS (what happened to Targa?) ;-)
[02:41] <KaiL> Nafallo: that doesn't mean anything. does it work?
[02:41] <allee> mjg59: thx, grub menu back  (nice experience)
[02:41] <Nafallo> KaiL: ofcourse ;-)
[02:42] <KaiL> so who cares....:)
[02:42] <KaiL> at least you don't need to fight with the pre-alpha drivers, windows users need to use :))
[02:43] <Nafallo> hehe
[02:43] <Riddell> mdz: do I still need to fix the permissions issue on the kubuntu seeds?
[02:43] <KaiL> btw. is there any hardware known not to work, except those SDIO stuff and some broken ACPI tables?
[02:43] <Nafallo> this amd64-laptop shipped with winxp home 32-bit for what it worth ;-)
[02:44] <mdz> Riddell: yes, please
[02:45] <allee> kail reminded me that it's late in europe ;)
[02:45] <allee> Nite
[02:46] <Riddell> mdz: hopefully that's it fixed now
[02:47] <mdz> Riddell: testing
[02:48] <mdz> Riddell: nope
[02:48] <mdz> drwxr-sr-x  3 jriddell warthogs 4096 Jun  3 01:46 /home/warthogs/archives/kubuntu-devel@lists.ubuntu.com/seeds/seeds--breezy/seeds--breezy--0/patch-6/++revision-lock/
[02:48] <mdz> Riddell: that stuff all needs to be group-writable; set your umask to 00x
[02:49] <mdz> and chmod -R g+w on the tree to fix what's already there
[02:50] <mdz> jbailey: regarding #11135, you said you had a patch ready for review?
[02:52] <KaiL> allee: good idea ;)
[02:53] <allee> KaiL: yeap
[02:59] <Riddell> mdz: do I change it on my local repository or on chinstrap?
[02:59] <mdz> Riddell: on chinstrap
[03:00] <Riddell> mdz: ah
[03:00] <mdz> chmod -R g+w  /home/warthogs/archives/kubuntu-devel@lists.ubuntu.com
[03:00] <mdz> to fix the immediate problem so that I can commit
[03:01] <Riddell> mdz: done
[03:01] <mdz> Riddell: much better, thanks
[03:02] <mdz> for the more permanent fix, make sure ~/.bashrc specifies umask 002
[03:02] <Riddell> mdz: do I have to set the umask somewhere so it stays like that?
[03:02] <Riddell> yep :)
[03:02] <mdz> though I thought that was the default
[03:03] <Riddell> mdz: it seems to be the default, but I had checked it out onto my laptop where the umask isn't set for group writable
[03:03] <mdz> oh, that might explain why this has happened to a few others as well
[03:03] <mdz> if the local permissions affect it
[03:03] <mdz> but I don't see why that would be
[03:04] <mdz> bob2: is that possible?
[03:04] <Riddell> I'll put a note on the seed page on the wiki
[03:06] <bob2> if you check it out, and the umask changes the permissions on-disk, they can get checked back in
[03:07] <mdz> bob2: in this case, the issue is permissions on files in the archive itself (like ++revision-lock)
[03:08] <infinity> mdz : pong.
[03:08] <mdz> infinity: can I get a retry of antlr?
[03:08] <mdz> it should succeed this time
[03:08] <Clint> mdz: acls can make those problems go away
[03:08] <infinity> mdz : Did the seeds get shuffled? :)
[03:08] <mdz> Clint: acls can replace those problems with, newer, harder to debug problems ;-)
[03:08] <mdz> infinity: seeds, versions, yeah
[03:09] <infinity> mdz : Alright, bouncing it.
[03:09] <mdz> I'm trying to unravel this chain of java stuff
[03:10] <mdz> if antlr builds, that should unblock gjdoc, which should unblock junit, which should unblock (I think) libant1.6-java, which unblocks the world
[03:10] <mdz> then we can move the lot into main and have free java apps in main
[03:11] <mdz> wasabi says eclipse is nearly there
[03:11] <wasabi> eclipse *is* there.
[03:11] <wasabi> or will be when the buildds get sorted out. ;)
[03:12] <wasabi> (and somebody uploads my java-common to main since I can't)
[03:12] <infinity> mdz : Actually, it looks like ecj-bootstrap binaries are stuck somewhere (queue/new?).  It's in state "Uploaded", not "Installed"... If it was in the archive, antlr would have built already (I set the right dep-waits last night)
[03:12] <wasabi> http://kyoto.larvalstage.net/~wasabi/java-common
[03:12] <wasabi> oh my apache is segfaulting. heh.
[03:15] <wasabi> http://akita.larvalstage.net/~wasabi/java-common
[03:16] <mdz> infinity: very weird
[03:16] <mdz> it's sitting in queue/unchecked
[03:16] <mdz> but queue/REPORT says it was accepted
[03:16] <mdz> ecj-bootstrap_3.0.1-3_i386.changes
[03:16] <mdz> ACCEPT
[03:16] <mdz> ecj-bootstrap_3.0.1-3_all.deb
[03:16] <mdz>   to pool/universe/e/ecj-bootstrap/ecj-bootstrap_3.0.1-3_all.deb
[03:16] <mdz> Accepting.
[03:17] <infinity> mdz : Neat.  (the one in unchecked may have just landed there, I reuploaded it thinking it may have got lost)
[03:17] <mdz> infinity: oh, so it did
[03:17] <mdz> no idea where the previous one went, but we'll see how this goes
[03:19] <infinity> Apparently, jackass has been acting up recently, and katie's been doing goofy things.  So, I'm not even willing to hazard a guess as to what might have gone wrong. :)
[03:19] <infinity> (A couple of source uploads seem to have been accepted, had mail go to -changes, then head off into NerverNeverLand too)
[03:19] <Traveler1> Does somebody knows where can I find some kind of documentation about multiseat package?
[03:20] <mdz> infinity: possibly a side effect of all my teri'ing lately
[03:21] <wasabi> mdz, do you mind uploading the java-common i posted?
[03:21] <mdz> infinity: is there anything else that needs attention in the libant1.6-java chain?  I want to move it, junit, libjaxp1.2-java, junit into main today
[03:23] <mdz> wasabi: done
[03:23] <wasabi> thanks you
[03:27] <mdz> daniels: what do you think is the best way to handle starting the X server on thin clients?  ltsp currently uses /etc/inittab, but I think a daemon with an init script would probably be more appropriate
[03:27] <mdz> daniels: can we (ab)use one of the display managers for this purpose?
[03:29] <infinity> mdz : Do we need to sync a new libant1.6-java from Debian still?  The one in the archive predates all this java buggery by several weeks.
[03:30] <wasabi> Nope.
[03:30] <wasabi> The ant in Ubuntu is radically different than the one in Debian.
[03:30] <wasabi> Just waiting for Sarge to be altered in Debian.
[03:31] <infinity> wasabi : Kay.  We've been pulling the other stuff from experimental, we don't want the experimental ant?
[03:31] <wasabi> Uh. Um....
[03:31] <wasabi> I can't say.
[03:31] <wasabi> I don't know if they merged my changes into there.
[03:31] <wasabi> Let me take a look.
[03:31] <wasabi> They're even differnet source packages.
[03:31] <infinity> Yay.
[03:32] <wasabi> 'ant' in ubuntu produces libant1.6-java, 'libant1.6-java' in Debian produces libant1.6-java, in main
[03:32] <wasabi> And a seperate 'ant' package in contrib does the rest.
[03:32] <infinity> Ahh.
[03:32] <wasabi> I unified them into one package, all compiled with free VMs
[03:33] <wasabi> It's just too big of a change for Sarge, so Debian wants to hold off.
[03:33] <infinity> Thanks, I was looking at the wrong source package. :)
[03:33] <wasabi> It doesn't look like the right stuff is in experimental.
[03:34] <infinity> wasabi : Is the current 'ant' build failure on breezy yet another "needs a newer gjdoc" dep-wait, then?
[03:34] <wasabi> Hadn't noticed. Looking.
[03:36] <infinity> Probably owing to the fact that there's been very little java stuff in main, historically. :)
[03:38] <wasabi> excuse my slowness, my computer is falling apart at the seems.
[03:38] <wasabi> epiphany now refuses to launch, along with nautilus. ;)
[03:39] <gentoon> Breezy ready for testing yet?
[03:39] <gentoon> I have a 40gb partition open
[03:39] <Amaranth> gentoon: Are you comfortable in a shell (no X)?
[03:40] <Amaranth> gentoon: Can you figure out complex linux problems without help?
[03:40] <infinity> gentoon : It's always ready for testing, but definitely not ready for every day use.  (so, not ready for regular "end user testing")
[03:40] <gentoon> No one can do that, thats why we have bugs and forums and docs
[03:40] <gentoon> I know thats why i asked if its ready for "testing"
[03:41] <Amaranth> #ubuntu users using breezy right now have agreed to not help people running breezy as a way of stopping casual users from trying to use it
[03:41] <wasabi> Nobody?
[03:41] <gentoon> And i work in pure CMI everyday all day at work
[03:41] <gentoon> Thats very anti-productive
[03:42] <gentoon> everyone thats testing it should talk to eachother
[03:42] <wasabi> Good, so they shouldn't be testing, and they won't have to be talked to!
[03:42] <gentoon> so you don't want people to test it yet
[03:42] <Amaranth> gentoon: Users who know what they are doing and plunge ahead anyway are the ones that can figure out what is going on and file bugs.
[03:43] <gentoon> okay thats what i was asking
[03:43] <Amaranth> gentoon: It needs lots of testing, just not from casual users right now.
[03:43] <gentoon> Well yea, but when testing new stuff, its nice to see known bugs, so ya dont run in circles that have been ran in my someone else
[03:44] <tseng> eh it needs more testing of universe CXX apps
[03:44] <tseng> and less people moaning about xorg
[03:44] <gentoon> Now you have insulted me, by calling me a 'casual' user, i have been coding since before you were alive probably
[03:44] <Nafallo> tseng: and monodeps ;-)
[03:44] <tseng> gentoon: who called you that?
[03:44] <tseng> we made a generalization that had nothign to do with you
[03:44] <gentoon> Anyways, i offered i guess its back to SID
[03:44] <tseng> but the situation in general
[03:45] <gentoon> see ya guys
[03:45] <gentoon> and good luck
[03:45] <wasabi> bye! =)
[03:46] <infinity> wasabi : So, the ant failure is...? :)
[03:47] <wasabi> I don't know.
[03:47] <wasabi> I can't even open a browser heh
[03:47] <infinity> mdz : Feh.  ecj-bootstrap is still stuck in Uploaded (again).  We may need an elmo if someone else with rights to the archive can't figure out where it went.
[03:48] <infinity> wasabi : wget http://people.ubuntu.com/~lamont/buildLogs/a/ant/1.6.2-2ubuntu3/ant_1.6.2-2ubuntu3_20050602-0708-i386-failed.gz
[03:49] <wasabi> Hmm. Looks like an actual problem.
[03:49] <mdz> 2005-06-03 02:20: ecj-bootstrap_3.0.1-3_i386.changes libchipcard_0.9.1-7_sparc.changes
[03:49] <mdz> ecj-bootstrap_3.0.1-3_i386.changes
[03:49] <mdz> REJECT
[03:49] <mdz> Rejected: ecj-bootstrap_3.0.1-3_i386.changes: a file with this name already exists in the Accepted directory.
[03:49] <mdz> Rejected: ecj-bootstrap_3.0.1-3_all.deb file already exists in the Accepted directory.
[03:49] <mdz> Rejecting.
[03:49] <wasabi> But it's caused by a bug in java-gcj-compat.
[03:49] <wasabi> Which I thought I fixed.
[03:49] <infinity> mdz : Okay, so it was properly accepted, just never got installed from accepted.  Fun.
[03:49] <wasabi> Can it be built again?
[03:49] <mdz> infinity: I'll kill it from accepted if you can reupload
[03:50] <infinity> mdz : I can.
[03:50] <infinity> wasabi : I can toss it back if you think it'll build on the second go.
[03:50] <mdz> infinity: ready when you are
[03:50] <wasabi> one second, making sure
[03:50] <infinity> mdz : Always ready.
[03:51] <mdz> infinity: killed
[03:51] <infinity> There, re-uploaded.
[03:51] <wasabi> Hmm. Actually I thought doko fixed this.
[03:51] <wasabi> doko, ping. Was java-gcj-compat updated to depend on ecj?
[03:52] <wasabi> infinity, it won't rebuild until I know what happened to the fix to java-gcj-compat, so don't retry it.
[03:53] <infinity> wasabi : It depends on ecj-bootstrap (the version we're trying to get installed right now) | ecj
[03:53] <wasabi> errr.
[03:53] <wasabi> Oh I see. Heh.
[03:54] <infinity> So, should I dep-wait ant on ecj-bootstrap? :)
[03:54] <wasabi> I uploaded a temporary fix to Ant because I couldn't upload a real fix to java-gcj-compat (because it was in main)
[03:54] <wasabi> There should still be a valid ecj-bootstrap binary.
[03:54] <infinity> Not the version that java-gcj-compat currently depends on. :)
[03:54] <wasabi> OH.
[03:55] <wasabi> and then doko fixed java-gcj-compat with a bad version. ;)
[03:55] <infinity> (Not the new version of j-g-c anyway)
[03:55] <wasabi> So, I think where we're at is that j-g-c needs to be fixed, and I think doko was doing that.
[03:55] <mpt> mdz: I was told yesterday that the wiki spec for LaunchpadIntegration is wildly out of date now -- Who should I be pestering to update it? Apparently I'm supposed to be helping with the design
[03:55] <infinity> So, yeah.  If you tell me it will all magically work when ecj-bootstrap is rescued from oblivion, I'll retry ant shortly.
[03:55] <wasabi> Java is so incestuest.
[03:55] <mdz> mpt: truthfully? sabdfl ;-)
[03:56] <mpt> ehhh
[03:56] <infinity> wasabi : Don't feel bad.  Most big package sets are like this.  Heck, even X build-depends on itself.
[03:56] <mdz> mpt: he has specified that he wants to add two items to the Help menu of applications which have a Help menu
[03:56] <mpt> mdz: That's what I suggested originally, and got shot down :-P
[03:57] <wasabi> We could fix j-g-c
[03:57] <wasabi> I just can't upload it, because it's in main now.
[03:57] <infinity> wasabi : What fix does it require?
[03:57] <wasabi> So I was waiting for doko to do so.
[03:58] <wasabi> IT needs to build-depend on ecj-bootstrap and runtime ecj-bootstrap.
[03:58] <wasabi> I think it's missing a runtime depend.
[03:58] <infinity> wasabi : No, the runtime dep is there.  It's just versioned too high to be useful right now.
[03:59] <infinity> wasabi : S'what I said up there.  <point>
[03:59] <wasabi> that's odd how my apt-cache showsrc doesn't show that
[03:59] <mpt> mdz: In other news, I'd really really like to be contributing to fix the Ubuntu Help content (you were in the CC of a message I sent about it earlier), but *nobody* in the docteam can give me a step-by-step guide to how to modify the table of contents etc, other than pointing to scrollkeeper which seems to have almost zero documentation itself ... Who can I ask about that?
[03:59] <infinity> wasabi : It'll sort itself when our poor wayward ecj-bootstrap binary finds its way to the archive.
[03:59] <wasabi> Okay, just tear the version off.
[03:59] <infinity> wasabi : showsrc doesn't show binaries. :)
[03:59] <infinity> wasabi : Try just "show"
[03:59] <wasabi> err. duh. ;)
[04:00] <wasabi> Yes. Strip the versioned depend off.
[04:00] <infinity> wasabi : But we need the new bootstrap anyway, according to doko...
[04:00] <infinity> wasabi : So we may as well just fix that.
[04:00] <infinity> (Or do we really?)
[04:00] <wasabi> i dont' even remembe rwhat it fixed.
[04:00] <wasabi> this is killing my brain
[04:01] <lifeless> Kamion: for your scrollback edification, base-config is now pending a pysvn patch which I'm sending doko & elmo now
[04:01] <wasabi> It doesn't. We don't really need it.
[04:01] <mdz> mpt: there are practical problems with that UI, such as the fact that it involves modifying a huge number of applications, and some applications don't even have a help menu
[04:01] <mdz> mpt: but it's what he wants at this point
[04:01] <mdz> mpt: if you have already thought about it along those lines, feel free to update the spec accordingly
[04:01] <lifeless> mdz: re baz COTM in hoary ...
[04:01] <mpt> ok
[04:02] <wasabi> ecj-bootstrap was updated to correct a build failure (because of an update to gcj-4.0.1)
[04:02] <wasabi> And it was also synched with debian.
[04:02] <wasabi> The existing binary works.
[04:02] <mdz> mpt: regarding documentation, I don't know any more about that than you do, I expect.  it'd be a yelp question, try seb128. he should be able to point you to someone who knows, if he doesn't
[04:02] <mpt> ok, thanks for your time mdz
[04:02] <lifeless> mdz: uhm, its your call really. as upstream I would advise against, because IMO there is a difference between 'latest stuff that is pipelining in ' and 'CRACKOFTHEMOMENT"
[04:02] <infinity> wasabi : Okay.  Then we have a few source packages with needlessely inflated build-deps/deps.  antlr is also waiting on that version, I think.
[04:03] <wasabi> Yeah, I believe doko did that on accedent.
[04:03] <infinity> mdz : Any sign of progress on that binary?
