[12:02] <pasc> mirak: stick a ./ in front of the file name
[12:03] <mirak> mmm
[12:03] <seb128> yes, that works too :p
[12:13] <pitti> elmo: still here?
[12:14] <_rene_> pitti: besides the fact that 1.1.3-2 is b0rked, 1.1.2dfsg1-3 and 1.1.3-2 have the l10n-* changes
[12:14] <elmo> pitti: yeah
[12:14] <pitti> _rene_: heh, nice
[12:14] <pitti> elmo: klogd     5192  0.1  0.2  2560 1472 ?        Ss   00:12   0:00 /sbin/klogd -P /var/run/klogd/kmsg
[12:14] <pitti> elmo: ^^^^^   another item gone from your list :-)
[12:15] <elmo> I thought you couldn't do that ?
[12:15] <pitti> elmo: it's night, remember?
[12:15] <pitti> the DEROOTIFICATOR strikes back
[12:15] <pitti> elmo: no, of course I had to cheat
[12:16] <pitti> elmo: klogd is now run as unprivileged user right from the start
[12:16] <jdub> azeem: your new multisync has Depends: evolution (<< 1.5) on the evo plugin
[12:16] <pitti> elmo: I added an option to read an alternative /proc/kmsg location (a pipe /var/run/klogd/kmsg)
[12:16] <jdub> pitti: man, you are rocking on the derootification
[12:16] <pitti> elmo: and have a root process (a mere 'dd') which pipes /proc/kmsg to /var/run/klogd/kmsg
[12:17] <pitti> elmo: I know it's an evil hack, but it works great
[12:17] <pitti> elmo: and so we separated the root task from all the parsing stuff which is now done purely as user
[12:17] <pitti> jdub: thanks :-)
[12:17] <elmo> pitti: rock on
[12:18] <pitti> well, bedtime now
[12:18] <pitti> I do some further cleanups and tests and upload tomorr
[12:18] <pitti> ow
[12:19] <fabbione> YEAH
[12:19] <fabbione> applying patch sparc64-syslog-register to ./ ... ok.
[12:20] <fabbione> the kernel starts to take shape
[12:20] <fabbione> MUHA MUHA MUHA
[12:20] <pitti> fabbione: how far is the porting effort?
[12:20] <fabbione> pitti: i am building the "golden debs"
[12:20] <pitti> fabbione: btw, will we all get a free sparc64 when you are done?
[12:20] <pitti> :-)
[12:20] <fabbione> basically ubuntu on top of ubuntu
[12:20] <pitti> fabbione: "golden"?
[12:20] <pitti> ah
[12:21] <fabbione> and i am hacking around the kernel now
[12:21] <pitti> fabbione: wait, tomorrow the kernel will run as normal user :-)
[12:22] <pitti> fabbione: btw, the platform supported by Debian is "only" sparc32?
[12:22] <pitti> fabbione: I had woody on a sparc64 once; you are now building the whole userland for 64 bit?
[12:23] <fabbione> pitti: ETOOMANYQUESTIONS
[12:23] <fabbione> pitti: oh yeah.. kernel as user ehhehe
[12:23] <fabbione> pitti: it depends from the kernel.
[12:24] <fabbione> but the userland is always 99.9% 32bit
[12:24] <fabbione> it's just faster
[12:24] <sjoerd> fabbione: btw. alsa is now working on my U5, you might want too look at the patches for that (as your doing kernel stuff)
[12:25] <fabbione> sjoerd: sure.. right now i need to get the packages to compile
[12:25] <sjoerd> :)
[12:25] <fabbione> and give some love to the config
[12:28] <fabbione> sjoerd: do you have idea why sparc have nsdiwrapper support?
[12:28] <fabbione> isn't that a i386/m$ thingy?
[12:28] <pitti> Solaris drivers for Linux/sparc?
[12:29] <sjoerd> fabbione: huh ?
[12:29] <Md> NDIS is a windows technology, I can't see how solaris enters the picture
[12:29] <pitti> Md: just guesssing :-)
[12:29] <elmo> because our kernel patch isn't very ipcky
[12:30] <fabbione> elmo: ok thanks
[12:30] <Md> does ubuntu have some modem autodetection script? I'm pondering about what to do with /dev/modem
[12:30] <fabbione> Md: nope...
[12:33] <Md> I suppose it's possible to send ATIwhatever to all serial ports and check if something comes back, but doing this at every boot could take too much time. and I do not even want to think about what happens if something which is not a modem is connected to the port...
[12:34] <elmo> anyone here use apt-move ?
[12:35] <fabbione> good night guys
[12:35] <fabbione> cya in 5/6 hours
[12:35] <pitti> night fabbione 
[12:41] <_rene_> Md: wvdialconf does something like that...
[12:41] <_rene_> Md: (just used it some time ago)
[12:51] <pitti> good night everybody!
[12:55] <lamont> evening
[01:22] <eruin> any of you have an idea as to what might be causing either a very long wait time after failing to insert pciehp/shpchp (operation not permitted) or a complete system halt (on boot)?
[01:22] <eruin> I see rp_filter. after it the times it doesn't actually halt my system (which is about 1/2 of the time), then setting up network
[01:22] <eruin> I've got one connected eth and one not connected, and I'm thinking this is more an issue with that fact than hotplug after reading bug #1869
[01:25] <haggai> eruin: long waits are often caused by unreachable dns servers.  Did you try commenting out all nameservers in /etc/resolv.conf?
[01:26] <eruin> no, but I will do
[02:02] <__daniel> good night
[02:10] <zul> evening
[03:08] <lamont> if someone rolled isdnutils forward from libstool 1.4, then we could move it to universe...
[03:08] <lamont> (it's ftbfs on amd64)
[03:40] <sladen> pitti: sweet klogdness!
[07:52] <jdub> doko: woo! :)
[07:54] <doko> :) morning, did the build start?
[08:02] <fabbione> morning guys
[08:02] <jdub> doko: i don't think i get to see running logs
[08:03] <fabbione> hey jdub
[08:03] <jdub> yo fab :)
[08:03] <fabbione> jdub: i think soon you will be able to run hoary on sparc
[08:08] <jdub> rad :)
[08:08] <jdub> how's kernel stuff?
[08:08] <fabbione> building now
[08:09] <fabbione> it still needs love, i am only checking the patches right now
[08:09] <fabbione> there is a few things that needs to be added properly
[08:09] <fabbione> the major blocker atm is silo
[08:10] <fabbione> because it build-dep on gcc-2.95
[08:10] <fabbione> it compiles fine with other gcc's
[08:10] <fabbione> but apparently it doesn't always work
[08:10] <fabbione> so someone will have to test it for me
[08:10] <fabbione> (because i only have ONE sparc that it is actually building)
[08:11] <fabbione> jdub: do you feel brave enough to give it a shot?
[08:11] <jdub> yeah, maybe on my U5
[08:11] <fabbione> jdub: cool
[08:12] <jdub> then on the 220R :)
[08:12] <fabbione> jdub: just grab silo from there
[08:12] <fabbione> don't use the other debs, because they are not all of them yet
[08:12] <jdub> i should try silo with a sarge install?
[08:13] <jdub> or...?
[08:13] <fabbione> jdub: it doesn't matter on which distro
[08:13] <fabbione> it's just recompiled with gcc-3.3 instead of 2.95
[08:13] <jdub> oh right
[08:13] <fabbione> there are no package or code changes
[08:14] <jdub> oh, sorry, missed your "only one sparc" bit :)
[08:14] <fabbione> ehehe
[08:14] <jdub> i'm tempted to put my sata disks in the sparc now...
[08:15] <jdub> no idea if it'll boot off them though. i suppose it wouldn't.
[08:15] <fabbione> nope
[08:15] <fabbione> obp is pretty anal
[08:15] <fabbione> either IDE or SCSI
[08:18] <jdub> doko: http://people.ubuntulinux.org/~lamont/buildLogs/o/openoffice.org/1.1.3-2.3ubuntu2/openoffice.org_1.1.3-2.3ubuntu2_20041126-0708-i386-failed
[08:18] <jdub> doko: hrm
[08:37] <HcE> ops :$ wrong #
[08:38] <fabbione> ehehhe
[08:38] <fabbione> HcE: go and take your piss :P
[08:38] <HcE> :P
[08:39] <HcE> norwegian should be a far off language only I understood
[08:39] <HcE> and of course the persone I was supposed to talk to
[08:39] <fabbione> HcE: i only understand a bit of danish/swedish/norvegian
[08:39] <fabbione> but that "bit" is pretty common
[08:39] <fabbione> to all of them
[08:40] <HcE> your not dannish?
[08:40] <fabbione> nope
[08:40] <fabbione> i am italian
[08:40] <fabbione> i live in dk
[08:40] <HcE> only live in .dk?
[08:40] <HcE> ah
[08:40] <fabbione> yeah
[08:41] <HcE> can't go to the toilett before I've check new packages from Debian
[08:42] <HcE> the days big highlight
[08:42] <fabbione> haha
[08:42] <HcE> happy to see wxWidget 2.5.3 as a .deb finally
[09:14] <pitti> Morning
[09:22] <fabbione> morning pitti
[09:45] <mdz> pitti: here?
[09:45] <fabbione> morning mdz
[09:45] <mdz> morning
[09:51] <Mithrandir> fabbione: I ran out of disk space last night, redoing build now.
[09:51] <fabbione> Mithrandir: ah cool
[09:51] <fabbione> thanks
[09:51] <fabbione> Mithrandir: 8012 -rw-r--r--   1 sparcbuildd sparcbuildd  8191682 Nov 26 08:04 linux-image-2.6.8.1-3-sparc64_2.6.8.1-17.1_sparc.deb
[09:52] <fabbione> ;)
[09:52] <Mithrandir> yay
[09:52] <jdub> woo :)
[09:52] <fabbione> Mithrandir: we might have problems with silo
[09:52] <fabbione> as i explained to jdub
[09:52] <fabbione> it needs testing :(
[09:52] <fabbione> otherwise we will have to pull in gcc-2.95 for sparc in main
[09:52] <seb128> morning
[09:55] <fabbione> hey seb
[09:57] <Mithrandir> ew, ok.
[10:05] <fabbione> ouch
[10:05] <fabbione> something is not properly working...
[10:05] <fabbione> the merge doesn't create kernel-image udebs
[10:23] <pitti> Hi mvo_, hi daniels!
[10:24] <pitti> daniels: btw, I can test the new radeon_drv if you want
[10:25] <daniels> pitti: hey dude
[10:25] <daniels> pitti: ah rad -- http://people.ubuntu.com/~daniels/xorg/powerpc-radeon_drv.o
[10:25] <pitti> daniels: anything I should test in particular?
[10:26] <pitti> daniels: shall I change anythign in Xorg.conf?
[10:28] <fabbione> daniels: hey kid
[10:28] <fabbione> daniels: libao needs to build-dep on libxau-dev, can you take care of it?
[10:29] <mdz> night all
[10:29] <fabbione> night mdz
[10:29] <pitti> Night mdz!
[10:29] <daniels> fabbione: k
[10:29] <daniels> mdz: night dude
[10:30] <daniels> pitti: ditch Option "UseFWPLL", that's about it
[10:30] <fabbione> daniels: thanks
[10:31] <mvo_> night mdz 
[10:31] <pitti> daniels: I already tried the new driver
[10:31] <pitti> daniels: still with the PLL option
[10:31] <daniels> pitti: still works?
[10:32] <pitti> daniels: it behaves exactly like the very old one
[10:32] <daniels> pitti: because I threw away the usefwpll patch
[10:32] <daniels> pitti: rockin!
[10:32] <pitti> daniels: external works, internal one flickers
[10:32] <pitti> daniels: so, "no"
[10:32] <daniels> pitti: oh, right
[10:32] <daniels> hold on, there's an option for that
[10:32] <pitti> daniels: the last one you sent me worked fine
[10:32] <pitti> daniels: "very old one" == the version shipped in 6.8.1-0.2
[10:33] <daniels> pitti: does 'Option "LVDSProbePLL"' help?
[10:33] <daniels> pitti: if not, does 'Option "NoLVDSProbePLL"' help?
[10:33] <pitti> daniels: first I try to remove UseFWPLL
[10:33] <pitti> daniels: then the other two
[10:33] <pitti> daniels: brb, I have to disconnect my monitor
[10:33] <daniels> pitti: usefwpll should have no effect with the new radeon_drv (the one I put up last night)
[10:34] <daniels> k
[10:34] <pitti> daniels: oh, then the option should not make any difference?
[10:35] <daniels> right
[10:35] <pitti> indeed, no difference, still flicker. Let's try the other one
[10:36] <azeem> jdub: euh
[10:37] <pitti> daniels: no, neither option makes it work
[10:37] <daniels> pitti: and it worked with my patch and UseFWPLL?
[10:37] <pitti> daniels: yes
[10:37] <daniels> heh
[10:38] <pitti> daniels: it worked with the 2.3 MB driver you sent me 
[10:38] <pitti> daniels: md5sum 2b44a78a...
[10:38] <pitti> daniels: 2327838 bytes
[10:38] <pitti> daniels: ^^ this certainly included debugging symbols?
[10:38] <daniels> could you please email benh@kernel.crashing.org, CCing daniel.stone@canonical.com, send him your logs from both with LVDSProbePLL and with NoLVDSProbePLL, tell him it worked with my patch that pulls the PLL values from the registers, and just quickly describe your problem?
[10:39] <daniels> pitti: yeah, it's only ~100kB stripped
[10:39] <pitti> daniels: 230 kB for me
[10:39] <pitti> daniels: anyway, do you want any logs?
[10:39] <pitti> daniels: shall I set any value for above option?
[10:39] <pitti> "true" or so?
