[12:43] <jdub> woo, time to upgrade to gusty!
[12:44] <grayman> hah
[12:44] <grayman> good call
[12:49] <jmg> sharms: you mean LCARS
[12:49] <sharms> yeah :)
[12:50] <jmg> http://themes.freshmeat.net/projects/lcarsaccess441/
[12:53] <jmg> (iii) make any copies of the PV Drivers or use the PV Drivers other than in conjunction with licensed Products.
[12:53] <jmg> oops
[12:53] <jmg> http://www.youtube.com/watch?v=YwXBPjLdJnU
[12:54] <sharms> ha that looks bad ass
[02:19] <\x-it> I know this is not a support channel but, basically, no one knows wtf to do in #linux/#ubuntu/#debian or anywhere and I am wondering if one if you would help me in a /msg? It is concerning wheel/su/chmod. Once I restart, I will not be able to login to root anymore (ever), basically.
[02:19] <\x-it> And I don't feel like reinstalling ><
[02:20] <\x-it> I can bribe? ;p
[03:08] <sharms> \x-it: #1 it isn't really proper etiquette to PM, or ask in channels which are not for the chosen topic.  Fortunately for you, I am the bad boy of ubuntu, so I will give you 120 seconds of my time
[03:09] <sharms> \x-it: there is no password for root by default
[03:09] <\x-it> Hmm?
[03:09] <\x-it> I know.
[03:09] <sharms> \x-it: with that stated, ask your question in a specific manner, and you have 92 seconds left
[03:09] <\x-it> It's ok
[03:10] <sharms> https://help.ubuntu.com/community/RootSudo  -- that should really cover anything related to root
[03:56] <johanbr> xq: Did you get it sorted out?
[04:20] <sid> root@cookooland:~# apt-get install linux-restricted-modules-2.6.20-15-generic, E: Couldn't find package linux-restricted-modules-2.6.20-15-generic, When I started restricted-manager... it tells me it can't run without that program. But this program isn't installable... anyone have a link to the deb? or know how to install it?
[04:20] <sid> I don't see it listed here: http://us.archive.ubuntu.com/ubuntu/pool/main/l/ .. Is this a bug?
[04:24] <grayman> erm
[04:24] <grayman> that's really the wrong place to ask that and you're doing it wrongly
[04:27] <sid> grayman: synaptic or aptitude don't show the package either.(my sources.list are here: http://rafb.net/p/hhRH1612.html ) .. I removed this package to mess with madwifi and openhal, but the module doesn't work with my chipset(yet), so I did make uninstall in my madwifi directory, and I am not able to get linux-restricted-modules-2.6.20-15-generic installed again. I'm running feisty with all default settings(except removing linux-restricted* for experim
[04:27] <sid> I think it's a bug, I'm pretty sure. Can anyone confirm, try to purge linux-restricted-modules-2.6.20-15-generic and install it again?
[04:28] <grayman> ah you're under root. it is there in my mirror
[04:28] <grayman> might be missing in yours
[04:29] <sid> grayman: So you were able to purge it and than install it?
[04:29] <grayman> i see it in the list and it's already installed
[04:29] <sid> grayman: You have a link?
[04:29] <grayman> to what?
[04:29] <sid> The deb?
[04:33] <sid> hmm, it was in restricted. It's strange that a package comes on the maincd, but I'm not able to remove/install it, since the sources.list doesn't have restricted by default.
[04:34] <sid> Should make it so you're able to remove/install any package that is installed via the disc when formatting with Ubuntu.
[04:34] <sid> thanks a lot for the help grayman, I appreciate it.
[04:34] <grayman> yes well, next time it's really a question for @ubuntu
[04:34] <grayman> s/@/#
[05:08] <jugo23> How exactly can I become involved with helping develop ubuntu?
[05:09] <jmg> jugo23: start by filing/fixing bugs
[05:09] <lifeless> jugo23: http://www.ubuntu.com/community/participate
[05:12] <jugo23> Where do I find reported bugs that need fixing?
[05:13] <pygi> jugo23: www.launchpad.net
[05:13] <jugo23> Thank you.
[05:15] <pygi> yw
[05:16] <jugo23> Another question: What programming language do you think assists in the overall development of ubuntu?
[05:16] <jmg> Python
[05:16] <jugo23> That's what I would have guessed.
[05:17] <jmg> or perl, c
[05:17] <jmg> equally valuable
[05:20] <imbecile> I just had to come in here to say that fiesty is by far the most pleasant linux distro I have ever used.. thank you all very much for your hard work.. I LOVE IT!!!!!
[05:20] <xq> johanbr: Yes, thank you very much
[05:21] <johanbr> I didn't really do anything, but you're welcome. :)
[05:21] <xq> Just asking and being polite is worth a thank you for asking :)
[05:38] <bryce> jugo23: there's several projects hosted at launchpad.net; make sure to click on the 'ubuntu' first, then search on projects
[05:38] <xq> Also be sure to join your local team if there is one :)
[05:42] <Hobbsee> morning all
[05:46] <pygi> hi Hobbsee 
[05:46] <Hobbsee> :)
[05:51] <pygi> what are you upto at this early morning? :)
[05:54] <Hobbsee> it's 2pm
[05:56] <pygi> or 5:55AM, depends on how you put it
[06:00] <Hobbsee> true
[06:18] <ion_> Morning.
[07:11] <Mithrandir> Lutin: sure.
[07:12] <fabbione> morning
[07:12] <ajmitch> morning fabbione, Mithrandir 
[07:13] <Mithrandir> morning ajmitch, fabbione 
[07:19] <Hobbsee> hi Mithrandir, fabbione 
[07:30] <Burgundavia> morning Mithrandir, fabbione
[07:31] <Mithrandir> hiya Burg
[07:36] <pitti> Good morning
[07:37] <Hobbsee> hi pitti!
[07:38] <Mithrandir> morning pitti
[07:41] <fabbione> hey pitti
[07:41] <Hobbsee> er, :)
[07:42] <ion_> Hi pitti
[07:42] <ion_> pitti: I finally got around to setting my jabberd back up; i also added your JID to my contact list.
[07:43] <pitti> ion_: ah, nice; I'm not on jabber ATM, I will be again when I'm back at my workstation (gf is still asleep :) )
[07:43] <ajmitch> hey pitti 
[07:43] <ion_> My JID is ion@heh.fi in case anyone else feels like adding it. :-)
[07:45] <Lathiat> yeh i like irc
[07:45] <Lathiat> jabber beats msn
[07:45] <ion_> pitti: Me too. But jabber is nice for talking with people for whom IRC is too complex, such as my parents.
[07:45] <pitti> ion_: right
[07:46] <Hobbsee> Lathiat: it's group chats are a bit weird though
[07:47] <ion_> ajmitch: Helping your dad configure stuff on his computer is a lot easier with a text-based system where you can write commands and he can paste their output. :-)
[07:47] <Lathiat> ion_: yeh thats what msn is for :)
[07:47] <Lathiat> most of my non 'geek' friends are on msn
[07:47] <ion_> I should set up a VPN from his computer to my network so i could just log in to his box, though.
[07:48] <ion_> MSN randomly drops messages. At least gaim shows an error (after a long delay), but the official Messenger doesnt even do that. :-)
[07:48] <Lathiat> i never have a problem with that
[07:48] <Mithrandir> there was a "TCP-over-gaim" SoC project which looked kinda interesting.
[07:48] <Lathiat> gaim actually sends me *too many* errors
[07:49] <Lathiat> half the time i send a message in msn on gaim it sends an error back but they got the message
[07:49] <Lathiat> usually the first message i send
[07:49] <Lathiat> but yeh i have sen errors come back like 10 minutes later, like . once or twice.
[08:01] <dholbach> good morning
[08:02] <ion_> Hi
[08:02] <dholbach> hey ion_
[08:33] <JussiP> I wrote a wiki page describing why Linux should support fully relocatable binaries. Read it here and tell me what you think: https://wiki.ubuntu.com/RelocatableBinaries
[08:34] <pitti> JussiP: is that -fPIE?
[08:34] <JussiP> No.
[08:34] <Mithrandir> pitti: noe
[08:34] <Mithrandir> nope
[08:34] <pitti> ah, no
[08:34] <phratman> Hello.
[08:34] <phratman> Could someone else look into this as well to verify that some sort of bug in kickstart -> d-i translation is in effect? Step a) Make a simple kickstart setup (doesn't really matter what it looks like as long as it has a post installation script that untars a .tar file) Step b) Boot up an Ubuntu Feisty Alternate Installation CD with the ks=<path to the kickstart configuration file> (Hit F6 and add ks=http://somewebserver
[08:34] <JussiP> Relocatable as in "move it anywhere within your filesystem and it still works".
[08:35] <phratman> If a "bad number" error is displayed in /var/log/syslog on the installation CD, then the error has been successfully been reproduced. I've tried it on two different machines (albeit with the same make and model, but physically two different machines) and I have reproduced the error at least two or three times over.
[08:35] <Mithrandir> JussiP: your claim that you can't run two apaches in parallell is false.
[08:35] <phratman> (on each)
[08:35] <JussiP> It was an example.
[08:35] <Mithrandir> JussiP: use examples which are true, then.
[08:36] <phratman> Also, I've noticed another bug with kickstart but I haven't filed it with Launchpad just yet. Is it faster to just talk to a developer about it here... ?