[04:03] <mdz> lifeless: can we get 'latest stuff that is pipelining in', then?  currently our only options are COTM and releases
[04:03] <wasabi> Or maybe he didn't. I don't know. ;)
[04:03] <lifeless> mdz: as we release monthly, I don't see how - 
[04:03] <mdz> ecj-bootstrap_3.0.1-3_i386.changes
[04:03] <mdz> ACCEPT
[04:03] <mdz> ecj-bootstrap_3.0.1-3_all.deb
[04:03] <mdz>   to pool/main/e/ecj-bootstrap/ecj-bootstrap_3.0.1-3_all.deb
[04:03] <mdz> Accepting.
[04:03] <mdz> Accepted 1 package set, 907 KB.
[04:03] <lifeless> but I could drop someone a mail that '1.5~200506021435 looks unbroken'
[04:04] <lifeless> for instane
[04:04] <lifeless> bob2: is 1.4.1 uploaded to debian yet ?
[04:04] <wasabi> infinity, looks like ecj-bootstrap just fixed itself. ;)
[04:04] <lifeless> :)
[04:04] <mdz> lifeless: er, monthly?
[04:04] <wasabi> rebuild j-g-c
[04:04] <mdz> lifeless: bazaar in breezy is 3 months old
[04:04] <wasabi> Actually it should kick itself off shouldn't it?
[04:04] <lifeless> mdz: yes, and is the release-before last
[04:05] <mdz> lifeless: ok, then scratch my previous request, and "can we please track the latest releases in breezy?"
[04:05] <lifeless> mdz: sure thing, love to.
[04:05] <lifeless> mdz: fabbione said that they sync via debian
[04:05] <infinity> wasabi : It'll kick off when ecj-bootstrap is actually installed.  katie lies a bit when it says it's moved files to pool/... It actually means it's mapped them and scheduled them to be moved on the next cron.daily.
[04:05] <lifeless> mdz: but if you were to pull directly, that would reduce some latency
[04:05] <mdz> lifeless: debian has the same old version
[04:06] <wasabi> "daily"?
[04:06] <lifeless> yes, thus my ping to bob2
[04:06] <infinity> wasabi : A misnomer from Debian.  It's daily in Debian, every 15 mins (I think?) with Ubuntu.
[04:06] <bob2> mdz: debian has 1.3.2, which was the latest until last week
[04:06] <infinity> wasabi : But still called cron.daily, just cause.
[04:06] <mdz> bob2: lifeless just said it was the release-before-last
[04:06] <bob2> lifeless: not yet, doing it now
[04:06] <lifeless> mdz: it is
[04:06] <lifeless> 1.3.2 released a week after hoary
[04:06] <lifeless> actually, 1.3 did.
[04:07] <mdz> a week after hoary was 2 months ago
[04:07] <mdz> <-- confused
[04:07] <lifeless> was it that long ? crap
[04:07] <infinity> mdz : Can you tell if cron.daily is running on jackass?  Apparently it was taking forever and a day to finish its runs yesterday (something about elmo adding a second CPU to the box which slowed it down immensely)
[04:07] <lifeless> 'monthly' is a little loose.
[04:07] <mdz> infinity: yes, it is running right now
[04:08] <wasabi> around here 'monthly' could mean 'every other hour'
[04:08] <infinity> wasabi : Now you're starting to catch on. :)
[04:09] <mdz> wasabi: where 'daily' means 'twice per hour'? ;-)
[04:09] <wasabi> 4 times it seems
[04:10] <infinity> Don't take my word for it.  I don't have direct access to jackass and I have a horrible memory.  COuld be every 30 mins. :)
[04:10] <daniels> mdz: should be able to, yes
[04:10] <bob2> mdz: hoary has 1.2, debian has 1.3.2
[04:11] <mdz> bob2: right, and breezy has 1.3.2
[04:11] <bob2> right
[04:11] <mdz> wasabi: twice, :03 and :33
[04:12] <mdz> infinity: ecj-bootstrap installed
[04:12] <mdz> cron.daily looks to have finished
[04:13] <infinity> Yup.  Everything's back on track.
[04:13] <infinity> Huzzah.
[04:13] <infinity> wasabi : Kicking back ant now.
[04:13] <wasabi> I think what really caused all this problem is every package involved got updated, all at the same time. ;)
[04:14] <wasabi> And they all broke in some way
[04:14] <wasabi> And then got moved to main.
[04:14] <wasabi> Not in that order.
[04:16] <infinity> wasabi : antlr doesn't need the new ant, does it?  (Please say no, cause they just started building in the other order)
[04:16] <Amaranth> I think you guys are tackling more in the 6 months for breezy then red hat did for fedora core 4. :P
[04:16] <wasabi> hmmm.
[04:16] <wasabi> Shouldn't. ;)
[04:16] <infinity> Good.  Looks like it was successful.
[04:17] <wasabi> I can't wait to see Eclipse build. ;)
[04:17] <wasabi> Oh, can you kick eclipse?
[04:17] <wasabi> It's dep waiting on j2sdk1.3
[04:18] <infinity> wasabi : Do you have a list of packages waiting to be untangled?
[04:18] <infinity> (other than eclipse, which I'll smack around right now)
[04:18] <wasabi> No, I don't.
[04:18] <wasabi> I think we're pretty good off now.
[04:18] <calc> how is gnome 2.11 coming along, i noticed 2.11.3 should be out next week
[04:19] <calc> and no source tarballs for any 2.11
[04:20] <wasabi> frick. i seem to have an unsolvable mess here on my hands.
[04:20] <wasabi> I am confident if I exit X it won't launch again. ;)
[04:21] <calc> yea X has been broken in various ways under hoary for several weeks
[04:21] <wasabi> breezy.
[04:21] <wasabi> X hasn't been broken in hoary has it? heh
[04:21] <calc> erm yea oops ;)
[04:21] <wasabi> ok, phew.
[04:21] <dooglus> I discovered today that 'grep' is broken in hoary
[04:21] <mdz> infinity: I want libant1.6-java, junit, libjaxp1.2-java and xerces-j
[04:22] <dooglus> is there any way to get it fixed officially?
[04:22] <mdz> dooglus: those are both pretty vague
[04:22] <infinity> dooglus : Define "broken"?
[04:22] <calc> from what i can tell the only two install bugs left are the xkbcomp and libxt issue
[04:22] <dooglus> infinity: it uses all available CPU for as long as you let it in some cases.  it crashed my laptop 3 times today due to overheating.
[04:22] <dooglus> infinity: try this:
[04:22] <dooglus> yes | head -1000000 | LC_ALL=C grep . > /dev/null
[04:23] <wasabi> uh.
[04:23] <dooglus> it should run pretty quickly, right?
[04:23] <wasabi> no.
[04:23] <wasabi> yes never exits dude.
[04:23] <dooglus> wasabi: it does when the 'head' tells it to
[04:23] <infinity> dooglus : That's run a very fast infinite loop.
[04:23] <wasabi> head doens't tell it to.
[04:23] <wasabi> why would it?
[04:23] <wasabi> yes outputs forever.
[04:23] <Amaranth> whoa, why did no one tell me Alt+arrow reorders tabs in xchat?
[04:23] <wasabi> head buffers, waiting for the last 100000 lines
[04:23] <dooglus> head exits, which means the 'yes' gets a sigpipe
[04:24] <wasabi> which of course never get there, as yes never stops
[04:24] <dooglus> but that's not the point
[04:24] <wasabi> when does head exit?
[04:24] <dooglus> head -1 exits after printing one line...
[04:24] <wasabi> print  the first N lines instead of the first 10; with the lead
[04:24] <wasabi>               ing -, print all but the last N lines of each file
[04:24] <dooglus> try it:  "yes | head -3"
[04:24] <dooglus> you'll see, it's not an infinite loop
[04:24] <Amaranth> wasabi: it doesn't infinite loop, i've tried it
[04:24] <infinity> wasabi : Erm, yeah.  It does get a sigpipe.  Duh.
[04:24] <wasabi> oh...
[04:24] <wasabi> heh. ;)
[04:24] <infinity> Anyhow.
[04:25] <dooglus> so anyway...  of these 2 commands, the first works, very quickly, and the 2nd hangs, almost forever, eating all available CPU:
[04:25] <dooglus> yes | head -1000000 | LC_ALL=C grep . > /dev/null    <- OK
[04:25] <dooglus> yes | head -1000000 | LC_ALL=en_US.UTF-8 grep . > /dev/null   <- bad
[04:25] <dooglus> can you reproduce that on breezy too?
[04:25] <Amaranth> dooglus: so you've figured it out more?
[04:25] <Amaranth> nice
[04:25] <wasabi> java is eatting my mind
[04:25] <wasabi> for some reason i can't comprehend that right now
[04:26] <dooglus> Amaranth: it's a very old bug.  It just didn't get fixed in ubuntu as far as I can tell
[04:26] <calc> wasabi: switch to mono?
[04:26] <infinity> dooglus : grep in UTF8 has always been slow as a dog, no?
[04:26] <Amaranth> using C took 0.5 real time, using en_US.UTF-8 took..well, it's still running
[04:27] <Amaranth> infinity: Either other distros cheat and set C or this got fixed somewhere.
[04:27] <mdz> dooglus: "grep is slow with UTF-8 locales" is not the same as "grep is broken"
[04:27] <dooglus> infinity: yes, but that's not the problem here either - the problem is that due to a bug in the grep code, the algorithm takes a time propertional to the *square* of the size of the input file.
[04:27] <mdz> at any rate, there's already a bug open about the performance issue
[04:27] <calc> istr something about locales and something being slow before ;)
[04:27] <dooglus> mdz: but it's stupidly slow - grep should be take linear time
[04:27] <dooglus> that's what's broken about it.
[04:27] <Amaranth> wow, en_US.UTF-8 is still running
[04:27] <calc> of course that isn't very helpful, i recall redhat had some performance issues when they switched to using utf8
[04:27] <dooglus> Amaranth: it will run all year if you let it
[04:27] <Amaranth> red hat fixed them somehow
[04:28] <dooglus> I think I have a one line workaround.
[04:28] <Amaranth> as long as it doesn't cheat and set LC_ALL to C that's good
[04:28] <mdz> dooglus: https://bugzilla.ubuntu.com/show_bug.cgi?id=1148
[04:28] <dooglus> the problem at the moment is that for every match of every character, grep looks through the whole rest of the file to see which characters are multi-byte.
[04:29] <dooglus> that's where the O(n squared) comes from.
[04:29] <mdz> if you have further analysis that's not in that bug, please add comments there
[04:29] <mdz> infinity: yes, it's a very old problem
[04:29] <dooglus> mdz: that bug is badly out of date
[04:30] <infinity> dooglus : Please update it, then.
[04:30] <Amaranth> dooglus: congrats on figuring this out, was driving me nuts trying to understand it
[04:30] <dooglus> infinity: could I ask you though - how is this _supposed_ to work?
[04:30] <dooglus> it's a GNU application, but you get it from debian, and you also hack on it yourself.
[04:31] <dooglus> and the redhat are hacking on it too - how do it all work - or how _should_ it all work?
[04:31] <infinity> dooglus : We can fix it in breezy, push the changed to Debian, and the Debian maintainer will push the changes upstream.  If anyone in that chain decides the changes suck, we'll just carry them ourselves.
[04:31] <dooglus> the fedora guys fixed it ages ago (see patch, here: http://cvs.fedora.redhat.com/viewcvs/devel/grep/grep-2.5.1-egf-speedup.patch )
[04:31] <infinity> dooglus : Alternately, report the bug (and fix) to GNU, and we'll get the changes trickled down eventually.
[04:32] <dooglus> redhat released an advisory about it, over a year ago.  but nothing happened to the GNU code
[04:32] <infinity> dooglus : Odds are that RedHat would have pushed their changes upstream already, so there's a fair chance the GNU maintainer decided it wasn't elegant/correct enough.  That wouldn't necessarily stop us from carrying the patch, if it seems to work well enough and not cause regressions elsewhere.
[04:32] <dooglus> infinity: that's what I don't understand - the GNU guys have been sitting on the bug for over a year too, doing nothing - same as the ubuntu guys.
[04:33] <dooglus> here's the GNU bug: http://savannah.gnu.org/patch/?func=detailitem&item_id=3803
[04:34] <dooglus> as far as I can see there's no comment from the GNU people at all
[04:34] <dooglus> oh, my bad.  that's March this year - so they've only sat on it for 3 months.
[04:34] <dooglus> Amaranth: did your grep finish yet?
[04:35] <Amaranth> dooglus: I quit it.
[04:37] <dooglus> is there anything useful I can do to help with breezy?  I've got a lot of debugging experience and nothing better to do with my time...
[04:38] <wasabi> irssi. =/
[04:38] <wasabi> Did it all get in?
[04:39] <schweeb> dooglus: you could join #ubuntu-motu and help out w/ universe
[04:39] <schweeb> dooglus: fix pkgs w/ build errors, etc...
[04:39] <Amaranth> dooglus: filing bug reports is important
[04:40] <dooglus> Amaranth: I'm a little disillusioned about that.  Last week I filed 3 and all 3 were shouted down, even though they were valid.  Now I see this massive problem with grep sitting in a bug report not being fixed...
[04:41] <Amaranth> links to the 3?
[04:42] <infinity> wasabi : Depends on what "it" was. I saw nothing from you before you dropped.
[04:42] <wasabi> "it all" = "java"
[04:42] <infinity> wasabi : Oh.  Still working on babysitting it.
[04:43] <infinity> wasabi : ant and antlr are uploaded, waiting for them to hit the archive before another wave of stuff that depends on them can start.
[04:43] <dooglus> Amaranth: they were all gentoo related.  Basically I was trying to install it by following the installation handbook, and filing documentation bugs whenever I found that the command they told me to type didn't work.
[04:43] <jbailey> mdz: I've attached the patch to 11135 that I put into breezy.  I don't have a Hoary system handy that I can butcher atm, but I tested it when I wrote it.
[04:43] <Amaranth> dooglus: oh, you're talking about another distro
[04:44] <Amaranth> dooglus: MOTU loves getting bug reports ;)
[04:44] <dooglus> Amaranth: yes, sorry.  I didn't make that clear.
[04:44] <dooglus> what is MOTU?
[04:44] <Amaranth> dooglus: It's about 20 guys managing 12000 packages, bugs tell them what to look at
[04:44] <dooglus> master of the universe?
[04:44] <Amaranth> yeah
[04:44] <dooglus> heh
[04:44] <schweeb> yep
[04:44] <Amaranth> #ubuntu-motu
[04:46] <wasabi> jbailey: don't suppose you heard a bout eclipse?
[04:47] <jbailey> wasabi: Only your mention earlier that you uploaded a new version to multiverse.
[04:47] <wasabi> gnome-panel needs to be rebuilt.
[04:47] <wasabi> ... at least that it what I'm seeing
[04:47] <wasabi> An actual corrupt .deb file?
[04:47] <jbailey> Eh.  Because of eclipse?
[04:47] <jbailey> Or did you context switch?
[04:47] <wasabi> unrelated
[04:47] <wasabi> context switch
[04:47] <mdz> what about gnome-panel?
[04:49] <wasabi> Hmm. Don't know if it's my fault or not... not sure how it could be.  dpkg-deb fails, subprocess killed by signal (Broken pipe)  --fsys-tarfile returned error exit status 2
[04:49] <mdz> infinity: ant meaning libant1.6-java?  (we do have an entirely separate source package called ant)
[04:49] <wasabi> mdz libant1.6-java, the source package, is deprecated and should be removed.
[04:50] <mdz> wasabi: ...
[04:50] <mdz> I thought it was meant to supersede ant
[04:50] <mdz> openoffice.org2 build-depends on libant1.6-java
[04:50] <wasabi> No, exactly backwards.
[04:50] <mdz> we had ant before we had libant1.6-java
[04:50] <wasabi> source packages ant super ceeds source package libant1.6-java
[04:50] <infinity> mdz : libant1.6-java is built from "ant"
[04:50] <wasabi> source package ant produces BINARY package libant1.6-java
[04:51] <mdz> infinity: yes, but libant1.6-java produces a libant1.6-java binary package which is newer
[04:51] <wasabi> What? It does?
[04:51] <infinity> wasabi : Is anything on your system NOT crashing?  (are you having hardware issues?)
[04:51] <mdz> so currently, libant1.6-java builds libant1.6-java in breezy
[04:51] <wasabi> not good.
[04:51] <wasabi> shouldn't do that.
[04:51] <wasabi> infinity: X issues.
[04:52] <infinity> mdz : Erm, no, ant should be newer.. 1.6.2-2ubuntu3
[04:52] <infinity> mdz : Though, only barely.
[04:52] <mdz> hmm, did they swap again recently?
[04:52] <wasabi> libant1.6-java needs to be removed. I don't know how that happened.
[04:53] <mdz> I swear I recall someone (jbailey? doko?) telling me that we were intentionally superseding libant1.6-java with the libant1.6-java source package
[04:53] <wasabi> that was me actually.
[04:53] <wasabi> but that's not what I said
[04:53] <wasabi> few months ago I think.
[04:53] <mdz> yes, I see our conversation in March
[04:53] <wasabi> Heck I think Hoary's copy works. ;)
[04:53] <mdz> Mar 21 18:23:48 <mdz>   then once the dust settles, you can request removal of libant1.6-java
[04:54] <wasabi> Dust never settled, haha.
[04:54] <infinity> The dust must be settling right now?
[04:54] <mdz> ...months pass...:-)
[04:54] <wasabi> Yeah, this is the dust right here.
