[01:01] <enrico> www.gocc.gov.  Tell ${limi} that they may have a problem with the default settings for the page footer :)
[02:18] <Matt|> Is there a problem with the Nimbus Roman font? There is no bold option, and also I have got appalling performance from the pdf converter in OOo. Does anyone know anything about this?
[02:19] <Matt|> example is here: http://mdke.mine.nu/images/pdf_problems.png
[02:24] <Matt|> might that be related to #3138?
[02:26] <Matt|> if anyone has any advice on this I would really appreciate a /msg or /memo. Thanks X
[02:26] <Matt|> nite
[03:10] <sm> hi all
[03:10] <sm> I'm having trouble getting /dev/cdrom symlinks working.. have searched bugzilla
[03:10] <sm> I had to add ide-cd to /etc/modules, since I upgraded from debian
[03:11] <sm> so the hd device is there.. tried restarting udev.. but cannot get cdrom symlinks nohow
[03:11] <sm> any ideas ?
[03:22] <jdub> sm: one for #ubuntu i think
[03:29] <sm> no luck there
[03:30] <sm> ok
[07:06] <fabbione> morning guys
[07:06] <Rene_S> Close enough
[08:19] <stuNNed> evolution doesn't have d/l public key from keyserver like moz-thunderbird has
[08:20] <fabbione> eheh
[08:20] <fabbione> night lamont
[08:21] <lamont> fabbione: it does feel good to be triumphant, you know. :-)
[08:22] <fabbione> lamont: oh yeah
[08:22] <fabbione> i discovered that i am very good at painting jobs
[08:22] <fabbione> but i suck at plumbing
[08:23] <lamont> painting is much more fun
[08:23] <fabbione> i will win that too one day :-9
[08:23] <fabbione> well it depends...
[08:23] <fabbione> the preparation of the walls is a royal pain
[08:23] <lamont> houses and such.
[08:23] <fabbione> but i am good at that too
[08:23] <lamont> what sort of prep?
[08:24] <fabbione> spartling i think is the english name
[08:24] <fabbione> put up the filt
[08:24] <fabbione> refiller
[08:24] <fabbione> sandpapering
[08:24] <fabbione> all that kind of things before paiting
[08:24] <lamont> "slinging mud."
[08:24] <lamont> or "taping"
[08:24] <fabbione> eheh
[08:24] <lamont> that's not too bad.
[08:25] <fabbione> nah it just takes a lot of time mostly due to the drying time
[08:25] <fabbione> between each different thing
[08:25] <lamont> but it does get boring.  Of course, if you know the right tricks, you can fill nails in 1 coat, sand, and be done with that.  Much better than 2-3 coats...
[08:25] <lamont> yeah.  lots of drying time.
[08:26] <fabbione> yeah i know.. but i don't know the tricks..
[08:26] <fabbione> so it's like all doubled
[08:26] <lamont> first time we painted our house, my wife was very fabbergasted that my idea of prepping a room to paint was throwing a drop cloth on the floor.  no taping off windows or anything.  really perplexed her.
[08:26] <fabbione> + the house is quite old and the walls are really uneven
[08:26] <lamont> I've just found it's less work if I just don't put paint on the windows.
[08:27] <fabbione> ahha
[08:27] <lamont> ask me in mataro, and I'll tell you how to patch stuff faster.
[08:27] <fabbione> i use a big piece of paper for the floor
[08:27] <fabbione> sure i will
[08:27] <lamont> I need visual aids for this one..
[08:27] <lamont> :-)
[08:27] <fabbione> but i still tape the windows because they are brand new :-)
[08:27] <fabbione> ahhaa
[08:28] <fabbione> we had to change the old ones because the wood was rotten
[08:28] <lamont> fabbione: taping windows takes too much _work_..  But then,  keeping paint off of them does require some practice first...
[08:29] <fabbione> lamont: yeah.. or to be very fast with a wet cloth
[08:38] <lamont> night for real.
[10:59] <Matt|> hi guys. I've had some problems with the nimbus roman font. When converting it to pdf in openoffice it comes out really badly. example is here: http://mdke.mine.nu/images/pdf_problems.png Does anyone have any advice?
[11:27] <amu> Matt|: if you print the file, it looks also bad?
[11:27] <Matt|> amu, not sure
[11:27] <Matt|> amu, i don't have a printer
[11:27] <Matt|> :(
[11:29] <amu> if you change your desktop fonts to nimbus, the look in a same way compared to the doc?  
