[00:11] <mathiaz> slangasek: so IIUC, if a new openssl package with tls extension enabled is uploaded, we can either hope that packages that depends on it won't break or rebuilt all the package that depends on libssl0.9.8 ?
[00:13] <cjwatson> OpenSSL is a package very commonly used by third-party packages, and we can't make them rebuild
[00:14] <mathiaz> cjwatson: ok. So what would be the options ? two source packages ?
[00:15] <cjwatson> mathiaz: try not to break the ABI :-)
[00:15] <cjwatson> very, very hard
[00:15] <cjwatson> er, I mean "try very, very hard"
[00:16] <slangasek> cjwatson: well, then the package would need to be renamed when enabling tls extensions
[00:16] <slangasek> because it changes the size of several non-opaque structs
[00:16] <slangasek> and if something links against openssl which is both a) third-party and b) dumb, this may result in segfaults
[00:17] <cjwatson> fun
[00:18] <cjwatson> could provide a separate binary package with the extension, perhaps?
[00:18] <cjwatson> which definitely ought to involve coordination
[00:19]  * slangasek nods
[00:20] <mathiaz> which parties would be involved in the coordination effort ?
[00:20] <slangasek> Debian maintainer + you :)
[01:15] <el_cubano> When I run 'apt-file update' on Dapper, I get "ERROR 501: Not Implemented."  Is this a known problem?
[01:20] <el_cubano> Also, does anyone know why packages.ubuntu.com is not responding?
[01:27] <pochu> el_cubano: for the latter, http://blog.djpig.de/2008/02/01#packages-updates
[01:29] <el_cubano> pochu: Perhaps that information should be posted somewhere prominent on the main page?
[01:30] <pochu> Maybe
[01:31] <el_cubano> Uggh...You have searched for the contents of lyx in dapper, architecture i386.
[01:31] <el_cubano> Can't find that package, at least not in that distribution and on that architecture.
[01:32] <el_cubano> Perhaps someone can answer my question.  I am trying to figure out a package build failure on Dapper.  Which package contains the /usr/bin/lyx binary?
[01:32] <StevenK> lyx
[01:32] <el_cubano> Not on Dapper it doesn't
[01:35] <pochu> !info lyx dapper
[01:35] <ubotu> lyx (source: lyx): High Level Word Processor. In component universe, is optional. Version 1.3.7-0ubuntu4 (dapper), package size 17 kB, installed size 48 kB
[01:36] <pochu> So it exists.
[01:36] <el_cubano> pochu: After 'aptitude install lyx'
[01:36] <el_cubano> the command 'which /usr/bin/lyx' produces no output.
[01:36] <el_cubano> The file is *not* present.
[01:37] <pochu> el_cubano: what's the output of "dpkg -l lyx" ?
[01:37] <el_cubano> ii  lyx            1.3.7-0ubuntu4 High Level Word Processor
[01:38] <el_cubano> dpkg -S /usr/bin/lyx
[01:38] <el_cubano> dpkg: /usr/bin/lyx not found.
[01:38] <pochu> No idea then, sorry.
[01:38] <el_cubano> This is most puzzling.
[01:39] <el_cubano> Of course, since apt-file is broken and the packages.ubuntu.com search (even on djpig's alternate host) is also broken, I have no way of finding out which package has that binary :-(
[01:39] <StevenK> I think because lyx is an alternative?
[01:39] <StevenK> Hrm
[01:39] <el_cubano> In that case, though, shouldn't installing it setup the alternative?
[01:41] <StevenK> On dapper, lyx-qt or lyx-xforms provides the lyx-{qt,xforms} binaries
[01:41] <StevenK> And then lyx is an alternative that points to either one, I guess
[01:41] <el_cubano> So, then there is no such thing as /usr/bin/lyx on Dapper?
[01:42] <el_cubano> The dependency resolution for 'aptitude install lyx' pulled in lyx-qt.
[01:43] <mjg59> Does lyx-qt contain an alternative for lyx?
[01:44] <el_cubano> mjg59: Nope.
[02:39] <Hobbsee> argh.
[02:39] <Hobbsee> what's the apache upload?
[02:40] <StevenK> Who uploaded it?
[02:40] <Hobbsee> can't tell - it's not on changelogs.ubuntu.com
[02:40] <emgent> [USN] [USN-575-1] Apache vulnerabilities - http://www.ubuntu.com/usn/usn-575-1
[02:40] <emgent> this ?