[04:54] <mdz> well, if ant's libant1.6-java is newer now, we should be able to remove libant1.6-java source
[04:54] <mdz> that's sufficient settling
[04:54] <wasabi> It might not be newer, but it is "good enough".
[04:54] <mdz> or did ant not build until just now?
[04:54] <wasabi> We're going to have to maintain a fork of it until Debian adopts it.
[04:54] <mdz> newer meaning "higher version number"
[04:55] <wasabi> (sarge, blah blah blah)
[04:55] <infinity> mdz ant just built and uploaded.
[04:55] <infinity> Put a colon in there somewhere.
[04:55] <wasabi> I was totally hoping to avoid that. ;)
[04:55] <wasabi> They both come from the same upstream sources, so we can always just, upgrade to make it superceed properly.
[04:56] <mdz> I've requested that libant1.6-java be removed
[04:56] <mdz> then we can just continue tracking ant
[04:57] <infinity> Oh, FFS.
[04:57] <infinity> wasabi : ant needs a new upload.
[04:57] <wasabi> It's main now. ;)
[04:57] <wasabi> volunteers?
[04:57] <bob2> lifeless: mdz http://crumbs.ertius.org/~rob/baz/
[04:57] <wasabi> what's wrong with it?
[04:57] <bob2> (in a few minutes)
[04:57] <infinity> wasabi : 1.6.2-2ubuntu3 is a lower version than 1.6.2-2.1
[04:58] <mdz> germinate seems to be choosing libant1.6-java's libant1.6-java over ant's libant1.6-java
[04:58] <wasabi> Oh. Ya know. I sear back in the day I made an upgrade of libant1.6-java with a change log that clear said what was going on.
[04:58] <wasabi> I guess that must have been the crack.
[04:58] <mdz> infinity: isn't that what I was saying earlier?
[04:58] <infinity> mdz : Yes, and I either can't read or... Can't read. :)
[04:59] <wasabi> Debian made a NMU and that made it into Ubuntu.
[04:59] <infinity> wasabi : So, we definitely want ant over the libant1.6-java?... So, I should just upload an identical ant package with a bunped version number? :)
[04:59] <infinity> bumped, even.
[04:59] <wasabi> Well, then you won't have merged the Debian changes.
[04:59] <infinity> And then we need libant1.6-java blacklisted from autosyncs.
[05:00] <infinity> wasabi : Well, if you want to merge the stuff from -2 -> -2.1, let me know. :)
[05:01] <wasabi> I don't. Heck I can't right now. Can you upload a new copy, bumped version, to -2.1ubuntu0 or something silly.
[05:01] <wasabi> Actually I guess it doesn't matter.
[05:01] <wasabi> ubuntu1
[05:01] <infinity> wasabi : The NMU was a 3-line patch.  I can pull it into our ant package, if that's desireable.
[05:02] <wasabi> It is.
[05:02] <wasabi> I need my GUI back.
[05:02] <infinity> Alright, will do.
[05:12] <emrysk> Mono is going to be in the Breezy base install, right?
[05:13] <ajmitch> emrysk: yep, it most likely will be
[05:13] <Amaranth> in main, don't think it'll be in desktop seed
[05:14] <Amaranth> might not even be in ship seed, that cd is bulging
[05:14] <ajmitch> Amaranth: depends if beagle or other apps get put in desktop seed
[05:14] <jsgotangco> "if"
[05:14] <Amaranth> smeg should be in desktop seed ;)
[05:15] <ajmitch> Amaranth: if anyone ever uses it ;)
[05:15] <jsgotangco> haha
[05:15] <Amaranth> ajmitch: heh, i think i'm covered there
[05:15] <ajmitch> probably
[05:15] <Amaranth> typical #ubuntu converstation: "<foo> how do i edit my menus? <bar> use smeg"
[05:15] <Amaranth> :D
[05:16] <ajmitch> I don't know if system tools is the best place in the menu for it
[05:16] <Amaranth> well, i wasn't too sure about that
[05:16] <ajmitch> it can be moved 
[05:16] <Amaranth> it made more sense when it had a root mode
[05:16] <Amaranth> it still does, if you run it with --root
[05:23] <infinity> mdz : Can we get libant1.6-java (the source package) blacklisted from syncs?  I've just uploaded the shiny new (correctly-versioned) ant right now.
[05:34] <jbailey> w
[05:42] <drbyte> ogra: ping
[05:42] <drbyte> or actually, does anyone know where ogra keeps the code for his hardware database thing?
[05:43] <ajmitch> grab the source package? I don't know if he has the code repository online
[05:44] <drbyte> ajmitch: hmm, ok. what is it called? i dont actually have access to an ubuntu system atm (shock, horror)
[05:45] <ajmitch> drbyte: hwdb-client
[05:45] <drbyte> ajmitch: is there a hwdb-server too?
[05:45] <ajmitch> not yet
[05:45] <ajmitch> you'd have to ask him
[05:46] <drbyte> http://packages.ubuntu.com/hoary/base/hwdb-client that's good. sweet
[05:46] <jsgotangco> hehe
[05:46] <drbyte> jsgotangco: dagger?
[05:47] <jsgotangco> dagger = knife or something
[05:47] <drbyte> you wanna knife me?
[05:47] <drbyte> erps, why?
[05:47] <jsgotangco> drbyte ajmitch: hmm, ok. what is it called? i dont actually have access to an ubuntu system atm (shock, horror)
[05:47] <jsgotangco> hehe
[05:47] <jsgotangco> just kidding
[05:47] <drbyte> jsgotangco: hehe. right
[05:48] <ajmitch> jsgotangco: remember he's still stuck with fedora
[05:48] <jsgotangco> there's no rush
[05:49] <drbyte> thanks ajmitch, jsgotangco 
[05:49] <drbyte> gotta talk to ogra later i guess
[05:49] <drbyte> or ogra, ping bytee if i'm not around. should be disconnecting soon
[05:50] <jsgotangco> ok see ya
[05:57] <dampjam> the i386 install: http, and torrent are corrupt on http://us.releases.ubuntu.com/releases/5.04/
[05:57] <dampjam> the install fails on the bsdutils package
[05:58] <dampjam> I am not sure where to report it
[05:58] <jsgotangco> mako, ping?
[06:02] <mako> jsgotangco: yes
[06:02] <mako> was about to run to sleep.. but go ahead, you caugth me
[06:04] <jsgotangco> oh i already sent it on email, i want to give it to the LugRadioLive guy but needs review as well
[06:04] <jsgotangco> also had jdub on cc since this is Ubuntu@conferences stuff
[06:10] <jsgotangco> brb
[06:14] <fabbione> morning
[06:14] <fabbione> daniels: you around?
[06:15] <wasabi> dampjam: have you checked the md5 sums?
[06:17] <mdz> infinity: I've already requested that it be removed entirely
[06:18] <fabbione> morning mdz
[06:18] <mdz> fabbione: morning
[06:18] <dampjam> I have
[06:18] <mdz> infinity: is the stack ready to move over into main?
[06:19] <dampjam> I then downloaded the images from carroll.cac.psu.edu, they worked without a flaw
[06:20] <mdz> infinity: ant junit xerces-j libjaxp1.2-java
[06:21] <mdz> hmm, and classpath
[06:22] <infinity> mdz : junit is building right now.  Checking on the others.
[06:23] <wasabi> classpath isn't really neccassary when using gcj
[06:23] <wasabi> Which we are doing.
[06:23] <wasabi> gcj includes a complete native classpath compilation.
[06:24] <Amaranth> nice
[06:24] <Amaranth> so breezy is getting native eclipse in main?
[06:24] <infinity> wasabi : xerces-j build-deps on classpath...
[06:24] <wasabi> Among other things.
[06:24] <wasabi> Does it?
[06:24] <wasabi> It probably needs to be fixed then.
[06:24] <Amaranth> rock
[06:24] <wasabi> What is xerces-j for?
[06:24] <Amaranth> mono in main, java things in main, this release is going to blow people away
[06:25] <infinity> wasabi : Heckifiknow. :)
[06:25] <Amaranth> it's an xml parser
[06:26] <wasabi> I've never had my hands on it.
[06:26] <wasabi> What depends on it?
[06:26] <infinity> mdz : Your list all looks ready exceit for junit which has just been uploaded.
[06:26] <Amaranth> libant1.5-java
[06:26] <wasabi> That is deprecated.
[06:26] <wasabi> mdz?
[06:27] <mdz>    [Reverse-Build-Depends: libant1.6-java] 
[06:27] <wasabi> Heh.
[06:27] <wasabi> I suspect that's the old ant package?
[06:27] <mdz> if ant doesn't use it, then we won't need it
[06:27] <mdz> that's the libant1.6-java source package
[06:27] <Amaranth> i'm looking at local stuff
[06:27] <wasabi> Yeah, the last update this thing had, other than what you did to it recently, was in 2002.
[06:27] <mdz> but ant uses it too
[06:27] <wasabi> does it.
[06:27] <mdz> oh, it uses xerces2
[06:28] <mdz> from xerces2-j
[06:28] <wasabi> yeah.
[06:28] <wasabi> xerces-j i think was being dropped by Debian
[06:28] <wasabi> I remember hearing something about it
[06:28] <mdz> once everything is installed, I'll re-germinate and see how things look
[06:28] <mdz> hopefully germinate will choose ant's libant1.6-java now, and things should be clearer
[06:28] <wasabi> http://java.debian.net/    a wiki page MovingJavaToMain
[06:35] <wasabi> X really is pretty screwed up isn't it. ;)
[06:38] <Lathiat> heh
[06:44] <mdz> bob2: how is the ~rob/baz/ stuff different from what's in breezy?
[06:44] <mdz> wasabi: working OK for me
[06:44] <mdz> I had to fix up that mouse business in xorg.conf a few days ago, but there have been updates since then and it's probably fixed
[06:47] <mdz> wasabi: hmm, the picture is significantly more complex now
[06:48] <mdz> and kaffe is back
[06:49] <mdz> and jython
[06:49] <wasabi> aheh.
[06:49] <mdz> and xerces-j is still here
[06:49] <wasabi> where'd they come from?
[06:49] <Micksa> anyone care to help me out with a hoary PXE boot problem?
[06:49] <Micksa> breezy is working fine but not hoary
[06:50] <wasabi> mdz: the big X move is killing me.
[06:50] <wasabi> Thing won't even start.
[06:50] <mdz> http://people.ubuntu.com/~mdz/temp/java-main.txt
[06:50] <mdz> that's the dependency chain leading back from openoffice.org2
[06:51] <mdz> ant seems to pull in a *lot* more stuff than libant1.6-java was
[06:51] <wasabi> Yeah, it does.
[06:52] <mdz> jlex at least needs fixing
[06:52] <wasabi> The old libant1.6-java was super minimal.
[06:52] <mdz> and libbsf-java
[06:52] <mdz> jlex pulls in kaffe, libbsf-java pulls in jython
[06:52] <wasabi> what program are you using to generate this?
[06:52] <mdz> anastacia
[06:52] <wasabi> jython should be fine.
[06:53] <mdz> it compares the current state of the archive to what germinate says it should be
[06:53] <wasabi> What's it pull in?
[06:53] <mdz> jython is way behind the times (python2.1); we don't want it
[06:53] <wasabi> I remember these two.
[06:53] <wasabi> Hmm.
[06:53] <wasabi> bsf might require it.
[06:53] <mdz> that would be most unfortunate
[06:53] <wasabi> bsf is the bean scripting framework. It is a generic scripting framework for Java.
[06:53] <mdz> someone was saying that jython was being updated for modern python
[06:54] <wasabi> It includes a number of scripting engines, javascript (rhino), and jython
[06:54] <wasabi> It might be disablable.
[06:54] <wasabi> I remember being on this road before.
[06:54] <mdz> why does javax-servletapi2.3 need bsf?
[06:54] <wasabi> Hmmm. Not sure.
[06:54] <mdz> that seems weird
[06:54] <wasabi> Yeah it does.
[06:56] <jblack> Is the netpbm / netpbm-free guy here, or somebody that knows a bit about those two packages? 
[06:57] <mdz> I know a bit about them; best to just ask
[06:58] <infinity> Indeed, ask away.
[06:58] <jblack> Ok. We've already done an info file for "netpbm". Theres also a info file for "netpbm-free". There doesn't seem to be a netpbm-free package in ubuntu.
[06:58] <jblack> (though there is a netpbm) 
[06:59] <infinity> netpbm-free is the source package.
[06:59] <jblack> It also has the same cvsroot and cvsmodule as netpbm. That means to me that its a duplicate. 
[06:59] <infinity> The netpbm-free source package generates the netpbm binary package.,
[06:59] <infinity> Is that the confusion? :)
[07:00] <wasabi> Wait.
[07:00] <wasabi> mdz, servlet2.3 doesn't depend on bsf that I can see.
[07:00] <mdz> hmm, germinate doesn't usually lie about these things
[07:00] <jblack> infinity: The confusion is that there's two info files that point to the same place. 
[07:01] <mdz> I don't see it either
[07:02] <infinity> jblack : Well, there hasn't been a "ntpbm" source package since 1998... So, "netpbm-free" is the canonical version.
[07:02] <mdz> wasabi: oh, that's backwards
[07:02] <mdz> wasabi: it's bsf that build-deps on servlet2.3
[07:02] <wasabi> That makes sense.
[07:02] <mdz> yes, much more so
[07:03] <mdz> is free-java-sdk a sane thing?  I haven't seen it before
[07:03] <jblack> infinity: ?? Isn't the canonical package in ubuntu is called netpbm ? 
[07:03] <wasabi> free-java-sdk is some stupid thing that the Debian guys are into.
[07:03] <wasabi> It's like java-gcj-compat, but it chooses a random JVM or something.
[07:03] <infinity> jblack : Not the source package, no.  You do imports based on source package names, not binary, right?
[07:03] <wasabi> Because they don't want to discriminate.
[07:03] <mdz> infinity: yes
[07:03] <wasabi> random build deps, uh huh.
[07:03] <jblack> infinity: typically, we try to, yes. 
[07:03] <infinity> jblack : apt-cache show netpbm, it's from the netpbm-free source.
[07:03] <mdz> it seems to be committed to sablevm at the moment
[07:03] <wasabi> oh?
[07:04] <mdz> Depends: jikes-sablevm, fastjar, sablevm, classpath-tools
[07:04] <wasabi> guess he won. ;)
[07:04] <wasabi> Gadek (teh sable guy) is very loud spoken. Hehe.
[07:04] <mdz> we currently have sablevm in main, though I don't recall why
[07:04] <wasabi> I'd discard free-java-sdk actually.
[07:04] <mdz> I think berkeley db or something build-deps on it
[07:04] <wasabi> And discard sable, and fix bdb to build on gcj, in the long run.
[07:05] <jblack> As it happens, I called it netpbm because ubuntu calls it netpbm, freshmeat calls it netpbm.
[07:05] <mdz> yeah, db4.2 build-depends on sablevm
[07:05] <infinity> jblack : Ubuntu calls the source package netpbm-free, though (same as Debian)
[07:05] <mdz> sablevm AND gcj AND libgcj6-dev
[07:05] <jblack> I notice at freshmeat.net there's a program called "netpbm-free" that is a CLI-interfaced image creation and manipulation program. is that what you're thinking. 
[07:05] <wasabi> Well, I think our idea is that we'll have one "official supported JVM"
[07:05] <wasabi> haha.
[07:05] <wasabi> Yeah, tons of Debian packages do that.
[07:05] <mdz> why?
[07:05] <wasabi> They build Java pieces with one JVM.
[07:05] <wasabi> Then build native versions with GCJ.
[07:05] <infinity> jblack : The source package could be named back, since there's no netpbm-nonfree anymore, but I doubt anyone cares enough to do it. :)
[07:05] <wasabi> The GCJ compiler traditionally SUCKS.
[07:06] <wasabi> It's not really usable for much.
[07:06] <jblack> also, sourceforge calls it netpbm, and the homepage calls it netpbm too. 
[07:06] <wasabi> But we're using ECJ, which is supurb.
[07:06] <jblack> even the cvs module is netpbm.
[07:06] <mdz> jblack: it's always been netpbm upstream
[07:06] <mdz> jblack: but in Debian it was split into netpbm-free and netpbm-nonfree
[07:06] <infinity> jblack : Yes, it /IS/ netpbm.  Our source package was just split back in 1998 to -free and -nonfree, because of GIF patents.
[07:06] <wasabi> Free Java is marred from years of having 10 VMs all at various levels of completion.
[07:07] <fabbione> that are actually expired btw
[07:07] <wasabi> Debian of course hilights that because all 10 VM developers are d-d's
[07:07] <infinity> jblack : They were reintegrated in 2004 (when the patent expired), but the source package is still called -free, for hysterical raisins (or sheer laziness of people who didn't want to push a new source through queue/NEW)
[07:07] <jblack> *click*. Ok. importd isn't going to be able to handle netpbm-free for you guys, because right now its dependant upon a different source for every product. 
[07:07] <mdz> jblack: hmm?  it should be fine, there is now a 1:1 correspondence between the netpbm-free source package and the netpbm product
[07:07] <wasabi> The kaffe fans always build with kaffe, the sable guys always use sable. A few people enjoy gcj. Some people use JamVM now too.
[07:08] <mdz> fun
[07:08] <wasabi> Each of those VMs include their own copy of classpath.
[07:08] <wasabi> So, then you have the Classpath + VM guys.
[07:08] <infinity> jblack : There is no netpbm source package at all, so as mdz says, there's a 1:1 mapping.