[10:39] <bob2> hrm, my laptop doesn't wake up from sleep
[10:39] <bob2> shockingly
[10:39] <daniels> pitti: if you could just throw them all (with LVDSProbePLL and with NoLVDSProbePLL) in the email to myself and benh, that'd be great, thanks
[10:40] <daniels> pitti: nope, just Option "LVDSProbePLL" or Option "NoLVDSProbePLL"
[10:40] <pitti> daniels: benh.get_email() == ?
[10:40] <bob2> tommorow I'd guess
[10:40] <daniels> pitti: benh@kernel.crashing.org
[10:40] <daniels> pitti: see the line just above 'yeah, it's only ~100kB stripped'
[10:41] <pitti> oh, missed that one
[10:41] <pitti> I'll do
[10:41] <pitti> daniels: how I shall refer to this driver I just downloaded?
[10:42] <pitti> daniels: is it from any official version?
[10:44] <fabbione> this is SOOOO evil!
[10:44] <fabbione> one of the (2000) Makefiles clean targets remove without any check all files called kernel-image!
[10:44] <fabbione> that of course.. disable the generation of the kernel-image*_udeb
[10:49] <daniels> pitti: tell him it's with his patch backported to the 6.8.x branch
[10:56] <lupus_> The following packages have been kept back:
[10:56] <lupus_>   openoffice.org-debian-files
[10:56] <lupus_> what does it misses?
[10:56] <lupus_> I did dist-upgrade
[10:57] <azeem> lupus_: do apt-get install openoffice.org-debian-files and see
[10:58] <mvo_> lupus_: it looks like not all of openoffice has hit the archive yet
[10:58] <lupus_> so those 2 packages aren't available yet?
[10:58] <mvo_> it appears so
[10:59] <mvo_> openoffice.org is still 1.1.2 
[11:05] <haggai> doko asked elmo to sync -debian-files once openoffice.org was ready, looks like he perhaps did it early
[11:08] <azeem> jdub: fixed now
[11:08] <azeem> (hopefully)
[11:27] <fabbione> lupus_: these kind of questions are more appropriate for #ubuntu
[11:34] <lupus_> k sorry
[11:34] <lupus_> but I thought hoary = devel so :)
[11:35] <lupus_> but I missed the point of this channel I guess :)
[11:35] <azeem> lupus_: yeah, but 'apt-get dist-upgrade does not work' is not a good error report
[11:36] <azeem> in any case, stuff will break occasionally and people will be aware usually. If it is broken for a longer time, reporting it is fine I guess
[11:42] <jdub> thom: GUS RELEASE!
[11:44] <daniels> jdub: ... gravis ultrasound?
[11:44] <jdub> daniels: gnome-user-share
[11:45] <Keybuk> jdub: what does that do?
[11:45] <daniels> jdub: NICE!
[11:45] <daniels> Keybuk: http://cvs.gnome.org/viewcvs/gnome-user-share/README?rev=1.2&view=auto
[11:45] <daniels> i so need to prise my mouldavia rewrite off trinity
[11:45] <jdub> Keybuk: user-policy sharing with apache+webdav :)
[11:46] <Keybuk> ah, cool
[11:46] <fabbione> jdub: apache1.3 or 2?
[11:46] <fabbione> or it doesn't care?
[11:46] <jdub> fabbione: 2 man, 2 :)
[11:46] <jdub> it's *rad*
[11:46] <fabbione> cool
[11:46] <fabbione> less bugs from lusers for me :P
[11:47] <daniels> jdub: mouldavia was still cooler
[11:47] <fabbione> daniels: please libao :-)
[11:48] <daniels> fabbione: yes
[11:50] <daniels> fabbione: wtf, dude, are you sure?
[11:50] <daniels> daniels@catsby:~/canonical/libao/libao-0.8.5% egrep -r '#include.*X' *
[11:50] <daniels> configure:#include <X11/Intrinsic.h>
[11:50] <daniels> configure:#include <X11/Intrinsic.h>
[11:58] <fabbione> daniels: apt-get --purge remove libxau-dev
[11:58] <fabbione> and build it
[11:58] <fabbione> it will fail building one of the libraries
[11:58] <fabbione> the makefile is retarted and doesn't notice
[11:59] <fabbione> until it starts creating the deb
[11:59] <fabbione> where it will fail becuase ls .libs will return empty
[12:00] <fabbione> there.. logs are on the way
[12:00] <fabbione> repeatable on i386
[12:06] <daniels> what the hell?  i can't see that this package uses Xau at all
[12:07] <fabbione> checking for XauFileName in -lXau... no
[12:07] <daniels> yes
[12:07] <fabbione> and it FTBFS without :-)
[12:07] <fabbione> checking for X... libraries /usr/X11R6/lib, headers 
[12:07] <daniels> the only place XauFileName is referred to is in autoconf files
[12:07] <fabbione> this line will change once you install Xau
[12:08] <daniels> i know that.
[12:08] <daniels> but it's not actually using any X headers or functions AFAICT
[12:09] <fabbione> ah crap! first 2 spam mail via @canonical.com
[12:10] <smurfix> can't escape them :-(
[12:38] <jvw> who do I mail for inquiries re sponsoring (the upcoming conference)? Mako? some role account?
[12:39] <daniels> jvw: mako@canonical
[12:39] <jvw> ok, tnx
[12:43] <daniels> Kamion: ping
[12:46] <_rene_> Kamion, jdub: http://cvs.debian.org/oo-deb/debian/rules.diff?r1=1.231.2.26&r2=1.231.2.27&diff_format=h&cvsroot=debian-openoffice
[12:48] <lucas_> hi
[12:48] <lucas_> Kamion: you there ? I'm now successfully generating iso images. I'd like to be able to replace some packages with my own now
[12:48] <fabbione> Mithrandir: how is going the build?
[12:48] <lucas_> I'd like to use the LOCALDEBS way
[12:49] <Mithrandir> fabbione: it built.
[12:49] <fabbione> Mithrandir: cool.. what about the udebs?
[12:49] <lucas_> but I don't understand which directory structure debian-cd expects in the local repository
[12:49] <Mithrandir> fabbione: there are udebs with non-zero size. :)
[12:50] <fabbione> Mithrandir: that sounds cool
[12:50] <fabbione> it build.. it works :-)
[12:50] <fabbione> i just need to wait for Kamion to confirm ppc
[12:50] <fabbione> even if the version you built has 2 bugs
[12:50] <fabbione> i fixed them locally
[12:50] <Mithrandir> ok
[12:51] <fabbione> it doesn't generate kernel*.udeb
[12:51] <fabbione> and as a consequence it miscalculate the dependencies
[12:51] <Mithrandir> oh, ok.
[12:51] <fabbione> but they look good here
[12:52] <Mithrandir> we want kernel*udeb?
[12:52] <fabbione> debdiff is happy
[12:52] <fabbione> Mithrandir: it's a missing $(touch modules/${arch}/kernel-image
[12:52] <fabbione> )
[12:52] <Mithrandir> heh
[12:53] <fabbione> but the missing file is related to something else
[12:53] <fabbione> it must be one of the clean target that wipes it away
[12:53] <fabbione> because it always disappear
[12:55] <Kamion> fabbione: damnit, I ran out of disk space :-(
[12:55] <fabbione> Kamion: shit
[12:55] <Kamion> fabbione: can you use a porting box in the lan or something?
[12:55] <Kamion> daniels: pong
[12:55] <fabbione> sorry
[12:56] <fabbione> Kamion: than wait before you start
[12:56] <fabbione> Kamion: i will give you another diff on top of 17
[12:56] <fabbione> Kamion: i fixed a couple of problems
[12:56] <fabbione> (see above)
[12:56] <Kamion> _rene_: dunno - you'd have to ask elmo/lamont whether lsb-release is installed on the buildds
[12:56] <Kamion> lucas_: mmkay, that's a little complex, let me see
[12:56] <_rene_> well, we build-dep on it, so ;-)
[12:57] <_rene_> that's the control.in part ;)
[12:57] <Kamion> _rene_: ah
[12:57] <Kamion> lucas_: LOCALDEBS is probably what you want, certainly
[12:57] <lucas_> Kamion: yes
[12:57] <Kamion> lucas_: do you have tla installed?
[12:57] <lucas_> Kamion: I tried to use apt-ftparchive to generate the packages and all
[12:57] <fabbione> Kamion:
[12:57] <fabbione> The following lines in the control files differ (wdiff output format):
[12:57] <fabbione> ----------------------------------------------------------------------
[12:57] <fabbione> Version: [-2.6.8.1-18-]  {+0.67ubuntu3+}
[12:57] <fabbione> Section: [-devel-]  {+debian-installer+}
[12:57] <fabbione> Maintainer: Debian [-kernel team <debian-kernel@lists.debian.org>-]  {+Install System Team <debian-boot@lists.debian.org>+}
[12:57] <lucas_> Kamion: yes
[12:57] <fabbione> Source: [-linux-source-2.6.8.1-]  {+linux-kernel-di-i386-2.6+}
[12:57] <_rene_>   * debian/control.in:
[12:57] <_rene_>     - add Builddep on lsb-release [RE] 
[12:57] <_rene_>     - change kdelibs4-dev builddep to kdelibs4-dev | ubuntu-artwork to
[12:57] <_rene_>       disable it for Ubuntu [RE] 
[12:57] <_rene_>   * debian/rules:
[12:57] <_rene_>     - use lsb_release -is to differentiate whether we build for Debian or
[12:57] <fabbione> [ new ]  { old }
[12:57] <_rene_>       Ubuntu and disable the KDE stuff when building on Ubuntu, thanks
[12:57] <_rene_>       Matthias Klose for the lsb_release hint. Honour
[12:57] <_rene_>       DEB_BUILD_OPTIONS=withkde to enable build with KDE on Ubuntu [RE] 
[12:57] <_rene_> is the changelog entries
[12:58] <Kamion> fabbione: you do need Section: debian-installer
[12:58] <Kamion> fabbione: may have to change kernel-wedge to ensure that
[12:58] <fabbione> Kamion: yes.. 
[12:58] <fabbione> i was looking into it
[12:59] <Kamion> fabbione: just add '$pkg{Section}'="debian-installer";' on line 154
[12:59] <Kamion> fabbione: of commands/gen-control
[12:59] <fabbione> ehhee
[12:59] <Kamion> lucas_: tla register-archive http://people.ubuntulinux.org/~cjwatson/archives/colin.watson@canonical.com--2004
[12:59] <fabbione> i was doing the same :-)
[12:59] <jdub> _rene_: cool :)
[01:00] <Kamion> lucas_: tla get colin.watson@canonical.com--2004/cdimage--mainline--0
[01:00] <Kamion> lucas_: bin/update-local-indices may help you
[01:00] <jdub> i have just uploaded the coolest 2.9 feature EVER
[01:00] <jdub> gnome-user-share
[01:00] <lucas_> Kamion: cool thanks a lot :)
[01:00] <jdub> it makes me very horny
[01:00] <Kamion> lucas_: except you'll need an apt-ftparchive.conf, hmm
[01:00] <jdub> that is all
[01:01] <lucas_> Kamion: I wrote one, but I'm not sure about it
[01:01] <seb128> jdub: rock !
[01:01] <fabbione> Kamion: do you want to upload kernel-wedge with it?
[01:02] <daniels> FRIGGING HELL LIBAO
[01:02] <Kamion> lucas_: http://cdimage.ubuntulinux.org/code/apt-ftparchive.conf is what I use
[01:02] <jdub> seb128: it's *really* cool
[01:02] <Kamion> lucas_: (I think the content-type's broken for that file, sorry)
[01:02] <lucas_> it's ok
[01:02] <Mithrandir> Kamion: text/plain is fine for that file?
[01:03] <daniels> calc: libao is broken to hell, dude
[01:03] <Kamion> fabbione: doing
[01:04] <Kamion> Mithrandir: hm, strange, the last time I looked at it in w3m the line-wrap was broken, but now it's fine
[01:04] <Mithrandir> Kamion: wget claims it's text/plain at least.
[01:04] <lucas_> Kamion: btw, I'm documenting what I'm doing. no problem with the tla archive going public ?
[01:04] <Kamion> lucas_: not at all
[01:05] <Kamion> lucas_: I'm not sure what the licence ought to be on my code there; TBH there probably isn't an awful lot that's copyrightable anyway
[01:05] <Kamion> but I'd have to check with Mark if you're creating derived works of it
[01:05] <lucas_> ok
[01:07] <lucas_> well I hope I won't have to modify it too much ; I'm not that interested in hacking debian-cd ;)
[01:09] <Kamion> fabbione: 0.25ubuntu5 uploaded
[01:09] <fabbione> Kamion: cool!
[01:11] <Kamion> (and change committed to d-i svn)
[01:13] <lamont> _rene_: Kamion: chroot build/chroot-hoary dpkg -l lsb-release
[01:13] <lamont> No packages found matching lsb-release.
[01:13] <lamont> it's not build-essential
[01:13] <_rene_> but we build-depend on it anyway, so no problem ;-)
[01:14] <jdub> elmo: ping
[01:14] <elmo> jdub: ?
[01:15] <jdub> elmo: can you kill g-u-s in NEW? i'm going to upload another one
[01:15] <jdub> then could you let it through? :)
[01:16] <fabbione> Kamion: http://people.ubuntulinux.org/~fabbione/17_to_18.diff <- NEW with bug fixes
[01:16] <fabbione> including the debian/control thing you mentioned yesterday
[01:16] <elmo> rejected
[01:17] <Kamion> fabbione: I'm still not going to have disk space to build it though ;)
[01:17] <fabbione> Kamion: i guess i will have to simulate a multi kernel thingy here than
[01:18] <fabbione> i need to verify that the archs with more than one kernel d-i set works
[01:19] <Kamion> fabbione: like I suggested earlier, can you get an account on a box in the LAN to do this?
[01:19] <Kamion> we must have a powerpc porting system
[01:19] <jdub> $ dput ubuntu gnome-user-share_0.3-1_source.changes
[01:19] <jdub> Already uploaded to upload.ubuntu.com
[01:19] <jdub> Doing nothing for gnome-user-share_0.3-1_source.changes
[01:19] <jdub> ^ wtf?