[02:40] <Hobbsee> it's a security upload
[02:40] <Hobbsee> yeah, probably
[02:40] <emgent> jdstrand ++
[02:40] <StevenK> Ah
[02:41] <emgent> Hobbsee, if you like know all secuirty update see u-hardened
[02:41] <Hobbsee> emgent: i don't usually care - it's just that that has broken drupal
[02:41] <cjwatson> it's on the -changes lists too
[02:41] <emgent> uhm ?
[02:41] <Hobbsee> dep wise
[02:42] <emgent> broken drupal ? :°
[02:42] <Hobbsee> The following packages will be REMOVED:
[02:42] <Hobbsee>   apache2-mpm-itk drupal5 libapache2-mod-php5 php5 php5-gd php5-mysql
[02:42] <Hobbsee> The following packages will be upgraded:
[02:42] <Hobbsee>   apache2-utils apache2.2-common linux-libc-dev
[02:42] <Hobbsee> 3 upgraded, 0 newly installed, 6 to remove and 0 not upgraded.
[02:43] <emgent> Hobbsee, gutsy?
[02:43] <Hobbsee> emgent: yes
[02:43] <Hobbsee> hm, apache2-mpm-itk probably needs a rebuild
[02:43] <emgent>  possible
[02:43] <Hobbsee> that's the breaking dep, on apache2.2-common
[02:44] <scolytus> Hi! I've asked in #ubuntu, but didn't get an answer and i want to support you. I've installed Ubuntu 7.10 Server. There was a bug related to GRUB/MBR. What do you need for a bug report?
[03:02] <emgent> keescook, ping
[03:06] <emgent> sabdfl, heya
[03:07] <sabdfl> hi emgent
[03:15] <Hobbsee> hi sabdfl, how's it going?
[03:16] <sabdfl> Hobbsee: good thanks, am on holiday in Rio, for Carnival, then off to work in CT for two days
[03:16] <sabdfl> how's life down under?
[03:16] <ion_> http://etymonline.com/index.php?term=carnival “to remove meat”
[03:16] <Hobbsee> sabdfl: wow! they give you holidays?  Rio sounds like fun
[03:17] <Hobbsee> sabdfl: wet.  and thieves suck.
[03:17] <emgent> ehehe
[03:17] <sabdfl> sounds like rio, but somehow everybody manages to have fun nonetheless
[03:17] <slangasek> ion_: a disputed etymology, fwiw
[03:22]  * jdong waves at sabdfl :)
[03:22] <sabdfl> hi jdong
[03:24] <jdong> ok, can some Moin expert explain to me why "wiki.ubuntu.com/interdiff" does not suggest "wiki.ubuntu.com/Interdiff", but instead suggests things like "HintedFonts" and "Littleiffel"???
[03:25] <jdong> though HintedFonts was a very informative read.
[04:36] <TheMuso> c
[04:37] <TheMuso> wrong tab...
[06:26] <pitti> Good morning
[06:32] <StevenK> Morning pitti
[06:35] <TheMuso> Morning pitti.
[06:43] <dholbach> good morning
[06:57]  * pitti hugs dholbach
[06:58]  * dholbach hugs pitti back
[08:19] <stgraber> moin
[08:21] <slytherin> Can anyone please sponsor the debdiff on bug 183164. I am waiting for it to fix the FTBFS for lucene2.
[08:21] <ubotu> Launchpad bug 183164 in w3c-dtd-xhtml "Wrong path for entity sets" [Undecided,Confirmed] https://launchpad.net/bugs/183164
[08:56] <Company> bryce: are you the right person to poke about https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/147846 ?
[08:56] <ubotu> Launchpad bug 147846 in xserver-xorg-video-ati "resolution out of sync with version 2:1.3.0.0.dfsg-12ubuntu8" [High,Incomplete]
[08:59] <seb128> Company: he might be sleeping now, you can try on #ubuntu-x though
[08:59] <Kalamansi> seb128 : sebastian?
[08:59] <Company> seb128: will do
[08:59] <seb128> Kalamansi: no
[09:03] <seb128> Kalamansi: did you have a question out of mispelling my name on irc? ;-)
[09:03] <lucas> q!
[09:03] <lucas> oops
[09:04] <Kalamansi> seb128 i thought you are sebastian
[09:04] <Kalamansi> seb - sebastian
[09:06] <seb128> did you have a question or you just try to guess people names there?
[09:07] <Kalamansi> actually yes
[09:07] <Kalamansi> i have a Q
[09:07] <seb128> you should ask then
[09:07] <Kalamansi> any good sources for traffix shapping+blocking porn sites+filtering?