[07:08] <jblack> mdz: its there as "netpbm" as the upstream name (we're supposed to use upstream names) 
[07:08] <wasabi> Who pick a VM, and build with the VM against the classpath package instead of hte included one.
[07:08] <wasabi> It's a real big mess. ;)
[07:08] <jblack> Oh, I getcha. You guys'll branch it. 
[07:09] <mdz> jblack: at this point, now that netpbm-nonfree is gone, it's simply a matter of the source package being named differently from upstream
[07:09] <mdz> nothing more insidious than that
[07:09] <jblack> I'll drop lifeless an email letting him know that you'd like it renamed. 
[07:10] <mdz> like what renamed?
[07:10] <mdz> wasabi: what do we do about classpath?
[07:10] <jblack> Never mind. 
[07:10] <wasabi> We ignore it.
[07:10] <wasabi> And remove it from our deps.
[07:11] <jblack> Thanks for the time guys
[07:11] <wasabi> The plan is to support one official VM. At this point it's GCJ.
[07:11] <mdz> and gcj includes its own copy of classpath?
[07:11] <wasabi> Yes.
[07:11] <mdz> ok, so in that case we also have the issue that libgnujaf-java and libjaxp1.2-java depend on classpath
[07:11] <wasabi> So, all these uploads of Java crud I was doing a few months ago, basically I was just converting packages to gcj.
[07:12] <mdz> and there are still more
[07:12] <wasabi> Apparently.
[07:12] <wasabi> I didn't have the cool tool you're using. ;)
[07:13] <mdz> we ought to publish its output someplace regularly, if we don't already
[07:13] <wasabi> I shall quickly patch up those two.
[07:14] <mdz> are the other packages in that list sane already?
[07:15] <mdz> if you could spend a little time writing a short howto for doing java packages the right way in Ubuntu, we could distribute this workload a bit
[07:15] <mdz> or does something like that already exist?
[07:16] <wasabi> Not really.
[07:16] <wasabi> The problem is Java itself has no such howto.
[07:16] <wasabi> C comes in autoconf, or a makefile.
[07:16] <wasabi> Java comes in ... sometimes ant.
[07:16] <wasabi> Free java more than not comes in a hacked together custom weird thing. ;0
[07:17] <wasabi> We need a comprehensive policy, really.
[07:17] <wasabi> Lots of ideas floating around, but few are written down.
[07:18] <mdz> I mean along the lines of what you just told me
[07:18] <mdz> use this compiler, this JVM, which packages to depend and build-depend on
[07:18] <wasabi> Debian suffers from the same problem really. Just not many Java hackers.
[07:18] <mdz> we have even fewer, I'm sure
[07:18] <wasabi> JavaPackagingProcess maybe
[07:19] <mdz> but given a clear direction to go in, we have quite a few people who are interested in getting packages into line with policy so that everything fits together nicely
[07:19] <wasabi> I never updated that hting. =/
[07:19] <mdz> of course, they have their hands mostly full with the C++ transition at the moment
[07:20] <mdz> JavaPackagingProgress?
[07:20] <mdz> I don't see a ...Process
[07:20] <mdz> yeah, Progress has a bit of the right stuff
[07:20] <wasabi> jeff had a page before, trying to find it
[07:21] <wasabi> JavaIntegration. Out of date.
[07:22] <Micksa> pfff, #ubuntu is no help
[07:22] <Micksa> okay, I have a new laptop which won't run my existing install
[07:22] <Micksa> and I don't wanna reinstall
[07:22] <Micksa> I've narrowed it down to modules that need to be loaded in the initrd
[07:22] <Micksa> I just gotta figure out which ones :)
[07:23] <wasabi> only ones that should matter are the ones for hte storage device, right?
[07:24] <Micksa> yeah. and loading other modules from the fs, but I think that's probably covered :)
[07:24] <Micksa> that's what's broken incidentally. the existing kernel isn't detecting the hdd.
[07:26] <wasabi> brb hope i got x fixed
[07:30] <Micksa> who is working on eclipse? how is it going?
[07:31] <infinity> wasabi : Better?
[07:31] <wasabi> thank goodness
[07:31] <wasabi> yeah.
[07:31] <wasabi> reinstalled all of X from hoary.
[07:31] <wasabi> There has got to be an easier way to figure out what consitutes "all of X"
[07:32] <Micksa> haha
[07:32] <infinity> As it becomes more modular, it'll be harder, not easier (but that's what dependencies are for)
[07:32] <Micksa> "dpkg -l"gre[ ^x" 8)
[07:32] <Micksa> er
[07:32] <Micksa> "dpkg -l|grep ^x" 8)
[07:33] <wasabi> I guess having a program that you can say "give me all of xserver-xorg" and it would spit out it's complete dependency chain
[07:33] <wasabi> that might help
[07:33] <Micksa> er, or even...
[07:33] <Micksa> oh, forget it
[07:33] <Micksa> depends on what you mean by "X" :)
[07:35] <wasabi> mdz uploaded a fixed libgnujaf. What was the other one?
 ok, so in that case we also have the issue that libgnujaf-java and libjaxp1.2-java depend on
[07:37] <infinity>           classpath
[07:40] <lifeless> hi 
[07:40] <lifeless> I've found a bug in tk8.3
[07:41] <lifeless> causes a broken build of gnu-smalltalk-2.1.x
[07:41] <lifeless> whats the best way to get that fixed for breezy, and is there any point in doing a fix for hoary? (I'd like to get a fixed gnu-smalltalk in hoary-universe if possible)
[07:47] <infinity> lifeless : File a bug, with a patch if you have one.
[07:48] <lifeless> infinity: haven't found the cause...
[07:49] <lifeless> but the vuild tkConfig.sh (/usr/lib/tk8.3/tkConfig.sh) has a CINCLUDE path it offers of '# no special path is needed'
[07:49] <lifeless> ^^ WTF CRACK^^
[08:00] <wasabi> Those two packages were uploaded. I'm going to bed now.
[08:02] <Micksa> does init= get ignored by initrd or something?
[08:05] <Mithrandir> init is done post-initrd
[08:05] <Micksa> what if I want a shell inside initrd?
[08:06] <Mithrandir> then you should set DELAY to something in mkinitrd.conf and press enter
[08:06] <Micksa> kay
[08:10] <Micksa> how do I get the intstaller to do the initrd modprobe detection, without doing a complete install?
[08:12] <Micksa> dammit, lsmod would be handy :)
[08:19] <Micksa> how do you continue booting after the shell?
[08:20] <lifeless> Micksa: these are more #ubuntu style questions
[08:21] <Micksa> aw geez
[08:21] <Micksa> I end up having to jump between here and there cos noone on there seems to know this stuff
[08:21] <crimsun> this is a devel channel, please keep support questions in there :)
[08:22] <lifeless> Micksa: if you are offering a ptch, or figuring out how to patch it, its on topic here ;)
[08:22] <lifeless> if you are essentially just using it ...
[08:26] <lifeless> bugs in main : are they still bugzilla ?
[08:26] <Micksa> okay, I got one :)
[08:27] <Mithrandir> lifeless: yes
[08:27] <Micksa> I'm looking at the initrd scripts and by the look of it there IS no clean way to continue boot after the initrd shell
[08:27] <Micksa> cos it gets exec'd, and in the middle of /linuxrc, which seems to do half the work
[08:28] <Mithrandir> Micksa: then you do the rest of the stuff that /linuxrc does and execs /sbin/init from there.  Or you fix the problem, then reboot.
[08:30] <daniels> fabbione: pong
[08:31] <fabbione> daniels: what's the status with xorg -21?
[08:31] <fabbione> broken xbase-clients is causing FTBFS in the buildd
[08:34] <daniels> fabbione: within the next 24h, hopefully
[08:34] <fabbione> daniels: ok
[08:36] <lifeless> infinity: could you read https://bugzilla.ubuntu.com/show_bug.cgi?id=11444 and tell me if I've made sense ?
[08:36] <lifeless> daniels: did X at any point tell apps that its include path should be '# no special path needed' ?
[08:37] <daniels> lifeless: no.
[08:37] <lifeless> daniels: ah well, not that source then ;)
[08:37] <daniels> lifeless: best practice has been #include <X11/foo.h> (or <X11/extensions/foo.h, or ...) forever, and most apps are supposed to have been using -I/usr/X11R6/include since R6 (about ten years ago)
[08:38] <lifeless> heh
[08:39] <lifeless> tk8.3 has a spethial config value
[08:39] <lifeless> see https://bugzilla.ubuntu.com/show_bug.cgi?id=11444
[08:39] <lifeless> aw f*ckage
[08:39] <lifeless> malone hates me
[08:40] <infinity> lifeless : That's a pretty... Sparse bug report. :)
[08:40] <lifeless> so I can't file the bug on gnu-smalltalk
[08:40] <infinity> lifeless : What does that break, and how?
[08:41] <lifeless> infinity: everything that builds against tk8.3 in the officially proscribed manner
[08:41] <lifeless> infinity: such as the gnu-smalltalk blox environment
[08:41] <infinity> Does it cause them to FTBFS, or produce broken binaries?
[08:41] <lifeless> depends on the package
[08:41] <Treenaks> yay for braindead upstream?
[08:41] <lifeless> packages that /require/ tk will FTBFS
[08:42] <lifeless> packages that have optional features will have those features silently disappear thanks to autoconf being nice
[08:42] <lifeless> the latter is what lead me down this path
[08:42] <infinity> lifeless : Alright, find one of those and link up the build log, svp.  Then someone might be able to make sense of why it's a bug. :)
[08:42] <infinity> ("one of those" meaning "one that's FTBFS")
[08:42] <lifeless> infinity: uh
[08:42] <lifeless> why ?
[08:42] <infinity> Alternately, point us at a broken binary.  I don't care.  Either.
[08:42] <lifeless> try this : export CFLAGS="# this is not valid"
[08:42] <lifeless> then build *anything*
[08:43] <lifeless> it. wont. build.
[08:43] <infinity> Okay, fair enough.  But I wasn't aware until just now that packages actually use tkBuild.sh to produce CFLAGS lines.
[08:43] <lifeless> I've rebuilt tk locally via debuild, it generates a good tkConfig.sh file
[08:43] <lifeless> so I can't give a source package patch.
[08:44] <infinity> Keen.  And it almost certainly would have in the past as well.
[08:44] <lifeless> I have to presume that tk's configure script went haywire and build crud, and thats had a knock on effect once it landed in the archive
[08:44] <infinity> Alright, let me poke at it.
[08:45] <lifeless> what I think is the right thing is a simple rebuild of tk on all platforms with broken file, and then a rebuild of anything that people report bugs on.
[08:45] <lifeless> i.e. I'd like to have gnu-smalltalk rebuilt, as I know its broken - thats how I got into this path.
[08:45] <infinity> it's right there in unix/tcl.m4.... Curious.
[08:45] <daniels> lifeless: no-one uses smalltalk anyway
[08:45] <lifeless> daniels: pfft
[08:46] <Mithrandir> infinity: tcl.m4 makes modern automakes cry.  Loudly.
[08:46] <lifeless> infinity: fuxked upstream then... thats NOT a valid path for XINCLUDES
[08:46] <lifeless> infinity: right, line 1769
[08:47] <infinity> Mithrandir : I'm not surprised.  Have you read it recently?... It's making me cry.
[08:47] <lifeless> infinity: I can give a patch to set that to "" rather than "# no special path needed"
[08:47] <Mithrandir> infinity: no, I try to get rid of it from my system when I can.
[08:48] <Mithrandir> s/infinity/lifeless/
[08:48] <infinity> lifeless : The problem is a bit deeper than that.  Have you read tcl.m4?
[08:48] <lifeless> infinity: read the related bit just now
[08:48] <Mithrandir> lifeless: the day I could purge my system of tcl would be a very happy day.
[08:48] <lifeless> well, gnu-smalltalk 2.1.10 has gtk support
[08:48] <infinity> lifeless : It's a fundamental difference between what tcl.m4 thinks that variable will be used for and what it actually IS used for, AFAICT.
[08:48] <lifeless> which I'm about to enable in the package, so yay. but 2.1.8 is whats in hoary
[08:49] <lifeless> google tells me that it is meant to be a -I line
[08:49] <lifeless> and meant to be a valid one
[08:49] <infinity> It is in tkConfig.sh, yes.
[08:50] <infinity> (is meant to be, that is)
[08:50] <lifeless> tkConfig.sh.in is where tkConfig comes from
[08:50] <lifeless> I think this is a fallow bug and that tk was built with X headers in /usr/include or a pre-set CFLAGS or some such
[08:51] <lifeless> and we've simply finally exposed it
[08:51] <lifeless> would you like a patch to correct tcl.m4
[08:52] <infinity> Yeah, I see what they were trying to do.
[08:52] <infinity> And why they're retarded.
[08:52] <lifeless> lol
[08:52] <daniels> i'm all about exposing crappy code that used to work
[08:53] <infinity> I assume someone was hoping for the script to come out looking like:
[08:53] <infinity> TK_XINCLUDES='' # no special path needed
[08:53] <infinity> But just wasn't really paying attention to reality.
[08:54] <daniels> Mithrandir: i thought it was all called 'pkgconfig' these days
[08:54] <Mithrandir> daniels: nah, I'm moving it all to pkg-config from pkgconfig
[08:54] <lifeless> ok, I have patch
[08:55] <infinity> lifeless : Just changing it to a blank variable if it = "nope", I assume?
[08:55] <lifeless> and init to nothing
[08:55] <lifeless> theres about 4 places to change
[08:56] <infinity> Or, not "nope", but empty.  Whatever. :)
[08:57] <infinity> I only see two to change.
[08:57] <infinity> The "nope" instance should stay.  It's breakage if there are NO X includes found at all.
[08:58] <lifeless> uyhm
[08:58] <lifeless> no, AC_MSG_RESULT is the feedback for that
[08:58] <lifeless> if its meant to be a failure, that would be AC_MSG_FAIL
[08:59] <lifeless> well, either way
[08:59] <lifeless> the important bit is the two lines.
[08:59] <infinity> Oh, it should fail.  Sure.  And tcl.m4 should be completely rewritten too.  I'm not going to fix either. :)
[09:00] <infinity> But I'll fix the obvious breakage.
[09:01] <lifeless> ok.
[09:01] <infinity> Then I'll cheat and patch configure, cause i really don't want to see what an autoconf run will do to this.
[09:02] <lifeless> so the million dollar question .. will this rebuild get done for hoary ?
[09:02] <lifeless> as an update ?
[09:03] <infinity> Oh, nevermind.  I wouldn't be the first to rerun autoconf on this source recently.
[09:03] <infinity> Guess I won't cheat.
[09:04] <infinity> How many packages in hoary did it affect that would also need to be rebuilt?
[09:04] <infinity> You might want to put such a list together, and slap that into the bug report.
[09:04] <infinity> I'll fix it in breezy, though.
[09:04] <jdub> GOOOOOOOD MORNING FREEDOM LOVERS
[09:06] <Treenaks> hi jdub 
[09:06] <Mithrandir> what's scary is jdub is actually saying that approximately 90 minutes after I've gotten up, for the last week.
[09:07] <Mithrandir> and he's been travelling around half the world.
[09:07] <lifeless> whats the reverse-depends tool ?
[09:08] <fabbione> hey jdub 
[09:08] <infinity> lifeless : apt-cache rdepends <package>
[09:08] <Treenaks> Mithrandir: IRC scripting for fun & profit
[09:08] <lifeless> apt-cache rdepends tk8.3
[09:08] <lifeless> there you go.
[09:08] <Mithrandir> Treenaks: I doubt it, since he says stuff afterwards which is jdub-responses to stuff people say to him.
[09:09] <fabbione>   * Add support for KPKG_ARCH override required for powerpc64 kernels.
[09:09] <Treenaks> Mithrandir: scary
[09:09] <fabbione> ok.. now..
[09:09] <fabbione> who will pay beer for it? :)
[09:09] <Mithrandir> fabbione: svenl will?
[09:09] <Treenaks> fabbione: come & get it ;)
[09:11] <pitti> morning
[09:12] <pitti> fabbione: so you'll upload rhcluster to universe first?
[09:12] <fabbione> pitti: no.. it can't be uploaded to universe first...
[09:12] <fabbione> germinate will pull in automatically in main
[09:12] <fabbione> due to the b-d for lvm2
[09:12] <fabbione> so it doesn't not matter where i upload
[09:12] <fabbione> it will be sucked in main automatically
[09:13] <fabbione> at least the source
[09:13] <fabbione> and libdlm-dev libdlm1
[09:13] <fabbione> the others will land in universe and it's fine for me
[09:13] <jdub> heh
[09:13] <jdub> nutballs :)
[09:14] <fabbione> pitti: in any case it won't happen before the end of next week
[09:14] <fabbione> pitti: some init scripts are missing.. even upstream :)
[09:16] <pitti> fabbione: ah, ok. Well, I'm not in a hurry with it
[09:16] <fabbione> neither am i
[09:16] <fabbione> it's a quite delicate package to install all together
[09:17] <fabbione> since it affects the boot process in an early stage
[09:17] <fabbione> and there are some services startup order that still needs to be evaluated
[09:20] <lifeless>  |blt
[09:25] <jsgotangco> hehe yeah
[09:29] <jsgotangco> mvo, hi
[09:33] <pitti> Hi carlos
[09:33] <carlos> morning
[09:35] <mvo> hey jsgotangco, morning all
[09:43] <pitti> Riddell: here?
[09:51] <doko> mvo: ping
[09:52] <mvo> doko: pong
[09:56] <jdub> shit, anyone read polish?