[08:36] <pitti> JussiP: two hald instances will fight each other badly, for example
[08:36] <pitti> or dbus
[08:37] <JussiP> But you can stop one and start the other one. Just as with self-compiled versions.
[08:37] <pitti> of course
[08:38] <JussiP> Which is pretty much the entire point.
[08:38] <JussiP> Replaced apache with foobar on the page.
[08:38] <Mithrandir> I'm not sure what problem you're trying to solve.  For instance, by far most of the software I come across runs just fine out of the compilation directories.
[08:38] <phratman> Hello?
[08:39] <Hobbsee> Mithrandir: i presume the sysadmins know that the wiki is dog slow, and has been for days?
[08:39] <phratman> Ah, overlap noted.
[08:39] <grayman> yup
[08:39] <Mithrandir> phratman: #ubuntu-installer might be a better channel to ask in.
[08:39] <phratman> Jeez, you people are even better organizers than I presumed :)
[08:39] <JussiP> The problem is that if you have a .so file and it has its own conf file, what should it pass to fopen to read it.
[08:40] <Mithrandir> Hobbsee: I don't know.
[08:40] <mdke> Hobbsee: best thing is to ask them. I wouldn't assume they do
[08:40] <Burgundavia> Mithrandir: really old laptop testing team results. Keep or nuke??
[08:40] <Mithrandir> JussiP: shared objects shouldn't have configuration files.
[08:40] <Mithrandir> Burgundavia: really old as in breezy?
[08:40] <Burgundavia> Mithrandir: and earllier
[08:40] <Hobbsee> mdke: who are they?
[08:40] <JussiP> Mithrandir: why not?
[08:40] <Mithrandir> Burgundavia: nuke them, IMO.
[08:41] <Mithrandir> JussiP: encoding the soname in the name of the configuration file is painful and makes migration hard when you upgrade, for a start.
[08:41] <mdke> Hobbsee: https://launchpad.net/~canonical-sysadmins / #canonical-sysadmin 
[08:41] <Burgundavia> anybody else got an opinion?
[08:42] <phratman> Hehe, I was told this channel isn't very active... apparently someone meant to say #ubuntu-installer isn't very active.
[08:42] <JussiP> And that would be different from executable files how?
[08:42] <Mithrandir> JussiP: do you know what a soname is and why we have them?
[08:42] <Hobbsee> mdke: great, thanks.  i'll take the duncecap for not picking the smart answer :P
[08:42] <mdke> np
[08:43] <JussiP> Sorry, I don't know linker magic all that much.
[08:45] <Mithrandir> JussiP: so, each time you break backwards compatibility, you change the name of the library.  Usually, that is done by incrementing a digit at the end of the name (libc6 is the sixth or seventh libc ABI for instance).  This is done so you don't have to recompile all your applications in one go and libraries with different sonames can be installed in parallell.
[08:45] <Mithrandir> we don't do anything like that for most applications, and when we do, we don't migrate settings, which is bad and leaves the user out in the cold.
[08:45] <Mithrandir> like apache 1 => apache 2; they're coinstallable, but no attempt at migrating the settings from a1 to a2 is done.
[08:48] <Mithrandir> doko: good morning.
[08:48] <Mithrandir> doko: gcc-4.1 ftbfs on ppc and sparc. :-/
[08:49] <JussiP> I know that, but can't formulate something suitable for an answer. So let's stick to executables then. Same question as above but s/a .so file/an executable file/.
[08:49] <doko> Mithrandir: seen, trying to track this down. was afk most of the weekend
[08:49] <Mithrandir> doko: thanks.
[08:52] <Mithrandir> JussiP: I actually think those use cases are, for the most part, better solved by using solutions such as zeroinstall or klik.
[08:55] <JussiP> Zeroinstall's faq says: "The main requirement is that the program doesn't use hard-coded paths." What I'm suggesting is a system-wide and standardised way to achieve this for any program.
[08:56] <Mithrandir> that's slightly incorrect, it just needs to use relative paths.
[08:58] <mvo> JussiP: klik should not have this requirement
[08:58] <Hobbsee> who said we're getting klik?
[09:01] <JussiP> It also says "a shell script would use `dirname $0`/mydata rather than /usr/share/myprog/mydata". The dirname part in shell code is exactly equivalent to having a function as I describe on the wiki page.
[09:03] <Mithrandir> you want a function that is essentially dirname(argv[0] ), you're aware of that?
[09:04] <JussiP> Except that argv[0]  is almost always just 'foo'.
[09:04] <JussiP> Not '/usr/bin/foo' or whatever.
[09:04] <pitti> dirname(/proc/self/exe), then? :)
[09:05] <pitti> well,l readlink()
[09:05] <JussiP> Please read this: http://autopackage.org/docs/binreloc/
[09:05] <phratman> Oh, I heard rumors that Ubuntu was going to get off Debian and get onto some sort of alternative (autopackage was a term that was thrown around). Is there any truth to that?
[09:05] <JussiP> pitti: yes, exactly except that a) it is not portable (I think) and b) does not work for libraries (as mentioned on the link above).
[09:05] <Mithrandir> phratman: no.
[09:05] <phratman> I thought Ubuntu and Debian had really close ties so I dispelled such rumors.
[09:06] <Burgundavia> phratman: last time I checked today is not April 1st,
[09:06] <phratman> Burgundavia: Point taken.
[09:07] <Hobbsee> phratman: if you read it off the forums, and not off ubuntu-devel ML or something, it's probably not true
[09:07] <pitti> JussiP: hm, I don't really understand why a library should behave differently when it is run from different paths
[09:07] <phratman> Hobbsee: It was by word of mouth.
[09:08] <shawarma> pitti: Is the automatic apport retracer thing not running anymore?
[09:08] <pitti> JussiP: people often use LD_LIBRARY_PATH to use a locally built version for testing, or to install a needed lib on computers where they don't have root etc., but that should often not alter their behaviour
[09:08] <JussiP> pitti: suppose it has to access some data files such as icons that get installed along with it. What file name should it pass to fopen?
[09:08] <pitti> shawarma: no, it isn't, we need to fix python-launchpad-bugs first; the latest LP rollout broke it
[09:09] <pitti> JussiP: right, I see such use cases
[09:09] <shawarma> pitti: Ah, ok. Thanks.
[09:09] <pitti> JussiP: my point is, that this should not *always* be the case
[09:10] <pitti> JussiP: i. e. developers often start e. g. gnome-panel with a locally built libgnomevfs.so, but that doesn't mean that they want the library to use /home/foo/icons (which doesn't exist)
[09:10] <JussiP> Fair enough, but it does have uses and is currently impossible (without hacking).
[09:10] <pitti> JussiP: or I want to install and use libfakeroot.so on a server where I'm not root, etc.
[09:11] <shawarma> pitti: Hm... So the a local "apport-retrace <bugnumber>" won't work either, I suppose?
[09:11] <pitti> JussiP: I'm not convinced that there is a central and nonintrusive solution TBH; this needs to be carefully implemented per-library
[09:11] <pitti> shawarma: I'm not sure whether it works with -o or -s; you can try
[09:11] <pitti> shawarma: I just disabled it quickly to not loose the bug tags until this is resolved
[09:12] <pitti> JussiP: I fully agree to 'being useful sometimes'
[09:12] <JussiP> Yes, but adding the *possibility* to do that to the system is a very small addition (since you can dig out the information if you really, really want to).
[09:12] <shawarma> pitti: Oh, it's the upload-the-new-traces-to-launchpad bit that's broken in python-launchpad-bugs?
[09:13] <JussiP> In other words: adding this function will break 0 applications and libraries. Having it available gives coders a lot more freedom and capabilities.
[09:13] <pitti> shawarma: I'm really not sure, I didn't have time to look into it yet; I hope that I can fix this today
[09:14] <shawarma> pitti: Ok. Thanks.
[09:16] <JussiP> Even more simplified: having this makes OSX app bundle type stuff possible. And thus Mac users have to find a new thing to boast about.
[09:17] <pitti> JussiP: they don't have apt-get install <anything you want> :-P
[09:21] <JussiP> Neither do we. There's a lot of stuff that's not packaged. In blue sky dreams having fully relocatable LSB binaries app devs could just create one of those and it would work on any distro, in any path. [add sufficient smileys here] 
[09:34] <delmorep> http://digg.com/linux_unix/Compile_the_Audacity_1_3_2_Beta_with_Ubuntu_Feisty
[10:07] <pitti> mvo: can I nag you about the apport and tzdata feisty-proposed verifications?
[10:07] <pitti> hi slomo!
[10:08] <slomo> hi pitti 
[10:08] <mvo> pitti: sure, its on my list for today, let me check if its available in the archive now
[10:08] <pitti> slomo: I will write some MIRs for texlive today, so that we can do the tetex->texlive transition as early as possible
[10:08] <pitti> mvo: I just checked, it is
[10:08] <pygi> hi pitti and mvo 
[10:09] <pitti> hi pygi
[10:09] <mvo> hey pygi!
[10:10] <slomo> pitti: sounds good :) i already planned that soon too and the versions currently in debian are working good for my stuff ;)
[10:11] <seb128> hi slomo
[10:11] <pitti> slomo: do you happen to know if they moved all the tetex-patches to texlive?