[09:07] <India7> http://forceindiaf1.freeblog.hu/
[09:08] <Kalamansi> India7 : i have better than than http://www.osnews.com/images/comics/wtfm.jpg
[09:08] <Kalamansi> traffic*
[09:08] <seb128> Kalamansi: you should try #ubuntu for user questions
[09:13] <Kalamansi> thanks
[11:20] <mvo> pitti: sudo has changed in hardy and now filters out everything in the environment that is not in the whitelist. this breaks the current way that gksu passes the proxy (via http_proxy)  bug #181999 seems to be caused by this. I would like to add http_proxy to the environment whitelist, is that ok with you?
[11:20] <ubotu> Launchpad bug 181999 in apt "[ubuntu hardy alpha 3] update behind proxy - connections time out " [High,Confirmed] https://launchpad.net/bugs/181999
[11:21] <pitti> mvo: can we fix gksu instead to explicitly keep that variable?
[11:21] <mvo> pitti: fix it in what way?
[11:22] <mvo> pitti: gksu talks to gconf, extracts the proxy, sets the environment and then calls sudo
[11:22] <mvo> pitti: this way the proxy information is preserved for the app running as root (e.g. synaptic)
[11:24] <pitti> mvo: hm, I thought sudo had an option to keep a particular environment variable
[11:24] <cjwatson> it doesn't seem to
[11:24] <cjwatson> I was looking for that the other day
[11:24] <pitti> seems you can either use -E (not advised)
[11:24] <pitti> or sudo VAR="$VAR" command
[11:25] <pitti> it should have a "sudo --keep-env VAR command" option
[11:25] <cjwatson> shame -E -e -K -k -P -p are all already used
[11:26] <pitti> mvo: ATM I think calling it with http_proxy=$http_proxy is the easiest method
[11:27] <mvo> pitti: its a bit ugly, but I can do that
[11:28] <Riddell> pitti: coudl you give back kdegames-kde4 on amd64
[11:28] <pitti> Riddell: done
[11:30] <Riddell> thanks
[11:57] <jdstrand> Hobbsee, emgent: apache2-mpm-itk is a universe package and needs to be rebuilt
[11:57] <jdstrand> (apparently)
[11:57] <jdstrand> I'll get to it today
[11:58] <geser> yes, apache2-mpm-itk needs a rebuild for every apache2 upload
[12:01] <Hobbsee> jdstrand: cool, thanks
[12:01] <jdstrand> Hobbsee: sorry about that. I'll make a note in our qa-regression-tests on apache2 so this won't happen again
[12:03] <Hobbsee> ok
[12:08] <abarbaccia> hey all - im wondering how to appropriately file a bug in the ubuntu kernel - both gutsy and hardy have a problem finding all 4gb of ram on a desktop
[12:09] <abarbaccia> HIMEM4G only lets you access 2GB of the ram but if you use the server kernel or roll your own using HIMEM64G then you get the full 4. It shouldn't be working like this...
[12:09] <abarbaccia> So my better question is how do i label the bug so it gets identified correctly in launchpad
[12:11] <persia> abarbaccia: I'd recommend reading https://wiki.ubuntu.com/KernelTeamBugPolicies and asking this type of question in #ubuntu-bugs.
[12:12] <sistpoty|work> abarbaccia: iirc himem4g will indeed limit userspace to 2g (memory split), that should be normal (but I may be wrong... haven't digged into this for quite some time)
[12:12] <abarbaccia> sistpoty|work: yes, userspace and kernel space should each be 2GB - but when you cat /proc/meminfo you should see all 4gb
[12:13] <sistpoty|work> abarbaccia: yes, indeed
[12:48] <Mithrandir> asac: do you happen to know if there are any plans to allow for allowing invalid security certs from the browser itself in ff3?  Having to go to preferences and add an exception is quite tedious.
[12:59] <asac> Mithrandir: yes ... next upload will contain it
[13:00] <asac> if you ask: when? ... i am waiting for beta3 which should be released pretty soon
[13:00] <Mithrandir> asac: ok, cool, that's fine for me.  I don't mind waiting a bit, I just didn't want us to end up with a usability problem.
[13:03] <seb128> asac: will that also fix the epiphany issue?
[13:05] <asac> seb128: unfortunately not ... i plan to fix epiphany features after FF (as upstream apparently doesn't care much 'bout 1.9 :()
[13:06] <seb128> ok
[13:14] <\sh> moins
[13:27] <\sh> pitti, octave3.0 transition, I think you forgot some packages from the buglist ;) could I send them to you on irc, so you can sync them directly, and the bug is closed? :)
[13:48] <pitti> \sh: yes, fine for me
[13:51] <\sh> pitti: cool...
[14:02] <ScottK> Is the tag "apport-crash" enough for something to get retraced or to I need to add another tag to the bug?
[14:03] <seb128_> ScottK: need-I386-retrace or the corresponding architecture variant
[14:03] <warp10> Hi all!