[09:57] <jsgotangco> dzien dobry
[09:57] <jsgotangco> hehe
[09:58] <jsgotangco> opi is polish but he's nowhere now
[10:03] <jsgotangco> chmj, hey how's it going
[10:03] <robitaille> jdub,  #ubuntu-pl  (not sure if anyone is there...)
[10:07] <pitti> Hey seb128 
[10:07] <seb128> hey pitti!
[10:08] <seb128> how are you? 
[10:08] <robitaille> jdub,  forget about #ubuntu-pl...it's empty (http://www.ubuntulinux.org/wiki/LoCoTeamList is out of date...)
[10:21] <pitti> hey second seb
[10:22] <seb128_> hey hey pitti :)
[10:22] <seb128_> I just have my daily IP change at 10:15 now, grrrr
[10:24] <pitti> seb128_: you can't change that time?
[10:24] <seb128_> I can, I just forget to reconnect before going to bed
[10:24] <pitti> reconnecting at 4 am should be a good time...
[10:25] <seb128_> yep, but I'm not awake at this hour
[10:25] <pitti> seb128_: right, that's why it is a good time :-)
[10:25] <seb128_> bah, I'm not the only one
[10:25] <pitti> Hi vuntz_ 
[10:25] <seb128_> hey vuntz_ vuntz :)
[10:26] <vuntz_> hi all
[10:34] <pitti> daniels: do you know about the repetitive postinst failure of xbase-clients because of some xkbcomp weirdness?
[10:38] <vuntz_> pitti: isn't it the "yes, xbase-clients is broken.  no, it's not fixed yet." part of the topic?
[10:38] <pitti> oops, right
[10:38] <pitti> sorry
[10:39] <vuntz_> :-)
[10:44] <doko> ogra: ping
[10:44] <doko> pitti: fabbione did give him a deadline ... ;-)
[11:03] <fabbione> doko: no i didn't :)
[11:03] <fabbione> but i am sure he will love to update asap, before strange things will happen to his house :P
[11:06] <pitti> fabbione: like a 20 meters high solid metal "X" falling on his roof?
[11:08] <fabbione> pitti: ahahha
[11:10] <doko> we should sponsor an ad in a local newspaper: free beer at daniels until X is fixed ;-)
[11:12] <Treenaks> :D
[11:22] <jdub> whoa, libmono in main
[11:24] <jsgotangco> JaneW, hi
[11:35] <jk24> Hi, i think there is a problem with xfonts-base : when it install it does a /usr/X11R6/bin/mkfontscale (i don't know where in the postinstall) but xutils install mkfontscale on /usr/bin, then after a dist-upgrade, Xorg cannot found the fixed font
[11:35] <jdub> jk24: see topic :)
[11:36] <jk24> jdub, not xbase-clients, xfont-base !
[11:36] <jk24> :)
[11:37] <pitti> seb128: I now have a prototype that dumps a /tmp/crashrep.<program>.<time>.txt with a crash report after a sigsegv/sigill/sigfpe
[11:38] <seb128> pitti: rock
[11:39] <seb128> pitti: is that public somewhere?
[11:39] <vuntz> pitti: will this replace bug-buddy or is it only for non-gnome programs?
[11:41] <pitti> seb128: I package it soon
[11:42] <pitti> vuntz: depends, this is still to be decided, I guess
[11:43] <vuntz> poor seb128... you will now have to handle all the GNOME crashes of ubuntu users...
[11:45] <vuntz> bugzilla really has a scaling problem
[11:45] <Kamion> s/scaling //
[11:46] <Kamion> actually s/has/is/
[11:47] <Mithrandir> Kamion: don't be mean to bugzille. :P
[11:47] <Kamion> aww
[11:47] <Mithrandir> uhm, s/lle/lla/
[11:47] <pitti> vuntz: believe it or not, but seb128 bugged me to write this stuff :-)
[11:48] <pitti> s/bugged/asked kindly/ :-)
[11:48] <seb128> vuntz: what I want is automatic debug backtrace, because replying to 80% of the bug to say "install package1-dbg package2-dbg, read the debug page and send a debug backtrace" is not really funny
[11:48] <vuntz> seb128: nod
[11:48] <pitti> seb128: right now the package does not fetch any debug information from the net (well, there is none so far)
[11:48] <pitti> seb128: once there is, I can enhance it to do so
[11:48] <seb128> rock
[11:48] <Riddell> pitti: hi
[11:49] <pitti> Hi Riddell
[11:49] <vuntz> pitti: great!
[11:49] <pitti> Riddell: I wanted to ask you something about KDE langpacks
[11:50] <pitti> Riddell: in the spec it was decided to split up the langpacks into -gnome, -kde, and -other
[11:50] <Riddell> yep
[11:50] <pitti> Riddell: so what should we do about kde-i18n-foo?
[11:50] <pitti> Riddell: treat them specially and make them a dependency of l-p-kde-<lang>?
[11:51] <pitti> Riddell: or just strip them normally and integrate the translations into the kde langpack?
[11:51] <pitti> Riddell: the latter has the nasty effect that we have a bunch of empty kde-i18n debs around
[11:51] <pitti> although, we don't 
[11:51] <pitti> Riddell: kde-i18n also contains translated doc and help, right?
[11:52] <Riddell> pitti: they do
[11:52] <Riddell> pitti: if they weren't stripped they couldn't contain files from rosetta
[11:52] <pitti> Riddell: yes, I would like to strip them, too
[11:52] <pitti> Riddell: so that we only need to update the langpacks, not the kde-i19n packages after a release
[11:53] <pitti> erm, ECOUNT
[11:53] <Riddell> :)
[11:53] <Riddell> pitti: how would the langpacks be updated without updating kde-i18n since that's where the updates would come from
[11:53] <pitti> Riddell: I meant updates from Rosetta
[11:54] <Riddell> pitti: but no updates from kde?
[11:54] <pitti> Riddell: the upstream updates are imported into Rosetta automatically throught the stripping and import processs
[11:54] <Riddell> pitti: but that process is only when a new package is uploaded?
[11:54] <pitti> Riddell: yes, right
[11:55] <pitti> Riddell: well, for Gnome I had a special script that pulled new translations from gnome cvs and put it into the langpacks
[11:55] <pitti> Riddell: if we need sth similar for KDE, this can be done, I think
[11:55] <Riddell> pitti: that sounds not too difficult
[11:56] <Riddell> pitti: what happens with all the other gnome programmes?  if gnome 2.10.x comes out does ubuntu get the updates language packs from that?
[11:56] <pitti> Riddell: if I crank up my script, we can do that
[11:56] <pitti> however, not automatically
[11:57] <pitti> unless the Rosetta guys (hi carlos) want to integrate this script into Rosetta proper
[12:12] <carlos> pitti, we will integrate a way to get translations from upstream after an Ubuntu release, but we don't have an dates yet
[12:13] <pitti> oh, cool
[12:18] <pitti> seb128: http://people.ubuntu.com/~pitti/crashrep/ , testers appreciated
[12:18] <seb128> thanks
[12:18] <Treenaks> pitti: whazzit?
[12:18] <pitti> seb128: basically, just install the two debs, and every newly started app should then generate a report
[12:18] <pitti> (when it crashes, that is)
[12:18] <seb128> should I drop bug-buddy?
[12:18] <Treenaks> pitti: ooh.. will it ask if I want to submit that? :)
[12:19] <pitti> seb128: I'm still testing the b-b interaction and if it is bad, then I add a conflict
[12:19] <pitti> Treenaks: there is code, but I disabled it for now
[12:19] <pitti> Treenaks: I set up a preliminary server on http://debcrash.piware.de
[12:19] <pitti> Treenaks: it already has some test reports, but we need to spec this out
[12:19] <pitti> Treenaks: for now you'll only get a flat file in /tmp
[12:19] <jdub> pitti: couldn't we integrate the two?
[12:20] <pitti> jdub: two?
[12:20] <pitti> ah, b-b and crashrep
[12:20] <jdub> b-b and crashrep
[12:20] <seb128> use bug-buddy as the UI
[12:20] <pitti> seb128: but b-b is ugly for beginners...
[12:20] <seb128> what is so ugly with it?
[12:20] <pitti> anyway, we should talk about the gui once the report system is stable
[12:20] <pitti> right now it isn't
[12:21] <pitti> it fails to generate stack traces sometimes
[12:21] <Treenaks> pitti: where does it get the symbols?
[12:21] <pitti> and we should have debug symbols before we actually throw this at people
[12:21] <pitti> Treenaks: right now it can't
[12:21] <pitti> Treenaks: later, from debug.ubuntu.com or so
[12:21] <Treenaks> coolness!
[12:21] <pitti> Treenaks: we want to strip the symbols on the buildds and put them there
[12:22] <Treenaks> pitti: yeah, I heard something like that
[12:22] <pitti> seb128: the main problem is that b-b is restricted to desktop applications, doesn't work with servers and for text mode
[12:23] <Treenaks> pitti: how about reportbug then?
[12:23] <pitti> seb128: and we have to modify it heavily to work with our db instead of submitting to gnome
[12:23] <pitti> seb128: I think modifying it is possible
[12:23] <seb128> for my part of the issue, which is "stop flooding us and upstream with zillion of useless bugs" that's fine
[12:23] <pitti> seb128: but crafting an easy pygtk interface should be easier
[12:23] <pitti> seb128: the code already has a small pygtk gui
[12:23] <seb128> you want to send all the crasher to bugzilla.ubuntu?
[12:23] <pitti> no
[12:23] <seb128> do you have an idea of the number of crashers/week send by ubuntu users? :)
[12:24] <pitti> seb128: they should becollected in something like http://debcrash.piware.de (that actually works right now!)
[12:24] <pitti> seb128: nope
[12:24] <seb128> nautilus get about 7-8 ubuntu crasher/days
[12:24] <seb128> and that's one module
[12:24] <pitti> seb128: if you uncomment the commented code in debcrash's main and disable the file_report, you'll get the gui
[12:24] <seb128> I guess that's some hundreds crashers/week 
[12:24] <seb128> vuntz|miam: any idea on this number?
[12:25] <pitti> we shouldn't flood bz with that, at least not unfiltered
[12:25] <seb128> how do you want to filter that?
[12:25] <seb128> automatically dup?
[12:25] <mvo> we talked about filters in the spec, it's a must IMHO :)
[12:25] <pitti> well, filter by package, version, signal, and maybe by the function the crash occurred in
[12:26] <seb128> a filter doesn't lower the number of bugs
[12:26] <seb128> automatic close of dup is probably not a piece of cake
[12:27] <seb128> especially than dups are often useful to get new informations on a bug sleeping as NEEDINFO
[12:28] <pitti> Treenaks, seb128: http://people.ubuntu.com/~pitti/crashrep-gui.png
[12:28] <pitti> it is *ugly* right now, but a demo
[12:28] <tseng> white screen of death!
[12:28] <pitti> the information needs to be presented in a much better form, but the gui shouldn't get any more complicated
[12:29] <seb128> right
[12:29] <seb128> bug-buddy is not really complicate
[12:31] <seb128> pitti: what is bad about the crash dialog from bug-buddy? That's rather like your screenshot
[12:32] <pitti> seb128: it's not really bad, but b-b has several screens of which most of them aren't interesting (gnome db update and such)
[12:32] <pitti> seb128: if we want to use our own db we need to rewrite half of the code, and I don't like to do it in C
[12:32] <pitti> seb128: but of course I will be fine with modifying bz if sb wants to do that
[12:32] <pitti> s/bz/bb/
[12:33] <pitti> seb128: still, we should first get the file dump perfect
[12:33] <seb128> I don't really mind, but jbailey made the xmlrpc changes to bug-buddy just before hoary
[12:33] <seb128> so it can speak to bugzilla.ubuntu.com by xmlrpc by example
[12:33] <seb128> right
[12:33] <pitti> yes
[12:34] <pitti> seb128: files should already help us a lot, and we should get the debug stuff up and running
[12:34] <pitti> seb128: without the symbols the traces are utterly useless
[12:34] <seb128> that's my concern, I want debug backtraces out of the box
[12:34] <seb128> the other points are bonus after that
[12:34] <pitti> right
[12:35] <pitti> seb128: let's try to get this file thing working for breezy
[12:35] <seb128> yep
[12:35] <pitti> seb128: did you already talk to lamont?
[12:35] <seb128> I'll have a look on the buildd changes soon
[12:35] <seb128> nop, planned for this afternoon
[12:35] <pitti> cool
[12:36] <pitti> seb128: I do some more testing now; the python code for trace acquisition looks rather ugly right now, but the much simpler version with Popen just doesn't work
[12:39] <jdub> dear daniels, i am using windows until X is fixed. love, jdub.
[12:39] <seb128> pitti: k
[12:39] <seb128> pitti: another topic, are #11157 and #11242 for you? (audio hangs on pause by example)
[12:40] <pitti> jdub: "use the text console, Luke"
[12:40] <pitti> seb128: ah, on rhythmbox?
[12:40] <seb128> yep
[12:40] <seb128> pause play pause
[12:40] <seb128> == hang
[12:40] <seb128> dunno if that's a alsasink issue or an alsa one
[12:41] <pitti> seb128: didn't dig into that so far
[12:42] <pitti> seb128: some other guy on the ML mentioned that directly using alsa will cause problems, so the only thing that is left is polypaudio
[12:42] <jordi> is polypaudio going for breezy again?
[12:42] <pitti> jordi: Erik de Castro told me that he fixed some major bugs in polypaudio
[12:42] <seb128> we have alsa 1.0.9
[12:42] <jordi> not that there's a lot of change wrt 1.0.9rc3 anyway
[12:43] <pitti> well, we have rc3 so far
[12:43] <jordi> ah.
[12:43] <jordi> pitti: I'll see if I can release stuff to exp during the evening
[12:43] <pitti> hehe, thanks
[12:43] <jordi> Debian maintainers are worthless
[12:43] <pitti> yeah, they are such slackers
[12:43] <jordi> clueless, can't even mail mitre to get CAN assigned :)
[12:44] <pitti> jordi: uh, yeah, it arrived :-)
[12:45] <jordi> pitti: yup, I updated my changelog last night
[12:46] <pitti> baz add *
[12:46] <pitti> Failed to add id for path '{arch}'
[12:46] <pitti> bah, can't this be fixed eventually?
[12:46] <jordi> it's still not accessible from the CVE webpage tho
[12:47] <pitti> no, that usually lasts a while
[12:47] <seb128> anybody here has tried serpentine so far?
[12:48] <seb128> pitti: k, that's a no :)
[12:48] <jordi> *H*
[12:48] <seb128> pitti: have you heard about it?
[12:48] <jordi> err. s/H/G/
[12:48] <ogra> pitti, i think we have a bug in g-v-m that affects the burning app that use the nautilus backend...
[12:48] <pitti> seb128: not really
[12:48] <seb128> pitti: k
[12:48] <ogra> apps even
[12:48] <pitti> ogra: $ serpentine
[12:48] <pitti> Traceback (most recent call last):
[12:48] <pitti>   File "/usr/bin/serpentine", line 34, in ?
[12:48] <pitti>     from serpentine import SerpentineApplication, SerpentineError, operations, gtkutil
[12:48] <pitti>   File "/usr/lib/python2.4/site-packages/serpentine/__init__.py", line 29, in ?
[12:48] <pitti>     from mastering import AudioMastering
[12:49] <pitti>   File "/usr/lib/python2.4/site-packages/serpentine/mastering.py", line 24, in ?
[12:49] <seb128> ah ah
[12:49] <pitti>     import operations, audio, xspf, constants, gtkutil
[12:49] <pitti>   File "/usr/lib/python2.4/site-packages/serpentine/xspf.py", line 25, in ?
[12:49] <pitti>     from xml.xpath import Evaluate
[12:49] <pitti> ImportError: No module named xpath
[12:49] <seb128> thanks pitti :)
[12:49] <ogra> ugh
[12:49] <seb128> what I was saying, 0 feedback
[12:49] <seb128> I'm sure there is load of bug, it doesn't work for me neither
[12:49] <pitti> "add dependencies, Luke"
[12:49] <seb128> yep
[12:49] <ogra> seb128, i know of about 10 people that tested it successfully now
[12:49] <ajmitch> I've used it & it seemed to work
[12:49] <seb128> make them mailing a list?
[12:50] <seb128> so other people have the feedback too
[12:50] <seb128> we have 0 bugzilla entry about it
[12:50] <ogra> but g-v-m 1. blocks the access to libnautilus-burn and 2. slows down ripping like hell
[12:50] <seb128> and there is a ~10 mails post about the move to main
[12:50] <seb128> how so?
[12:50] <pitti> ah, now it starts
[12:51] <pitti> ogra: add Depends: python-xml, please
[12:51] <ogra> seb128, you have to disable autostart for the CD to speed up ripping it seems
[12:51] <seb128> pitti: I'm the maintainer :p
[12:51] <seb128> ogra: that's weird
[12:51] <pitti> oh
[12:52] <ogra> seb128, and  have no idea et what blocks the burning.... my hal/g-v-m journey starts next week... this week is screensavr :)
[12:52] <vuntz> seb128: 109 bugs open by ubuntu users in one week
[12:52] <pitti> seb128: looks rather nice
[12:52] <seb128> vuntz: k, thanks :)
[12:53] <pitti> but a lot of exceptions
[12:53] <vuntz> seb128: most of them are NEEDINFO
[12:53] <ogra> pitti, which ones ?