[01:19] <elmo> rm g-u-s*.upload
[01:19] <fabbione> Kamion: oh yeah... i didn't read the scrollback
[01:20] <fabbione> Kamion: but i can simulate it locally
[01:20] <jdub> elmo: thanks
[01:20] <fabbione> no big deal ;)
[01:20] <fabbione> on the other side.. i want to be 200% sure
[01:20] <fabbione> elmo: what is the new ppc porting box?
[01:20] <daniels> fabbione: davis
[01:20] <fabbione> daniels: ok
[01:21] <fabbione> elmo: can i get linux-source build-dep + kernel-wedge ?
[01:21] <lupus_> jdub, wouldn't it be better if you open websites that they are openend in a new tab by default?
[01:21] <fabbione> (version -4 is ok)
[01:21] <fabbione> for kenrel-wedge
[01:21] <jdub> lupus_: i think that's a good option for advanced users to enable
[01:22] <lupus_> why advanced users?
[01:22] <lupus_> if you give the new tab autofocus
[01:22] <jdub> because they'll know their page opened in the other window in a tab, instead of popping up straight in front of them (as reasonably expected)
[01:23] <elmo> fabbione: done
[01:23] <fabbione> elmo: thanks!
[01:24] <lupus_> if firefox is already open the current website is just changed, firefox doesn't get caption
[01:24] <lupus_> and as I user I would wonder why my other page went
[01:25] <jdub> not with the current default settings
[01:25] <jdub> a new window is opened
[01:25] <jdub> (unless we've had a reversion of those settings in hoary)
[01:25] <jdub> (which is not unlikely)
[01:25] <lupus_> ic sorry I have maybe some old settings lying around :p
[01:33] <fabbione> Kamion: how much did you diverge linux-kernel-di from debian kernel-di ?
[01:34] <Mithrandir> fabbione: how far from being debootstrappable is sparc?
[01:35] <fabbione> Mithrandir: it will take a while still
[01:36] <fabbione> Mithrandir: because i am building ubuntu on top of ubuntu
[01:36] <fabbione> Mithrandir: but sparc64 kernels are ready
[01:36] <fabbione> Mithrandir: i am going to drop sparc32 on the floor for now
[01:36] <fabbione> if someone wants it, he will provide patches
[01:37] <fabbione> theoretically i could upload the kernel in hoary
[01:37] <fabbione> but since i have like 3 branches at the moment
[01:37] <fabbione> i would prefer to go for one at a time
[01:37] <fabbione> first udeb integration
[01:37] <fabbione> 2.6.9
[01:37] <fabbione> sparc integration
[01:38] <fabbione> Mithrandir: i would say that i should be able to give you all the debs in a week or so
[01:38] <fabbione> it's only question of time basically
[01:40] <Mithrandir> fabbione: goodie.
[01:41] <fabbione> + i am catching a bunch of FTBFS due to the xlibs split that we didn't notice in the first run
[01:42] <fabbione> so that's even better
[01:42] <Kamion> fabbione: debdiff, man :)
[01:42] <fabbione> daniels: 4132
[01:42] <Kamion> fabbione: a bit
[01:42] <fabbione> Kamion: ?
[01:42] <Kamion> fabbione: firmware, extra modules, etc.
[01:42] <Kamion> fabbione: Debian uses linux-kernel-di-*, not kernel-di-*; using debdiff to see the differences is trivial
[01:42] <fabbione> AHHH
[01:43] <fabbione> ok
[01:43] <fabbione> sure
[01:43] <daniels> fabbione: eh?
[01:44] <fabbione> daniels: #4132 <-
[01:44] <daniels> yeah
[01:44] <daniels> any specific reaosn why I won it?  or do you need it urgently or something?
[01:44] <fabbione> daniels: because it's a FTBFS?
[01:46] <elmo> whee, expect is FUBAR in hoary
[01:57] <Kamion> pitti: isn't the LSB init stuff in lsb-base, not lsb-release?
[01:57] <Kamion> pitti: (re your anacron changelog)
[01:59] <lucas_> Kamion: tools/scanpackages generates errors when running apt-ftparchive :
[01:59] <lucas_> E: Could not open file /space/ubuntu-custom/scratch/tmp/warty-i386/CD1/dists/warty/local/binary-i386/Packages.gz.new - open (2 No such file or directory)
[01:59] <pitti> Kamion: indeed
[01:59] <lucas_> E: Could not open file /space/ubuntu-custom/scratch/tmp/warty-i386/CD1/dists/warty/local/binary-i386/Packages.new - open (2 No such file or directory)
[01:59] <lucas_> E: Error Processing directory dists/warty/local/binary-i386/
[01:59] <pitti> Kamion: I copied that from another package
[01:59] <pitti> Kamion: so there is at least one other faulty package
[01:59] <Kamion> lucas_: might need to create those directories by hand
[01:59] <pitti> Kamion: I'll fix it
[01:59] <Kamion> pitti: ta
[02:00] <Kamion> lucas_: er, not the directories under CD1, though
[02:00] <lucas_> mmh
[02:00] <Kamion> lucas_: actually FWIW I get the same error in the official builds, it's harmless as far as I know
[02:01] <lucas_> ok
[02:01] <lucas_> :-)
[02:01] <Kamion> I should probably investigate it at some point
[02:01] <lucas_> the script responsible for copying the Packages to CD1/ should create the dir
[02:01] <lucas_> or they should be created before
[02:02] <Kamion> well, not sure; local is weird in a few places
[02:02] <Kamion> do you actually have anything in your local tree yet? the official builds (obviously) don't
[02:03] <lucas_> I do
[02:03] <lucas_> I rebuilt base-config
[02:03] <Kamion> hm, ok, it used to work fine for us though
[02:03] <lucas_> but debian-cd doesn't pick my local version
[02:03] <Kamion> did you increase the version of base-config?
[02:03] <lucas_> yup
[02:04] <jdub> fabbione: do you recommend doing "libx11-dev | xlibs-dev, libice-dev | xlibs-dev" and so on, or just depending on xlibs-dev?
[02:04] <fabbione> jdub: nothing should Build-Dep on xlibs-dev
[02:04] <daniels> jdub: the former
[02:04] <fabbione> you need to just use the real Build-Dep
[02:05] <daniels> fabbione: makes backports hard oh shit 0% battery
[02:05] <fabbione> libx11-dev, libice-dev, libdanielscrack-dev
[02:05] <Kamion> lucas_: try 'tools/apt-selection cache show base-config'
[02:05] <fabbione> jdub: still...
[02:05] <lucas_> correct version
[02:05] <lucas_> but I think I understand
[02:05] <Kamion> lucas_: your modified version?
[02:06] <lucas_> I mkdir -p /space/ubuntu-custom/scratch/tmp/warty-i386/CD1/dists/warty/local/binary-i386/
[02:06] <lucas_> then apt-ftparchive is able to generate the Packages on CD1
[02:06] <daniels> jdub: you only want xlibs-dev if you want to be compatible with pre-4.3
[02:06] <lucas_> so I just need to add the creation of this dir swhere in scanpackages
[02:06] <Kamion> lucas_: the issue is that nothing post-install expects local to be there
[02:06] <jdub> daniels: oh
[02:06] <jdub> bugger that, then
[02:06] <daniels> yes!
[02:07] <lucas_> Kamion: it doesn't really matter for me
[02:07] <Kamion> lucas_: so you can generate a CD that way, but base-config probably won't get installed
[02:07] <Kamion> lucas_: yes, it does :-)
[02:07] <lucas_> oh
[02:07] <daniels> fabbione: and -static-dev, and -static-pic
[02:07] <lucas_> so the best solution would be not to use local at all
[02:08] <fabbione> daniels: these are a bit more hard to kill.. but yes
[02:08] <Kamion> lucas_: you might find that easier
[02:08] <Kamion> I guess I should try using local for something again to make sure it still works the way I remember
[02:09] <lucas_> when apt-ftparchive finds 2 packages with the same name but different versions
[02:09] <lucas_> I expect it only writes the newer one in Packages ?
[02:09] <Kamion> lucas_: but really the only reason we have LOCAL turned on at all is that in the early days of Ubuntu development the archive wasn't quite usable for building installable CDs, but I had a few local hacks that made it more or less work so I used those temporarily
[02:09] <Kamion> lucas_: right, should do
[02:10] <lucas_> ok, it should do the trick
[02:10] <Kamion> I've just turned LOCAL off for our builds
[02:10] <Kamion> let's see if it keeps working
[02:10] <lucas_> it probably will
[02:10] <lucas_> it WFM before I enabled LOCAL
[02:11] <Kamion> fabbione: probably has to wait for a full Debian release cycle though; that's a *lot* of source packages to change
[02:11] <Kamion> fabbione: xlibs-dev is really handy too
[02:11] <Kamion> for users
[02:12] <daniels> Kamion: i dislike the entire idea of xlibs*
[02:12] <pitti> Kamion: stupid me...
[02:12] <daniels> Kamion: (not a rhetorical question, largely because I don't know the answer) do we have a gnome-dev?
[02:12] <pitti> Kamion: anacron does depend on lsb-base, I just made the error in the changelog
[02:13] <fabbione> Kamion: yes i know.. :( but i plan to kill it in etch if Branden will allow me
[02:13] <bob2> daniels: gnome-devel
[02:14] <fabbione> Kamion: more than users.. we are talking about lazy maintainers
[02:14] <fabbione> ;)
[02:14] <daniels> hm
[02:14] <Kamion> daniels: gnome-devel and gnome-core-devel, yeah
[02:14] <daniels> well, yeah
[02:14] <daniels> maybe we should keep xlibs-dev and make it illegal to b-d on them (make dpkg-source break)
[02:14] <Kamion> fabbione: users are more important in the long run though
[02:15] <daniels> either that or just rename it to x-libraries-devel
[02:15] <daniels> -> users keep the metapackage to install, b-ds on xlibs-dev break and I get to go fix them all
[02:15] <Kamion> I dislike the idea of creating a billion serious bugs just because you don't like a particular package
[02:15] <Kamion> that's SO not fun
[02:15] <fabbione> Kamion: it's not like it will happen in sarge
[02:15] <Kamion> fabbione: so?
[02:16] <fabbione> there is an entire release cycle to fix that
[02:16] <Kamion> fabbione: I don't want everything to go to shit after sarge either
[02:16] <Kamion> modifying that many packages is HARD
[02:16] <Kamion> we did it for /usr/share/doc, and it truly sucked
[02:16] <Kamion> it took two release cycles to get all the packages changed
[02:16] <fabbione> Kamion: yes i understand that...
[02:16] <Kamion> and it sucked up huge amounts of developer effort that was just a total waste of time
[02:17] <Kamion> you're basically proposing deliberately wasting developer time
[02:17] <jdub> fabbione: yeah man, stop punching me in the face!
[02:17] <fabbione> not really.. i am proposing to kill a meta package that is there because maintainers are lazy to figure out their build-dep
[02:17] <fabbione> i know it's not "nice"
[02:18] <fabbione> but imho it's cleaner to know on what you really depend on
[02:18] <fabbione> instead of just pulling a gazillion of libs for nothing
[02:18] <Kamion> $ grep-dctrl -nsPackage -FBuild-Depends,Build-Depends-Indep xlibs-dev /var/lib/apt/lists/riva_debian_dists_unstable_main_source_Sources | wc -l
[02:18] <Kamion> 795
[02:18] <Kamion> I'm sorry but I don't think that's acceptable
[02:19] <Kamion> you must have a transition plan
[02:19] <smurfix> fabbione: the shlibdeps already tell that. Maybe we can use them to automate the transition.
[02:19] <Kamion> and removing xlibs-dev and fixing everything up afterwards isn't one
[02:19] <daniels> Kamion: fixing everything up before?
[02:19] <fabbione> Kamion: xlibs-dev is already a transition to avoid breakage from X < 4.3
[02:19] <Kamion> daniels: that would be less bad
[02:20] <fabbione> so i think it will be time with X > 6.8 to shake the dust away
[02:20] <elmo> I have to agree with kamion here - you're imposing a massive transition for aesthetics, not technical reasons and then trying to justify by blaming maintainers for being lazy
[02:20] <Kamion> fabbione: it's obviously not a completed transition, because it's still being used
[02:20] <Kamion> fabbione: realise that maintainers aren't AWARE that you think xlibs-dev is a lazy thing to use
[02:20] <daniels> i think 'lazy' isn't the right justification
[02:21] <daniels> having properly articulated b-ds is a win
[02:21] <elmo> daniels: how?
[02:21] <fabbione> Description: X Window System client library development files transitional package
[02:21] <Kamion> although it is oldlibs - but I don't think it's been particularly widely announced, e.g. wishlist bugs on everything that b-ds on it
[02:21] <fabbione> This transitional package is
[02:21] <fabbione>  only depended upon by packages that haven't yet corrected their dependencies to
[02:21] <fabbione>  reflect the new library arrangement.
[02:21] <elmo> and how does this win counter-balance the massive transition effort involved?
[02:21] <Kamion> fabbione: how about you go file the wishlists to move away from it now?
[02:21] <elmo> fabbione: dude, I have packages that depend on X and have done for years - what makes you think I would have read that description since it changed?
[02:21] <daniels> elmo: eh, I'd rather like people to know what they actually use
[02:22] <elmo> daniels: how is that a technical win, dude
[02:22] <fabbione> Kamion: it will make more sense immediatly after sarge and do it properly, once, with Xorg split
[02:22] <Kamion> browser-history, certainly, I haven't uploaded that for ages
[02:22] <elmo> don't hand wave.  give me cold, hard, technical benefits that don't involve aesthetics
[02:22] <daniels> rather than just a rough sledgehammer 'oh right, so I'll install everything from the base authentication through to four toolkits'
[02:22] <jdub> elmo: with the Big X Split coming up, it will matter more
[02:22] <daniels> elmo: not dragging in 200MB of crap you don't need
[02:22] <daniels> and yes
[02:22] <elmo> jdub: why?