[14:03] <seb128_> hey warp10
[14:03] <ScottK> seb128_: Thanks.
[14:03] <warp10> hey seb128_!
[14:04] <seb128_> SCottK: you are welcome
[14:10] <jdstrand> Hobbsee: apache2-mpm-itk rebuilt and uploaded
[14:10] <jdstrand> https://launchpad.net/ubuntu/+source/apache2-mpm-itk
[14:10] <Hobbsee> jdstrand: \o/
[14:20] <pitti> ScottK: hm, this seems to happen quite often now; broken python-launchpad-bugs, I assume (or people have old versions installed)
[14:21] <ScottK> pitti: OK.  I've added the tag, so hopefully it'll produce something useful.  A crashing spamassassin isn't something I want to leave lying around.
[14:25] <pitti> ScottK: SA? isn't that perl?
[14:25] <pitti> or spamd?
[14:26] <ScottK> It's perl.
[14:26] <dholbach> bdmurray: can you prepare debian/changelog entries, so we can get new revisions of pylpbugs and bughelper uploaded?
[14:26] <pitti> ScottK: oh, a crash in perl itself then?
[14:26] <ScottK> pitti: It failed retrace.  It's Bug #189232 - are there any additional packages the reporter should install to get a better retrace if it recurs?
[14:26] <ubotu> Launchpad bug 189232 in spamassassin "spamassassin crashed with SIGSEGV" [Undecided,New] https://launchpad.net/bugs/189232
[14:26] <ScottK> So it appears.
[14:30]  * soren hugs the sidebar patch for Mutt
[14:31] <jdong> soren: what! what sidebar! why don't people tell me this stuff?
[14:35] <geser> jdong: apt-get install mutt-patched
[14:37] <soren> jdong: You didn't ask.
[14:38] <geser> jdong: I hope you don't mboxes listed in mailboxes, as it does reread them on every occasion :( maildirs are better
[14:38] <jdong> geser: I currently use imaps://, but I do have local access to the maildirs behind that :)
[14:40] <geser> jdong: I use a mix of mboxes (for low-volume lists) and maildirs, every moving in the inbox made mutt to check all mailboxes for new mail
[14:40] <jdong> hmm
[14:40] <jdong> I had issues with maildirs in that mutt refused to check for new messages in the main folder listing view
[14:40] <jdong> which is eventually why I hacked a dovecot server around it
[14:52]  * pitti laughs -- "TypeError: communicate() takes at least 14155777 arguments (1 given)"
[14:52] <pitti> now, *that's* what I call a rich API :)
[14:53] <Keybuk> sounds like a relationship -- 14 million arguments
[14:53] <dholbach> pitti: ROCK :-)
[14:53] <dholbach> haha
[14:53]  * pitti is still quite stunned what to do with that report
[14:54] <cjwatson> sounds like a python-internal explosion to me :)
[14:56] <pitti> yeah, I reassigned it to python2.5 as incomplete (needs reproducer)
[14:56] <pitti> it'll probably just time out itself
[15:03] <Lutin> Mithrandir: got a minute for a pkg-config question ?
[15:05] <Mithrandir> Lutin: if it's quick, sure.
[15:06] <Lutin> Mithrandir: yep :) . I was wondering .. now that Requires.private exists, what are the use cases for Requires ?
[15:06] <Lutin> in .pc files
[15:07] <emgent> CVE-2006-3636
[15:07] <ubotu> Multiple cross-site scripting (XSS) vulnerabilities in Mailman before 2.1.9rc1 allow remote attackers to inject arbitrary web script or HTML via unspecified vectors. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2006-3636)
[15:08] <Mithrandir> Lutin: you might want to link in a static library, for instance.
[15:08] <Mithrandir> (think the mozilla embedding thing)
[15:09] <Lutin> Mithrandir: but requires.private is looked at when building static libs, isn't it ?
[15:09] <Mithrandir> it is, but not the other way around.
[15:09] <Mithrandir> and you can't include non-fPIC code in a shared library.
[15:10] <Mithrandir> and you might have languages which don't have shared libraries or where they work differently than for C-based languages.
[15:11] <thekorn> pitti: hi, I don't think Bug 163051 is a problem in py-lp-bugs,
[15:11] <ubotu> Launchpad bug 163051 in python-launchpad-bugs "apport-retrace crashed with LPUrlError in _safe_urlopen()" [Undecided,New] https://launchpad.net/bugs/163051
[15:11] <thekorn> I guess the reporter wanted to retrace a private bug,
[15:11] <thekorn> and you can not even get the description of a private bug
[15:11] <thekorn> when you are not logged in
[15:12] <Lutin> Mithrandir: ahh ok. so if I got that correctly, when linking C code with shared libs, Requires.private can completely replace Requires ?