[11:29] <amu> s/the/they
[11:31] <Matt|> not really: the problem with the doc is the inconsistency
[11:31] <Matt|> there is something weird about the nimbus font in ubuntu
[11:31] <Matt|> there is also no bold of that font
[11:36] <amu> Matt|: what hapen if you change the font to an bitstream?
[11:37] <Matt|> amu, lemme try it
[11:39] <Matt|> amu, it seems reasonable
[11:44] <amu> the font's at your desktop are the bitstream, and they look fine?   
[11:45] <Matt|> i've tried putting the document in bitstream and converting it to pdf
[11:49] <amu> you have a crt or lcd display?  
[11:53] <amu> i'm not sure if oof use now his own fontconfig, probably run a dpkg-reconfigure fontconfig and enable subpixeling from the computer-desktopsettings-fonts
[11:59] <Matt|> amu, ok i'll try that
[12:00] <Matt|> amu, how about the "autohinter module"?
[12:03] <amu> Matt|: Depending on your display and on which fonts you use, they can look better or worse when using the autohinter module. Enable it if you happen to prefer the look it gives to your fonts.
[12:03] <Matt|> amu, same deal after enabling subpixeling
[12:05] <Matt|> amu, i have an idea
[12:07] <Matt|> amu, ok solved. The problem was that that part of the document was actually flagged as bold, although it didn't appear when i looked at the document as the nimbus roman font has no bold. Thanks for your help
[12:13] <pitti> Hi carlos
[12:16] <amu> Matt|: hehe, np
[12:26] <carlos> pitti: hi
[01:24] <trulux> hey guys
[01:24] <trulux> lamont, hey, my gcc  pkgs are almost ready
[01:24] <trulux> (compiling)
[02:07] <trulux> what gcc version is using ubuntu?
[02:08] <trulux> is it 3.3-3.3-4-13 ?
[02:09] <thom> gcc version 3.3.4 (Debian 1:3.3.4-9ubuntu5)
[02:09] <thom> (on warty)
[02:09] <trulux> thom, thanks
[02:10] <trulux> is/are the maintainer/s of gcc packages here?
[02:10] <trulux> thom, thanks again ;-)
[02:10] <trulux> doko, hey
[02:11] <trulux> doko, do you have a few minutes?
[02:11] <trulux> seems away
[03:21] <kagou> I have a "Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8029(AS)" it seem's that ubuntu load the "8390  11424  1 ne2k_pci" module but i can't get it working at 100 (just at 10)
[03:21] <kagou> how can i force the 100 speed
[03:21] <tseng> man ifconfig
[03:22] <tseng> search for media type
[03:22] <kagou> thanks tseng
[03:24] <trulux> tseng, hey
[03:24] <tseng> hi
[03:24] <trulux> tseng, the pkgs take long time to finish
[03:24] <trulux> but it's on the enf right now
[03:27] <Mithrandir> the build log for ooo is 53MB..
[03:27] <Mithrandir> that's just wrong.
[03:31] <kagou> where can i put parameters for modules loaded automatically by ubuntu (i mean i wan't to add :  full_duplex=1 to the ne2k_pci module)
[03:56] <amu> kagou: /etc/modules.conf
[04:12] <haggai> Mithrandir: you are more than welcome to split it up (c:
[04:12] <haggai> Mithrandir: I already got a proof of concept to do it, and approved by the Sun developers, but like all things OOo it needs a lot of time..
[04:13] <haggai> Mithrandir: ..and I have a lot of bugs to sort out before I do things like that
[04:18] <fabbione> Mithrandir: you around?
[04:19] <fabbione> ehhe
[04:19] <fabbione> probably
[04:19] <fabbione> Mithrandir: i will see you tomorrow in BCN
[04:20] <fabbione> Mithrandir: let's meet up immediatly outside where you grab the lagguages
[04:22] <fabbione> and i am off
[04:22] <fabbione> cya tomorrow
[04:23] <amu> have fun fabbione this night 
[04:23] <fabbione> amu: thanks
[04:23] <fabbione> i am planning get some good amount of danish beer ;-)
[04:24] <amu> hehe  
[04:27] <haggai> doko: please bring my jacket :)
[04:29] <doko> trulux: pong
[04:30] <trulux> doko, hey!
[04:30] <trulux> doko, i suppose you already know about all the proplice/ssp stuff, so, let me talk to you about the final implementation i've made
[04:30] <trulux> it's smooth, first :)
[04:30] <amu> if i make a kde-ppc live CD you'll kill me right ? 