[02:23] <jdub> elmo: (i'm not taking sides here, but that is a valid point)
[02:23] <elmo> AFAIK xlibs-dev could still exist in a X split world, AFAICS
[02:23] <daniels> yes, it could
[02:23] <jdub> elmo: because those libs will be released separately
[02:23] <Kamion> fabbione: and for all of this we have to figure out how to get the changes into testing, which takes a lot of time in the real world
[02:23] <daniels> but it's just a craptastic idea imo
[02:23] <fabbione> elmo: don't get me wrong.. i understand why xlibs-dev exists...
[02:23] <daniels> elmo: your packages don't depend on X, BTW
[02:23] <elmo> daniels: bzzt
[02:24] <daniels> elmo: they likely depend on the general client-side X library, which is up for replacement
[02:24] <Kamion> there's also random bits of documentation out there that say "to get X libraries on Debian, install xlibs-dev"; that documentation doesn't just go away
[02:24] <elmo> I maintain a KDE-based package - I think that's likely to depend on X
[02:24] <jdub> if you have to depend on a specific version, you'd have to switch or add a specific depends anyway
[02:24] <daniels> elmo: 'depend on X' is handwavey random crap
[02:24] <jdub> daniels: i don't get why gtk-dev doesn't have the right x dev depends
[02:24] <daniels> jdub: it doesn't?
[02:24] <jdub> no
[02:24] <jdub> that just bit me with GUS
[02:24] <fabbione> Kamion: X usually moves in one shot in testing..
[02:24] <elmo> daniels: you know what I mean
[02:25] <daniels> jdub: then it needs updating
[02:25] <Kamion> fabbione: I'm not talking about X
[02:25] <elmo> so basically, the only technical reason you guys have is "so we install less as build-depends" .. mmk
[02:25] <daniels> elmo: yes, I assume you depend on the general X11 client-side library, which is up for replacement (thus chaining through to the X authorisation library), and you probably depend on the library that provides a few extensions
[02:25] <Kamion> fabbione: I'm talking about nearly 800 other source packages, many of which will have fun interdependencies
[02:25] <fabbione> Kamion: but i see the complexity.. i am just saying that xlibs-dev was is already a transitional package and that can be killed doing thing properly
[02:25] <daniels> elmo: possibly the I Can't Believe It's Not IPC library
[02:25] <fabbione> Kamion: yes.. i grok that
[02:26] <jdub> elmo: when the big X split happens, the difference between the libs will matter
[02:26] <daniels> elmo: but probably not all of four toolkits
[02:26] <daniels> elmo: and whatever random crap was in 'xlibs'
[02:26] <Kamion> fabbione: if the transition had been done properly I wouldn't be seeing nearly 800 build-deps in unstable right now
[02:26] <jdub> elmo: you will be depending on different versions
[02:26] <Kamion> (and that's only main)
[02:26] <Kamion> jdub: surely if the versions are being allowed to get wildly out of sync you lose anyway
[02:26] <jdub> elmo: up til now, it's always been one great big galumphing monolithic release
[02:26] <jdub> Kamion: with x, no, because it's been api/abi stable for years anyway
[02:27] <Kamion> jdub: if xlibs-dev depends on say x-dev and libx11-dev, how does it make any difference if you write out the build-dependencies separately?
[02:27] <Kamion> jdub: metapackages and writing out build-deps separately are semantically identical
[02:27] <jdub> Kamion: because you may need libx11-dev 56 and libice-dev 23
[02:27] <daniels> elmo: this is what you are build-depping on --
[02:27] <jdub> Kamion: and which package owns xlibs-dev in future? and which versions does it depend on?
[02:27] <Kamion> jdub: uh-huh, so if you do you can say so
[02:27] <fabbione> Kamion: i don't have enough historical backgroud to tell what was done to inform developers that xlibs-dev is a transitional package, so i cannot comment on it. On the otherside this doesn't not mean that things can be still done
[02:27] <daniels> daniels@catsby:~/canonical/xorg/arch/source/xorg-6.8.1/build-tree/xc/lib% ls -ld * | wc -l && du -sh . | tail -1
[02:27] <daniels> 64
[02:27] <daniels> 701M    .
[02:28] <Kamion> fabbione: right, but it means that dropping xlibs-dev for etch is a highly ambitious goal
[02:28] <jdub> Kamion: so you end up switching anyway :)
[02:28] <daniels> elmo: if you're unable to further articulate your build-deps, then I'm sorry, but that's just crap
[02:28] <elmo> daniels: dude, grow the fuck up
[02:28] <Kamion> jdub: sure, in the unlikely event that say browser-history needs to
[02:28] <elmo> I'm not unable to articulate anything
[02:28] <daniels> elmo: i understand that this is a change
[02:28] <daniels> elmo: and that previously there was just xlibs
[02:28] <jdub> daniels: the "you are lazy" reasoning doesn't help your case
[02:28] <daniels> elmo: but what I'm trying to say here is that was *wrong*
[02:28] <fabbione> Kamion: i never said it was going to be easy, did I? ;)
[02:28] <daniels> agh
[02:28] <daniels> elmo: not meant to imply that this is your fault or that you're lazy or whatever
[02:29] <daniels> a non-specific, hand-wavey 'you'
[02:29] <daniels> Kamion: (x-dev is an aberration anyway)
[02:29] <elmo> daniels: sorry, but it doesn't matter whether you're accusing me or the n hundred maintains of being "lazy or whatever", all it's doing is convincing me you have only religious reasons for this
[02:29] <Kamion> daniels: so ... what am I supposed to do to get X11/Xos.h and X11/Xatom.h?
[02:30] <daniels> Kamion: the selection of files for x-dev is ... odd
[02:30] <Kamion> daniels: this isn't particularly helping your case either you know :)
[02:30] <daniels> Kamion: sort of randomly picked from whichever files didn't neatly fit into whichever categories were picked as a first glance
[02:30] <daniels> Kamion: i know
[02:31] <daniels> i'm just saying, that if I were designing it from the ground up right now, I know how I'd package it (largely because I already have packaged it, from the ground up)
[02:31] <Kamion> if maintainers get the impression that X is going to randomly move files around between development packages, they're fully entitled to say "screw it, I'll just b-d on the metapackage and that'll work no matter what they rearrange"
[02:31] <daniels> and that is not bundling 64 libraries with a 701MB build-tree into a single package
[02:31] <daniels> and that the result is rather a horror and I'd like it to go away
[02:32] <daniels> Kamion: understood.  that's why I would rather like to reorganise it sooner rather than later, and then come out with a transition document (or, hell, even a semi-automated 'this is what you probably need to do for package <foo>' generator) describing it
[02:32] <Kamion> on the scale of 11 years of Debian development, and by comparison to say libc6-dev, X *does* randomly move files around, and expecting every maintainer to keep up with it when it's not clearly documented for them is a bit much
[02:33] <fabbione> Kamion: usually it happens only on major releases
[02:33] <fabbione> Kamion: but that's also an upstream decision
[02:33] <smurfix> Kamion: so now's the chance, with the X split, to actually document it (and, dare I say, freeze the stuff...)
[02:34] <fabbione> Kamion: like the last library split. it was not our decision
[02:34] <daniels> Kamion: right now, it's so utterly random because at the time we were doing the split, neither Branden nor I had an idea of what most of them did or belonged
[02:34] <Kamion> sure
[02:34] <fabbione> Kamion: upstream changed a bunch of libs from static to shared...
[02:34] <daniels> Kamion: i did part of the xlibs split, so I'm thus partially to blame for the current craptastic situation
[02:34] <Kamion> I'm just saying that preserving build-dep compatibility *is* important
[02:34] <Kamion> even (or especially) if it's your (collective) mistake
[02:34] <daniels> Kamion: the xlibs-static-* stuff was upstream realising that saying the Xinerama API was 'too fluid for a shared library' was arse
[02:35] <daniels> Kamion: understood
[02:35] <daniels> Kamion: i'm just saying, i will be happy when no packages b-d on xlibs-dev
[02:36] <fabbione> Kamion: i don't disagree.. but if some changes might require other maintainers to do some work, i am not going to ask the buildd to pull 200 libs because people can't change a line in debian/control
[02:36] <fabbione> ^^missing: well.. they can move their butt and do it
[02:37] <elmo> ARGH
[02:37] <elmo> the buildd couldn't CARE LESS
[02:37] <elmo> it has the 200 packages cached
[02:37] <Kamion> I'm just wondering in what way you expect the release team to accept this
[02:37] <elmo> the time-to-install is noise on the big scale
[02:38] <elmo> please do not use that as an argument, it's total BS
[02:38] <Kamion> it's entirely possible that britney might in the future decide to hold X out of testing until it stops breaking build-deps
[02:38] <daniels> Kamion: neither of us are arguing to break B-Ds at the moment
[02:38] <Kamion> since having britney enforce that sort of thing is high on the wishlist for post-sarge
[02:38] <fabbione> elmo: i am not talking only about install time...
[02:39] <elmo> so what else is left?
[02:39] <elmo> removal time?
[02:39] <elmo> as I said, on any Debian buildd, the 200  packages will be cached locally
[02:39] <fabbione> elmo: space (s390), bw 
[02:39] <elmo> s390 needs it's space fixed for other reasons - that's not an argument
[02:39] <elmo> it's all virtual disks anyway
[02:39] <Kamion> (like kde)
[02:39] <daniels> elmo: should we merge x back into a single package, then?
[02:40] <daniels> i just fundamentally fail to see how arbitrarily connecting a bunch of disconnected libraries can be a good thing
[02:40] <elmo> daniels: nothing I said suggested that
[02:40] <elmo> but it's a nice strawman
[02:40] <daniels> xlibs-dev as a concept is fundamentally dead anyway
[02:41] <fabbione> elmo: as soon as we reupload X and the packages needs to be refreshed.. you are starting from the beginning, but yeah.. we have resources
[02:41] <daniels> given that no libraries added after 4.2.x have been added to xlibs-dev, it is no longer 'the clientside X libraries'
[02:41] <daniels> given that two toolkits are in there, where the distinction was drawn in the first place is a mystery, anyway
[02:41] <Kamion> fabbione: it amortises
[02:42] <daniels> so right now it's just an arbitrary grouping of packages that is kept alive because of inertia
[02:42] <Kamion> fabbione: as soon as one package fetches some set of libraries it's free for all the others
[02:42] <fabbione> Kamion: yes.. i agree on that..
[02:43] <fabbione> Kamion: but right now you pull 200 libs each time you ask fro xlibs-dev (and yes they get cached)
[02:43] <daniels> the other point is that the X packaging is still horrifically broken
[02:43] <Kamion> (I'm fixing most of my packages now; can't fix groff at the moment though)
[02:43] <fabbione> Kamion: perhaps with a proper b-d there will be less than 10
[02:43] <daniels> and the damage was only barely just started to be unravelled in 4.3.0, and mistakes were made along the way
[02:43] <fabbione> Kamion: clearly these are just fake numbers
[02:43] <fabbione> Kamion: but there is still a benefit, even if minimal
[02:44] <Kamion> source: xfree86 only generates 26 -dev packages ;)
[02:44] <fabbione> Kamion: check x.org
[02:44] <fabbione> there must be a bunch more
[02:44] <Kamion> fabbione: still not 200
 Kamion: clearly these are just fake numbers
[02:44] <Kamion> daniels: X is certainly allowed to unravel its mistakes, and that's a good thing
[02:44] <fabbione> 200 was to give the idea
[02:45] <Kamion> daniels: the point is that the rest of the distribution has to be allowed to catch up gradually, not in a flag day
[02:45] <daniels> Kamion: absolutely, I agree with you here
[02:46] <Kamion> daniels: if you guys agree on that, then we seem to be in agreement, except perhaps about what "gradually" means. :)
[02:46] <daniels> Kamion: (also, bear in mind that my ideal world has ~70 library source packages)
[02:46] <daniels> Kamion: rather violent agreement, it seems
[02:46] <fabbione> Kamion: but none of us ever wrote "from one day to another"
[02:46] <fabbione> at least.. if my short memory doesn't betrade me
[02:46] <Kamion> fabbione: removing xlibs-dev before all its b-ds are fixed counts as a flag day
[02:47] <daniels> Kamion: agreed
[02:47] <fabbione> Kamion: neither i wrote to do it that way
[02:47] <fabbione> i just expressed the whish to kill xlibs-dev
[02:47] <fabbione> wish even
[02:48] <fabbione>  fabbione would really like to see xlibs-dev vanishing
[02:48] <Kamion> ok
[02:48] <fabbione> that was all i wrote
[02:48] <fabbione> in terms of future plans
[02:48] <Kamion> I understood the stuff about etch to be "come hell or high water" :)
[02:48] <fabbione> i didn't say to do it now or how
[02:48] <Kamion> alright, sorry then :)
[02:49] <fabbione> well clearly i would like to do it asap, but asap has a very large relative terms in debian
[02:49] <daniels> Kamion: basically, I know how I would like X to be packaged, and that doesn't involve xlibs*
[02:49] <fabbione> Kamion: don't be sorry.. i love to discuss
[02:49] <fabbione> Kamion: and i like to hear different opinions on stuff
[02:49] <fabbione> Kamion: other than daniels' one that we know are crap :P
[02:49] <Kamion> still, has the bonus that I have a bunch of my packages to fix, some of which haven't seen an upload in two years
[02:50] <fabbione> it's a good opportunity to refresh the timestime on the inodes :)
[02:50] <fabbione> timestamps
[02:51] <zul> morning
[02:52] <elmo> I did a madison patch once that showed you the last upload time - I should resurrect it
[02:53] <elmo> almost everyone I showed the output to for their packages went "oh, yeah, I forgot about <foo>"
[02:53] <fabbione> libcaca <- i wonder who invented this name
[02:53] <fabbione> literaly translated from italian is: libtakeashit
[02:53] <Kamion> elmo: cool
[02:53] <smurfix> elmo: sometimes that's actually a good thing ;-)
[02:54] <smurfix> fabbione: what does it do?