[15:12] <pitti> thekorn: ah, good point
[15:13] <pitti> thekorn: thanks for poitning out
[15:13] <Mithrandir> Lutin: if you only have -l and -L flags, yes.  If you have -pthread, no.
[15:14] <mynameisdeleted> hi
[15:14] <mynameisdeleted> question question
[15:14] <Lutin> Mithrandir: ok. thanks a lot for the explanation :)
[15:14] <mynameisdeleted> my ide drive appears to be scsi in kernel, and wont let me use hdparm
[15:15] <mynameisdeleted> I'd really like to be able to check dma status and readahead etc
[15:15] <Mithrandir> Lutin: thanks for the question, I hadn't really realised how obsolete Requires is.
[15:15] <Mithrandir> mynameisdeleted: please ask support questions in #ubuntu.
[15:15] <Lutin> Mithrandir: :)
[15:18] <mynameisdeleted> could the fake scsi driver for ide force it in nodma mode?
[15:19] <Chipzz> mynameisdeleted: 16:15 < Mithrandir> mynameisdeleted: please ask support questions in #ubuntu.
[15:20] <Chipzz> mynameisdeleted: and I think I heard something about sdparm
[15:20] <mynameisdeleted> thats all I wanted
[15:22]  * Chipzz wonders just how dificult it is to follow channel rules though - *especially* after they are explicitely stated
[15:32] <davmor2> can anyone tell me the command to run oem complete please the icon appears to be missing from the desktop.
[15:41] <cjwatson> davmor2: if you're running into bug 188240, chances are it's simply not installed
[15:42] <ubotu> Launchpad bug 188240 in ubiquity "[Hardy] OEM install from Desktop CD doesn't work" [High,Fix committed] https://launchpad.net/bugs/188240
[15:42] <Chipzz> dholbach: why did you mark https://bugs.launchpad.net/bugs/125247 as incomplete?
[15:42] <ubotu> Launchpad bug 125247 in vim "Apache config files in /etc/apache2/sites-available and /etc/apache2/sites-enabled do not alwyas have proper syntax highlighting" [Undecided,Incomplete]
[15:43] <Chipzz> I think all necessary information is provided?
[15:43] <Chipzz> although I am willing to make a debdiff though
[15:43] <davmor2> cjwatson: yes what's the package name and I'll finish this test?
[15:44] <cjwatson> davmor2: oem-config-gtk but it will be broken if you just do that and it's really not worth trying
[15:44] <cjwatson> you would have to do a fair bit of other setup duplicated from ubiquity
[15:44] <dholbach> Chipzz: that'd be great
[15:46] <davmor2> cjwatson: ah right okay np I'll just go for a standard install then and see if some of the gvfs/gio bugs have been worked out :) ta for the help
[15:46] <Chipzz> dholbach: though, really, it's a one-line fix and the change is obviously correct; I don't see the point in making a debdiff ;)
[15:47] <dholbach> Chipzz: it will speed up getting it included
[15:54] <davmor2> cjwatson: just tried it to save reinstall and it worked just installing oem-config-gtk
[15:58] <cjwatson> davmor2: mm, ubiquity does do more than that though
[15:58] <cjwatson> I think it's great if it works, but it's not really a valid test :)
[15:58] <cjwatson> ubiquity sets up autologin for the oem user for example
[16:00] <davmor2> cjwatson: yes the autologgin worked.  It was just the means to save the changes and setup the new user I was missing.
[16:01] <Chipzz> dholbach: debdiff needs to be against vim hardy version?
[16:01] <dholbach> Chipzz: yeah, that'd help
[16:28] <asac> doko: j2se1.4-amd64 doesn't build in hardy anymore :/
[16:33] <doko> blackdown ...
[16:37] <asac> doko: ... rules is broken as well ... a perl expression failed like: http://paste.ubuntu.com/4207/
[16:38] <asac> doko: whats the state of icedtea in hardy? is it a viable replacement for 1.4 - so we can drop that?
[17:24] <ScottK2> pitti: Do you have a moment for a postgresql-8.3 question?
[17:28] <ScottK> Nevermind.  Figured it out.
[17:48] <Riddell> pitti: please give back kdebase-workspace on amd64
[17:54] <warp10> software released as "public domain" can be synced in universe, right?
[17:54] <cjwatson> yes
[17:55] <warp10> cjwatson: ok, thanks
[17:55] <cjwatson> it depends slightly on the country, as in some countries you're not allowed to relinquish your copyright
[17:55] <cjwatson> some people choose to add something that basically says "or the nearest approximation available" (but in better wording)
[17:58] <ln-> it would be an interesting move from an PD author to sue people for copyright violation in such countries.