[04:32] <trulux> my gcc pkgs have only 3 patches: protector (updated and re-compressed) from IBM (SSP/ProPolice), a patch to make gcc able to read specs fles from an environment variable
[04:32] <doko> the work by Etoh?
[04:32] <trulux> yeah
[04:32] <trulux> the third is the libssp one
[04:32] <trulux> how it works?
[04:33] <trulux> simple, it makes gcc able to use standard -fstack-protector option, but reads and uses ssp symbols from a shared library 
[04:33] <trulux> libssp
[04:33] <trulux> so, we use -fstack-protector -lssp
[04:33] <trulux> and we will have the SSP, adn also we can modify SSP and its functions without having to recompile the whole bucnh of glibc or gcc (if ssp is there)
[04:34] <trulux> so, we have this:
[04:34] <trulux> lorenzo@dunruin:~/LIBSSP/libssp-0.2$ gcc -fstack-protector -lssp vuln-stack.c  && ./a.out 1234567890123456
[04:34] <trulux> a.out: stack smashing attack in function main()
[04:34] <trulux> ----
[04:34] <trulux> main=0x80000914 __guard_setup=0x4001daf0 __guard=0x22324fa2
[04:34] <trulux> ppid=10622 pid=12190 uid=1000 euid=1000 gid=1000 egid=1000
[04:34] <trulux> ----
[04:34] <trulux> Abortado
[04:34] <trulux> lorenzo@dunruin:~/LIBSSP/libssp-0.2$
[04:34] <doko> my concern is, that the patch is only available for 3.4, and we loose the option to switch to gcc-4.0 as the default compiler, if we introduce a patch like this. Etoh did update the patch for 3.4, I'll ping him about 4.0 and put you on the CC.
[04:35] <trulux> doko, i've backported it also to 3.3
[04:35] <trulux> lorenzo@dunruin:~/LIBSSP/libssp-0.2$ ls /lib | grep ssp && ls /usr/lib/ | grep ssp
[04:35] <trulux> libssp.so.0
[04:35] <trulux> libssp.so.0.2
[04:35] <trulux> libssp.a
[04:35] <trulux> libssp.so
[04:35] <trulux> lorenzo@dunruin:~/LIBSSP/libssp-0.2$
[04:35] <trulux> doko, ok, thanks
[04:35] <trulux> i think we could have this done in a really small time manner
[04:37] <doko> could you open a bug report for it and/or send me the updated patch?
[04:41] <doko> trulux: what's your email address?
[04:42] <trulux> lorenzo@gnu.org or lorenzo@debian-hardened.org
[04:42] <trulux> doko, sure
[04:43] <trulux> doko, where's ubuntu bug tracker/ bugzilla?
[04:43] <doko> http://bugzilla.ubuntu.com/
[04:44] <trulux> ok, thanks
[04:47] <martink> doko, should openoffice.org (debian dir and ooo-build from cvs) be buildable on hoary?
[04:50] <Riddell> amu: no
[04:51] <doko> martink: yes according to http://people.ubuntu.com/~lamont/buildLogs/o/openoffice.org/1.1.3-2.3ubuntu4/
[04:52] <haggai> martink: what's up?
[04:53] <martink> haggai, doko, it didn't find libXinerama_pic (formerly in xlibs-static-pic). (This is on powerpc)
[04:53] <martink> I'll look at the build log
[04:54] <amu> Riddell: *g* you can help to force the mepis guy to release his sources, he compiled his cool stuff against qt-free, my mailrequest and question where i can get the source of it. are refused :(     
[04:54] <haggai> martink: hmm, that hasn't been changed and probably needs to be changed
[04:54] <haggai> martink: I thought its time we introduce an Ubuntu patchset target
[04:55] <trulux> doko, what pkg to choose for the report? gcc, gcc-3.3 either gcc-3.4?
[04:55] <doko> haggai: no, that's changed. how would you syncronise the debian and ubuntu changeset?
[04:56] <doko> trulux: gcc-3.4
[04:56] <trulux> doko, ok
[04:56] <martink> haggai, I'm sure an Ubuntu patchset will be handy, but this xinerama stuff is supposed to be autodetected, isn't it?