[02:54] <fabbione> smurfix: no idea
[02:54] <smurfix> fabbione: hopefully something unrelated to its name ...
[02:54] <daniels> oh wow, I still maintain dbtcp
[02:54] <elmo> ^-- like that :P
[02:55] <fabbione>  libcaca is the Colour AsCii Art library.
[02:55] <stratus> fabbione, hah it doesn't sound good in portuguese too.
[02:56] <daniels> elmo: hey, I don't have any bugs to my name in debian
[02:57] <infinity> daniels : Only because you sneak around co-maintaining things, so the bugs accumulate on other addresses. :)
[02:58] <infinity> (adconrad@debian.org doesn't have any open bugs either!)
[02:58] <daniels> infinity: i don't see *you* fixing apache2 :P
[02:58] <infinity> daniels : Yeah, yeah.  I'll get another upload in before I go on vacation.
[02:59] <infinity> Maybe throw something in to shut up all the "I'm too stupid to set up SSL without handholding" whiners.
[02:59] <infinity> I dunno.
[02:59] <daniels> heh
[02:59] <thom> heh
[02:59] <daniels> 'morning thom
[02:59] <fabbione> ihih
[02:59] <elmo> if apache's SSL handling wasn't so insanely fragile it might garner less whiners :-P
[02:59] <fabbione> hey thom
[02:59] <fabbione> thoom: still jetlagging?
[02:59] <daniels> elmo: dude, you should give 2.0.16 a shot
[03:00] <daniels> elmo: those were they days where they sent out a different 100kB 'this might work, hey actually it bombs out on compile' SSL patch
[03:00] <elmo> "WAH WAH, you omitted one section of the config file - I'm not going to tell you about it, I'm going to just break silently, hahaha, good luck.  p.s. don't try loglevel debug, it won't help, kthxbye"
[03:00] <thom> elmo: well, i think they reasonably assume that someone wanting ssl on should have a cert attached to that vhost
[03:00] <elmo> thom: I'm not arguing it shouldn't break, I'm arguing it should break BETTER
[03:00] <thom> agreed
[03:00] <infinity> More spectacularly, even.
[03:01] <infinity> A lot of apache could do with being more verbose.
[03:01] <infinity> Or differently verbose.
[03:01] <infinity> Some of the error messages are a bit... Odd.
[03:01] <fabbione> infinity: we can still add a "poweroff" if apache fails to start
[03:01] <thom> some of them are just genius
[03:01] <thom> but they've gotten better
[03:02] <infinity> I love watching my requests randomly disappear into the ether too.
[03:02] <thom> fabbione: and yes, my timezones are broken
[03:02] <thom> infinity: did you ever play with that "dav eats my files" bug?
[03:02] <infinity> thom : I looked atit a bit when it came in, but not since.
[03:03] <infinity> thom : I was going to look at it again before the next upload.
[03:03] <infinity> thom : The logic in there seems... Broken.
[03:03] <jdub> thom: btw, what do you think about checking the "lock screen after X minutes" setting before locking on lid down?
[03:03] <infinity> thom : If you follow it, it's definitely wrong.  I'm just undecided as to what the RIGHT behaviour is.
[03:03] <thom> jdub: i think it should lock instantly
[03:03] <infinity> thom : (Right now, if a move fails, it seems to want to cover its tracks by deleting... both copies..)
[03:04] <thom> infinity: yes, that is very broken indeed
[03:04] <jdub> thom: i'm thinking if you have that turned off, it shouldn't lock
[03:04] <daniels> ah, the log stuff is shit
[03:04] <seb128> jdub: what should we do about #3043 ?
[03:04] <thom> jdub: bleah, maybe
[03:05] <elmo> jdub: eww, please don't do that
[03:05] <daniels> 'the directory you specified for one of your logs doesn't exist, so I'm going to bail now, and not even print a \n, because I'm such a hateful bastard'
[03:05] <elmo> I want my laptop locked on resume
[03:05] <daniels> infinity: !
[03:05] <elmo> please at least me allow to specify that easily
[03:05] <jdub> elmo: ideally, this would be power-management-user-policy stuff
[03:05] <jdub> but for now, perhaps we should hook it to that preference
[03:06] <infinity> jdub : A certain other OS that shall remain nameless makes a clear separation between "locking the screen" (ie: screensaver settings) and "lock on resume).
[03:06] <thom> yeah, if that's turned off in xscreensaver settings, it makes sense
[03:06] <jdub> elmo: it would work that way by default even with this change
[03:06] <jdub> infinity: that's the ideal, yes.
[03:06] <thom> daniels: yes, that's an RoUS special
[03:06] <jdub> seb128: oh
[03:06] <jdub> seb128: i think i have a viable bounty hunter/team for that
[03:07] <daniels> thom: WOO!
[03:07] <seb128> jdub: should I reassign it to you so ?
[03:07] <jdub> seb128: yeah, thanks
[03:07] <seb128> np, thank you :)
[03:07] <thom> g'night mate
[03:08] <infinity> thom : G'night.
[03:09] <daniels> infinity: night dude
[03:17] <elmo> gar
[03:17] <elmo> tcl 8.4.6-1 works, tcl 8.4.7-1 breaks
[03:22] <fabbione> still 2 and 1/2 kernels to build :(
[03:22] <elmo> you should do amd64 instead
[03:23] <elmo> still mutliple variants, but it'll build a bucket load faster
[03:23] <fabbione> elmo: not for the udebs part
[03:23] <Kamion> elmo: powerpc has the three sets of udebs
[03:23] <fabbione> amd64 creates only one set of udebs
[03:23] <elmo> ah I see
[03:24] <fabbione> at this speed i will have time to integrate sparc too :-)
[03:24] <fabbione> elmo: i have the kernels btw
[03:25] <fabbione> i am doing another round of build to be 100% sure
[03:27] <Matt|> daniels, there's a guy in ubuntu-it who has a problem with his touchpad and xfree, have you got a few minutes for him?
[03:27] <daniels> Matt|: yeah, sure
[03:27] <jdub> ok
[03:27] <jdub> who's running hoary and i386 and wants to try some SEXY new gnome love?
[03:27] <tseng> jdub: pick me, pick me!
[03:27] <smurfix> jdub: ?
[03:28] <Matt|> jdub, are you appealing for regular users?
[03:28] <jdub> hell yeah
[03:28] <tseng> hah.
[03:28] <jdub> this is maximum SASS
[03:28] <stratus> jdub, i! :)
[03:28] <stratus> break my hoary!
[03:28] <jdub> http://people.ubuntulinux.org/~jdub/gnome-user-share_0.3-2_i386.deb
[03:28] <jdub> install that
[03:28] <Matt|> daniels, thanks a lot, I've asked him to pm ya
[03:28] <jdub> and run gnome-file-share-properties
[03:28] <stratus> doing..
[03:28] <Matt|> jdub, i'm running 686 kernel it is cool?
[03:28] <jdub> then look in Computer > Network :-)
[03:28] <daniels> Matt|: thanks
[03:29] <Matt|> daniels, thank YOU
[03:29] <jdub> Matt|: hrm?
[03:29] <daniels> jdub: how come it's not in the archive?
[03:29] <tseng> its going that way
[03:29] <stratus> jdub: wait, it needs libhowl0, installing...
[03:29] <jdub> daniels: g-u-s?
[03:29] <stratus> *sigh*, apache2 and mdnsresponder?
[03:29] <stratus> it needs to be really cool
[03:29] <jdub> it is
[03:29] <daniels> jdub: yeah
[03:29] <daniels> jdub: you should upload it, dude
[03:30] <jdub> daniels: you're fixing why it doesn't build right now, aren't you? :)
[03:30] <zul> jdub, what is it?
[03:30] <Matt|> jdub, too many dependencies for me
[03:30] <stratus> still installing...
[03:30] <jdub> Matt|: it's not...
[03:30] <daniels> jdub: yeah :)
[03:30] <Matt|> jdub, apache2?
[03:30] <jdub> zul: user level zeroconf/webdav file sharing for gnome
[03:31] <jdub> Matt|: just update-rc.d -f apache2 remove after you've installed it
[03:31] <stratus> hey jdub
[03:31] <stratus> is libhowl0 >= 0.9.8-1 around?
[03:31] <jdub> in hoary, yes
[03:31] <jdub> this is a hoary package
[03:31] <jdub> for testing and so on
[03:31] <stratus> updating packages now...
[03:31] <zul> jdub, its installed
[03:31] <smurfix> jdub: Syntax error on line 39 of /usr/share/gnome-user-share/dav_user.conf: Invalid command 'MinSpareServers', perhaps mis-spelled [...] 
[03:32] <jdub> smurfix: aha, which mpm do you have installed?
[03:32] <smurfix> jdub: sh
[03:32] <tseng> works for me
[03:32] <tseng> except, we still dont have howl in gnome-vfs?
[03:32] <smurfix> jdub: should grey out the password when it's "never" too
[03:32] <jdub> smurfix: yeah
[03:33] <jdub> tseng: yeah, we do
[03:33] <smurfix> jdub: the wrong one ;-)  -- apache2-mpm-worker
[03:33] <tseng> jdub: i dont see myself in network
[03:33] <thom> um, worker is the default? :-)
[03:33] <jdub> it should work with worker
[03:33] <pitti> Kamion: I'm currently merging the new udev 0.046, we need it for several reasons
[03:34] <jdub> hrm, i have prefork
[03:34] <pitti> Kamion: I merged all changes but some details of the udeb creation
[03:34] <smurfix> jdub: should, but apparently doesn't
[03:34] <stratus> I see the same error message about dav_user.conf
[03:34] <pitti> Kamion: the new udev Debian package already has provisions for building the udeb, however, some files are missing that you ship with 0.042
[03:34] <jdub> smurfix: yeah
[03:34] <jdub> okay dudes, install apache2-mpm-prefork
[03:35] <pitti> Kamion: could you take a look at it and finish the merge?
[03:35] <jdub> i'll figure out how to get it working with the other ones
[03:35] <daniels> libapache2-mod-dav, innit?
[03:35] <stratus> jdub, ok
[03:35] <jdub> daniels: a) it's not separate anymore, b) to run apache, you need an apache server :)
[03:35] <smurfix> jdub: I saw that error after running gnome-file-share-properties -- it should have appeared at install time too
[03:36] <jdub> smurfix: nah, it shouldn't have
[03:36] <stratus> smurfix, the same here.
[03:36] <daniels> jdub: well, yeah, heh
[03:36] <tseng> jdub: works!
[03:37] <jdub> alternatively, you can comment the two *Spare* lines in dav_user.conf
[03:37] <fabbione> gordian:/var/lib/dpkg/info# for i in `ls *-dev.*`; do grep ldconfig $i; done
[03:37] <fabbione> gordian:/var/lib/dpkg/info# 
[03:37] <fabbione> daniels: nobody calls ldconfig in -dev
[03:37] <daniels> why not just grep ldconfig *-dev.* ?
[03:37] <jdub> tseng: got computer > network love?
[03:37] <fabbione> grep ldconfig *-dev.*
[03:37] <fabbione> gordian:/var/lib/dpkg/info# 
[03:37] <tseng> jdub: indeed.
[03:37] <fabbione> daniels: better now?
[03:38] <jdub> sexy, huh? :)
[03:38] <tseng> jdub: using perfork
[03:38] <tseng> hot as hell
[03:38] <tseng> and not buggly like eppitance
[03:38] <jdub> heh
[03:38] <daniels> fabbione: i wasn't saying the results were wrong, just a wasteful method
[03:38] <mvo_> jdub: what is the future for gamin (the fam replacement)? will it stay or is that a dependencies that gnome would like to drop long-term?
[03:38] <jdub> hooray for reusing code from leading free software projects :)
[03:38] <daniels> anyway, I do not see how what we have can even possibly work
[03:38] <jdub> mvo_: it'll stay
[03:38] <daniels> elmo: would really love your input on this, please
[03:38] <mvo_> so I can safly use it in the update notifier tray icon?
[03:39] <jdub> yes, that's a great idea
[03:39] <jdub> mvo_: gamin is binary compat with a subset of fam, so it should work with either
[03:39] <stratus> jdub, but huh what's up with that default password?
[03:39] <mvo_> jdub: great, thanks :)
[03:40] <daniels> Any package installing shared libraries in one of the default library directories of the dynamic linker (which are currently /usr/lib and /lib) or a directory that is listed in /etc/ld.so.conf[42]  must use ldconfig to update the shared library system.
[03:40] <jdub> stratus: there is no default password, is there?
[03:41] <tseng> ya there is.
[03:41] <Kamion> pitti: sure, Md and I were talking about that
[03:41] <stratus> jdub, i see four asterisk when running 'gnome-file-share-properties' at password field
[03:41] <stratus> smurfix, don't you?
[03:41] <pitti> Kamion: http://people.ubuntu.com/~pitti/udev/
[03:41] <smurfix> stratus: same here
[03:42] <stratus> jdub, :)
[03:42] <pitti> Kamion: it contains the current status of the merge and the Ubuntu diffs that I did not yet merge (udeb-relevant)
[03:42] <jdub> stratus: you'll always see four asterisks
[03:42] <jdub> stratus: however long your password is
[03:42] <elmo> daniels: say what?
[03:42] <jdub> stratus: determine the length of the password as stored in ~/.gnome2/user-share/passwd
[03:42] <stratus> jdub, but i didn't write nothing i just started g-f-s-p
[03:42] <daniels> elmo: ok, so here's the scenario
[03:42] <daniels> libx11-6 installed, along with libX11.so.6 and libX11.so.6.2
[03:42] <daniels> calls ldconfig in postinst
[03:42] <jdub> stratus: even if the passwd is zero length, you will see four asterisks
[03:43] <daniels> -> libX11.so.6 registered with ld
[03:43] <daniels> --later--
[03:43] <daniels> libx11-dev installed, with libX11.so
[03:43] <daniels> ldconfig not called in postinst
[03:43] <stratus> jdub, oh my failure.