[18:00]  * warp10 hates, hates, hates debian/copyrgiht
[18:00] <Robot101> ln-: how about a code user in such a country sueing the author because it's unfit for purpose? :)
[18:03] <slangasek> warp10: except for sauerbraten-wake6, which has to go to multiverse because of its /dependencies/, not because of its license...
[18:04] <warp10> slangasek: indeed. I always check deps when I request syncs, but I missed that sauerbrauten is in multiverse :(
[18:06] <slangasek> warp10: well, it would be exceptional that a package in Debian contrib should be synced to universe instead of multiverse, without there being a categorization bug on one side or the other :)
[18:08] <warp10> slangasek: I am currently looking at glam2, in contrib, that is licensed as public domain and doesn't either depends or build-depends on software in multiverse or restricted. Shouldn't it go on universe?
[18:10] <slangasek> warp10: er, glam2 isn't in contrib :)
[18:13] <warp10> slangasek: damn... my flu confounds both my mind and my eyes. My apologies :)
[18:56] <smagoun_> Can someone with a hardy system install totem-gstreamer and tell me whether the .so's are linked against libtotem-plparser-mini.so.10?
[18:58] <jdong> can't that be checked by unpacking said deb and using ldd?
[19:06] <smagoun_> jdong: I suppose so...
[19:21] <pochu> smagoun_: http://paste.ubuntu.com/4211/plain/
[19:23] <zyg2> sorry for asking here (ubuntu+1 seems idle) but have there been any package upgrades since alpha 4 release?
[19:23] <pochu> zyg2: there are package updates everyday, but there haven't been a new release
[19:24] <zyg2> hmm, that's what I find strange
[19:24] <zyg2> I expected lots of packages rebuilt everyday
[19:24] <zyg2> but I don't see any ...
[19:24] <smagoun_> pochu
[19:25] <zyg2> ahhh
[19:26] <zyg2> I got it - I was using local country mirror that does not seem to pull hardy yet
[19:27] <smagoun> pochu: thanks for the info
[19:31] <pochu> Who is the hppa porter? lamont?
[19:34] <ScottK2> Yes
[19:35] <pochu> Thanks. I'll wait until he shows up.
[19:49] <Riddell> Mithrandir: could you give back kdebase-workspace on amd64
[20:17] <zul> soren: ergh
[20:17] <soren> zul: Which version of mysql-doc-5.0 is installed right now?
[20:18] <zul> 5.0-0ubuntu1
[20:19] <soren> If a package has "Replaces: mysql-doc-5.0 (<< 5.0.56-0ubuntu1)", how does it make sense that it says "trying to overwrite `/usr/share/man/man1/mysql_config.1.gz', which is also in package mysql-doc-5.0", when trying to install it and mysql-doc-5.0 is at 5.0-0ubuntu1?
[20:20] <soren> Shouldn't the "Replaces:" tell dpkg that it's fine for the package to overwrite it?
[20:21] <mathiaz> Hum.. I've also updated the package version mysql-doc
[20:21] <mathiaz> It was mysql-doc-5.0-0ubuntu1 - now mysql-5.0-5.0.56-0ubuntu1
[20:22] <soren> Er..
[20:22] <soren> How much of that is package name, and how much is version?
[20:22] <Mithrandir> Riddell: done
[20:22] <soren> slangasek: You're clever... Can you provide enlightenment here?
[20:23] <mathiaz> soren: old version: 5.0-0ubuntu1 - new version: 5.0.56-0ubuntu1
[20:23] <mathiaz> soren: see https://launchpad.net/ubuntu/+source/mysql-doc-5.0/
[20:23] <soren> mathiaz: Ok.
[20:23] <soren> mathiaz: I fail to see the point.
[20:24] <luisbg> http://www.cs.helsinki.fi/u/hprajani/phun/disk.jpg
[20:24] <mathiaz> soren: I don't have any point. It was just another comment regarding things that have changed.
[20:24] <soren> mathiaz: Oh, ok.
[20:25] <zul> ill try the replace+conflict we were discussing earlier
[20:26] <slangasek> soren: the Replaces should indeed tell that it's ok to overwrite it; from what's said above, I don't see why this should fail
[20:27] <mathiaz> Is there a way to tell which iso has been used to install a machine once the machine has rebooted ?
[20:27] <soren> slangasek: Thanks so much. I thought I was going nuts. :)
[20:27] <soren> mathiaz: Yes.
[20:27] <soren> mathiaz: /var/log/installer/syslog
[20:27] <soren> mathiaz: Or something to that effect.