[12:53] <pitti> ogra, seb128: http://www.rafb.net/paste/results/aFBEBx35.html
[12:54] <vuntz> seb128: 19 UNCONFIRMED, 1 FIXED, all others are DUPLICATE/NEEDINFO/INCOMPLETE...
[12:54] <ogra> pitti, thats intresting, since it worked for the other testers in here
[12:55] <seb128> vuntz: do you have something to generate such stats or that's just bugzilla queries?
[12:55] <vuntz> seb128: just queries
[12:55] <ogra> pitti, i havent seen any tracebacks yet.... only bugs that stopped it from burning and were g-v-m related
[12:55] <vuntz> seb128: I also chose a random week, but I suppose it's nearly always the same
[12:55] <seb128> ogra: put an not-empty CD and try to record
[12:56] <seb128> vuntz: ok, thanks
[12:56] <ogra> pitti, could you open a bug ? 
[12:56] <pitti> ogra: sure
[12:56] <ogra> seb128, heh, thats obviously a usecase we didnt have yet :)
[12:56] <pitti> ogra: but it doesn't really crash, it just looks a bit scary
[12:57] <ogra> pitti, i'll look into it and talk with upstream about error messages ;)
[12:57] <ogra> assign it to me ...
[12:57] <pitti> ogra: so what's the bug title? "serpentine: various exceptions"?
[12:58] <ogra> heh, yep
[12:58] <seb128> ogra: my point is "why is there no public discussion" ?
[12:58] <seb128> apparently you have pinged a bunch of user
[12:58] <seb128> but made no feedback public
[12:58] <pitti> bah, there is no "serpentine" package in bz
[12:58] <ogra> seb128, they pinged me, in here
[12:58] <ogra> pitti, t waits for python-gnome2-extras
[12:59] <seb128> what about python-gnome2-extras?
[12:59] <ogra> it is not in main
[01:00] <pitti> ogra: #11447
[01:01] <ogra> pitti, thanks
[01:13] <jordi> pitti: hey, we've decided to upload 1.0.9 to unstable :)
[01:13] <pitti> cool
[01:14] <Lathiat> hmm shtoom wants twisted2 and thats not in hoary, d'oh.
[01:37] <bod> Kamion: that sync the other day from experimental, is that now tracked?
[01:38] <seb128> pitti: the rhythmbox hangs seems to be fixed with the new gst-plugins0.8, so probably an alsasink issue
[01:38] <pitti> ah, cool
[01:38] <pitti> seb128: will check after a dist-upgrade, it occurs to me, too
[01:38] <seb128> k
[01:38] <seb128> jbailey: have you read https://bugzilla.ubuntu.com/show_bug.cgi?id=11146 ?
[01:39] <bod> if not, could you sync again please?  5.8.7 in experimental
[01:39] <bod> elmo: about?
[01:40] <pitti> seb128: hm, I don't even have this package installed
[01:40] <pitti> seb128: still, rhythmbox hangs
[01:40] <seb128> pitti: that's the source package, gstreamer0.8-alsa is the binary
[01:40] <pitti> ah
[01:40] <seb128> and I've just uploaded like 5 min ago
[01:40] <jbailey> seb128: I have.  Now that cdbs is building again, I can look at the cdbs bugs.
[01:40] <seb128> 0.8.9
[01:40] <seb128> jbailey: k, thanks
[01:40] <pitti> seb128: ah, that's why I still ecperience it, but I'm up to date
[01:40] <seb128> yep, wait a bit
[01:40] <seb128> get a lunch or something :)
[01:41] <pitti> jbailey: btw, I just uploaded a new cdbs with a fixed gnome.mk
[01:41] <jbailey> seb128: Do you need it for today?
[01:41] <seb128> jbailey: not at all
[01:41] <jbailey> pitti: Already uploaded or about to?
[01:41] <Kamion> bod: no, experimental isn't tracked automatically
[01:41] <pitti> jbailey: last time I erroneously put in an old version
[01:41] <pitti> jbailey: uploaded
[01:41] <jbailey> pitti: Ah well.  (Sorry sb, I tried. *g*)
[01:41] <Kamion> elmo: please sync perl 5.8.7-1 from incoming
[01:41] <pitti> jbailey: bad?
[01:42] <seb128> jbailey: np :)
[01:42] <seb128> pitti: no, was just about #11146 I guess
[01:43] <jbailey> pitti: No worries, as seb128 said.  If it was about to, I'd have gotten you to add the 1 line patch from 11146 =)
[01:44] <bod> Kamion: thanks.  Got the impression that it wasn't automatically synced in general, but wasn't sure if the sync from experimental was "sticky" if done once
[01:44] <fabbione> thom: ping?
[01:45] <Kamion> bod: nah, syncs from unstable happen automatically if there are no Ubuntu changes (technically, if the version in breezy doesn't match /ubuntu/), all else is manual
[01:45] <bod> k
[01:55] <Kamion> elmo: please sync libgetargs-long-perl from unstable
[02:01] <lifeless> jbailey: ping
[02:01] <jbailey> lifeless: here.
[02:02] <lifeless> can you fix /usr/lib/tk8.3/tkConfig.sh 
[02:02] <jbailey> lifeless: Finally finished your workday? =)
[02:02] <jbailey> 'fix' ?
[02:02] <lifeless> )
[02:03] <lifeless> https://bugzilla.ubuntu.com/show_bug.cgi?id=11444
[02:03] <lifeless> theres a patch there for installed tk's ;)
[02:03] <lifeless> or by hand, line 46 is corrupt
[02:03] <lifeless> should be TK_XINCLUDES=''
[02:04] <infinity> lifeless : It's already been fixed in breezy.
[02:04] <doko> jbailey: rebuild with the "xorg in transition state" should fix it
[02:04] <lifeless> infinity: I know
[02:04] <lifeless> infinity: this is on his ia64, so unless its actually hit the archive
[02:04] <lifeless> ...
[02:05] <jbailey> lifeless: The ia64 is on debian...
[02:05] <lifeless> jbailey: heh, has the same problem though
[02:05] <infinity> Ahh.
[02:06] <Amaranth> whew
[02:06] <Amaranth> ubuntuguide doesn't tell people to use marillat anymore
[02:06] <lifeless> infinity: could you file that particular patch in debian, or would you like me to ? ( I want to unbreak gst there too)
[02:07] <infinity> lifeless : Too late for sarge, unless you can argue that it's so incredibly RC is should delay the release by a few days.
[02:08] <infinity> lifeless : And good luck with that argument. :)
[02:08] <lifeless> infinity: did I mention sarge ?
[02:09] <infinity> Oddly, no.  I'm just very used to people in the last two weeks begging for last minute bugfixes.
[02:09] <lifeless> noone uses debian stable
[02:10] <infinity> Say... Why is your beloved app compiling against an obsolete tk version anyway?
[02:11] <infinity> Didn't everyone transition to 8.4 in, like, 2002?
[02:12] <lifeless> infinity: well now, thats an interesting question
[02:12] <lifeless> I'll see if its just maintainer slackness
[02:12] <jbailey> lifeless: That was one of the questions I had in the packaging. =)
[02:12] <jbailey> That and why the useless {post,pre}{rm.inst} scripts. =)
[02:13] <infinity> lifeless : tk8.4 has had this bug fixed for ages, it looks like.
[02:13] <infinity> lifeless : Anyhow, I'll push the tk8.3 fix up sometime, but I doubt anyone cares much, since it is obsolete.
[02:14] <lifeless> I'll see about updating the package then
[02:22] <mvo> ping enrico 
[02:26] <enrico> mvo: pong!
[02:26] <enrico> mvo: just back from lunch
[02:26] <enrico> mvo: what can I do for you?
[02:54] <pitti> Hi Nafallo 
[03:02] <jdong> hey guys
[03:02] <Amaranth> hey
[03:02] <jdong> just wanted to bring to your attention that blackbox and fluxbox conflict
[03:02] <jdong> bsetroot's manpage is in both....
[03:02] <jdong> i don't know how you guys would handle it; but it's definitely a packaging issue :)
[03:03] <Amaranth> jdong: don't suppose you could accept smeg and my new gnome-menus in backports :)
[03:03] <Amaranth> jdong: switching gnome-menus to the backports style version (~5.04ubp) makes it not break gnome-devel
[03:04] <jdong> Amaranth: does your gnome-menus break ABI compatibility, though?
[03:04] <Amaranth> not sure if the two functions removed are available externally
[03:04] <Amaranth> i didn't remove them, they were removed on the 2.10 cvs branch
[03:04] <Amaranth> so i'd say no
[03:04] <jdong> Amaranth: why isn't smeg in universe yet, btw
[03:04] <Amaranth> jdong: takes 48 hours
[03:05] <jdong> Amaranth: oh, so you just tried to get smeg in universe?
[03:05] <Amaranth> yeah, 12 hours ago
[03:05] <jdong> Amaranth: well, if everyone around here is fine with the new gnome-menus, fine, backports it is, as soon as I see the universe packages :)
[03:06] <Amaranth> MOTUNewPackages shows that it was uploaded, you can just be faster than breezy ;)
[03:06] <jdong> lol
[03:06] <jdong> k, where's you gnome-menus?
[03:06] <Amaranth> the gnome-menus package is basically what 2.10.2 will be, if/when it's released
[03:07] <Amaranth> http://dev.realistanew.com/gnome-menus
[03:07] <jdong> k
[03:07] <Amaranth> src dir has diff.gz, .dsc, and .orig.tar.gz so you can build it yourself
[03:07] <fabbione> elmo, thom: i really really need one of you 
[03:07] <enrico> _mvo_: welcome back
[03:08] <enrico> _mvo_: if you want to tell me about libtagcoll1 problems, I'm here
[03:08] <jdong> Amaranth: k
[03:08] <mvo> enrico: my network is unstable :/
[03:08] <jdong> Amaranth: 0.7.4?
[03:09] <jdong> Amaranth: I'm gonna use  ubp-build.py http://dev.realistanew.com/smeg/latest/smeg.tar.gz
[03:09] <jdong> Amaranth: diff.gz patches, location?
[03:10] <jdong> never mind again :)
[03:10] <Amaranth> http://dev.realistanew.com/smeg/0.7.4/
[03:10] <Amaranth> i made a forum thread about this ;)
[03:11] <jdong> Amaranth: uploading :)
[03:11] <Amaranth> cool
[03:16] <Lathiat> jdong: plz to have sources
[03:17] <jdong> Lathiat: huh?
[03:17] <Lathiat> jdong: arent you the backports guy?
[03:17] <jdong> Lathiat: yes
[03:17] <Lathiat> jdong: Then, please to have source archives for backports :)
[03:18] <jdong> Lathiat: this will happen naturally when I transition to the Ubuntu archive system
[03:18] <Amaranth> Lathiat: he will, when he moves to ubuntu servers ;)
[03:18] <Lathiat> oh you are?
[03:18] <Lathiat> sweet
[03:18] <jdong> Lathiat: for now, I see it as a unnecessary waste of space, time, and bandwidth
[03:18] <Lathiat> jdong: i see it as a possible license violation
[03:18] <jdong> Lathiat: all the sources are exactly the same, minus the version number change
[03:18] <Lathiat> jdong: so, thats a change :) depends what license debian/ is under. :)
[03:19] <Lathiat> and more to the point, its convenient for me when i want the sources to something :)
[03:19] <Lathiat> but if your migrating to shit and then youo'll have it then rock on
[03:19] <jdong> Lathiat: why not deb-src Breezy?
[03:19] <Lathiat> jdong: well not all your stuff is from breezy for example
[03:19] <Lathiat> (at least, if you include extras)
[03:19] <jdong> Lathiat: either way; I will have sources soon....
[03:19] <Lathiat> jdong: like i said, cool, rock on :)
[03:19] <jdong> Lathiat: extras will have to be introduced into universe
[03:20] <jdong> Lathiat: what packages, specifically ,are you interested in?
[03:20] <Lathiat> jdong: isnt that an issue for some of the media stuff ?
[03:20] <Lathiat> or java or something
[03:20] <jdong> Lathiat: Extras will continue for restricted formats
[03:20] <Lathiat> (im not sure what the issues are there)
[03:20] <Lathiat> jdong: ah ok
[03:20] <jdong> Lathiat: though I'm scared what legal issues I can run into :)
[03:20] <Lathiat> just host it in sweden :)
[03:21] <jdong> Lathiat: lol!
[03:22] <Lathiat> (that was mostly serious)
[03:22] <ogra> jdong, btw, do you build your packages already in a pbuilder ? you should make yourself familiar with it
[03:22] <jdong> Lathiat: combine forces with debian-marillat
[03:22] <Lathiat> jdong: sounds like a plan...
[03:22] <jdong> ogra: pbuilder who?
[03:22] <ogra> jdong, the build envoronment that imitates a buildd....
[03:23] <jdong> ogra: yeah, reading docs. What's the difference between this and a chroot?
[03:23] <ogra> it checks your dependencys ;)
[03:24] <jnc> psst, amd64 devs;  bug #11453
[03:24] <jdong> ogra: isn't that just an dpkg-buildpackage and apt-get build-dep ordeal?
[03:24] <ogra> jdong, http://www.ubuntulinux.org/wiki/PbuilderHowto
[03:25] <jdong> ogra: this looks like a wrapper around what I do already :)
[03:25] <ogra> jdong, the nice thing is that if you know it builds reliable in a pbuilder, you can very much rely n the fact it will build on the buildd too.... where using a chroot that already holds dependency packages might fool you
[03:25] <jdong> ogra: ah, ok
[03:26] <jdong> ogra: I see how that works. So it's more useful if you're making a new package, right?
[03:26] <ogra> i assume you dont rebuild your chroot everytime ;)
[03:26] <jdong> ogra: once every week or two, I do
[03:26] <jdong> ogra: I have a few scripts for that ;)
[03:26] <ogra> nope, if you build any source package :)
[03:26] <jdong> ogra: what's the chances deps for Firefox will change from 1.0.3 to 1.0.4?
[03:26] <ogra> jdong, was just a hint, makes the life a lot easier if you work with buildds
[03:27] <jdong> ogra: ok, thanks :). I'll see how it'll fit in with my build scripts :)
[03:27] <ogra> jdong, that a thom question 
[03:27] <jdong> ogra: laziness is keeping me at my scripts now
[03:27] <ogra> yep
[03:27] <jdong> so ogra, can I relay that fluxbox/blackbox bug to you guys?
[03:28] <ogra> jdong, just file it
[03:28] <jdong> ogra: I'm having trouble using malone
[03:28] <jdong> ogra: system error
[03:28] <Amaranth> yeah, malone was broken last night too
[03:29] <jdong> screen -r
[03:29] <Amaranth> sudo pbuilder build file.dsc and away it goes
[03:29] <jdong> ugh wrong terminal
[03:29] <ogra> jdong, then defer it :) in any case all bugs should end up on one of our bugtrackrs :)
[03:29] <Amaranth> drops the results in /var/pbuilder/results/, iirc
[03:29] <jnc> err who does ia32 libs?
[03:30] <jdong> Amaranth: ubp-build.py packagename and away it goes... Drops it into the Ubuntu Backports mirror system iirc ;)
[03:30] <Amaranth> heh
[03:30] <Amaranth> can i have that script ;)
[03:31] <Nafallo> jnc: seems to be doko atm :-)
[03:31] <jnc> bwahaha
[03:31] <jnc> let him deal w/ it
[03:31] <jdong> Amaranth: http://ubuntubackports.org/ubp/projects/ubp-tools/ubp-build/
[03:31] <jdong> Amaranth: still need a svn password to attack me, though ;)
[03:31] <Amaranth> nice
[03:32] <Amaranth> you even use psyco :)
[03:32] <jnc> doko: hey buddy, somebody broke ia32-libs *cough*  bug #11453
[03:32] <jdong> Amaranth: of course :) Why not?
[03:32] <Amaranth> jdong: i'll modify this to use pbuilder
[03:32] <jdong> Amaranth: pysco backport, too ;)
[03:32] <jdong> Amaranth: awesome, that'd be interesting to have
[03:32] <jdong> Amaranth: not really worth the effort though, depending on how long it'll be before I get on the the Ubuntu system
[03:33] <doko> Nafallo, jnc: it's not broken, it's waiting
[03:33] <jdong> Amaranth: which I assume the uploading procedure (at the very least ) will be very different
[03:33] <Amaranth> why did you write your own argument parser?
[03:33] <Amaranth> well, this could be useful for me
[03:33] <jdong> Amaranth: I didn't know how else to parse arguments
[03:33] <jdong> Amaranth: I'm a VERY traditional argv person :)
[03:33] <Amaranth> hmm, maybe optparse wouldn't be useful there
[03:34] <jdong> Amaranth: not like there's any arguments that need complicated parsing
[03:34] <jdong> alright guys, I have to head back to class
[03:34] <jdong> have fun :)
[03:35] <jnc> doko: okay want to mark that bug invalid or should i do it?
[03:35] <Nafallo> doko: I know :-). but you're still the one that cares for it atm, and that was the question ;-)
[03:35] <doko> jnc: mark it pending
[03:35] <jnc> thanks
[03:37] <jnc> if a dep is broken, does that make it worthy of "blocker" status?
[03:37] <jnc> as a gentoo dev i usually get annoyed when users mark things with high priority
[03:38] <Amaranth> lmao
[03:38] <Amaranth> uh_die()
[03:38] <jnc> our deps are not broken as often as packages stop building for no apparent reason
[03:39] <jnc> so, had to ask
[03:41] <Kamion> very little justifies blocker status
[03:41] <jnc> 10-4
[03:41] <Kamion> we'd prefer that only to be used as a release management tool; we generally downgrade anything from blocker immediately
[03:42] <tepsipakki> is ACPI enabled on the hoary live-dvd?