[10:12] <mvo> pitti: http://archive.ubuntu.com/ubuntu/dists/feisty-proposed/main/binary-i386/Packages.gz <- is empty for me
[10:12] <pitti> slomo: the thing we currently still carry is the 'patch for not crashing for 0.5 hour timezones'
[10:12] <pitti> mvo: urgh, they are on drescher
[10:12] <slomo> pitti: no idea... and your poppler patch is still not there as poppler 0.5 is still not in unstable...
[10:12] <pitti> mvo: that url is not empty for me, hm
[10:12] <slomo> hi seb128 
[10:12] <pitti> slomo: that's easy enough to carry ourselves for now
[10:13] <slomo> pitti: i would assume that the timezone patch is upstream but shouldn't be hard to verify ;)
[10:13] <pitti> right
[10:13] <mvo> pitti: it is for me even with wget --no-cache, strange
[10:13] <Nafallo> not empty for me either
[10:13] <pitti> mvo: hm, you might have a different mirror which is out of date
[10:14] <pitti> mvo: try de.archive
[10:14] <mvo> pitti: I will try to reach the sysadmins
[10:14] <Mithrandir> pitti,mvo : archive.u.c was being hammered over the weekend, so I suspect it's just not up-to-date.
[10:15] <Nafallo> Mithrandir: wget said it wasn't empty for me :-)
[10:15] <mvo> pitti: thanks, de.archive.u.c works
[10:16] <pitti> mvo: the two apport bugs have hints for testing, btw
[10:18] <mvo> pitti: thats good, thanks
[10:22] <saispo> hi folks
[10:23] <saispo> debootstrap under edgy is not updated for gutsy ?
[10:23] <saispo> ./usr/lib/debootstrap/scripts/gutsy: Not found in archive
[10:26] <Fujitsu> saispo: It won't be... You always need the one from the development release.
[10:27] <elena_g> can someone tell me if this 56k modem is supported by Feisty?:     01:07.0 Modem: ALi Corporation SmartLink SmartPCI563 56K Modem
[10:27] <elena_g> opps hello all :)
[10:27] <saispo> Fujitsu: i must upgrade my edgy to feisty for having a debootstrap with gutsy ?
[10:27] <Fujitsu> elena_g: Try #ubuntu
[10:27] <elena_g> Fujitsu, ok :)
[10:27] <Fujitsu> saispo: No, you must upgrade debootstrap to gutsy's version.
[10:27] <Fujitsu> Come to think of it, I don't think it's even in gutsy yet.
[10:28] <saispo> hmmmm strange
[10:28] <Mithrandir> Fujitsu: given that gutsy is frozen until we have a toolchain, yes.
[10:28] <Fujitsu> How's it at all strange, saispo?
[10:28] <saispo> Mithrandir: ok, i understand know :)
[10:28] <Fujitsu> Mithrandir: Well, tzdata got through.
[10:28] <saispo> Fujitsu: because before, i can use beootstrap for feisty under an edgy :)
[10:28] <Fujitsu> But I guess that doesn't really need to build all that much.
[10:28] <Mithrandir> Fujitsu: yes, tzdata was the first bit of the bootstrap.
[10:28] <saispo> i must wait for toolchain :)
[10:28] <Fujitsu> Do we backport debootstrap or something?
[10:29] <saispo> Fujitsu: no
[10:29] <Mithrandir> then binutils, then glibc, then gcc-4.1, but the gcc-4.1 build failed on powerpc and sparc so we need to fix that first.
[10:29] <saispo> ok
[10:29] <saispo> thanks Fujitsu and Mithrandir 
[10:29] <saispo> i will move my edgy to feisty
[10:29] <Fujitsu> How is tzdata particularly important for bootstrapping? Isn't it just a set of timezone files which can be updated arbitrarily?
[10:41] <saispo> i see updates for feisty but not available on repository, it's "normal" ?
[10:41] <seb128> saispo: what updates?where?
[10:44] <saispo> euh... works now :/ update for capplet, apport and some other things
[10:44] <saispo> maybe the mirror was not up2date :)
[10:59] <mvo> siretart: do you plan uploads for edgy, dapper for #104332 as well? 
[11:23] <siretart> bug 104332 was the rdesktop update, no?
[11:23] <ubotu> Malone bug 104332 in rdesktop "Segmentation Fault (core dumped)" [High,Confirmed]  https://launchpad.net/bugs/104332
[11:23] <siretart> right.
[11:23] <siretart> mvo: I can do it. However, it appeared as a consequence of an security upload
[11:24] <siretart> mvo: are machines with -security but not -updates common?
[11:24] <mvo> siretart: sometimes, but I guess not common (but I have no figures)
[11:24] <mvo> siretart: I'm just asking because it has a dapper and edgy tasks
[11:25] <siretart> mvo: honstely, I'm not exactly sure if dapper and edgy are affected at all
[11:25] <mvo> eh, edgy only
[11:25] <siretart> if they are affected, then it's because of a security related change in libx11-6
[11:26] <mvo> ok, I leave it as it is then, I do not have anything to test rdesktop against
[11:34] <pitti> dholbach: argh @ the multitude of branches without descriptions on https://code.launchpad.net/bughelper
[11:35] <pitti> dholbach: if I do the attachment upload fix on the bughelper.0.1 branch, will you merge it to trunk?
[11:39] <dholbach> pitti: yes
[11:39] <dholbach> pitti: will add descriptions
[11:44] <pygi> ho dholbach 
[11:45] <dholbach> hi pygi
[11:48] <pitti> dholbach, seb128: ah, I have the python-launchpad-bugs fix
[11:48] <seb128> pitti: rock on
[11:49] <ion_> Hi pygi
[11:49] <dholbach> pitti: super! let me know when you committed it
[11:56] <pitti> dholbach: bug 109213
[11:56] <ubotu> Malone bug 109213 in bughelper "recent LP rollout broke Bug.add_comment()" [High,In progress]  https://launchpad.net/bugs/109213
[11:56] <pitti> dholbach: I added the branch, tasks, etc.
[11:57] <dholbach> saw the bug - thanks pitti
[12:21] <pitti> alright, retracers are happy and running again
[12:22] <pitti> with a huge work pool ;)
[12:23] <pitti> oh, need to do ppc retracer as well
[12:36] <ogra> now that i have a working implementation in initramfs it appears immensely slower than v3 :/
[12:55] <racarr> Questions about UDS: 1. Does the Hotel provide breakfast? 2. What do people do for lunch?
[12:58] <Treenaks> racarr: 1. yes, like most hotels (it costs extra, no idea if it's included if you're sponsored); 2. no idea.
[01:06] <racarr> Treenaks: Mrgh. Ok. I guess I will send an email to find out if it is covered with sponsorship then...
[01:06] <racarr> Thanks.
[01:06] <pygi> racarr: they usually eat for lunch :P
[01:06] <racarr> pygi: Har har. 
[01:06] <Amaranth> The only one I've been to had lunch from Google so... :)
[01:07] <Treenaks> I'
[01:07] <Treenaks> We went to the hotel restaurant in Montreal
[01:07] <Treenaks> and there was Death in a Bag in mataro (only for sponsored people)
[01:08] <Mithrandir> lunch tends to be on-site, dinner is usually out in the city somewhere, most of the days you're on your own, but some days there are big group dinners.
[01:08] <Amaranth> Death in a Bag?
[01:08] <Mithrandir> I'm not sure if this is how it'll be done this time around as well, but that's how it's usually been done.
[01:08] <Mithrandir> Amaranth: I doubt we'll have those every again.
[01:08] <Amaranth> In Mountain View my dinner ended up being whatever came out of the hotel vending machine :P
[01:09] <Amaranth> What is this 'Death in a Bag'?
[01:09] <thom> *giggle*
[01:09] <Treenaks> Amaranth: oskuro.net/blog/freesoftware/mataro-halfway-2004-12-10-12-29
[01:09] <pygi> Amaranth: a cat
[01:09] <pygi> :P
[01:09] <racarr> Mithrandir: And do you know about Breakfast being included in sponsorship?
[01:09] <Mithrandir> we had some unfortunate lunches in mataro which quickly got the nickname "death bags"
[01:10] <Mithrandir> racarr: I would assume so, yes.
[01:10] <Amaranth> Treenaks: that makes me download an empty file
[01:10] <StevenK> Hum. oskuro.net looks broken
[01:10] <Treenaks> Amaranth: google cache for it works
[01:10] <Mithrandir> racarr: afaik, when people are sponsored their travel + hotel + food is sponsored, but I haven't been sponsored for years so I'm not sure if this has changed.
[01:11] <racarr> Mithrandir: Ok. Thanks.
[01:12] <Lutin> Mithrandir: thanks for rejecting allegro and SRUing cinepaint :)
[01:12] <Amaranth> Food from the local jail sounded like a better alternative? Yikes. :)
[01:12] <Treenaks> Amaranth: that was Mataro :) The hotel in Montreal was much much better
[01:13] <pygi> dholbach: yay, message sent :P
[01:13] <Treenaks> (also, there were lots of places where you could eat.. Subway, etc.)
[01:14] <Mithrandir> Lutin: oh, my pleasure.  Would you mind pushing cinepaint through the rest of the process so we could get it into -updates?  (I did it mainly because we needed a test package for -proposed)
[01:16] <Lutin> Mithrandir: sure, I'll do that :)
[01:16] <Mithrandir> thanks
[01:23] <Adri2000> pitti: have you received my email (requestsync patch)?