[20:28] <mathiaz> soren: ok. Thanks
[20:28] <soren> zul: Could you please "dpkg -I mysql-server-5.0-blahblah.deb" and triplecheck that the Replaces field is set properly?
[20:29] <zul> soren: Replaces: mysql-doc-5.0 (<< 5.0.56-0ubuntu1),
[20:30] <soren> slangasek: We had a discussion earlier today about the necessity of adding the corresponding Conflicts: field (when a file moves from one package to another).. I couldn't come up with a good explanation as to why it's added, but looking through the archive, it's *very* common.
[20:32] <slangasek> soren: there are two possible explanations for this. a) widespread misunderstanding of the meaning of the fields, b) if you have a package that Replaces: another but does not Conflicts:, and you install the replacing package and then remove it, the replaced package will still be 'installed' but will have some of its files magically missing
[20:32] <soren> slangasek: Ah, yes, of course!
[20:32] <soren> slangasek: Excellent point.
[20:34] <slangasek> soren: I don't think the latter is likely to be the real explanation for why it's done, though, except perhaps by way of cargo-cult copying, because very few of the people I've discussed this with had recognized this possibility :)
[20:35] <slangasek> (and even once recognizing it, many maintainers seem indifferent to it)
[20:35] <soren> slangasek: Heh :)
[20:53] <smagoun> seb128: Hi, totem-gstreamer doesn't function without libtotem-plparser10. You closed bug 188133 on the issue saying it's already been reported. I couldn't find another bug on the issue though. When do you expect to push a fix to Hardy?
[20:53] <ubotu> Launchpad bug 188133 in totem "totem lacks dependency: libtotem-plparser10" [Medium,Invalid] https://launchpad.net/bugs/188133
[20:56] <stgraber> Does one of you know what could generate this : "[unknown] ASSERT: "i >= 0 && i < height()" in file image/qimage.cpp, line 1712" ?
[20:56] <stgraber> Riddell: ^
[20:56] <stgraber> it's from italc which I'm trying to integrate for the next Edubuntu and it's like the only remaining bug we have
[20:57] <stgraber> (this one causes a segfaul on the client when the teacher computer start a demo)
[21:00] <slangasek> seb128: do you know if anyone's working on bug #187328, then?
[21:00] <ubotu> Launchpad bug 187328 in seahorse "seahorse: misbuild on 64-bit architectures due to missing ldap prototypes" [High,Confirmed] https://launchpad.net/bugs/187328
[21:00] <seb128> smagoun: bug #181794
[21:00] <ubotu> Launchpad bug 181794 in totem "Xubuntu's instance of totem fails (dup-of: 182231)" [Low,Invalid] https://launchpad.net/bugs/181794
[21:00] <ubotu> Launchpad bug 182231 in gnome-desktop "[Hardy A3] libgnome-desktop-2.so.2 is missing" [Medium,Confirmed] https://launchpad.net/bugs/182231
[21:01] <stgraber> http://paste.stgraber.org/141 <--- The backtrace
[21:01] <seb128> smagoun: I'll have a look to it but it's low issue on my list since other things depends on the lib anyway in the standard installation
[21:02] <seb128> slangasek: is there any hurry? I'll forward it upstream soon and it'll likley be fixed with next tarball, but feel free to do an upload if you want to get it fixed now
[21:03] <slangasek> seb128: no particular hurry, it's just that seahorse is the last package in main depending on libldap2 and I'd feel really good about demoting or nuking that package after all these years :-)
[21:05] <seb128> slangasek: well, it didn't fail to build so it should make no difference there
[21:05] <seb128> no?
[21:05] <slangasek> seb128: sorry, what do you mean?  seahorse hasn't been rebuilt at all against libldap-2.4-2 yet, because doing so without this source fix would cause a misbuild on amd64
[21:05] <smagoun> seb128: Thanks for the info. I'm working on Ubuntu Mobile (which doesn't have the same set of packages as desktop), so I'd like to get a fix for Hardy. I can push a patch to LP, does that work for you?
[21:06] <seb128> slangasek: ah ok, reading the bug I though it had but the result was buggy
[21:06] <slangasek> nah, I reported it based on the detected misbuild on ia64/debian
[21:07] <seb128> smagoun: I'll likely look at it soon, I think the bug comes from debian but you are welcome to fix the issue
[21:21] <Biff> hi, dont know if this is the correct place to ask, but i have been quite annoyed at the command-not-found handler thingy which is implemented in python and is very slow (200-1000ms) to find a package, so i reimplemented it in c, and it is quite fast, is this interesting to get into ubuntu instead of the python version?