[03:42] <tepsipakki> by default
[03:42] <Kamion> should be
[03:42] <tepsipakki> ok, my fellow worker has had problems with FC on a Thinkpad T42p ;)
[03:43] <tepsipakki> so I burned him that dvd
[03:43] <tepsipakki> burnt, even
[03:44] <jnc> get teh aloe vera, 'cause someone got BURNED
[03:44] <Kamion> wow, I managed to break grub-installer with MountingHddFilesystems
[03:44] <Kamion> that was impressive
[03:44] <Kamion> s/Hdd/HDD/
[03:44] <Kamion> unintended-consequences-'r'-us ...
[03:45] <Treenaks> Kamion: WikiStyle!
[03:45] <Kamion> Treenaks: *shrug* I didn't create the page
[03:52] <doko> elmo, Kamion, mdz: gcc-4.0 needs NEW/main love for lib32z1-dev and lib32z1
[04:02] <SloMo_> Amaranth: I'm currently building your smeg stuff for ppc... when everything works fine I'll upload later to backports
[04:02] <Kamion> jnc: ... but trivial's for spelling mistakes and stuff, dependency problems would always be >= normal
[04:02] <Amaranth> SloMo_: cool
[04:02] <Amaranth> SloMo_: you mean gnome-menus, right?
[04:03] <Kamion> if in doubt, file bugs at normal (unless they're clearly enhancement) and let somebody else sort it out :)
[04:03] <Amaranth> SloMo_: the rest is platform independent
[04:03] <SloMo> Amaranth: yes
[04:08] <nanomad> just a question...xbase-clients (broken now, i know...) also affects keys?
[04:08] <SloMo> Amaranth: have you built a smeg*~5.04ubp1.deb? only found the breezy version... otherwise I'll just rebuild
[04:08] <Amaranth> SloMo: jdong did
[04:10] <SloMo> Amaranth: yeah right... I'll upload the gnome-menus stuff now, seems to work fine ;)
[04:10] <Amaranth> fixes many things too
[04:10] <Amaranth> if only it fixed layout support...
[04:11] <SloMo> layout support?

[04:11] <Amaranth> part of the menu spec
[04:11] <SloMo> ah ok
[04:11] <Amaranth> that's why you can't reorder things if you're using gnome 2.10
[04:15] <SloMo> it is fixed in gnome-menus HEAD, right? or is this fix for another problem:
[04:15] <SloMo> 2005-05-30  Mark McLoughlin  <mark@skynet.ie>
[04:15] <SloMo> 	Fix problem where menus and items mentioned in a <Layout>
[04:15] <SloMo> 	after a <Merge type="menus"> or <Merge type="files">
[04:15] <SloMo> 	Bug #305723
[04:21] <Amaranth> that's the fix for a problem with <Layout> in HEAD that i reported
[04:21] <Amaranth> but <Layout> is supported in HEAD, yes
[04:23] <jnc> doko, elmo, Kamion, mdz:  FYI the bug for the lib32z thing is #11453, and i have marked it as suggested.
[04:24] <SloMo> Amaranth: are you interested in german translations for smeg? ;) seems there is currently no support for gettext, but is it planned?
[04:24] <Amaranth> it's planned, once i figure out how all the other python projects do it
[04:24] <Amaranth> they all seem to have their own way
[04:25] <Amaranth> i'm collecting translations though, please send them to alleykat@gmail.com
[04:25] <Amaranth> make a list of the english text and the german for it
[04:25] <Amaranth> oh, and make sure you say it's german, i have one that i'm totally lost on because i wasn't told the language in the email
[04:29] <SloMo> Amaranth: ok, I'll send the translations later this day
[04:29] <Amaranth> cool
[04:31] <wasabi_> Hmm what happened to the better Ubuntu branded xscreensaver
[04:33] <dilinger> i'd like to request the bouncing cow be replaced with other 3d ubuntu-ish things.  Bouncing ubuntu logos, bouncing naked people...
[04:34] <thoreauputic> bouncing developers....
[04:34] <siretart> bouncing bounces...
[04:35] <maswan> bouncing elmo?
[04:35] <siretart> lol
[04:37] <SloMo> Amaranth: do you want them in .po-file style? or just a list with english/german translations?
[04:37] <ogra> maswan, do you know a good 3d designer who could build us an elmo ? i'd be willing to hack that in ;)
[04:38] <Amaranth> SloMo: just a list
[04:38] <JaneW> ajmitch:ping
[04:38] <Amaranth> jimmac? :)
[04:38] <maswan> not really, the 3d guys I know are rather people involved in modeling textiles in mathematical ways etc
[04:39] <maswan> or for that matter, doing realistic sand/gravel for simulators
[04:41] <ogra> hmm, what about a bouncing warthg ?
[04:41] <maswan> :)
[04:41] <ogra> warthog even....
[04:41] <dilinger> ogra: or badger?
[04:41] <ogra> could anybody be religious offended ?
[04:41] <ogra> dilinger, fear the badger cult *G*
[04:41] <opi> did I hear someone says badger?
[04:41] <opi> ;-)
[04:42] <maswan> should we fear the badger cult more or less than the elmo cult?
[04:42] <dilinger> ogra: for apr 1, some mushrooms could float across the screen horizontally
[04:42] <dilinger> or something :p
[04:42] <ogra> hehe
[04:44] <thoreauputic> bouncing snakes would be scary, if you go the badger route...
[04:49] <diego> ogra: is there anyone already writing the GraphicalConfigTools?
[04:51] <ogra> diego, not yet, but i havent done the detailed app spec... 
[04:52] <diego> ogra: hm..well i'm quite interested in helping with those. i'm not extremely experienced but i've worked with python/pygtk/glade enough that the 3 listed sound fairly easy
[04:53] <diego> (and i really want to help out ubuntu because it rocks hehe)
[04:53] <ogra> where i can, i will spec it out as a dbus driven app... you should probably have a look at that...
[04:54] <diego> ok, great
[04:55] <diego> ogra: can you give me an example of the functionality it would/could gain by being dbus-driven?
[04:55] <ogra> i.e. having everything as a commandline app that accepts dbus messages from a gui, so you have in fact two apps
[04:57] <diego> ogra: ah, i see. then there could be a curses-based interface too if someone wanted to write it?
[04:57] <ogra> yeh
[04:58] <ogra> so it could be usable for headless servers etc.
[05:07] <SloMo> Amaranth: you have mail ;)
[05:07] <Amaranth> cool
[05:08] <lamont__> Kamion: ping
[05:09] <Kamion> lamont__: pong
[05:20] <diego> ogra: ok, i've just been reading about dbus. what's the plan to move forward on the GraphicalConfigTools?
[05:21] <ogra> diego, wait a bit until i have written the detailed spec :)
[05:21] <diego> ogra: yeah, ok. about how long is "a bit" though?
[05:21] <ogra> we can discuss it then
[05:22] <diego> a couple weeks or so?
[05:22] <ogra> i think i'll come around to make it during the next week...
[05:22] <ogra> today i'm busy with xscreensaver hacking... 
[05:22] <diego> ok, great
[05:23] <diego> the detailed spec will go into the udu wiki, correct?
[05:25] <SloMo> Amaranth: have you read http://www.python.org/dev/doc/devel/lib/node322.html ? as far as i can see everything is explained there. if you're interested I can try to implement it
[05:28] <ogra> diego,yes
[05:46] <pitti> Morning mdz
[05:46] <mdz> pitti: morning
[05:47] <pitti> mdz: would you be opposed to introducing a tiny setuid root program in the eject package that maps a devmapper-device to its attached real device?
[05:48] <pitti> mdz: the goal is to make eject handle encrypted (i. e. device-mapped) devices as well
[05:48] <pitti> mdz: but without root privileges the mapping becomes very hackish
[05:48] <pitti> mdz: (i. e. string matching on device name and hoping that the user didn't change it, and so on)
[06:04] <fabbione> morning mdz
[06:04] <fabbione> elmo: are you around by any chance?
[06:04] <fabbione> or somebody with root on davis?
[06:06] <Amaranth> SloMo: looks interesting
[06:06] <Amaranth> SloMo: i'll have to figure out how to use it with glade files though
[06:07] <SloMo> Amaranth: pygettext doesn't support glade files... you have to use xgettext to extract the strings
[06:07] <Amaranth> SloMo: I am now in the realm of "no fucking clue". ;)
[06:08] <Amaranth> SloMo: If you want to try to do it that'd be great. :)
[06:09] <SloMo> Amaranth: and for displaying the translated string there is either something to set for glade in python or it works out of the box when you initialized gettext before loading the glade file. but i have no idea how one can say setup.py, that it has to build the .gmo files and install them later
[06:10] <Amaranth> I can always set the strings in python.
[06:10] <Amaranth> self.w('foo').set_text(_('File'))
[06:10] <SloMo> sure but glade can do this on it's own
[06:10] <Amaranth> yeah
[06:10] <Amaranth> i'll look at how straw does it
[06:14] <Amaranth> hmm, gtranslator
[06:15] <zAo^> lo all
[06:16] <Loevborg> hoary faling to set up /etc/hosts and /etc/network/interfaces correctly (to include loopback), is that a known problem? I guess it's rather difficult to figure out for a newbie.
[06:16] <mdz> pitti: how can it do better if it has root privileges?
[06:16] <pitti> mdz: it can use the libdevmapper functions
[06:16] <pitti> mdz: but these require root
[06:17] <mdz> pitti: they use /dev/mapper/control or something?
[06:18] <SloMo> Amaranth: hm, i think i understand how straw does it ;) they've written their own setup.py and for glade... you can set the gettext domain with glade.XML and everything will work ;) the only tricky part is the setup.py, everything else is really easy
[06:18] <zAo^> Loevborg, I installed several instances of Ubuntu, none failed to build /etc/hosts or /etc/network/interfaces. DHCP worked out-of-the-box
[06:19] <Loevborg> zAo^, it probably has to do with my skipping the "network configuration" part of the installaion (because my wlan card doesn't work with the detected prism54 drivers)
[06:19] <pitti> mdz: so far I use cryptsetup status, which gives an error when ran as non-root; now I'm writing a small app that uses libdevmapper directly
[06:19] <zAo^> Ah, I will try that on my test machine next week :)
[06:19] <pitti> mdz: probably /d/m/c, yes
[06:19] <pitti> mdz: I didn't find a /proc or /sys interface
[06:20] <Amaranth> SloMo: I'm glad one of us understands.
[06:24] <dcraven> Amaranth: You have one, how does a project qualify for a forum in "3rd Party Ubuntu Projects"?
[06:24] <Amaranth> dcraven: *shrug*
[06:24] <dcraven> heh
[06:24] <Amaranth> dcraven: smeg stuff was spamming the desktop forum and ubuntugeek made me a subforum
[06:24] <Amaranth> i think mine was the first, but backports might have been there first
[06:24] <dcraven> Amaranth: Ahh.. I see.
[06:25] <Amaranth> if the project looks like it has a future and is being created and used by the ubuntu community i'm sure ubuntugeek will give you a subforum
[06:25] <Amaranth> but you have to have something to show
[06:26] <dcraven> Amaranth: Yeah. I was just curious about that. Might be handy some day. Cheers.
[06:27] <mdz> pitti: it is a shame to add another setuid root program to the default install :-/
[06:28] <mdz> pitti: but if it is tiny and auditable and has a definite value to add, that seems reasonable
[06:28] <pitti> mdz: I think it's better than device string matching; I started with this, but it's really ugly
[06:29] <pitti> mdz: if you have a better idea, then I'd appreciate :-)
[06:33] <SloMo> Amaranth: the setup.py stuff is really complicated... hum... reminds me of automake/autoconf ;) I'll try to get sth working later or tomorrow and let you know
[06:33] <Amaranth> SloMo: ok, cool
[06:36] <\sh> can someone explain to me, how we can remove one package out of universe and replace it with another newer package from debian? 
[06:36] <Mithrandir> \sh: "sync"? :-)
[06:36] <\sh> the reason is: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=286700
[06:37] <\sh> Mithrandir: but i have to touch it anyways...
[06:37] <\sh> so the question is, how can i remove one package out of the repos?
[06:38] <Mithrandir> \sh: why do you need to remove it?
[06:38] <\sh> Mithrandir: gnuradio-0.9 is not maintained anymore, upstream dev has uploaded new version gnuradio-core-2.5 to debian
[06:38] <Mithrandir> \sh: the package name is gnuradio-0.9?
[06:38] <\sh> Mithrandir: gnuradio
[06:38] <Loevborg> \sh, why not just "dpkg -i" the package file
[06:39] <Mithrandir> \sh: then just upload a newer package?
[06:39] <\sh> Mithrandir: ok, and what about the old one?
[06:39] <\sh> oh yes
[06:39] <Mithrandir> \sh: it'll go away, since it's superseded.  As long as the package name doesn't change.
[06:39] <\sh> Conflicts: Replaces: 
[06:39] <Mithrandir> no, those are totally different things.
[06:39] <\sh> Mithrandir: but the name changed as well ;)
[06:40] <\sh> at least the source name as I can see right now
[06:40] <\sh> oh no...also the binary package name changed
[06:40] <Mithrandir> \sh: if the source package name changes, but the binary not, don't worry, it'll be caught by rene
[06:40] <\sh> Mithrandir: as I said, the binary package name changed as well
[06:41] <Mithrandir> \sh: doesn't matter.
[06:41] <mdz> pitti: I'm afraid I know very little about the dm crypto stuff
[06:41] <pitti> mdz: I'll explain it in a mail
[06:41] <\sh> Mithrandir: ok...then I will adjust the new package for breezy and source upload it then
[06:53] <pitti> bye folks, have to go. nice weekend everybody!
[06:54] <jnc> ehhh
[06:55] <jnc> this may be a dumb person asking a question...    is the fact that OpenOffice asks me to "auto save", and this pops up and interrupts my use of the desktop, is that a bug?
[06:56] <jnc> i want to really ask if that should be converted into ULTRA MEGA PULSATOR like gaim popup windows do
[06:59] <mdz> am I the only one whose firefox is horribly broken?
[06:59] <mdz> (in breezy, obviously)
[06:59] <zul> broken as in how?
[06:59] <mdz> mizar:[...nonical/seeds/ubuntu/breezy]  firefox
[06:59] <mdz> /usr/lib/mozilla-firefox/firefox-bin: error while loading shared libraries: libfreetype&sg.6: cannot open shared object file: No such file or directory
[07:00] <mdz> "libfreetype&sg.6" is interesting
[07:00] <zul> no problems here
[07:00] <mdz> just a couple of bits off from libfreetype.so.6
[07:00] <fabbione> that looks so much like a patch stolen from a web page without text conversion
[07:00] <mdz>         libfontconfig.so.1 => /usr/lib/libfontconfig.so.1 (0xb7602000)
[07:00] <mdz>         libfreetype&sg.6 => not found
[07:00] <mdz>         hibexpat,si.1 => not found
[07:00] <mdz>         libc,so.6 => not found
[07:00] <mdz> note the comma in the libc dep
[07:01] <dilinger> neat
[07:02] <mdz> 84593149de55f308d04ef3461826b6e5  /usr/lib/mozilla-firefox/firefox-bin
[07:02] <fabbione> mdz: what arch is that?
[07:02] <mdz> i386
[07:02] <seb128> 84593149de55f308d04ef3461826b6e5  /usr/lib/mozilla-firefox/firefox-bin
[07:02] <seb128> same
[07:02] <mdz> seb128: but your ldd output is sane?
[07:02] <seb128> correct
[07:03] <mdz> very weird
[07:03] <fabbione> 84593149de55f308d04ef3461826b6e5  /usr/lib/mozilla-firefox/firefox-bin
[07:03] <fabbione> mdz: ldd is ok here
[07:03] <mdz> oh, those are indirect deps of some shared library
[07:04] <fabbione> mdz: ld cache corruption?
[07:04] <Nafallo> seb128: where have you hidden "run programs"? :-)
[07:04] <mdz> not the bin itself
[07:04] <mdz> I need to find which one
[07:04] <fabbione> why i have the feeling that it will bring you to X?
[07:04] <seb128> Nafallo: alt-F2
[07:05] <Nafallo> seb128: aha. I need to put the keybinding back. thanx ;-).
[07:05] <seb128> np :)
[07:06] <mdz> what is linux-gate.so.1?
[07:06] <fabbione> mdz: nothing to be worried about
[07:06] <fabbione> it has been there forever
[07:06] <mdz> is it new?  I don't remmeber seeing it before
[07:07] <fabbione> but in previos versions of ldd it was just hiddeng
[07:07] <fabbione> hiding it
[07:07] <fabbione> mdz: i had the same concern when binutils was updated
[07:07] <seb128> I don't have a such file here 
[07:07] <fabbione> and kernel initrd generation was broken :)
[07:08] <mdz> mizar:[/tmp]  objdump -x /usr/lib/libfontconfig.so.1 | grep 'NEEDED.*,'
[07:08] <mdz>   NEEDED      hibexpat,si.1
[07:08] <mdz>   NEEDED      libc,so.6
[07:08] <mdz> 0ac9a59ec54622148a0ee49ae783d93e  /usr/lib/libfontconfig.so.1
[07:10] <fabbione> mdz: what version of fontconfig?