[01:23] <pitti> Adri2000: I got it, yes, but we cannot upload to gutsy yet, so I didn't look at it yet
[01:24] <Adri2000> ok
[01:24] <pitti> Adri2000: looks reasonable so far, although invoking $EDITOR would be cool :)
[01:25] <pitti> reading from stdin should be good enough for smaller diffs, though
[01:25] <Mithrandir> pitti: it should then just call sensible-editor
[01:30] <Demon012> hi everyone I am wishing to join in with development (I am new to developing open source software. Is this channel the correct one for me or is it the #ubuntu-motu ?)
[01:31] <Amaranth> Demon012: #ubuntu-motu might be a better place
[01:31] <Mithrandir> #ubuntu-motu is probably a better starting point.
[01:31] <harrisony> Demon012: if you want to package the software (that is used in synaptic MOTU is for you)
[01:31] <Amaranth> Demon012: But that's mostly for packaging software for Ubuntu (which requires some knowledge about how the software works)
[01:32] <Demon012> ok ty guys
[01:32] <Demon012> I wish to help with the actual coding though. But do you think I should start with packaging?
[01:33] <harrisony> Demon012: if you want to be a dev of ubuntu here is the right place
[01:33] <harrisony> (i dont dev i just admire the devs and all the cool people hang out here)
[01:34] <pygi> Demon012: go to launchpad, look at bugs, write patches, and attach them to bugs
[01:35] <Demon012> ok pygi will start with trying to do that 
[01:35] <Demon012> is there a standards list I should look at or anything for ubuntu code?
[01:35] <pygi> Demon012: what are you using most? What are your interests?
[01:36] <Demon012> like a standardised indentation style?
[01:36] <Demon012> I am a programming student
[01:36] <Demon012> have been programming properly for 4 years
[01:36] <pygi> Demon012: I mean what are your interests :) Music applications, education, ltsp, cd-recording apps, etc, etc?
[01:36] <Demon012> oh right sorry =)
[01:37] <Demon012> mmm music, education, communcation, user interface design
[01:37] <Demon012> and general program design for that matter
[01:38] <pygi> ok, well, go to #edubuntu and ask what can you help with some of the educational applications :)
[01:38] <Demon012> ok will do ty
[01:46] <racarr> Demon012: All the upstream projects will have their own style.
[01:47] <sjoerd> pitti: Thanks for your deb<->ubuntu merge mail for gvm.. little nag: please use the pkg-utopia-maintainers list next time :)
[01:58] <ogra> pitti, if i could manage to get hal to run on a thin client and manage to read the info for a blockdevice, mount it via ltspfs and then use set property to create a copy fr that device on the server but with the ltspfs mount as device node, do you think hal would survive and handle that ? 
[02:09] <pitti> re
[02:10] <pitti> sjoerd: oh, I will; is that free for anybody for posting?
[02:10] <mantiena-baltix> Hi all
[02:11] <sjoerd> Not 100% sure, but your allowed to post (as the hal changes discussion was partly on it)
[02:11] <pitti> ogra: you'd need a network dbus connection between the two, or use some other means of communication
[02:11] <pitti> ogra: but yes, it sounds doable
[02:11] <mantiena-baltix> tepsipakki: hi, do you have few minutes to talk about bug #89853 ?
[02:11] <ubotu> Malone bug 89853 in xorg-server "[regression]  7.2 broke vesa: "No matching modes found"" [High,Needs info]  https://launchpad.net/bugs/89853
[02:11] <pitti> sjoerd: ah, great; I can forward it to the list if you want me to
[02:12] <sjoerd> please do 
[02:12] <sjoerd> It's mostly me and mbiebl that worry about gvm, but it's good if the other guys can see it too
[02:12] <ogra> pitti, why is that ? if the device is there (in form of the device node) and all data is in the hal db i shoudlnt need direct communication, do i?
[02:12] <sjoerd> Which reminds me that i should make sure their on the list :)
[02:12] <tepsipakki> mantiena-baltix: shoot
[02:13] <mantiena-baltix> it seems this bug is solved only when I install both xserver-xorg-video-vesa_1.3.0-1ubuntu4.2_i386.deb and xserver-xorg-core_1.2.0-3ubuntu7.4_i386.deb packages from http://users.tkk.fi/~tjaalton/dpkg/ :)
[02:14] <tepsipakki> mantiena-baltix: that's just plain weird :)
[02:14] <mantiena-baltix> ;)
[02:14] <tepsipakki> I'll ask ajax if the backtrace is of any use
[02:14] <pitti> ogra: I mean, somehow the client has to tell the server hald about the new device?
[02:15] <mantiena-baltix> I just tested on Fujistu Amillo laptop with ATI Technologies Inc ATI Radeon XPRESS 200M 5955 (PCIE) video card
[02:15] <mantiena-baltix> tepsipakki: it boots fine on feisty final with ati driver, but doesn't boot if I change Driver "ati" to Driver "vesa"
[02:16] <ogra> pitti, yes, thats done already through an ssh tunnel (and sbalneav is working on an X based comm layer atm)
[02:16] <mantiena-baltix> But when I install xserver-xorg-video-vesa_1.3.0-1ubuntu4.2_i386.deb and xserver-xorg-core_1.2.0-3ubuntu7.4_i386.deb packages from http://users.tkk.fi/~tjaalton/dpkg/ then X starts fine even with vesa on that laptop ;)
[02:17] <ogra> indeed i need to get the data from the client to the server for hal-add-property :)
[02:17] <ogra> i just wanted to know if its possible to cheat hal in this way ;)
[02:17] <ogra> so we can move out all the ltspfs handling from gnomevfs to hal ... and Riddells users will be more happy in LTSP :)
[02:17] <mantiena-baltix> tepsipakki: should I write this in launchpad bug comments ?
[02:18] <tepsipakki> mantiena-baltix: you could do that, yes
[02:19] <pitti> ogra: I didn't look at it in detail yet, but it should be possible to encapsulate this as a hal addon, yes
[02:19] <ogra> cool
[02:20] <mantiena-baltix> tepsipakki: ok, btw, could you build  xserver-xorg-core_1.2.0-3ubuntu7.4_i386.deb without dbg symbols ? I have very slow internet connection and getting build-deps for  xserver-xorg-core would take a lot of time :( Please...
[02:20] <Mithrandir> mantiena-baltix: no.
[02:20] <Mithrandir> policy tells you to build with debug symbols.
[02:21] <tepsipakki> Mithrandir: actually, it's stripped by default
[02:21] <tepsipakki> that version isn't
[02:21] <pitti> erm, but they certainly should not appear in the debs by default
[02:22] <Mithrandir> yes, which is done by dh_strip or similar.
[02:22] <tepsipakki> I tried to backport the -dbg package stuff from debian, but thought that this was easier for testing (just skip dh_strip)
[02:22] <Mithrandir> it should be _built_ with debug symbols.
[02:22] <tepsipakki> ah :)
[02:22] <Mithrandir> they're just stripped out before the package is assembled.
[02:24] <pitti> sjoerd: sent; and this time with an actual patch attachment (brown paperbag)
[02:25] <mantiena-baltix> tepsipakki: I really need smaller xserver-xorg-core package and it seems you already have build-deps for this, so, maybe you could build stripped package (like standard ubuntu package) for me ?
[02:25] <sjoerd> pitti: thanks!
[02:25] <tepsipakki> mantiena-baltix: it's built in pbuilder
[02:25] <pitti> sjoerd: I merge gnome-mount now and do the same
[02:26] <sjoerd> nice work :)
[02:26] <tepsipakki> mantiena-baltix: and you have downloaded it already?
[02:27] <mantiena-baltix> tepsipakki: I don't have build-deps
[02:28] <tepsipakki> mantiena-baltix: why do you need to build it again?
[02:28] <mantiena-baltix> tepsipakki: I need to make small testing CD for testing Ubuntu Feisty for local computer shop and they told me, that lots of laptops with ATI X1xxx video cards doesn't work
[02:29] <tepsipakki> mantiena-baltix: right.. I'll roll a new version for you
[02:29] <mantiena-baltix> tepsipakki: your xserver-xorg-core_1.2.0-3ubuntu7.4_i386.deb takes about 40 MB after installation, while standart takes only about 10 ..
[02:30] <mantiena-baltix> tepsipakki: thanks, Lithuanians will not forget you ;)
[02:30] <tepsipakki> I guess there are a lot of X1xx users who don't forget me :P
[02:31] <Mirv> I witnessed that ATI X1xxx Mobility doesn's start even in VESA mode. Go ATI!
[02:32] <tepsipakki> ..with xorg-server-1.2
[02:37] <mantiena-baltix> tepsipakki: where you will upload stripped xserver-xorg-core_1.2.0-3ubuntu7.4_i386.deb packages ? On http://users.tkk.fi/~tjaalton/dpkg/ ?
[02:37] <cjwatson_> Hobbsee: don't see an obvious reason why you reassigned bug 109146 from debian-installer to ubiquityt
[02:37] <ubotu> Malone bug 109146 in ubiquity "unable to get anything installed" [Undecided,Needs info]  https://launchpad.net/bugs/109146
[02:37] <cjwatson_> -t
[02:37] <cjwatson_> (not that I want it moved back, but there's no evidence which of them it should be right now)
[02:37] <Hobbsee> cjwatson_: oh was it debian-installer?  I thought it was the kde upgrader?
[02:38] <cjwatson_> "Binary package hint: debian-installer"
[02:38] <Hobbsee> or couldnt tell what the heck it was, or where it would be?
[02:38] <Hobbsee> hmmm, point.