[21:26] <Biff> it uses the same databases and stuff as the old one, its just the bash hook that is in c
[21:27] <_MMA_> Biff: I would actually email the -devel and or the -devel-discuss mailing lists.
[21:27] <_MMA_> https://lists.ubuntu.com
[21:27] <Biff> and or? :-)
[21:29] <_MMA_> Your question looks appropriate for both. -devel-discuss will go throught faster as its unmoderated.
[21:29] <seb128> grrr at the linux network stack
[21:30] <seb128> slangasek: I've opened the seahorse bug on bugzilla now
[21:30] <Biff> ok, thanks
[21:31] <TheMuso> Good morning.
[21:34] <Biff> evening here
[21:35] <seb128> who was speaking about the totem depends issue?
[21:36] <Chipzz> slangasek: with my complete lack of official understanding of that rule, my guess would be the conflicts in most cases is a *versioned* conflict to force the upgrade to a version of the package that doesn't have the conflicting file anymore
[21:36] <seb128> Chipzz: you want to use Breaks in such cases
[21:36] <seb128> didn't read the question though
[21:37] <Chipzz> seb128: 21:30 < soren> slangasek: We had a discussion earlier today about the necessity of adding the corresponding Conflicts: field (when a file moves from one package to another).. I couldn't come up with a good explanation as to why it's added, but looking through the archive, it's *very* common.
[21:37] <seb128> Chipzz: or are you speaking about same contents in different binaries? in which case you want to use replaces
[21:37] <seb128> ah
[21:37] <Chipzz> seb128: conflict/replaces pair is common
[21:38] <seb128> yeah, it's controversial whether it's right or not
[21:38] <Chipzz> and the question was about what purpose the conflicts part served
[21:38] <seb128> it allow to downgrade without issue
[21:38] <seb128> which is not really a supported case so is not required
[21:38] <_MMA_> seb128: I think it was smagoun re: Totem issue.
[21:39] <Chipzz> seb128: but nice to have since you can downgrade manually I guess
[21:40] <ScottK2> Chipzz: I think the key issue is if you want to force the old package uninstalled or not.   For the common case of renamed binary package you do.
[21:40] <seb128> Chipzz: not really nice to have, it makes upgrades harder to calculate
[21:40] <seb128> ScottK2: in such cases you sue conflicts,replaces,provides which is correctly supported
[21:40] <Chipzz> seb128: so what would you do while testing a package you're changing? force stuff with dpkg --force-... ?
[21:40] <smagoun> seb128: totem depends was me
[21:41] <seb128> smagoun: ok, I've figured the issue, will upload tomorrow
[21:41] <seb128> Chipzz: what do you mean? replaces is enough to install the new version
[21:42] <smagoun> seb128: thanks, I appreciate it. was it just a missing libtotem-plparser10 dependency in the control file? (I tested such a patch locally and it worked for me)
[21:42] <slangasek> Chipzz: yes, they are generally versioned conflicts; but in that case, the question is why use replaces at all instead of just the conflicts :)
[21:42] <Chipzz> seb128: say you, as a package maintainer, are testing a new version of a package in which a file has migrated from one package to another; you can upgrade to test it, but if you want to subsequently downgrade to test a new version, what do you do?
[21:43] <seb128> I do apt-get install --reinstall binary=version where version is the version I want to reinstall
[21:43] <seb128> the issue is rather if you try to reinstall the package which has been replaced
[21:45] <seb128> another issue is that when you downgrade the package you install the files which were replaced will not be copied there again, so you need to reinstall the other package too
[21:45] <seb128> technically the conflicts are not required but they avoid those corner cases
[21:46] <seb128> smagoun: no, that's a dh_shlibdeps call in the rules which is lacking
[21:53] <seb128_> re
[21:53] <seb128_> smagoun: no, that's a dh_shlibdeps call in the rules which is lacking
[21:53] <seb128_> not sure if you received it before the disconnect
[22:07] <jdong> thanks, ~ubuntu-archive for accepting clutch :)
[22:16] <superm1> ~ubuntu-archive, why was wxformbuilder rejected?
[22:17] <ScottK2> superm1: Usually they mail you.
[22:17] <superm1> yeah i didn't get anything about it
[23:03] <Riddell> seb128, slangasek: after doing backports mind and pass them through new or unapproved, there were ones that have been waiting there for three weeks
[23:08] <Riddell> Hobbsee, pitti, Mithrandir: please give back kdeaccessibility-kde4 kdeartwork-kde4 kdemultimedia-kde4 kdenetwork-kde4 on amd64
[23:15] <TheMuso> Woohoo. More pulseaudio esound compat layer bugs. :)