[03:43] <daniels> would ldconfig need to be called there, to register libX11.so?  or is there something I'm missing?
[03:44] <elmo> daniels: you're breaking my mind.  you mean ld.so right?
[03:44] <elmo> ld(1) is very different
[03:44] <daniels> elmo: never mind me
[03:44] <daniels> elmo: i blame keybuk
[03:46] <Kamion> pitti: the first two hunks can be dropped
[03:47] <pitti> Kamion: indeed, these work fine
[03:47] <pitti> Kamion: however, the Debian udeb does not ship many files the Ubuntu one does
[03:48] <Kamion> pitti: yeah, I know, I'm looking at those
[03:48] <fabbione> elmo: dpkg-deb: building package `linux-image-2.6.8.1-3-sparc64-smp' in `../linux-image-2.6.8.1-3-sparc64-smp_2.6.8.1-17.1_sparc.deb'.
[03:48] <fabbione> that's all for you :P
[03:48] <pitti> Kamion: I don't think we need all of the remaining files, that's why I did not just take the Ubuntu version
[03:49] <thom> fabbione: my sparc wouldn't boot 2.6 last i tried :/
[03:50] <fabbione> thom: we will see later 2.4 support
[03:50] <fabbione> thom: right now i had to drop sparc32 too
[03:50] <thom> ahr
[03:50] <fabbione> because i really don't have one to test on
[03:50] <fabbione> and kernel needs love on sparc32
[03:50] <fabbione> sparc64 gets enough
[03:51] <fabbione> thom: don't you have an u30?
[03:51] <Kamion> pitti: indeed
[03:52] <thom> u10
[03:52] <Kamion> pitti: could you add links.conf to /etc/udev? I'm not absolutely sure about it but I'm not comfortable with leaving it out yet
[03:52] <fabbione> thom: did you try debian kernels or custom kernels?
[03:53] <jdub> tseng: at the bottom of your dav_user.conf, put:
[03:53] <pitti> Kamion: sure
[03:53] <jdub> <IfModule prefork.c>
[03:53] <jdub> MinSpareServers 1
[03:53] <jdub> MaxSpareServers 1
[03:53] <jdub> MaxClients 3

[03:53] <jdub> 
[03:53] <jdub> :-)
[03:54] <Kamion> pitti: can you drop that udev.startup thing from debian/rules too? I'll talk to Md about that; rootskel must be able to control d-i's startup, and /lib/debian-installer-startup.d is not a published interface for random packages to use
[03:54] <tseng> ok.
[03:54] <jdub> now it should build too
[03:54] <Kamion> pitti: (at least, not yet)
[03:55] <jdub> /bin/dd if /proc/kmsg of /var/run/klogd/kmsg <- fear ;)
[03:56] <thom> fabbione: debian
[03:56] <Kamion> pitti: other than that I *think* it's ok
[03:56] <pitti> jdub: I know, it's a hack, but it works :-)
[03:56] <Kamion> pitti: actually, hmm
[03:57] <Kamion> pitti: yes, never mind me, should be fine, might have to change rootskel slightly
[04:00] <pitti> Kamion: so the important things are only links.conf and the dropping of udev.startup?
[04:02] <Kamion> pitti: in fact, I misread the diff, links.conf is still there
[04:02] <jdub> lamont: please kick gus :)
[04:02] <smurfix> jdub: Hmmm, so what do I (or anyone else for that matter) need to get Nautilus to show the public folders? I'll freely admit that I've not played with zeroconf stuff before.
[04:02] <Kamion> pitti: so just drop /lib/debian-installer-startup.d/S02udev, please, rest should be fine
[04:02] <pitti> Kamion: okay thanks
[04:03] <jdub> smurfix: nautilus uses gnome-vfs's howl stuff
[04:03] <pitti> Kamion: pure curiosity, has d-i a something similar like /etc/init.d/?
[04:03] <jdub> smurfix: so it should just appear in Computer > Network
[04:03] <pitti> Kamion: this symlink looks like sysvinit-style
[04:03] <jdub> smurfix: you've run gnome-file-share-properties, and verified that gnome-user-share (and apache) are running?
[04:03] <Kamion> pitti: kinda sorta
[04:03] <elmo> I wish there was a way to do opportunistic gdb-ing or strace-ing, i.e. say "next time something called 'expect' starts, attach to it"
[04:04] <jdub> elmo: obviously you want dtrace.
[04:04] <Kamion> pitti: in fact, Md's startup script will break d-i because he didn't know how it worked ... whoopsie
[04:04] <pitti> Kamion: so this _is_ relevant for Debian, too
[04:04] <pitti> Kamion: well, the Debian version currently disables it anyway
[04:04] <jdub> smurfix: not working?
[04:04] <Kamion> pitti: ... actually maybe not, he installs it executable so it'll get executed not sourced
[04:05] <smurfix> jdub: not so fast :-/
[04:05] <Kamion> pitti: I'm still uncomfortable with /lib/debian-installer-startup.d not being entirely controlled by rootskel
[04:05] <Kamion> pitti: doesn't matter for Debian yet because Debian won't be using udev for a little while yet
[04:05] <smurfix> I'm typing with one hand
[04:05] <jdub> TMI
[04:05] <pitti> Kamion: if I understood that correctly, witht the current version udev would be started twice?
[04:06] <pitti> Kamion: once by rootskel, another time by this S02udev?
[04:06] <smurfix> jdub: it's temporary ...
[04:07] <eruin> what's been happening to gdesklets recently? it's more broken than what I'd be after a 90kph car-to-brick-wall-crash
[04:08] <Kamion> pitti: no ... rootskel already ships S02udev, so they'd clash and have to be resolved somehow anyway
[04:08] <pitti> eek
[04:09] <pitti> Kamion: okay, thanks for your help!
[04:09] <smurfix> jdub: Not seeing it on the desktop (which is the one publishing), or from the MacOS-X laptop. My ubuntu laptop's busy updating to Hoary at the moment.
[04:09] <Kamion> pitti: as it happens udev's would win because the initrd udebs get unpacked in alphabetical order
[04:09] <jdub> smurfix: install howl-utils to do some testing for me :)
[04:10] <pitti> Kamion: "r"ootskel > "u"dev ???
[04:10] <pitti> not in my alphabet...
[04:10] <pitti> Kamion: anyway, I removed it
[04:10] <smurfix> jdub: done
[04:11] <jdub> smurfix: type this:
[04:11] <jdub> mDNSBrowse _webdav._tcp
[04:12] <Kamion> pitti: udev's would win; it gets unpacked second and therefore overwrites
[04:13] <smurfix> jdub: sitting there, not doing much
[04:13] <pitti> Kamion: ah, I see. I thought it would be similar to dpkg where it just refuses to overwrite something
[04:13] <jdub> smurfix: hrm
[04:13] <smurfix> (mDNSBrowse, not me)
[04:13] <jdub> smurfix: and mdnsresponder is running?
[04:14] <Kamion> pitti: we use 'dpkg -x'
[04:14] <smurfix> jdub: 20205 ?        Ssl    0:00 /usr/sbin/mDNSResponder -f /etc/mdnsresponder/mDNSResponder.conf
[04:14] <smurfix> strace shows that it did respond to the browser's first packet
[04:15] <Kamion> pitti: actually, no, we don't, we use dpkg --unpack, but we also use --force-overwrite
[04:15] <pitti> ah
[04:15] <pitti> Kamion: why overwrite? Did this already occur on several other udebs?
[04:17] <Kamion> pitti: yes
[04:17] <Kamion> pitti: busybox-cvs-udeb and module-init-tools-udeb for a start
[04:17] <pitti> ugh, sounds like gambling
[04:18] <Kamion> sure, but there are some battles not worth fighting
[04:18] <pitti> I removed the script and the directory, I upload now. Let it break... :-)
[04:18] <Kamion> ok, I'll upload rootskel to match
[04:18] <Kamion> thanks
[04:18] <pitti> thanks to you too
[04:19] <pitti> Kamion: this version fixes several important errors, that's why I'm eager to get it :-)
[04:22] <smurfix> jdub: I just restarted the mdnsresponder with -d; mdnsbrowse now reports ...
[04:23] <smurfix> jdub: browse reply: Add Service 0x2 smurf's public files _webdav._tcp. local.
[04:23] <Kamion> Keybuk: are targets like 'install-libLTLIBRARIES' considered internal? i.e. can I use them from debian/rules?
[04:23] <smurfix> jdub: resolve reply: 0x2 smurf's public files _webdav._tcp. local. 192.109.102.35 53848
[04:23] <smurfix> jdub: ... but still doesn't return.
[04:24] <smurfix> ah, restarting Nautilus worked.
[04:24] <smurfix> jdub: that shouldn't be necessary.
[04:25] <thom> jdub: worth thinking about adding a "Branding" keyword to bugzilla?
[04:26] <smurfix> jdub: now I'd please like to rename that shared folder.  ;-)
[04:26] <jdub> "no." :)
[04:27] <smurfix> jdub: "Duh, I'll keep using MacOS then, it works there."  :-)
[04:27] <thom> jdub: um, to which :P
[04:28] <jdub> thom: not you - on phone
[04:28] <thom> ahr
[04:32] <smurfix> jdub: Interestingly the webdav share can be seen but not mounted from OSX. Not sure yet why.
[04:32] <jdub> knownish bug :)
[04:34] <smurfix> jdub: I take that to mean that you don't yet know why ..?
[04:35] <Keybuk> Kamion: very internal
[04:36] <Kamion> Keybuk: guess I just have to install and then remove everything but the library, then ...
[04:39] <RubenV> patched laptop-mode for LSB
[04:39] <RubenV> patch is in bugzilla
[04:40] <RubenV> my boot is now almost clean :)
[04:40] <thom> can you attach the patch to #1580
[04:40] <thom> since that's the canonical bug
[04:41] <RubenV> ah, great :)
[04:41] <RubenV> didn't knew that one
[04:41] <RubenV> there's also an alsa one from me somewhere
[04:41] <RubenV> i'll put it in 1580 too
[04:42] <thom> great
[04:43] <RubenV> voila
[04:50] <jdub> ok, back
[04:51] <jdub> thom: branding would be handy for me, yeah
[04:51] <jdub> thom: or you can just reassign viciously :)
[04:52] <jdub> smurfix: no, not known why
[04:53] <jdub> http://people.ubuntulinux.org/~lamont/buildLogs/g/gnome-user-share/0.3-3/
[04:53] <jdub> hoo-ray
[05:03] <smurfix> jdub: Apparently the stupid thing produces neither network traffic nor log file entries when it does that. :-(
[05:11] <stratus> humm news about gnome-user-share?
[05:12] <smurfix> stratus: it worked here after I restarted nautilus.
[05:12] <stratus> smurfix, it worked here before my lunch but i was asking about others mpm.
[05:13] <jdub> stratus: the updated packge should be in hoary by now
[05:13] <stratus> jdub, thanks i was lunching sorry for the noise.
[05:15] <smurfix> stratus: then you should ask more specific questions. ;-)
[05:16] <jdub> smurfix: which 'stupid thing'?
[05:22] <smurfix> jdub: The MacOS laptop. "It doesn't work" error messages are sub-standard.
[05:28] <smurfix> It seems that Apple does its own zeroconf file sharing with AFP, not WebDAV. I don't feel any particular urge sto start hacking netatalk.
[05:28] <zul> no?
[05:28] <zul> :)
[05:28] <smurfix> zul: The day has only so many weeks. :-/
[05:29] <zul> heh...i wouldnt either :)
[05:34] <smurfix> Is there a way to mount a WebDAV file system?
[05:34] <jdub> with hacky stuff like FUSE
[05:35] <smurfix> Well, that does seem to be on its way into the official kernel.
[05:37] <smurfix> It's just differently hacky than something that exists only in gnome.' 
[05:38] <daniels> not really
[05:38] <daniels> gnome has its own vfs stuff
[05:38] <daniels> fuse is kinda handwavey there-is-no-standard-file-interface-because-I-just-replaced-it
[05:38] <daniels> these are entirely separate ideas
[05:40] <smurfix> daniels: I know they're separate ideas, that's kind of my point.
[05:44] <smurfix> File system contents should be visible to all programs, not just those which happen to have a gnome GUI.
[05:44] <daniels> mmmmm
[05:44] <daniels> try this:
[05:44] <jdub> s/have a gnome GUI/use gnome-vfs/
[05:44] <smurfix> right
[05:44] <daniels> write a WebDAV server which sends 'Connection: keep-alive' unconditionally (because you wrote a buggy test)
[05:44] <daniels> sorry, Connection: close, unconditionally
[05:45] <daniels> then have it assume it's doing keep-alive, and keep the connection open forever
[05:45] <daniels> mount any volume from this WebDAV server on a Mac OS X machine
[05:45] <daniels> double-click on the volume, and watch Finder lock up solid
[05:45] <daniels> drag it over to the Dock to eject it, and watch the Dock lock up solid
[05:45] <daniels> watch anything that attempts to do file IO lock up solid
[05:46] <daniels> swear and reset your machine
[05:46] <daniels> i assure you, there are some *serious* issues with moving stuff like WebDAV down into the kernel
[05:47] <smurfix> daniels: FUSE isn't exactly doing WebDAV in the kernel.
[05:47] <daniels> understood, but the problem space isn't quite as simple as you make it out to be
[05:47] <daniels> i know, because I spent three or so days in a mount/reset cycle on the OS X machine
[05:48] <daniels> (afaik, the problem still isn't fixed)
[05:48] <smurfix> never said it was, it's just a *different* problem space, one I personally prefer.  ;-)
[05:53] <smurfix> Anyway, Apple certainly could do a lot better in that area. They still can't keep NFS mounts over wireless networks stable. If the IP address doesn't change, that's not supposed to be difficult to do.