[04:56] <haggai> martink: hmm, not sure how _rene_ did it
[04:57] <doko> haggai: no, it's a changed patch in ooo-build/
[04:57] <haggai> doko: easy, at the top of the apply file we have a list of distros and corresponding patches.  We pass the name of the patchset in from debian/rules, and we already know (using lsb_release) what we have there
[04:58] <haggai> doko: look in ooo-build/patches/OOO_1_1_3/apply and be in awe of the magic ;)
[04:58] <enrico> Hello.  I have a tiny nuisance to report with the live cd
[04:58] <enrico> a friend run the ubuntu live CD, then booted back on his sytem and he couldn't mount /home anymore
[04:59] <enrico> I told him to fsck his home; fsck did its job and now he has his home again
[04:59] <doko> haggai: known, but who assures, that you add a patch to the ubuntu set as well as the debian set?
[04:59] <enrico> Appearently, he turned of the live CD system with EXit on gnome, and his machine turned off
[05:00] <amu> enrico: he mounted his home, and turned off the computer? 
[05:00] <enrico> No: he used the "exit" command in gnome, and let the system shutdown and turn off
[05:00] <enrico> He mounted the partition with "automount in nautilus"
[05:00] <sladen> enrico: so, the situation is;  ext2 on /home, which is not mounted automatically and is not listed as fsckable in fstab?
[05:01] <amu> enrico: problem is with the init, it should unmount the filesystem, and than eject the CD, but it doent do that 
[05:01] <Riddell> amu: what isn't he releasing the source to?
[05:01] <enrico> . /home is not mounted automatically by ubuntu livecd
[05:02] <amu> Riddell: unfortunately not, ralf has this year a lot of fun, now ralf is lost in space, and he continue :(  
[05:03] <enrico> appearently, his system mounts it as ext2, but nautilus mounted as ext3.  fsck said it found a journal on it.
[05:04] <enrico> However, I seem to be going into sci/fi.  Unfortuntately, the report can't get any better than this vague
[05:04] <enrico> That is, he's 100Km from here and I have no access to his machine to investigate, and so on.  So, this can happily be dropped until it's reported by someone else
[05:07] <amu> enrico: the only solution atm is, umount the filesys manually before rebooting
[05:10] <haggai> doko: as I said, have a look at the magic :)  You have several sections, and you add an UbuntuOnly section for just ubuntu's patches, and you share the rest, just like Debian does with RH, SuSE, etc
[05:13] <Mithrandir> fabbione: ack, ok.
[05:14] <enrico> amu: right.  However, being it the ubuntu live CD, if it has a problem like that it should definitely be looked into (since it's a test system, it shouldn't allow to do mistakes, even by mistake).  Telling nautilus to mount read-only by default, for example, could be a good idea
[05:16] <martink> haggai, found the answer in the ubuntu diff.gz. In ooo-build/patches/OOO_1_1/xinerama-pic-on-all-archs.diff the s/-lXinerama/-lXinerama_pic/ change was removed
[05:18] <haggai> martink: hmm, that should be fine
[05:19] <haggai> martink: it's definately not auto-detected, looking at that patch :)
[05:20] <haggai> martink: there is a Build-Depends: libxinerame-dev in control, so it should have been ok if that part of the patch has been removed
[05:20] <amu> enrico: worth to discuss, a liveCD is something critical, and should be upgraded time by time. At least, if pitti release a USN
[05:24] <martink> haggai, yes, definitely not autodetected. And debian-only.
[05:25] <Treenaks> OK, I'm in the hotel
[05:25] <Treenaks> now whwrnow where do I find ubuntu people? :)
[05:25] <Mithrandir> Treenaks: you're in BCN?  fun! :)
[05:28] <sivang> Treenaks : how was the trip from barcelona to mataro?
[05:33] <Treenaks> sivang: the train smelled like piss, but the rest is nice :)
[05:36] <Treenaks> anyway, what I was asking.. how do I find Ubuntu people here?
[05:38] <sivang> Treenaks : eh growse, this is the train from barcelona to mataro?
[05:39] <Treenaks> sivang: yes, but it could'vebeen somebody's kid
[05:44] <RubenV> are there any known problems with ndiswrapper?
[05:44] <trulux> doko, hey
[05:44] <trulux> doko, https://bugzilla.ubuntu.com/show_bug.cgi?id=4374
[05:45] <doko> trulux: thanks
[05:45] <trulux> doko, ask me about anything you want related to that issue
[05:53] <lamont> probably just one.