[02:38] <cjwatson_> er, it's the installer - either debian-installer or ubiquity
[02:38] <tepsipakki> mantiena-baltix: it'll be there, but 7.5
[02:38] <cjwatson_> alternate => debian-installer, desktop => ubiquity
[02:38] <Hobbsee> ah, right.  sorry, didnt know they were the same
[02:38] <Hobbsee> cjwatson_: want me to change it back?
[02:38] <Hobbsee> ahhh
[02:38] <Hobbsee> oops
[02:38] <cjwatson_> no, leave it now, like I say there's no evidence which it is - was just wondering why you bothered :)
[02:39] <Hobbsee> cjwatson: thought debian-installer was some form of universe-y package which definetly didnt fit.  not sure why
[02:39] <cjwatson> d-i's definitely main :)
[02:39] <Hobbsee> thought it was some similarly named thing that wasnt :P
[02:40] <Hobbsee> bah.  didnt you drop d-i when you wrote ubiquity?  :P
[02:40] <cjwatson> er, hell no ;)
[02:40] <jsgotangco> haha
[02:40] <StevenK> ubiquity uses d-i, actually
[02:41] <Hobbsee> cjwatson: this is what happens when i run out of chocolate.  strange things happen.
[02:41] <Hobbsee> i believe it's the same as when humans run out of coffee.
[02:43] <pitti> Hobbsee: Loving chocolate and hating coffee, I must belong to your species, then
[02:43] <Hobbsee> pitti: hehe.  are you green?  or purple?
[02:43] <Hobbsee> and alien-like?
[02:44] <robertj> (here comes the furry talk again)
[02:44] <StevenK> Hrrm, I'm related too.
[02:44] <pitti> Hobbsee: rather brown, see my LP icon https://librarian.launchpad.net/7143516/Pittiplatsch-Kopf-64x64.png
[02:44] <Hobbsee> robertj: furry talk?
[02:44] <Hobbsee> pitti: haha
[02:45] <pitti> Hobbsee: I don't suppose you actually know this figure?
[02:45] <Hobbsee> pitti: not offhand no
[02:45] <StevenK> I think I've seen that some place before.
[02:45] <StevenK> It's ringing a vague bell.
[02:46] <pitti> it's 'Pittiplatsch', one of the main characters of a famous German evening show for kids
[02:46] <pitti> and my local nickname, of course
[02:48] <Treenaks> Pittiplatz? You have an entire square of your own now? :P
[02:48] <Hobbsee> ahhh
[02:49] <tepsipakki> mantiena-baltix: the package is ready
[02:49] <StevenK> pitti: Is it comparable to some other kids show, one that Hobbsee or I might know?
[02:49] <Treenaks> StevenK: http://en.wikipedia.org/wiki/Pi
[02:49] <Treenaks> uhr
[02:49] <Treenaks> StevenK: http://en.wikipedia.org/wiki/Pittiplatsch
[02:49] <StevenK> Heh
[02:49] <mantiena-baltix> tepsipakki: thanks
[02:49] <StevenK> And now that seems so obvious.
[02:50] <Hobbsee> actually, i may have seen it, or part of it, or similar
[02:50] <pitti> Treenaks: of course, seen in Sydney: http://people.ubuntu.com/~pitti/tmp/martinplace-pittstreet.jpg
[02:50] <Treenaks> pitti: cool :)
[02:51] <StevenK> pitti: I bet you got a kick of that. :-)
[02:51] <StevenK> kick out of, even
[02:51] <ion_> pitti: Hehe
[02:51] <Hobbsee> pitti: haha, yep
[02:54] <mantiena-baltix> tepsipakki: btw, will new xorg intel drivers (on  http://users.tkk.fi/~tjaalton/dpkg/ I've noticed 2.0.0-1ubuntu1) uploaded to ubuntu-backports ? 
[02:55] <tepsipakki> mantiena-baltix: remains to be seen
[02:58] <pitti> speaking about X drivers, 'deb http://mh21.piware.de/debian/ mh21 main' has packages for nouveau (amd64 packages only ATM)
[02:58] <pitti> ^ for the adventurous testers
[02:59] <tepsipakki> pitti: there are going to be drivers in debian experimental "soon"
[02:59] <pitti> tepsipakki: ah, sweet
[03:01] <pitti> hey mh21
[03:02] <ogra> hrm :(
[03:03] <tepsipakki> ogra: if you continue to have problems with it, you can ask on nfsv4@linux-nfs.org for help on the tricky details ;)
[03:04] <pitti> tepsipakki: is there any concrete work already? mh21 put his git trees there as well, maybe this is of some help
[03:05] <pitti> tepsipakki: he currently did a separate package for the kernel module; this would need to go into linux-source directly of course, but right now it's good enough for testing
[03:05] <tepsipakki> pitti: not that I know of, but it was discussed recently
[03:05] <mh21> tepsipakki: http://mh21.piware.de/git
[03:05] <tepsipakki> there were two people who volunteered to maintain it
[03:06] <tepsipakki> mh21: you could mail debian-x about that :)
[03:06] <tepsipakki> I'm off now ->
[03:06] <ogra> tepsipakki, well, it seems a tad buggy to me ... my ltsp clients use to create tpmfses for every file thy need writeable ... intrestingly only a bunch of them is ther and a bunch isnt ...
[03:07] <ogra> seems everything mounting related changes completely with nfsroot in nfs4
[03:08] <ogra> but i cant find a scheme why some tmpfs mounts work and some dont :/
[03:10] <StevenK> Right. Now that my Feisty machine has booted with 2.6.17, to find out how to make 2.6.20 boot.
[03:16] <StevenK> Ah ha. It wants irqpoll
[03:17] <robertj> are items included in universe considered "part of the normal operating system distribution" for licensing purposes?
[03:18] <pitti> robertj: we do not 'ship' those in the sense of 'push the bits to the users via CD'
[03:20] <robertj> pitti: so would linking against openssl be kosher only if it was included on the cd?
[03:21] <pitti> robertj: no, I don't think so
[03:21] <pitti> robertj: that distinction makes mainly sense for patent issues, not for GPL compatibility
[03:21] <robertj> pitti: the openssl faq seemed to indicate otherwise
[03:22] <robertj> "On many systems including the major Linux and BSD distributions, yes (the GPL does not place restrictions on using libraries that are part of the normal operating system distribution)."
[03:22] <pitti> robertj: ah; IANAL enough to interpret this properly, I'm afraid
[03:24] <robertj> pitti: debian-legal seemed to take a pretty strict stance against it in '01 or so, is their opinion gospel or does Canonical have its own suits?
[03:25] <pitti> robertj: so far I didn't see anyone care in Ubuntu, but we generally follow Debian's lead except when we explicitly discuss things and decide to deviate for some good reasons
[03:25] <robertj> netatalk is pretty useless as-is. No encrypted passwords, and no passwords longer than 8 characters
[03:25] <pitti> we don't have 'our own suits' so far at least
[03:26] <pitti> robertj: it doesn't work with gnutls?
[03:27] <robertj> pitti: AFAIK no. Mandriva seemed to claim they linked against gnutls but I looked & didn't see that happening
[03:28] <Keybuk> personally I think debian-legal are far too strict in this case
[03:28] <Demon012> #ubuntu-motu
[03:28] <Demon012> oops
[03:29] <pitti> neither we nor Debian are really consistent in that regard; e. g. a whole lot of GPL apps link against libpq5, which links against openssl
[03:29] <robertj> Keybuk: my suspicion is that for this particular case, rewriting would be cheaper than paying a decent lawyer to even look it over, but over the whole of main & universe that may not be true
[03:29] <robertj> (est. ~16hrs implementation by someone who had any prior experience in C)
[03:30] <pitti> robertj: I figure porting to gnutls should be much easier?
[03:30] <robertj> pitti: that is what I meant
[03:30] <robertj> pitti: I think there are some unimplemented features in the gnutls layer, mostly data structures
[03:31] <Keybuk> a) libssl is a dependency of ubuntu-minimal, so any reasonable interpretation of "the normal operating system distribution" would appear to apply here
[03:31] <Keybuk> b) the GPL doesn't cover dynamic linking (personal opinion :p)
[03:32] <robertj> Keybuk: mind if I attach point a to a launchpad bug on this particular package
[03:32] <iwj> The status of GPL-incompatible programs dynamically linked against GPL'd code is disputed.  Some (including Keybuk it seems) say that there is no problem.  Others (including me) say that it's forbidden unless the system library exception applies.
[03:32] <Keybuk> "attach point" ?
 a) libssl is a dependency of ubuntu-minimal, so any reasonable interpretation of "the normal operating system distribution" would appear to apply here
[03:33] <robertj>  b) the GPL doesn't cover dynamic linking (personal opinion :p)
[03:33] <robertj> (well just meant to be a) and not b) but anyway)
[03:33] <Keybuk> robertj: *shrug* it's not anything useful for a bug, just this developers opinion
[03:33] <iwj> But NB that the system library exception (what robertj refers to above) definitely cannot help when it's all on the same CD.  (Whether it can help if the components are together in the archive is unclear.)
[03:34] <Keybuk> why does the system library exception not count if it's on the same CD?
[03:35] <iwj> The system library exception is GPLv2 s3 2nd main para (after bullet (c)), 2nd half `However ...', right ?
[03:35] <iwj> But it says `... unless that component [the GPL'd library]  itself accompanies the executable [the GPL-incompatible bit] '.