[07:10] <fabbione> or libfontconfig
[07:10] <mdz> 2.3.2-1ubuntu1
[07:10] <mdz> libfontconfig1
[07:11] <fabbione>  bc04c04241ac216f50039c9cde89dff4  /usr/lib/libfontconfig.so.1
[07:11] <mdz> bc04c04241ac216f50039c9cde89dff4  usr/lib/libfontconfig.so.1.0.4
[07:11] <mdz> from the .deb
[07:11] <mdz> interesting, so I have file corruption
[07:12] <mdz> that's rather disconcerting
[07:13] <mdz> I have not even had so much as an unclean shutdown in a long time
[07:13] <fabbione> mdz: badblocks?
[07:13] <mdz> -rw-r--r--  1 root root 344 2005-05-05 18:58 /var/lib/dpkg/info/libfontconfig1.list
[07:13] <mdz> firefox definitely has worked since then, so it was corrupted on disk, not during install
[07:13] <mdz> very bad, considering I would never have modified that file
[07:14] <fabbione> mdz: i would suggest to check the hd for badblocks
[07:14] <mdz> debsums shows no other corrupt files
[07:14] <mdz> but it does find libfontconfig.so.1
[07:14] <mdz> cosmic rays perhaps
[07:15] <mdz> I will check badblocks
[07:15] <mdz> but I have no I/O errors in my logs, so I doubt it will find anything
[07:15] <fabbione> mdz: bad ram?
[07:16] <fabbione> i am sure of one thing.. it's not a kernel error :)
[07:16] <fabbione> mdz: what fs btw?
[07:16] <mdz> ext3
[07:16] <fabbione> .12rc5 ?
[07:16] <fabbione> or .10?
[07:16] <mdz> I have been tracking .12
[07:16] <mdz> 2.6.11.93-1.1 right now
[07:16] <fabbione> ok
[07:17] <fabbione> in that config i really doubt it's a kernel problem
[07:17] <fabbione> i386 + ext3 is one of the most tested bits
[07:17] <fabbione> and i am using it here everywhere too
[07:17] <mdz> yep, that's why I use it :-)
[07:17] <mdz> badblocks showed no errors
[07:17] <mdz> seems like probably memory corruption
[07:17] <fabbione> mdz: i blame GTK!
[07:18] <fabbione> yeah
[07:18] <fabbione> that's more possible
[07:18] <mdz> what a very interesting way for it to occur, though
[07:19] <fabbione> mdz: you seem very excited by it :)
[07:19] <fabbione> perhaps you want to debug it? :P
[07:21] <stern_type> her name?
[07:21] <mdz> wasabi_: I've updatedhttp://people.ubuntu.com/~mdz/temp/java-main.txt
[07:21] <mdz> fabbione: no, I have already reinstalled fontconfig and started using firefox again ;-)
[07:24] <infinity> mdz : Despite it being a weekend, I'll pop in and out to help wasabi massage that stuff through some more, if he's working on it today.
[07:24] <infinity> (Also, what am I doing still up at 3:30am?... I don't know either...)
[07:24] <fabbione> mdz: ehhehe
[07:25] <fabbione> infinity: you were supposed to go to sleep eons ago
[07:25] <infinity> fabbione : I know, but then I got caught up in that other distro that's a hobby in my off hours.
[07:25] <fabbione> infinity: oh damn :)
[07:25] <infinity> So, basically, I've been nerdy all day and all night, and I think I'm going to shoot myself.
[07:26] <fabbione> i am going out soon.. today is party :)
[07:26] <infinity> Celebrating elmo's birthday?  Man, everyone seems to be.
[07:26] <fabbione> now it's happenning exactly what i expected a few months back
[07:26] <infinity> (Or are you more of a doko fan?)
[07:27] <fabbione> infinity: i love both of them
[07:27] <infinity> :)
[07:27] <Micksa> so how does one debug STR problems?
[07:27] <infinity> For real this time.
[07:27] <fabbione> infinity: good night
[07:27] <Micksa> oh that's right
[07:27] <infinity> mdz : Tell wasabi to ping me via /msg if he needs help massaging his java chain.
[07:32] <stern_type> they love jesus-'idol' so much,they put him in a state of pain with her
[07:33] <dieman> ok
[07:33] <dieman> im sold
[07:33] <dieman> baz and the whole arch thing is the way to work
[07:33] <dieman> i just used bazaar to merge in changes from phorum into a locally hacked copy of forum
[07:33] <dieman> phorum, rather
[07:34] <dieman> it works super, super, super well.
[07:34] <Micksa> and it finished before you died too
[07:34] <dieman> i did the last one by had
[07:34] <dieman> hand 
[07:34] <dieman> with diff
[07:34] <dieman> it was very painful.
[07:34] <Micksa> hukhukhuk
[07:34] <dieman> this one had a whole 6 files conflicted. :)
[07:36] <Micksa> you've merged using CVS right?
[07:36] <dieman> yeah, thats painful too.
[07:37] <dieman> i never could figure out how to get the controtions to get cvs to work for what i needed in this case.
[07:37] <dieman> its probally possible, i guess.
[07:39] <dcraven> Amaranth: If you are looking for internationalization guidance, there is pygnome-hello in the gnome-python package. It uses GNU autotools though.
[07:46] <mdz> infinity: meaning you're *not* going to sleep?
[07:46] <infinity> mdz : Meaning I'll probably be up again in a few hours.
[07:47] <mdz> fun
[07:47] <infinity> (This is merely a clever AI autoresponder.  Pay no attention)
[07:47] <mdz> wasabi: looks like libxerces2-java is the last free-java-sdk package
[07:48] <Micksa> mjg doesn't respond well to being poked does he?
[07:49] <daniels> pittyeah, known issue, will be fixed tomorrow (i.e. saturday)
[08:00] <Micksa> I want STR love :(
[08:10] <Micksa> right
[08:10] <Micksa> I'll take a shot at this myself
[08:41] <mdz> daniels: around?
[08:41] <mdz> daniels: for ThinClientIntegration, I need to preseed xserver-xorg/config/inputdevice/keyboard/layout for reconfiguration, but this doesn't currently work
[08:57] <maswan> thom: want feedback on your httpd2.1 packages?
[08:59] <tseng> maswan: only if it requires him to do lots of menial work. in that case, he cant get enough.
[08:59] <maswan> tseng: How about I whine meaninglessly about how it sucks, doesn't work, and is bad, but not actually tell him anything useful in how it is broken?
[09:00] <tseng> oh, he loves that too
[09:00] <maswan> great!
[09:06] <maswan> thom: if you actually read scrollback, lots of /etc/httpd2.1/ was missing after install. Nothing worse (yet).
[09:07] <\sh> hmm...any g++ geeks around?
[09:08] <maswan> well.. with the exception that the large file support seems broken on amd64
[09:09] <\sh> if a class is deklaring another class as friend, is the friend able to call a private constructor?
[09:09] <\sh> s/deklaring/declaring/
[09:10] <ska-fan> I think it should.
[09:11] <jbailey> mdz: around?
[09:12] <\sh> ska-fan: but it fails
[09:12] <\sh> ska-fan: g++4 is complaining
[09:12] <ska-fan> \sh: I'm really anything but a c++ expert, sorry
[09:12] <ska-fan> \sh: try #c++
[09:13] <ska-fan> there are very smart people there.
[09:13] <ska-fan> ##c++ even.
[09:14] <\sh> thx
[09:14] <jdub> seb128: around?
[09:14] <chrissturm> it could even be a bug in gcc 4.0
[09:18] <jdub> mdz: around?
[09:18] <jbailey> jdub: The Jeff's are a hunting, I see. =)
[09:18] <jdub> heh
[09:19] <jdub> mdz is hacking on ltsp
[09:21] <Nafallo> jdub: morning :-)
[09:21] <Kamion> Nafallo: you pinged me last night?
[09:21] <jdub> hey Nafallo 
[09:21] <mdz> jdub: yes?
[09:21] <mdz> jbailey: yes?
[09:22] <Nafallo> Kamion: yea. the damn script wants to redownload the whole of main and restricted every sync when I use /pool/.
[09:22] <Nafallo> Kamion: you got time to take a quick look at the script? :-)
[09:23] <seb128> jdub: yep
[09:23] <jbailey> mdz: Do you need any hooks so far in the initramfs beyond the script you sent me?  I'm doing up a similar set for mdadm and friends right now.
[09:24] <mdz> jbailey: I need a hook wihch allows me to do what I was doing in that script, but at a later stage of the boot process
[09:24] <mdz> that's it
[09:25] <monolive> Anyone here familiar with ubuntu bounties?
[09:26] <jbailey> mdz: The run-init at the end of that script is what chains to init.  Can you describe where you need the hooks to go?
[09:26] <Kamion> Nafallo: well, uh, I tend to attack such problems with strace, so I'm not sure how much help I'll be :)
[09:26] <mdz> jbailey: that script was only meant to show you what I was doing; it's not meant to exist in that form
[09:26] <Kamion> Nafallo: definitely start by using rsync -v
[09:27] <mdz> jbailey: the way it should work is: 1. mount root filesystem using the stock nfs script, 2. unionfs magic runs, 3. run-init
[09:27] <Nafallo> Kamion: http://www.magicalforest.se/~nafallo/mirrorscript/
[09:27] <jbailey> mdz: Cool, thanks.
[09:27] <mdz> jbailey: so the unionfs magic would take whatever's on /root, slap a copy-on-write layer on top of it, and then proceed as normal
[09:27] <Nafallo> Kamion: there are the scripts, and s/script/log/ for the logs :-P
[09:27] <Kamion> (what's with the [[ ... ] ] ?)
[09:28] <Nafallo> Kamion: I'm using the same as you I believe.
[09:28] <Kamion> I would use [ ... ] 
[09:28] <Kamion> anyway
[09:30] <Nafallo> Kamion: stolen from maswan's script iirc ;-)
[09:32] <Kamion> Nafallo: nothing's jumping out at me as obviously wrong
[09:32] <maswan> I deny everything
[09:32] <Kamion> bear in mind I'm not a professional mirror operator though - that was just what I needed to do to get cdimage to work
[09:33] <Nafallo> Kamion: and it doesn't resync everything at your place? ;-)
[09:33] <Kamion> no
[09:33] <Kamion> but your script is fairly significantly different from mine, so that says nothing really
[09:33] <Kamion> for one thing I only do two rsync passes
[09:34] <Nafallo> Kamion: i used universe/ and multiverse/ in excludes only. but I don't think that should change anything significantly :-)
[09:34] <Kamion> mine does unnecessary resyncs of some files which are main ->/<- universe symlinks, like yours
[09:35] <Kamion> you sure that rsync -n is telling the truth?
[09:36] <Kamion> you could try it on smaller subdirectories of pool to experiment
[09:37] <Nafallo> Kamion: I tried without "pool/" and it does almost what it should. except that it plays with the daily-installerstuff.
[09:38] <Nafallo> Kamion: oh well. I should hack a bit more on it.
[09:38] <Nafallo> Kamion: I probably wouldn't want to use three sync against maswan's damn clusterstuff :-P.
[09:38] <maswan> sorry?
[09:39] <Nafallo> maswan: and _no_! you're _not_ innocent ;-)
[09:39] <Kamion> daily-installer-* contains symlinks within itself, hence the two rsync passes in my mirror script
[09:39] <Kamion> likewise installer-*
[09:39] <Nafallo> those are not hard-links?
[09:40] <Kamion> no, current -> <some directory>
[09:40] <Kamion> you can't hard-link directories
[09:41] <Nafallo> hmm, right. I read that yesterday in the middle of the night :-P.
[09:45] <Nafallo> maswan: as I mentioned earlier it's still slow for me to get the file lists from you.
[09:46] <maswan> Nafallo: does it take more than 5 minutes?
[09:47] <Nafallo> maswan: yes. could be that I'm on 440k ;-)
[09:48] <Nafallo> maswan: I thought you said it was the filesystem latency though...
[09:48] <Nafallo> Kamion: ahh, hardlinks for instance installer-i386/20050317ubuntu5/images/netboot/pxelinux.0? :-)
[09:49] <Kamion> cjwatson@jackass:~/ftp/dists/breezy/main/installer-i386/20050317ubuntu5/images/netboot$ ls -l pxelinux.0
[09:50] <Kamion> lrwxr-xr-x  1 katie katie 32 May 10 19:35 pxelinux.0 -> ubuntu-installer/i386/pxelinux.0
[09:50] <Kamion> Nafallo: that's a symlink
[09:51] <Nafallo> hmm. I run it with --copy-links (-L) and this: " => " shows up in the log.
[09:51] <Kamion>         -L, --copy-links            copy the referent of all symlinks
[09:51] <Kamion> I mirror dists/ with --links, not --copy-links
[09:52] <Kamion> (and --hard-links as well, although I don't think it's necessary)
[09:52] <Kamion> the reason for two passes in my script is so that I can use --links on dists/ but --copy-links on pool/
[09:52] <Nafallo> Kamion: yepp, I keep --hard-links everywhere.
[09:52] <Kamion> Nafallo: yes, but if you're doing --copy-links on dists/ you will mirror stuff unnecessarily
[09:53] <Kamion> --hard-links is pretty irrelevant, I think; can't think of anything in the Ubuntu archive that would be a hardlink
[09:54] <Nafallo> hmm, I might clean some space on my lappy and try without -n in the last step :-P.
[09:55] <Kamion> as I say, try in pool/main/a/a52dec/ or something
[09:55] <Kamion> rather than the entire pool
[09:56] <Nafallo> Kamion: will try that now
[10:04] <mvo> Kamion: I assume some python-apt code for #185424 wouldn't help you? i.e. a pyhton-apt based progress thing?
[10:05] <Kamion> mvo: as long as it works with only the minimal seed installed, it's fine
[10:05] <Kamion> although getting at python from base-config is a tad fiddly
[10:05] <Kamion> but I should be able to use debconf.py, I suppose
[10:05] <Kamion> mvo: I would like to see it in apt though
[10:05] <Kamion> and I need aptitude to be able to call it
[10:06] <mvo> Kamion: base-config is written in perl? or sh?
[10:07] <mvo> Kamion: and it then calls aptitude to do the install work?
[10:08] <Nafallo> Kamion: -n didn't lie. thanx anyway. I'm probably just need to hack on it a bit more :-P.
[10:08] <Kamion> mvo: mix of sh and perl; yes, it does
[10:10] <Kamion> bear in mind I really want to get this into Debian (it's a long-sought feature there too, and I want to avoid further base-config forkage since it's a nightmare at the moment), and requiring python would make life awkward there since python isn't in Debian's base
[10:11] <mvo> ok
[10:12] <Kamion> now obviously if big parts of apt move to python then that will have to change anyway, but I don't want it to have to happen *just* for base-config :)
[10:13] <mvo> apt has a very limited knowledge about what dpkg does right now
[10:13] <mvo> but I agree that it really should be done in apt itself :)
[10:14] <mvo> I have a fairly good idea how it could be done in python-apt, that's why I initially tried to ask about that ...
[10:34] <mdz> heh
[10:34] <mdz> the HP amd64 workstations which go out with FreeDOS have a "designed for Windows" sticker which looks like this: http://people.ubuntu.com/~mdz/temp/notwindows.jpeg
[10:35] <Treenaks> mdz: wow, you're on a mad ltsp hacking spree?
[10:35] <mdz> this has been my first chance to put some solid hours into it since UDU
[10:36] <mdz> I've got thin clients netbooting all the way up to a gdm login over xdmcp
[10:36] <sivang> mdz: wow cool :-)
[10:36] <Treenaks> mdz: rockage!
[10:36] <sivang> Treenaks: howdy btw 
[10:36] <Nafallo> mdke: lol
[10:36] <Treenaks> mdz: does it work /after/ that? :)
[10:36] <Treenaks> hi sivang 
[10:38] <mdz> Treenaks: yes, that part's easy
[10:38] <mdz> it's all server-side
[10:38] <sivang> mdz: the logo came crossed like this?
[10:39] <mdz> sivang: yes
[10:39] <sivang> mdz: amazing :-)
[10:39] <sivang> mdz: how much such a machine costs?
[10:39] <mdz> I'm not sure how much that one was, but an HP amd64 workstation can be had for under $1k USD
[10:40] <Treenaks> mdz: hmm.. nice
[10:42] <Treenaks> mdz: what's the type number?
[10:42] <Treenaks> uh
[10:42] <Treenaks> model
[10:44] <mdz> I don't know what the one in the photo is
[10:44] <mdz> but a650e would be one example
[10:44] <mdz> (pavilion)
[10:46] <Treenaks> they don't exactly go out of their way to promote freedos
[10:53] <mdz> jbailey: initramfs is now the only remaining piece of ltsp client setup which is not automated
[10:55] <Treenaks> mdz: you can get a dx5150 with an amd64, 256M, 48x CD drive for $559.99
[10:55] <Treenaks> ex. shipping
[10:56] <Amaranth> a what?
[10:56] <Treenaks> Amaranth: HP amd64
[10:57] <Amaranth> needs at least double the RAM, probably has a crappy HD, and integrated audio/video
[10:57] <Treenaks> http://h71016.www7.hp.com/dstore/MiddleFrame.asp?page=config&ProductLineId=429&FamilyId=2045&BaseId=13463&oi=E9CED&BEID=19701&SBLID=
[10:58] <Treenaks> Amaranth: no doubt.. but I'd buy it, install ubuntu and provide it as a Samba server to my clients.. if I had any ;)
[11:14] <\sh> someone working on g77 for 4.0?
[11:14] <\sh> ;)