[05:55] <Treenaks> to find Ubuntu people
[05:55] <trulux> lamont, check http://software.newsforge.com/comments.pl?sid=42361&cid=102743
[05:55] <Treenaks> though I have NO idea what they look like
[05:55] <trulux> lamont, sorry, wrong link :P
[05:56] <trulux> lamont, https://bugzilla.ubuntu.com/show_bug.cgi?id=4374 <- this is the right one
[06:18] <Kamion> elmo,thom: any chance that little could have an exception from archive's rsync limit somehow?
[06:18] <Kamion> elmo,thom: can't build new CDs at the moment because I can't rsync the archive
[06:20] <trulux> doko, what do you think about the report?
[06:21] <trulux> when do you think it could be handled?
[06:21] <trulux> i want to collaborate in ubuntu development in that terms , but i don't anything about the policy applied to development candidates
[07:28] <mdz> Treenaks: most folks don't arrive until tomorrow, I think
[07:28] <Treenaks> mdz: I know thom is here somewhere :)
[07:28] <mdz> yes, should be
[07:33] <Treenaks> I'm connected through the UBUNTU network though.. and it gets better if I move towards my window
[07:37] <Mithrandir> jump out the window and follow the beacon?
[07:38] <Treenaks> Mithrandir: *aaaaaaaaaghh* *splat*
[08:06] <Mithrandir> haggai/rene: for some reason, it builds with libgnomevfs, then without, then with again.
[08:24] <haggai> Mithrandir: that's _rene_'s hack
[08:25] <Mithrandir> haggai: shifting blame, eh? ;)
[08:29] <_rene_> Mithrandir: it should only build the gnomevfs stuff on the "normal" build and then immediately build the non-gnomevfs one
[08:29] <_rene_> I have no idea why it would build a gvfs version again...
[08:29] <Mithrandir> _rene_: it does, but it builds the gnomevfs one afterwards again
[08:29] <Mithrandir> timeskew problems?
[08:29] <Mithrandir> do you build on 2.6 boxes?
[08:29] <_rene_> yes, I do
[08:31] <haggai> _rene_: why don't you make sure the output goes into a different directory instead of overwriting the original version?
[08:31] <_rene_> uh?
[08:32] <_rene_> the second build nees to overweite the gvfs one
[08:32] <_rene_> since the first build is the gnomevfs one
[08:32] <_rene_> and instsetoo takes its stuff from the solver
[08:32] <Mithrandir> uhm, no, it doesn't, it builds in another solver directory
[08:32] <_rene_> it builds in another tools/ directory
[08:33] <_rene_> the library is then cp'ed into the solver in its normal name
[08:33] <_rene_> and the gvfs one renamed .gvfs before
[08:33] <haggai> _rene_: ouch :( I thought you did it the other way around
[08:33] <Mithrandir> right
[08:33] <Mithrandir> but when it then makes in solver again, it gets overwritten
[08:33] <haggai> _rene_: that's asking for trouble
[08:34] <_rene_> haggai: the problem is how to make it in ooo-build. either we don't apply the patch and then apply all those stuff manually or we do it this way
[08:34] <_rene_> hmm. we could copy it over and build the gvfs stuff in tools/gvfs and apply the GnomeVFS section manually there...
[08:34] <_rene_> erm, that patch
[08:35] <haggai> _rene_: with your soln, you have to do some patch application manually, so you mayaswell not apply the patch, and then apply the patch later and do a special build
[08:35] <_rene_> the problem is that we need to apply the whole section
[08:35] <haggai> _rene_: so make a new patch target
[08:35] <_rene_> where are patches not touching tools/
[08:35] <_rene_> haggai: hmm?
[08:36] <Mithrandir> _rene_: can't the patch be split, then?
[08:36] <haggai> _rene_: apply.pl --patchset=debianWithGnome or something
[08:36] <_rene_> Mithrandir: it already are some patches. we just unapply the one changing libtl
[08:36] <_rene_> hmm. will look
[08:37] <haggai> _rene_: even better, make those patches dependent on a preprocessor var
[08:37] <_rene_> uh? how?
[08:37] <haggai> #ifdefing stuff
[08:37] <_rene_> -DUSE_GNOMEVFS?
[08:38] <haggai> yup
[08:38] <trulux> #ifdef
[08:38] <haggai> and an envvar for the makefiles
[08:38] <_rene_> that would mean to a) need conditionalzing the whole stuff and then running configure thrice
[08:38] <_rene_> Hmm
[08:38] <trulux> _rene_, try a GCC wrapper
[08:38] <haggai> _rene_: I wasn't thinking about adding stuff to configure
[08:38] <haggai> _rene_: just to the makefiles and code
[08:39] <trulux> that reads special flags from a file, created with another tool who reads simple human-readable configuration labels
[08:39] <_rene_> ah, and then export in the environment?