[03:35] <Keybuk> right, and that only counts for distributing source code of one's program that derives from something included in the base
[03:35] <Keybuk> since we'd be distributing source code here, I don't see how that applies
[03:35] <iwj> So if you put both on the same CD they `accompany' each other.
[03:36] <Keybuk> that paragraph allows you to copy and paste code from glibc into your program
[03:36] <Keybuk> and then not distribute the source to that program
[03:36] <iwj> Err, distribution of just the source is not AFAICT restricted (unless some weird situation exists).  The potential violation is in the distribution of binaries.
[03:37] <iwj> Keybuk: That paragraph was originally intended to allow GPL'd code to be ported to systems with GPL-incompatible libcs.
[03:37] <Keybuk> iwj: true, but it's also worded loosely enough to allow the opposite
[03:38] <iwj> ion_: It's a left single quote too, in ASCII.  Just because ISO-646 isn't compatible with ASCII isn't my fault :-).
[03:38] <Keybuk> it certainly doesn't apply to the OpenSSL case
[03:38] <iwj> Keybuk: Indeed it applies to the opposite.
[03:38] <ion_>  :-)
[03:38] <Keybuk> iwj: IRC's standard character set is not ASCII or ISO-646 ;)
[03:38] <iwj> Keybuk: UTF-8 is defined as the same as ISO-646 for the codepoint `.  And ISO-646 was allegedly compatible with ASCII.
[03:39] <iwj> So you newfangled people are supposed to be compatible with my usage.  Go fix your standard :-).
[03:39] <Keybuk> iwj: UTF-8 is not the standard character set for IRC either
[03:39] <iwj> What character set do you claim to be using ?
[03:39] <ion_> So fonts should render ` as a ? What should one do when she actually wants a grave accent? :-)
[03:39] <Spads> the standard character set for IRC is now PETSCII.  I have decreed it.
[03:39] <Keybuk> IRC uses a bizarre finish character set
[03:39] <iwj> I bet it claims to be compatible with ASCII for the codepoint `.
[03:40] <Keybuk> this is demonstrable by the fact you can message {PUPPETS}Gone instead of [PUPPETS] Gone
[03:40] <Hobbsee> Spads: hah.  good luck with that
[03:40] <ogra> root@edubuntu:~# mount|grep fstab
[03:40] <ogra> tmpfs on /etc/fstab type tmpfs (rw)
[03:40] <ogra> root@edubuntu:~# cat /etc/fstab
[03:40] <ogra> # UNCONFIGURED FSTAB FOR BASE SYSTEM
[03:40] <ogra> root@edubuntu:~# echo "blupp" >/etc/fstab
[03:40] <Keybuk> err, s/Gone/Gonzo/
[03:40] <ogra> root@edubuntu:~# cat /etc/fstab
[03:40] <Keybuk> :p
[03:40] <ogra> # UNCONFIGURED FSTAB FOR BASE SYSTEM
[03:40] <ogra> GRRRR !!!!!
[03:40] <ion_> IRC originally used the bizarre Finnish character set, but AFAIR the spec doesnt actually specify anything else than that it is just octet streams.
[03:40] <iwj> For OpenSSL you can make the system library exception apply if `that component' (eg, OpenSSL) doesn't accompany `the executable' (here, the GPL'd library).
[03:40] <Keybuk> iwj: yes, but what good does the system library exception grant you?
[03:41] <Keybuk> the system library exception ONLY grants you the ability to not distribute sourece to either
[03:41] <iwj> Keybuk: It permits you to fail to licence the `whole work' (actual program + openssl + gpl lib) under GPLv2.
[03:41] <Keybuk> iwj: no it doesn't
[03:41] <Keybuk> it's very explicit in what it permits
[03:41] <Keybuk>  However, as a
[03:41] <Keybuk> special exception, the source code distributed need not include
[03:41] <Keybuk> anything that is normally distributed (in either source or binary
[03:41] <Keybuk> form) with the major components (compiler, kernel, and so on) of the
[03:41] <Keybuk> operating system on which the executable runs, unless that component
[03:41] <Keybuk> itself accompanies the executable.
[03:41] <Keybuk> -- 
[03:42] <Keybuk> it permits you to not need to distribute the compiler, libc, etc. with your source
[03:43] <Keybuk> (headers for those)
[03:43] <iwj> `The source code distributed need not include anything [OpenSSL]  that is normally distributed ... unless [OpenSSL]  accompanies the executable'
[03:43] <Keybuk> no
[03:43] <Keybuk> read the whole paragraph
[03:44] <iwj> I have read it (obviously).
[03:44] <Keybuk> it defines "source code" immediately above it
[03:44] <iwj> Yes.
[03:44] <Keybuk> and what it defines it as is not compatible with [OpenSSL] 
[03:44] <iwj> You think the source to OpenSSL isn't part of the `source code for a work' ?
[03:44] <geser> Seveas: could you fix the LP url for Daniel at https://launchpad.net/~ubuntumembers/+poll/cc2007-dholbach/ ? the c and h are transposed
[03:44] <Keybuk> yes
[03:44] <iwj> That depends on what you think `work' is.
[03:45] <Keybuk> unless you are arguing that OpenSSL is a "module contained" by the binary?
[03:45] <iwj> But if you are right then the system library exception is irrelevant since you're home free anyway.
[03:45] <iwj> I contend that `the work' is libssl + /usr/bin/program + libc + libgplthing
[03:45] <Keybuk> I content that the GPL doesn't say that
[03:45] <iwj> Yes, I know.
[03:45] <Keybuk> The source code for a work means the preferred form of the work for
[03:45] <Keybuk> making modifications to it.  For an executable work, complete source
[03:45] <Keybuk> code means all the source code for all modules it contains, plus any
[03:45] <Keybuk> associated interface definition files, plus the scripts used to
[03:45] <Keybuk> control compilation and installation of the executable.
[03:46] <iwj> You said that earlier./
[03:46] <iwj> Please stop pasting the GPL into channel.  I have it over -> there.
[03:46] <ogra> my OS ignores me :(
[03:46] <Keybuk> so is OpenSSL a module it contains?
[03:46] <iwj> I asserted that the status of this situation was disputed, which you must agree with ?
[03:46] <Keybuk> oh, I entirely agree it's disputed
[03:47] <iwj> And this dispute divides the two of us and I was carefully trying to avoid having to have it here.
[03:47] <iwj> But I also asserted that the system library exception is no help except in some subtle cases.
[03:47] <Keybuk> personally I believe that use of a library through its published interfaces cannot possibly be considered anything other than "use" of that library, so not covered by copyright
[03:47] <iwj> Sure you must agree with that ?
[03:47] <iwj> In particular it is no help in this case.
[03:47] <iwj> The reason you disagree with me is nothing to do with the system library exception.
[03:47] <Keybuk> I don't think the system library exception even applies without my interpretation
[03:47] <Keybuk> correct
[03:48] <pygi> oh well
[03:48] <iwj> You think the system library exception is pointless because the preconditions don't apply and I think it is useless because if the preconditions apply you usually find that the `accompany' bit bites you.
[03:48] <siretart> whats the problem? cdrtools?
[03:48] <Keybuk> but I also disagree that even assuming the "traditional" application of the GPL to dynamic linking, that the system library exeception doesn't apply here anyway
[03:48] <iwj> You disagree that it doesn't apply ?  Eh ?
[03:48] <iwj> Do you mean you think it applies or doesn't ?
[03:49] <Keybuk> since OpenSSL is none of the things listed in the paragraph where the exception to not distribute them is granted
[03:49] <iwj> I think you have misunderstood what you describe as `the "traditional" application of the GPL to dynamic linking'.
[03:49] <Keybuk> no, I don't think I have
[03:50] <pygi> siretart: you've got a pm :)
[03:50] <Keybuk> I think that if the GPL covers dynamic linking, then the work is inherently a derivation of OpenSSL
[03:50] <Keybuk> so OpenSSL is a consituent part of the work
[03:50] <iwj> The broad interpretation follows from the `work as a whole', 2nd main para of s2.
[03:51] <Keybuk> so thus isn't covered by the exception, which is only for auxiliary pieces needed to build it
[03:51] <iwj> I argue that if you distribute /usr/bin/program + openssl + libc + libgpl then the _combination_ is a `work as a whole'.
[03:51] <iwj> Now I agree that the wording isn't entirely clear and I can see that people might reasonably disagree.
[03:51] <Keybuk> I would agree there
[03:51] <iwj> But if you agree with that, then `work' in s3 means all of those four parts together.
[03:51] <Keybuk> and that those things are not covered by the exception
[03:52] <iwj> And OpenSSL is definitely contained in it.
[03:52] <Keybuk> yes
[03:52] <Keybuk> so work includes those parts
[03:52] <iwj> Right.
[03:52] <Keybuk> and the exception doesn't cover those parts
[03:52] <iwj> Oh, good, you seem to be agreeing with me now.  I mean, apart from the dispute about work as a whole.
[03:53] <iwj> I would say that you could use the system library exception for openssl but only if you don't distribute thje four pieces together.
[03:53] <iwj> Or at least there are some combinations where it might apply.
[03:54] <iwj> Eg, if you distribute only /usr/bin/program and rely on the system's libgpl and openssl.
[03:54] <iwj> If `you' isn't the OS distributor then you're probably covered by the exception.