[06:20] <elmo> seb128: ping WRT libzvt?
[06:26] <seb128> elmo: oh yes, I've forgotten this one. We don't need it in main
[06:27] <elmo> seb128: thanks
[06:27] <seb128> np
[06:59] <chrisa> jesus
[06:59] <chrisa> I was so confused when the postwoman handed me a box of packages
[06:59] <chrisa> Turns out they're all Ubuntu cds
[07:00] <bob2> haha
[07:00] <daniels> ordering lots of CDs tends to have that effect, yeah
[07:00] <Matt|> *laughs*
[07:01] <smurfix> chrisa: so how many did you order?
[07:02] <chrisa> smurfix: 200 for the oss group / lug
[07:02] <chrisa> and to also throw in random places on campus
[07:02] <Matt|> where are the cd's shipping from?
[07:02] <smurfix> cool
[07:03] <chrisa> No return address on the envelopes, no idea
[07:03] <Kamion> I use "request"
[07:03] <smurfix> Kamion: thanks, that's better
[07:04] <fabbione> Kamion: good that we did the ppc and that i integrated sparc :-)
[07:04] <fabbione> Kamion: i spotted 2 errors in one line :-)
[07:04] <chrisa> Neat, these have a live cd and install cd, as well as directions on them
[07:06] <Kamion> fabbione: heh
[07:06] <fabbione> Kamion: -18 will have sparc + udebs for i386/ppc/amd64/sparc
[07:06] <fabbione> Kamion: rocking :-)
[07:06] <fabbione> lamont: do you have anything from ia64? (kernel wise)
[07:07] <lamont> fabbione: not personally, no.
[07:08] <fabbione> lamont: what about t-bone?
[07:08] <fabbione>         install -D -m 644 debian/d-i-powerpc/boot/vmlinux-2.6.8.1-3-powerpc debian/kernel-image-2.6.8.1-3-powerpc-di/boot/vmlinux
[07:08] <fabbione>         install -D -m 644 debian/d-i-powerpc/boot/vmlinux-2.6.8.1-3-power3 debian/kernel-image-2.6.8.1-3-power3-di/boot/vmlinux
[07:08] <fabbione>         install -D -m 644 debian/d-i-powerpc/boot/vmlinux-2.6.8.1-3-power4 debian/kernel-image-2.6.8.1-3-power4-di/boot/vmlinux
[07:09] <fabbione> Kamion: i guess we are rocking :-)
[07:09] <lamont> fabbione: t-bone would be the one to ask
[07:09] <lamont> and/or dannf, but that's probably a monday thing
[07:09] <fabbione> lamont: ok.. i guess ia64 can wait :-)
[07:09] <fabbione> yeah
[07:10] <fabbione> Kamion: if you want to login on davis and take a look in my home...
[07:11] <fabbione> Kamion: you are more used than me watching ppc udebs (at least in terms of numbers/names)
[07:12] <fabbione> later
[07:13] <Kamion> fabbione: don't think I have an account
[07:13] <fabbione> Kamion: afaik everybody does
[07:13] <fabbione> on porting machines
[07:13] <fabbione> and daves is one of them
[07:13] <fabbione> anyway.. food
[07:14] <Kamion> my ssh key doesn't seem to work, and the password I think I remember doesn't work either
[07:14] <Kamion> elmo: ?
[07:14] <daniels> er, IIRC it's a specific thing
[07:14] <daniels> i didn't have access to the port boxes until I asked
[07:15] <daniels> Kamion: if you want me to pull specific things out, let me know
[07:15] <elmo> yes, port stuff is on request
[07:15] <daniels> or, even better, just ask elmo
[07:16] <elmo> Kamion: do you need an account or just the files?
[07:18] <zopi> hi
[07:18] <zopi> where I can find kernel-headers for unbuntu ?
[07:19] <tseng> apt-cache search kernel-headers
[07:19] <zopi> there is linux-kernel-headers but nothing in /usr/src
[07:19] <zopi> already done
[07:19] <zopi> there is none matching the kernel
[07:19] <zopi> 2.6.8.1-3 
[07:19] <zopi> even using universe repository
[07:19] <Matt|> zopi, are you talking about linux-headers?
[07:20] <thom> #ubuntu for support questions, please. (you want linux-headers)
[07:20] <zopi> no
[07:20] <lamont>  /bin/sh: ps: command not found
[07:20] <lamont> hehe
[07:20] <zopi> kernel-headers
[07:20] <Matt|> zopi, #ubuntu
[07:20] <zopi> ok
[07:20] <zopi> sorry so
[07:21] <zopi> Does it plan to add rescue mode on CD for the next release ? 
[07:22] <lamont> zopi: #ubuntu is the right place for that.  this is the right place to discuss your patch that implements it
[07:22] <Matt|> *laughs*
[07:22] <zopi> ok :)
[07:22] <zopi> lol
[07:22] <Kamion> elmo: just the files is fine
[07:23] <Kamion> zopi: rescue mode> maybe; I'd like it to exist but there is absolutely no code for that available for our installer as far as I know, so it'd be an interesting development project for somebody
[07:24] <zopi> just to boot
[07:24] <zopi> just like this rescue root=/dev/hd**
[07:24] <Kamion> yes, I know what you're referring to
[07:24] <zopi> ok :)
[07:25] <zopi> Kamion : and the code from Debian Woody ?
[07:25] <elmo> Kamion: rookery:~james/davis
[07:25] <Kamion> the trick is managing to make that work without requiring a big monolithic kernel
[07:25] <lamont> Kamion: I thought that's what alt-f2 was for... :-)
[07:25] <Kamion> zopi: code from woody is inapplicable because we totally rewrote the installer for sarge (and hence warty)
[07:25] <Kamion> elmo: thanks
[07:25] <zopi> Kamion : ah ok
[07:26] <zul> why?
[07:26] <Kamion> zul: our own packaging policies are likely to be pretty similar
[07:27] <Kamion> fabbione: I haven't looked through all the udebs, but those look OK to me
[07:27] <zul> i c..so why not rename it to ubuntu-policy?
[07:27] <lamont> Kamion: in fact, if ubuntu-policy does differ from debian-policy, it'd be nice if it was just a list of changes from debian policy, since it shouldn't vary _that_ much
[07:27] <Kamion> fabbione: with the exception of Section: devel => debian-installer, but I think you were using the previous version of kernel-wedge?
[07:28] <Kamion> zul: varies; we haven't renamed debian-installer for instance (and I'd be against doing so - the upstream name is "debian-installer", renaming it arrogates too much credit to ourselves)
[07:29] <seb128> daniels: here ?
[07:29] <Kamion> zul: ubuntu-policy might be a different case, I suppose; it might depend on how much we changed or whether we just took d-p verbatim
[07:30] <zul> Kamion, truebut at what point would that be
[07:30] <Kamion> not sure
[07:31] <zul> you might need 2 debs...not sure either
[07:31] <lamont> but speaking of policy... what's the official way to add a service to /etc/services?
[07:32] <lamont> because this sure isn't it (in Makefile...):
[07:32] <lamont>                 @(grep isdnlog $(SERVICEFILE) >/dev/null) || \
[07:32] <lamont>                 (echo "";echo "";echo "Add a line to the file $(SERVICEFILE)" ;echo "";echo ""; \
[07:32] <lamont>                 echo "isdnlog $(SERV_PORT)/tcp        isdnlog" >> $(SERVICEFILE))
[07:32] <daniels> seb128: sup
[07:33] <Kamion> lamont: you get the netbase maintainer to add it
[07:33] <Kamion> lamont: that code's so wrong it's not true
[07:33] <Kamion> lamont: isdnlog's already in netbase's copy of /etc/services though
[07:35] <Kamion> http://people.ubuntulinux.org/~lamont/buildLogs/d/debian-installer/20041118ubuntu4/debian-installer_20041118ubuntu4_20041126-1806-i386-failed - crap
[07:35] <seb128> daniels: do you if this is supposed to work ? "echo "Xcursor.size: 20" | xrdb -merge -"
[07:35] <seb128> +know
[07:36] <daniels> seb128: should do, but I don't know if you have to do specific known sizes like 32/64
[07:36] <daniels> seb128: also, you might need to do it before you start everything -- dunno that you can change it on the fly, but worth a shot
[07:36] <seb128> daniels: ok, because GNOME do this (with known size) to change between small/medium/big cursors and that doesn't work here
[07:36] <lamont> Kamion: doesn't seem to be in the hoary chroot's services, thouhg
[07:37] <lamont> Kamion: that'd be because /etc/services isn't there
[07:37] <Kamion> lamont: ah, no netbase
[07:37] <lamont> yeah, adding that too.
[07:37] <Kamion> I guess b-ding on netbase would be wrong? :)
[07:38] <lamont> don't see why not.
[07:40] <Kamion> lamont: netbase's postinst runs /etc/init.d/networking though ...?
[07:41] <lamont> ah, yes. that might be bad...
[07:41] <lamont> install phase finished...
[07:42] <lamont> Kamion: so how do I get those perl bitches about locale to go away without installing locales in the chroot, and leaving the locale in place outside the chroot???
[07:43] <lamont> is it as simple as stuffing LANG=C somewhere?
[07:45] <elmo> lamont: empty /etc/environment
[07:46] <seb128> daniels: ok, changing the cursor theme works but the size (12/24/36) doesn't
[07:46] <Kamion> or LANG=C before sbuild
[07:46] <elmo> the machines don't have locales installed - /etc/environment should be updated to match
[07:47] <Kamion> true
[07:47] <Kamion> keep forgetting you mean /etc/environment outside the chroots
[07:47] <lamont> elmo: doesn't work for me.
[07:47] <Kamion> lamont: you'd have to restart the buildd, it'll already have LANG=whatever surely?
[07:47] <lamont> or you mean outside the chroot?
[07:47] <elmo> lamont: on which machine?
[07:47] <lamont> my machine
[07:47] <elmo> yes, in base
[07:48] <elmo> and restart the buildd
[07:48] <lamont> doesn't that break me using a locale?
[07:48] <Kamion> lamont: on your own machine, just LC_ALL=C before building
[07:48] <elmo> oh, well, if it's not your machines, go with what kamion said
[07:48] <elmo> s/not//
[07:48] <Kamion> (LC_ALL in case you've set LC_FOO by hand)
[07:49] <lamont> elmo: any objection to sbuid just forcing LANG=C into the environment?
[07:49] <lamont> LANG= seemed to work...
[07:49] <Kamion> if you're going to use something as an override it should be LC_ALL=C
[07:50] <elmo> lamont: no, I suppose not
[07:50] <lamont> I'm more inclined to teach buildd to inject something into the environment from buildd.conf
[07:51] <Kamion> I. HATE. MKLIBS.
[07:51] <lamont> Kamion: LC_ALL doesn't change LANG... is that gonna bite me?
[07:52] <Kamion> no
[07:52] <Kamion> type 'LC_ALL=C locale'
[07:52] <Kamion> oh, true, that shows LANG=en_GB.UTF-8 for me; but nothing checks that
[07:53] <lamont> ah, ok
[07:53] <lamont> given %ENV and %ENVOVERRIDES, how do I get all of %ENVOVERRIDES into %ENV?
[07:53] <Kamion> lamont: perl (and pretty much everything else that cares about this kind of stuff) does setlocale(LC_ALL, "")
[07:54] <lamont> perl: warning: Falling back to the standard locale ("C").
[07:54] <Kamion> lamont: $ENVOVERRIDES{$_} = $ENV{$_} for keys %ENV;
[07:54] <lamont> actually, looks like 'C'
[07:54] <Kamion> #ifdef LC_ALL
[07:54] <Kamion>     if (! setlocale(LC_ALL, ""))
[07:54] <Kamion>         setlocale_failure = TRUE;
[07:54] <Kamion> #endif /* LC_ALL */
[07:54] <lamont> Kamion: actually the other way around, yes?
[07:54] <Kamion> lamont: oh, yeah, right
[07:57] <Kamion> $ENV{$_} = $ENVOVERRIDES{$_} for keys %ENV;
[07:57] <Kamion> bah, keys %ENVOVERRIDES
[07:59] <lamont> yeah - that falls so trippingly off the fingers, doesn't it...
[08:01] <fabbione> seb128: you around?
[08:06] <lamont> Kamion: it doesn't like me... dammit
[08:06] <lamont> env changes aren't making it into child processes
[08:07] <Kamion> %ENV is definitely exported (in shell terminology) by perl
[08:07] <lamont> yeah, well...
[08:07] <Kamion> I suspect a child process is resetting it
[08:07] <lamont> to what it was _outside_ the chroot?
[08:07] <Kamion> check exactly what %ENV looks like, too
[08:07] <Kamion> perl -d rocks
[08:08] <Kamion> is there an /etc/environment inside the chroot?
[08:08] <lamont> emtpy
[08:08] <lamont> but present
[08:09] <mdz> morning
[08:09] <Kamion> hi mdz
[08:09] <fabbione> hey mdz
[08:09] <zul> hey mdz
[08:10] <lamont> Kamion: that assuems that one knows what to do wiht perl -d.....
[08:10] <lamont> morning mdz
[08:11] <Kamion> lamont: type 'n' or 's' gdb-style at it until one gets to the line one's interested in, then 'x %ENV'
[08:13] <mdz> anyone seen amu lately?
[08:13] <Kamion> yeah, he was around on #canonical just an hour or two ago
[08:15] <zul> is anyone working on selinux support?