[08:39] <trulux> like gentoo eclasses and so on
[08:39] <trulux> :)
[08:39] <_rene_> but that will break full-gnomevfs builds
[08:39] <haggai> _rene_: that way, you have a new target which goes  source $ENVFILE; USE_GNOMEVFS=TRUE; OUTPUT=unxlngi4.gnome ; cd tools; build.pl
[08:39] <haggai> _rene_: and you get your output in a seperate directory
[08:40] <_rene_> yeah, I see
[08:40] <haggai> _rene_: and nothing gets overwritten so you can use make's magic filestamping to keep everything working
[08:40] <_rene_> but that will break "normal" builds
[08:40] <haggai> _rene_: no.  that's why I said to conditionalise it
[08:40] <_rene_> ooo-build builds that is
[08:40] <_rene_> and when you don
[08:40] <_rene_> ''t define USE_GNOMEVFS?
[08:40] <haggai> _rene_: oh, ok reverse the variable meaning
[08:41] <haggai> _rene_: add to the makefile.mk: USE_GNOMEVFS*=TRUE
[08:41] <_rene_> ok, I'll look
[08:41] <haggai> _rene_: that way you can unset it from the environment
[08:41] <_rene_> trulux: rule of thumb: a setence containing gentoo is good for me ignoring this sentence ;)
[08:42] <haggai> trulux: and is overengineering for the problem in this case
[08:43] <trulux> haggai, yep, just dreaming ;-)
[08:43] <trulux> _rene_, that  hurts me a bit, i'm a concerned gentoo user
[08:43] <trulux> and debian also
[08:43] <trulux> in fact -> Ubuntu
[08:43] <trulux> ;-)
[08:55] <wasabi> Hum. Would everybody poopoo me if I suggested making -dbg packages of most Big Gnome packages? Evolution, Nautilus, etc.
[08:56] <wasabi> Things that tend to crash a lot with no explanation. :)
[09:02] <haggai> wasabi: we tried to make -dbg packages for OOo and discovered there is a bug somewhere - you may hit it perhaps
[09:02] <wasabi> hmm?
[09:03] <_rene_> we tried to use --dbg-package= form dh_strip
[09:03] <_rene_> binutils somehow didn't like the libs
[09:03] <wasabi> Hmm.
[09:03] <_rene_> but only *sometimes*
[09:04] <haggai> _rene_: oh, ok
[09:07] <Mithrandir> _rene_: so you'r making the changes which should fix my problem? :)
[09:20] <mdz> _rene_: that's very interesting; I would expect the resulting stripped binary to be exactly the same, regardless of whether the debug info was saved to a file
[09:21] <_rene_> mdz: no, that was not the problem. it was binutils choking during dh_strip
[09:22] <_rene_> mdz: i.e. strip failing
[09:36] <quitte> i want to make xorg packages for debian. as xorg is in hoary i figured it was bestto get te source of the ubuntu packages to start with. can you tellme were i can find it?
[09:39] <mdz> ah
[09:47] <chrisa> quitte: I think several people have already done that individually, there were a bunch of debs last time I looked
[09:50] <mdz> quitte: the source for everything is in the archive
[09:50] <mdz> quitte: but talk to fabbione; I imagine he is planning to make official X.org packages for Debian
[09:52] <kylem> shaya did a set of them.
[09:55] <daniels> mdz: i imagine that would be a fair way off, by the time everyone from every architecture has sorted it out
[09:55] <mdz> daniels: for unstable, sure, but not for experimental
[09:55] <kylem> that reminds me to take a look at xorg on hppa.
[09:55] <daniels> mdz: IIRC, Branden didn't put 4.3.x into experimental until it was working on basically every architecture except hurd-i386 and sh.
[09:56] <daniels> kylem: if you want to throw me over a MANIFEST.hppa.in, and, for bonus points, *.install.hppa, I'll throw it in our packages
[09:56] <kylem> ok... err, does ubuntu intend to port to hppa or something?
[09:56] <daniels> not really
[09:56] <mdz> lamont is likely to do it at some point
[09:57] <daniels> but since they're currently basically the sole source for debian/ubuntu packaging, i figure having the hppa bling there would be neat ;)
[09:57] <kylem> fine, fine, heh.