[03:54] <Treenaks> *brain hurts*
[03:55] <iwj> And AFAICT you can distribute /usr/bin/program+libgpl even.  That's just the same as the FSF were doing with Solaris shellutils, really.
[03:55] <Keybuk> assuming you decide that OpenSSL is distributed "with" one of the major components like the compiler, or kernel, yes
[03:55] <iwj> Yes.
[03:55] <Keybuk> except then you get bitten because the compiler is distributed by the FTP archive
[03:55] <Keybuk> so arguably anything on the FTP archive is distributed "with" the compiler
[03:56] <iwj> Yes.
[03:56] <iwj> Which is why this exception is really hard to use _for the OS vendor_ which is entirely the idea AIUI.
[03:56] <iwj> Ie, the os library exception doesn't help Debian or Ubuntu.
[03:56] <Keybuk> ok, it sounds like we were arguing the same thing at different points, with different language :)
[03:57] <iwj> Good :-).
[03:57] <Treenaks> it only helps debian's or ubuntu's users, if they recompile $program with ssl support
[03:57] <iwj> Treenaks: Right.
[03:57] <Keybuk> (of course, I believe the above conclusion is bullshit, but it helps to be able to argue from other people's stand points)
[03:57] <iwj> And random 3rd parties.
[03:57] <Treenaks> (which is teh suck, as I want ubuntu, not gentoo, for a reason)
[03:58] <iwj> Keybuk: Well, yes.  But it's worth considering because _if_ the os library exception were useful it might avoid having to worry about `work as a whole'.
[03:58] <Treenaks> (I've been bitten by this thing quite a few times now.. *sigh* @ freeradius+{postgresql/libssl})
[03:59] <robertj> Treenaks: did you end up moving over to gnutls?
[03:59] <iwj> Treenaks: Welcome to politics, unfortunately.  Politics is about disagreements so obviously things turn out awkward.
[03:59] <iwj> Just be glad we're relatively civilised and that you don't have swordsmen or thugs settling these arguments nowadays.
[04:00] <jsgotangco> lol
[04:00] <Keybuk> the sword seems somewhat more humane than the court
[04:00] <Treenaks> robertj: freeradius + gnutls doesn't work, and still links in libssl through postgresql, which also doesn't quite work with gnutls, afaik
[04:01] <Keybuk> Treenaks: I have a more fun one
[04:01] <Keybuk> Debian haven't distributed a piece of my software because it "links with" OpenSSL
[04:01] <Treenaks> Keybuk: which piece of software?
[04:01] <Keybuk> live-f1
[04:02] <Treenaks> that still works? cool :)
[04:02] <Keybuk> except live-f1 doesn't link with OpenSSL
[04:02] <Keybuk> it links with the GPL'd libneon
[04:02] <Keybuk> which Debian happen to build linked against OpenSSL
[04:03] <Keybuk> I offered a Debian-specific exception to allow linking against things linked with OpenSSL
[04:03] <Keybuk> they didn't like that idea much either :)
[04:03] <Treenaks> which makes it debian-specific, which isn't allowed either.. argh :)
[04:04] <iwj> Why not NMU libneon to build against gnutls ? :-)
[04:04] <iwj> `fixes RC bug'
[04:05] <root___> exit
[04:05] <Keybuk> because that breaks things, iirc
[04:05] <root___> quit
[04:05] <iwj> Oh, how annoying.
[04:05] <Keybuk> I had a fun suggestion
[04:05] <Treenaks> iwj: because it might conflict with other (BSD-licensed) packages, that are linked to libneon? :P
[04:05] <Keybuk> since it is generally believed that the GPL *doesn't* cover use of command-line programs
[04:05] <Keybuk> write a program that exposes all of the OpenSSL API via a command-line tool
[04:05] <Treenaks> Keybuk: like.. 'openssl'!
[04:05] <Keybuk> and "thunks" all usage of OpenSSL through that helper
[04:05] <iwj> My laywers will see that dodge in court.
[04:06] <Keybuk> they might
[04:06] <iwj> Well, actually, they won't.  The Qt people already did it with dpkg and the FSF were wimps.
[04:06] <Keybuk> exactly
[04:06] <Keybuk> because the FSF already encountered this
[04:06] <Keybuk> when someone had their proprietary debugger used by emacs ;)
[04:06] <iwj> I have no idea why the FSF were wimps.  I think it's because RMS is stuck in the 1970s when `program' meant `executable text with a single symbol table'.
[04:07] <iwj> Keybuk: They had this with GCC and Objective C, didn't they ?
[04:07] <Keybuk> could be
[04:07] <iwj> They put the boot in there.  I don't think they've thought about these things clearly because (see above).
[04:07] <Keybuk> I agree
[04:08] <Keybuk> personally I think if you stick to the interfaces exposed by a library, then you're "using" not "deriving"
[04:08] <Keybuk> and if you add your own interfaces to satisfy your program, you're clearly deriving
[04:08] <Keybuk> likewise for command-line apps
[04:08] <Treenaks> Keybuk: what about _using _internal _functions()
[04:08] <iwj> TBPH I wouldn't be at all surprised if a court decided that Ubuntu was a `work as a whole'.
[04:08] <iwj> Which would be quite a PITA.
[04:08] <Keybuk> Treenaks: internal functions are arguably not part of the published interface, so derivation, depending on author preference
[04:08] <Keybuk> iwj: quite
[04:09] <iwj> I think that's the plain meaning of the text but custom and practice dictate otherwise.  Which leaves grey areas, well, grey.
[04:09] <iwj> The previous GPLv3 draft had a much better approach to this IIRC.  Unfortunately the latest one is an enormous mess.
[04:34] <[knap] > can anyone take a look at this bug report related to alsa? https://bugs.launchpad.net/ubuntu/+source/alsa-utils/+bug/92889
[04:34] <ubotu> Malone bug 92889 in alsa-utils "Low sound volume" [Undecided,Confirmed]  
[04:40] <crimsun> [knap] : a fix for that has been committed (some time ago, in fact - Wed Apr 4 23:50:32 2007 -0400)
[04:40] <crimsun> [knap] : it's also incorrectly filed against alsa-utils, but I'll triage it.
[04:41] <[knap] > crimsun i first noticied this bug in feitsy herd5 now i'm using feitsy final and this bug still exists
[04:41] <crimsun> [knap] : that's because the fix is not present in the kernel shipped with 7.04. It will be in a post-release kernel update.
[04:42] <[knap] > ok, can i do anything to fix it till the new kernel is shipped?
[04:42] <crimsun> [knap] : pastebin the output from ``lspci -vvn'' for me, and tell me the url
[04:43] <[knap] > ok
[04:43] <crimsun> [knap] : also, let's migrate to #ubuntu - this channel is strictly for development, not support.
[04:43] <[knap] > ok
[04:49] <pitti> seb128, Riddell: now that we have avahi by default, do you see any reason to keep the control center patches for disabling/enabling it?
[04:49] <pitti> seb128, Riddell: I'm merging avahi and wonder whether it is worth the trouble to keep enable_avahi and check_avahi
[04:49] <seb128> pitti: no
[04:50] <Riddell> pitti: might people not want an easy way to turn it off?
[04:50] <seb128> Riddell: do you have a services manager with KDE?
[04:50] <pitti> Riddell: you can still turn it off in /etc/default/avahi, question is why normal people would want to
[04:54] <Riddell> seb128: yes, that's a good alternative 
[04:55] <Riddell> pitti: I don't mind really
[04:56] <pitti> Riddell, seb128: ok, so let's make all our lives a bit easier and just drop this, shall we?
[04:56] <seb128> pitti: WFM
[04:58] <jwendell> seb128, hi!
[04:58] <seb128> hey jwendell
[04:59] <jwendell> seb128, can i set my repositories to gutsy in order to keep my work on a up-to-date system?
[04:59] <seb128> yes you can
[05:00] <seb128> it's not going to be stable though
[05:00] <seb128> lot of things are going to change quickly once it's open and we start syncs on Debian and new uploads ;)
[05:00] <pitti> now with etch being out of the door, the unstable merge will bring some new good (and bad) crack :)
[05:01] <ogra> yeah
[05:02] <Riddell> pitti: ok
[05:02] <hunger> When is the merge about to begin?
[05:03] <hunger> So when will g* hit the archive? After your conference I guess?
[05:09] <ogra> hunger, break it manually ? 
[05:09] <ogra> dont always rely on us to brek your box please !
[05:09] <ogra> *break
[05:11] <zul> hunger: do something interesting like take the ram out of the box and see if it boots
[05:11] <hunger> ogra: You are the professionals! I can only break stuff in a pretty amateurish way. And besides: If I break it, I know where to look to fix stuff:-)
[05:13] <ogra> hunger, just post your IP and ssh credential for your box ... then :P
[05:13] <pygi> ogra: you don't even need more then IP :P
[05:14] <hunger> ogra: I might use unstable ubuntu distributions on a "production box", but I am not that stupid!
[05:14] <Mithrandir> hunger: we had a small hiccup with gcc.  Assuming that builds fine we're looking at opening gutsy properly on Wednesday, I think.
[05:14] <ogra> heh
[05:14] <Mithrandir> maybe a bit earlier, but realistically not.
[05:15] <pygi> yay, by then I should have packages ready  :-D
[05:15] <muszek> hi... can someone please tell me where is Ubuntu Open Week held?  Some IRC channel?