[08:16] <Kamion> not among Canonical folks, dunno about anyone else
[08:16] <zul> i wouldnt mind taking a shot at it though
[08:18] <fabbione> mdz: http://people.ubuntulinux.org/~fabbione/changelog
[08:18] <fabbione> mdz: doing the last test now
[08:19] <fabbione> mdz: but that's basically what i would like to go for -18
[08:28] <mdz> zul: consensus is that rather than disrupt mainline development, it should be done in a derivative branch, and then merged in when it's reasonably stable
[08:29] <Kamion> lamont: please make debian-installer dep-wait on pciutils-udeb (>= 2.1.11-15ubuntu3)
[08:29] <Kamion> lamont: that is, assuming you can dep-wait on udebs (can you?); if not, the corresponding source
[08:29] <lamont> which source?
[08:29] <Kamion> pciutils
[08:29] <zul> mdz: sure no problem..once i have something i can put patches into the bugzilla cant i?
[08:30] <lamont> Kamion: d-i is PaS'ed on ia64?  or is the ia64 box just slow I wonder... Anyway, the 3 architectures you care about are d-w./
[08:31] <zul> mdz: nm...im not thinking
[08:31] <mdz> zul: certainly.  also, we'll be doing a lot of work to make it easier to work on derivatives, which we'll talk about in Spain
[08:31] <Kamion> lamont: ta
[08:31] <Kamion> lamont: won't work yet on ia64 anyway ...
[08:31] <Kamion> lamont: (lack of kernels)
[08:31] <lamont> right
[08:31] <zul> mdz: heh...i wish i could go to spain :)
[08:31] <lamont> bbl
[08:42] <fabbione> hmmmm
[08:42] <fabbione> pool/universe/x/xcompmgr/xcompmgr_1.1.1+cvs.20041109-0ubuntu2_i386.deb
[08:47] <fabbione> mdz: do you know why it has been moved to universe?
[08:47] <fabbione> or anybody?
[08:47] <mdz> fabbione: moved? was it in main previously?
[08:47] <fabbione> yes
[08:47] <mdz> the two things which control that are seeds and dependencies
[08:48] <mdz> if it's not in a seed, and isn't depended on by anything in main, it'll go into universe
[08:48] <fabbione> they are xclients.. so nothing depends on it
[08:48] <fabbione> (together with fdclock and transset
[08:48] <fabbione> )
[08:48] <fabbione> ok
[08:48] <fabbione> we might want to add them to a seed than
[08:48] <fabbione> but i will give daniels a remider 
[08:49] <fabbione> brb
[08:50] <Kamion> mdz: which reminds me; did you see my question about pciutils-udeb and usbutils-udeb?
[08:51] <elmo> fabbione: daniels and jdub discussed it yesterday
[08:51] <Kamion> 2004-11-25 11:53:03 GMT Daniel Stone <daniel.stone@canonical.com>       patch-31
[08:51] <Kamion>     Summary:
[08:51] <Kamion>       fdclock/xcompmgr/transset -> universe, per jdub
[08:52] <Kamion> mdz: I've created all the udebs now, just need to seed them and add the necessary support to hw-detect
[08:53] <mdz> Kamion: no, I don't think I did
[08:53] <mdz> my cablemodem shat itself for about 12 hours yesterday while i was out, so I didn't have scrollback
[08:53] <Kamion> 20:48 < Kamion> mdz: oh, do you mind if I add pciutils-udeb, usbutils-udeb, and libusb-0.1-udeb? I need pci.ids and usb.ids now that discover1-data is no longer around to provide the names of network
[08:53] <Kamion>                 interfaces for netcfg's UI
[08:53] <Kamion> 20:48 < Kamion> mdz: which I had totally not thought about in advance
[08:53] <Kamion> 20:49 < Kamion> mdz: and getting lspci and lsusb will probably be a bonus for debugging purposes, too
[08:53] <Kamion> ah
[08:54] <mdz> sounds fine
[08:54] <mdz> in general, it's fine with me if you add things to the installer seed at your own discretion
[08:55] <Kamion> ok, thanks
[08:55] <mdz> it doesn't have the same associated issues as adding things to base, desktop or ship which require that those be discussed first
[08:55] <Kamion> haven't quite decided yet whether I want to include usbutils-udeb in any of the initrds
[08:56] <Kamion> I suspect not; it's a fair size and we degrade relatively gracefully
[08:56] <seb128> fabbione: pong
[08:56] <Kamion> ooh, I have a cunning plan: I can use Enhances for a technical purpose!
[08:57] <Kamion> anna is one of the very few tools anywhere in Debian that actually pays attention to and uses Enhances
[08:57] <fabbione> seb128: control-center is a FTBFS due to type-handling :(
[08:57] <seb128> fabbione: what's wrong with type-handling ? it works on my box
[08:58] <fabbione> elmo, Kamion: DOH!
[08:58] <fabbione> seb128: it's from universe
[08:58] <seb128> depends from a main should go automatically in main, right ?
[08:58] <seb128> +package
[08:59] <fabbione> elmo, Kamion: do daniels and jdub realize that we did all this X.org sprint to get the new extensions in, like composite, ad we are taking away the only ONE available manager?
[08:59] <fabbione> seb128: yes, but we don't want type-handling in main :-)
[08:59] <Kamion> fabbione: no idea
[08:59] <seb128> fabbione: arg, any reason for that ?
[08:59] <seb128> what's the problem with it ?
[08:59] <elmo> seb128: not for type-handling - we've been trying to avoid bring that in, and have already removed it from several packages' build-depends
[09:00] <fabbione> seb128: i remember that we were told to remove it from everywhere
[09:00] <Kamion> seb128: ECRACK basically
[09:00] <seb128> ok
[09:00] <seb128> it has been added in some GNOME package because some BSD guys submitted patches IIRC
[09:01] <fabbione> yeah! another 2 and a half kernels to go
[09:01] <seb128> should I slay it on the debian side too ? :)
[09:01] <Kamion> seb128: I'll take a guess that that was the author of type-handling
[09:01] <fabbione> guys.. my apologizes if i was bitching on X compilation time
[09:01] <fabbione> i really really feel sorry for that
[09:01] <fabbione> the kernel is MUCH worst
[09:02] <fabbione> specially when you compile it in 200 flavours
[09:02] <fabbione> including the choccolate cake one with coffe.ko
[09:02] <seb128> Kamion: yep, just checked (ie: #272722)
[09:13] <Kamion> night all
[09:14] <fabbione> cya Kamion 
[10:08] <mdz> seb128: type-handling is fairly crackful
[10:09] <jdub> mdz: tried gnome-user-share?
[10:09] <mdz> nope
[10:09] <jdub> mdz: it will most likely be in gnome 2.9, and depends on apache2 and mdnsresponder. :-)
[10:09] <mdz> "easy user-level file sharing" is defiintely a gap that needs filling
[10:10] <mdz> it is not, however, something that warrants open ports by default :-)
[10:10] <seb128> mdz: ok, I've not played with it so I didn't know :)
[10:10] <jdub> :-)
[10:10] <Mithrandir> mdz: sounds crackful.
[10:10] <jdub> by default, it's not on
[10:10] <mdz> jdub: no problem, then?
[10:10] <jdub> but there's a black cloud hanging over mdnsresponder
[10:10] <jdub> which will be very un-zeroconf if it doesn't listen by default
[10:10] <mdz> mdnsresponder just needs some smarts so that it is only enabled if the user has asked for it
[10:11] <mdz> this is something where I think we are justified in being bull-headed
[10:11] <jdub> yeah
[10:11] <mdz> I need to read up on how it works so that I can understand the security model
[10:12] <mdz> but I'm ostensibly off today and will be heading out soon
[10:12] <mdz> if you have some pointers to resources on that subject, please send them to me
[10:12] <jdub> it's the kind of thing that could be run by the library when required, but unfortunately, it listens on :5353 udp
[10:12] <jdub> and there must only be one daemon because of that
[10:12] <seb128> jdub: GNOME used to have a preview of the mouse cursors, right ? 
[10:13] <jdub> seb128: not a proper one, no
[10:13] <jdub> seb128: it had a "small" and "large" picture, but it wasn't of the actual cursor
[10:13] <seb128> jdub: I swear I can remember a screen with cursor pictures
[10:13] <seb128> oh, perhaps that's it so
[10:13] <jdub> yeah, it was in the old mouse properties dialogue
[10:14] <jdub> we haven't had anything that handles cursor themes though
[10:14] <seb128> how old is the dialog ? I don't find the change in the ChangeLog
[10:14] <jdub> hrm, that probably disappeared in 2.4
[10:15] <seb128> ok
[10:16] <seb128> I'm doing some bug triage in the control-center's bugzilla, not always easy to understand the bug without the old capplet to compare :)
[10:17] <jdub> oh yeah, it's turkey day
[10:21] <fabbione> jdub: want to take a look at http://people.ubuntulinux.org/~fabbione/changelog before i upload?
[10:23] <jdub> fabbione: cool
[10:23] <jdub> exciting :)
[10:23] <Mithrandir> jdub: have you done any fun stuff on libuser?
[10:23] <jdub> Mithrandir: haven't had time :|
[10:24] <jdub> Mithrandir: perhaps we should beg lifeless to import it asap :)
[10:25] <jdub> fabbione: so i guess i should bring on the U5 :-)
[10:25] <Mithrandir> jdub: I guess so, so we havething to work off.. I hope I
[10:25] <Mithrandir> 'm able to do something next week, but exam on Monday, so..
[10:26] <fabbione> jdub: if it can boot 2.6
[10:26] <fabbione> thom claims his u10 can't
[10:27] <jdub> duck
[10:27] <jdub> er
[10:27] <jdub> suck
[10:32] <fabbione> accepted
[10:36] <jdub> ah
[10:36] <jdub> solaris 9
[10:37] <sjoerd> an U10 is just an U5 in another box right ?
[10:37] <jdub> i think so
[10:38] <fabbione> get ready for -19
[10:38] <fabbione> there was an error in debian/control
[10:39] <jdub> U5 is surprisingly noisy
[10:39] <sjoerd> the front fan is noisy 
[10:39] <sjoerd> my U5 runs 2.6.9 just fine so..
[10:40] <jdub> fabbione: so you reckon a sarge install and piecemeal upgrades to your packages is the best bet?
[10:40] <fabbione> no
[10:40] <fabbione> there is a much better way
[10:41] <fabbione> install sarge in the future swap
[10:41] <jdub> aha
[10:41] <jdub> and debootstrap into the / ?
[10:41] <fabbione> install debootstrap from ubuntu and patch it with debootstrap.diff from the website
[10:41] <fabbione> and than debootstrap it
[10:41] <jdub> nice
[10:41] <fabbione> you need to find a way to make the _all.deb available to your machine
[10:41] <fabbione> my archive is only _sparc.deb
[10:43] <fabbione> ok.. _19 is going up
[10:48] <lamont> albeit as normal, not serious
[10:49] <chrisa> hrm, I like the idea of putting windows versions of apps on the ubuntu live cd
[10:50] <fabbione> yeah.. but you can't symlink
[10:50] <lamont> jdub around?
[10:50] <fabbione> argh
[10:54] <lamont> seb128: you around?
[10:59] <fabbione> lamont: mind to babysit the kernel build?
[11:02] <seb128> lamont: yes
[11:12] <lamont> seb128: libglademm2.0 is yours, yes?
[11:12] <lamont> fabbione: bouncing around, but I can check on it, yets.
[11:12] <lamont> yes.
[11:13] <fabbione> lamont: it would be enough for me to know if it started the real build
[11:13] <fabbione> (compiling)
[11:14] <lamont> heh
[11:14] <lamont> any particular architecture?
[11:15] <fabbione> all of them? :)
[11:15] <lamont> i386:  CC [M]   fs/reiserfs/tail_conversion.o
[11:15] <lamont>   CC [M]   fs/reiserfs/journal.o
[11:15] <lamont>   CC [M]   fs/reiserfs/resize.o
[11:16] <lamont> ditto amd64 and ppc.
[11:16] <lamont> ia64 I just marked failed.
[11:16] <seb128> lamont: not really but I can work on it
[11:16] <fabbione> cool
[11:16] <seb128> lamont: I don't maintain the *mm
[11:16] <fabbione> lamont: well.. ia64 isn't supported yet
[11:17] <lamont> fabbione: true 'nuf
[11:17] <lamont> seb128: ah, ok
[11:18] <lamont> fwiw - i didn't assign you the bug.  if it's assigned, bz thinks you own it...
[11:18] <seb128> lamont: what's needed ? I can work on libglademm2.0
[11:18] <lamont> amd64? ftbfs
[11:18] <seb128> ok
[11:18] <lamont> 4144
[11:18] <lamont> and bz didn't give it to you.  istr the "please use 2.4" was assigned to you though...
[11:19] <seb128> "please use 2.4" ?
[11:21] <seb128> BTW I'm not official maintainer for gnome-mm stuff but all the "*gnome*" stuff seems to get assigned to me :)
[11:21] <lamont> seb128: your lucky then. :-)
[11:21] <seb128> sort of :p
[11:21] <lamont> 3621 is the one assigned to you "Please update to ver 2.4"
[11:23] <lamont> in the lamont-needs-to-file-a-bug camp: binutils, linux86, svgalib, zsh
[11:23] <lamont> I'm fixing zsh tonight if noone beats me to it, elmo knows about binutils
[11:23] <lamont> specifically "[Something went b0rken with tcl/expect in hoary in i386] "
[11:23] <lamont> or some such, iirc
[11:23] <seb128> hum, yes, 3621 has been assigned by Matt, not bz :)
[11:24] <lamont> right
[11:24] <lamont> knew that when I saw that 4144 hadn't been autoassigned
[11:36] <lamont> $conf::should_build_msgs ||= 1;
[11:36] <lamont> if $conf::should_build_msgs is 0 before that, what does that statement do?  (trying to figure out how it becomes 1...)
[11:39] <lamont> doh
[11:46] <Mithrandir> lamont: howso?
[11:53] <BradB> lamont: It does what it looks like it does.
[11:53] <BradB> lamont: e.g. what does x += 1 do in Python? :)