[09:57] <daniels> good man.
[09:57] <daniels> your leg is safe, at least for the time being. ;)
[10:04] <elmo> rookery (aka people.u.c)'s going down
[10:11] <elmo> back
[10:14] <chrisa> hrm, jeff said on planet that people who ordered a lot of Ubuntu cds received display packs
[10:14] <chrisa> I seem to have missed out on those
[10:16] <winkle> seb128: I think pygtk just broke due to new gtk or something.
[10:19] <seb128> yep
[10:20] <daniels> chrisa: >50
[10:21] <chrisa> daniels: I have 200 next to me
[10:21] <chrisa> well, about 180 left
[10:21] <daniels> hm
[10:21] <Mithrandir> hmm
[10:21] <chrisa> ah, I just searched the box. Some of the cases I haven't opened do have them
[10:21] <Mithrandir> I can reproducibly break nautilus
[10:21] <chrisa> Carry on :)
[10:23] <Kamion> python-parted's in the desktop seed?!
[10:23] <Kamion> sheesh
[10:23] <Kamion> guess I'd better fix it then, but I don't think it needs to be there
[10:24] <daniels> Kamion: yo, dude
[10:24] <daniels> Kamion: ide modules are in the bios, yeah?
[10:24] <daniels> Kamion: sideshow has an x40 here, and he can't install because it hasn't picked up his ide controller
[10:24] <daniels> Kamion: drivers/ide is suspiciously empty
[10:24] <daniels> Kamion: (this is a pxe install, mind)
[10:25] <daniels> Kamion: er
[10:25] <daniels> Kamion: imagine I never said 'ide modules are in the bios'
[10:25] <daniels> Kamion: are the ide drivers in the kernel (i.e. non-modular) in the installer?
[10:27] <Kamion> daniels: they're modular
[10:27] <Kamion> daniels: drivers/ide/ will be empty from the start of a netboot install; ide-modules is pulled in later over the network
[10:28] <daniels> Kamion: right, ta
[10:28] <daniels> Kamion: this doesn't seem to have happened
[10:28] <Kamion> daniels: what stage of the install has it got to?
[10:28] <elmo> Kamion: it was part of the python-world phase
[10:28] <daniels> Kamion: partitioning
[10:29] <Kamion> daniels: oh, also, what version of the installer is he using? if it's hoary and <= day or two ago, it won't work
[10:29] <daniels> Kamion: warty, yo
[10:29] <Kamion> hmm
[10:30] <Kamion> do me a favour, switch to tty2 and 'debconf-get mirror/suite'
[10:32] <Kamion> daniels: I'll confess that I haven't netbooted warty since like forever, but fabbione tested it pre-release I believe
[10:34] <Kamion> mkay, new python-parted uploaded, ubuntu-desktop/hoary should become installable again in a bit
[10:38] <daniels> Kamion: yeah, 'show I installed mine
[10:39] <daniels> Kamion: (sideshow is re-netbooting now, so we'll see if it's still arse when he boots from mine; if it is, diagnostics will be on the way)
[10:39] <Kamion> daniels: ta. I'm around for maybe another hour and a half
[10:42] <daniels> Kamion: rad
[10:42] <trulux> doko, i want to create an ubuntu-hardened livecd, what's the man in charge of livecds?
[10:42] <Kamion> it has just struck me how totally trivial it is to churn out bootable USB images that work the way I want
[10:43] <daniels> Kamion: it's working now
[10:43] <Kamion> hmm
[10:44] <Kamion> daniels: if it happens again I need /var/log/syslog
[10:45] <Kamion> gah, this sucks, can't build Ubuntu d-i images on Debian any more
[10:45] <Kamion> bloody parted
[10:45] <daniels> Kamion: will do, cheers
[10:45] <daniels> Kamion: oh, bugger.  abi change?
[10:45] <Kamion> uh-huh
[10:46] <daniels> sheez.
[10:46] <Kamion> ah well, I have a million chroots lying around
[10:49] <mdz> Kamion: we ought to get the installed system booting from USB correctly, too
[10:49] <daniels> mdz: yeah
[10:51] <Kamion> mdz: very painful on powerpc unfortunately
[10:51] <Kamion> mdz: (but yes)
[10:52] <mdz> Kamion: oh? why particularly on powerpc?
[10:52] <Kamion> mdz: IIRC joeyh has had debian-installer successfully booting from one USB stick and installing to another, so it shouldn't be too hard
[10:52] <mdz> installing to it is fairly straightforward, but getting it to boot afterward needs some work
[10:53] <Kamion> mdz: because ofpath (yaboot) doesn't have a clue how to find the Open Firmware path to the USB stick, and even when I looked at it by hand it required a certain amount of guesswork
[10:53] <mdz> either adding the necessary modules to the mkinitrd list, or moving hotplug into the initrd
[10:53] <Kamion> (and therefore I don't see an easy way to modify ofpath to fix that problem)
[10:53] <mdz> hotplug-in-initrd really should happen anyway
[10:53] <Kamion> hotplug-in-initrd sucks; found that out in d-i
[10:53] <mdz> oh?
[10:53] <Kamion> having udevd start up before you've got around to running pivot_root is not good
[10:54] <Kamion> if udevd weren't there it might just about be sane, although the race conditions still concern me
[10:54] <Kamion> I think hotplug would need to be disabled around the pivot_root
[10:54] <Kamion> and you'd have to wait for any running scripts to go away
[10:54] <mdz> I think hotplug-in-initrd is the way forward, though it has some issues to work out
[10:55] <Kamion> my solution in d-i was to install hotplug as /sbin/hotplug.real and only enable it once the system is in a roughly sane state
[10:55] <Kamion> I think you'd want to do something like that for hotplug-in-initrd; basically just use it as coldplugging, don't attempt true hotplug
[10:56] <trulux> lamont, ping
[10:56] <mdz> right
[10:56] <mdz> trulux: he's left for the airport
[10:57] <trulux> mdz, ok
[10:57] <mdz> it's just a nice way of centralizing the logic and getting rid of all that guesswork in mkinitrd to figure out what to load for the root filesystem
[10:58] <mdz> and as a bonus being able to pull the disk from one machine and put it into another, and have it still boot
[10:58] <trulux> mdz, how easy is to rebuild the whole ubuntu pkgs tree? wanna-build?
[10:58] <mdz> trulux: not very esay
[10:58] <mdz> easy
[10:58] <mdz> assuming you're talking about bootstrapping a new architecture
[10:59] <mdz> rather than just recompiling things for an existing one
[10:59] <mdz> recompiling for an existing one is fairly simple
[10:59] <daniels> (and if you are recompiling, why?)
[10:59] <trulux> mdz, why you say not "for an existing one" ?
[10:59] <Kamion> mdz: pull the disk> assuming that grub does the right thing ...
[10:59] <daniels> trulux: if you are recompiling for i386, amd64, or powerpc, it's very easy
[11:00] <daniels> trulux: if you are recompiling everything on an architecture where ubuntu doesn't already run, it's hard
[11:00] <amu> *gaehn* 
[11:01] <trulux> daniels, i mean i386
[11:01] <trulux> recompiling for i386
[11:01] <daniels> trulux: why are you recompiling for i386?
[11:02] <trulux> i'm doing a hardening improvement for Debian and mainly (now) Ubuntu
[11:02] <haggai> trulux: amu is the livecd man
[11:03] <trulux> kay
[11:03] <trulux> i'm going to make some noise on ubuntu bugzilla about the things i work on
[11:04] <trulux> i'm starting to love the ubuntu development "policy", Debian goes slower
[11:11] <Kamion> gah
[11:11] <Kamion> FABBIONE 
[11:11] <Kamion> fabbione: what happened to sk98lin-update.dpatch?
[11:13] <Kamion> oh, bleh, it's in the changelog ... but I'm missing the MODULE_DEVICE_TABLE
[11:13] <Kamion> this is the problem with hotplug, some complicated kernel merge leaves you with no hardware detection :(
[11:15] <trulux> mdz, what's the best way to recompile the pkgs?
[11:18] <haggai> hmm, the Wiki says Mataro is zone 3 but looking on the map it looks like it is in zone 4.  Anyone know for definite?
[11:21] <amu> oOo matrix3 on tv 
[11:27] <trulux> daniels, any idea about my last question?
[11:28] <Kamion> but, theoretically, sticking d-i netboot on a USB stick seems to work; will test further in Mataro
[11:33] <daniels> trulux: it's not easy -- you need to set up wanna-build and buildd with sbuild
[11:39] <carlos> haggai: yeah, it seems like it's zone 4
[11:39] <carlos> haggai: just ask for a ticket to Mataro
[11:39] <carlos> and that's it
[11:40] <trulux> daniels, our use debian-builder for each package
[11:40] <trulux> daniels, any documentation about that?