[05:15] <ogra> yes, should be written on the wikipage
[05:15] <pitti> muszek: https://wiki.ubuntu.com/UbuntuOpenWeek
[05:16] <muszek> pitti, ogra: it's not written there
[05:16] <harrisony> muszek: /j #ubuntu-classroom
[05:16] <pitti> muszek: 'joining in'
[05:16] <harrisony> and /j #ubuntu-classroom-chat NOW! jono is talking as we speek
[05:16] <muszek> harrisony: thanks
[05:16] <harrisony> *speak
[05:18] <ogra> muszek, its written here :) https://wiki.ubuntu.com/UbuntuOpenWeek/Rules
[05:18] <muszek> ogra: yep, I missed that.  Thanks and sorry.
[05:18] <ogra> (behind the bold link that says "please read" :) on the other page)
[05:20] <sharms> the openweek wiki entry is not intuitive at all
[05:20] <sharms> it would be much better to have the irc address on the main page
[05:20] <sharms> since inevitably everyone needs it
[05:21] <ogra> well, the whole text of the "please read" page should go on the first page right at the top imho
[05:22] <ogra> but jono will have had his reasons to put it like that 
[05:22] <sharms> thats what I figured so I left it alone
[05:24] <ogra> did the old mozilla get dropped completely in feisty ? 
[05:24] <LaserJock> I believe so
[05:24] <ogra> sbalneav, ^^^
[05:36] <ogra> could some KDE user check if its ture that gpm breaks kde like described here ?: https://lists.ubuntu.com/archives/ubuntu-users/2007-April/112362.html 
[05:36] <ogra> *true
[05:36] <nixternal> I will ogra 
[05:37] <ogra> thanks
[05:38] <nixternal> so the problem is, gnome-power-manager working/trying to work in kde, now I wonder if Guidance is also trying to work at the same time
[05:41] <ogra> ergh
[05:41] <ogra> you could have gnome with just installing g-p-m :)
[05:42] <ogra> g-p-m adds itself to /etc/xdg/autostart ... i assume KDE reads that
[05:45] <sbalneav> LaserJock: Who must I bribe to get it put back in Universe?
[05:45] <LaserJock> sbalneav: man, all about the bribery ;-)
[05:45] <sbalneav> I am willing to offer much beer and meat.
[05:45] <harrisony> and beer
[05:46] <LaserJock> sbalneav: maybe shoot an email to ubuntu-motu with the reasons you need it.
[05:46] <sbalneav> ubuntu-motu@ubuntu.com?
[05:46] <LaserJock> lists.ubuntu.com
[05:46] <sbalneav> ok
[05:46] <poningru> sbalneav: gnomefreak is working on getting seamonkey uploaded
[05:47] <sbalneav> I'll have to subscibe, I suspect
[05:47] <gnomefreak> not atm im not
[05:47] <poningru> keep in mind after 1.7.3 support for mozilla suite was dropped by mofo
[05:47] <poningru> or I guess not?
[05:47] <poningru> wth
[05:47] <gnomefreak> i will be uploading it as soon as i can
[05:47] <poningru> gnomefreak: is anyone working on that?
[05:47] <gnomefreak> poningru: i have to respin it for gutsy first it may be a couple of weeks
[05:47] <gnomefreak> im upgrading my chroot atm
[05:48] <gnomefreak> poningru: there are some merges that need to be done before iceape shows up in gutsy
[05:49] <poningru> oy ve
[05:49] <poningru> again with the iceape
[05:49] <gnomefreak> poningru: thats what it will be
[05:49] <poningru> yeah yeah I know
[05:52] <asac> poningru: so what is your point?
[06:16] <groo_> hi/2 all, any ubuntu dev awake? :)
[06:16] <nixternal> ogra: OK, I have it setup, and I don't see gnome-power-manager running at in when booted into Kubuntu
[06:16] <nixternal> I wonder if it is when I might switch users, and have one logged into KDE and one into GNOME
[06:18] <groo_> anyone is able to add a usb epson printer with kprinter?
[06:18] <groo_> kprinter has the local printer greyed out :(
[06:19] <nixternal> groo_: you have stumbled upon a devel channel, for support you need to try out #kubuntu or #ubuntu
[06:19] <nixternal> our devs are currently taking naps after the Feisty release :)
[06:19] <Zober> hey i know this isnt the right chat for this, but i am desperate need of assistance with VB6 for a final exam/project.  Can anyone help me with MDI forms in VB6 please?
[06:20] <ogra> nixternal, pfft, naps are for wimps ... we prepeare to break the world atm :)
[06:20] <pitti> Zober: this is about as far away from ubuntu development as you can get
[06:20] <groo_> nixternal: its a develop problem.. usb printer support is broken, also the libgphoto timeout is set too low (500ms)...
[06:20] <asac> Zober: ... this is not a wrong place to ask that, but a *bad* place
[06:20] <nixternal> ogra: I am waiting to break the world
[06:20] <nixternal> pitti: hahaha, I froze, I didn't know how to respond
[06:20] <Zober> i cant figure out how to define form sizes for MDI child objects, it seems to define them based on the size of the MDI form
[06:20] <Zober> I know guys, i am so sorry, the VB channel is lazy and no one knows what they are talking about
[06:21] <pitti> Zober: try using glade instead :)
[06:21] <Zober> I promise im normally an all out linux guy, but this is for a class i am required to take and VB6 is as retarded a language as can be
[06:22] <BenC> How do I request that a package in mom be reverted and sycned directly with debian, trashing all of our ubuntu changes?
[06:22] <Zober> >_< Professor knows nothing about any other programming other than VB so he forced us all to use VB6 to write a final project >_<
[06:22] <ogra> yeah, gtk/glade works fine on windows i heard
[06:22] <thom> Zober: please stop, you're so far off topic it's not funny
[06:22] <ogra> python as well ...
[06:23] <nixternal> Zober: I have to use VB.NET for a class I am taking, so I understand the grief
[06:23] <harrisony> !offtopic
[06:23] <ubotu> #ubuntu is the Ubuntu support channel, #ubuntu+1 supports the development version of Ubuntu and #ubuntu-offtopic is for random chatter. Welcome!
[06:23] <Zober> nixternal, have you had to deal with MDI forms? and child objects?
[06:23] <nixternal> #ubuntu-offtopic and we can chat
[06:24] <Zober> ok, once again guys sorry to ask questions so off topic
[06:44] <cjwatson> BenC: requestsync(1)
[06:45] <cjwatson> except maybe not so much with the (1) since it has no man page. /me glares at pitti
[06:45] <BenC> cjwatson: I was just about to say :) thanks
[06:45] <cjwatson> BenC: in any case, https://wiki.ubuntu.com/SyncRequestProcess
[06:55] <keescook> cjwatson: the publisher is on automatic again?  (i.e. will security updates go through without me needing to bug someone?)
[06:56] <pitti> cjwatson: manpage> *cough*, noted
[07:26] <cjwatson> cjwatson@drescher:~$ sudo -u lp_publish crontab -l | grep cron.daily
[07:26] <cjwatson> 3 * * * * /srv/launchpad.net/codelines/current/cronscripts/publishing/cron.daily
[07:26] <cjwatson> keescook: ^-- yes
[07:27] <keescook> cjwatson: okay, thanks.  :)
[08:26] <Seveas> geser, lp sez: Only polls that are not yet opened can be edited. As soon as a poll opens it can't be edited anymore.
[08:27] <robertj> iwj: so err, I read the scrollback, went home, took a nap, reread the scrollback, and it makes a bit of sense...so is there a consensus on whether we can link against libopenssl if its not shipped on the CD?
[08:28] <robertj> (it being the program linking against it, we assume openssl will be on the cd)
[08:29] <Mithrandir> robertj: we ship DVDs too, so please don't assume anything based on what happens to be on the CDs today.
[08:30] <robertj> Mithrandir: well that could be fixed by having an alternate version of the package available in multiverse right?
[08:31] <Mithrandir> robertj: no, if you have two free programs with incompatible licences where one links to the other it's undistributable, not non-free.
[08:33] <robertj> Mithrandir: I think its a bit more complicated than that here, read the scroll back from 9:17-10:08
[08:33] <robertj> (EST)
[08:34] <XOP> UBUNTU UR MUM
[08:36] <Mithrandir> oops
[08:36] <Mithrandir> there
[08:36] <Mithrandir> robertj: why do you think so?
[08:37] <psusi> these days X internally handles fonts instead of using the font server right?
[08:37] <Mithrandir> psusi: lots of apps uses client-side fonts, yes.
[08:38] <psusi> do gnome apps render true type fonts themselves rather than have the X server render them?  and non X apps like emacs just let the X server render the non true type fonts?
[08:38] <divrapier> hello. can a developer guess, why ubuntu 6.0.6 wont boot, and stop after selecting 'start or install' on a pentiumD805 with asus p5b?
[08:39] <mrsno> im not a developer but problems with the p965 chipset most likely divrapier 
[08:40] <mrsno> support questions aren't meant for here really so check in #ubuntu  , but dapper doesn't have a new enough kernel to boot 100% with those chipsets to the best of my knowledge
[08:40] <Mithrandir> psusi: yes, AIUI.
[08:40] <harrisony> divrapier: #ubuntu
[08:40] <sharms> should I file a bug on xorg for a memory leak?
[08:40] <sharms>  4923 root      15   0  916m 768m  18m S    5 37.9 342:52.82 Xorg          
[08:40] <psusi> I see... hrm....
[08:41] <psusi> sharms: what makes you think it is leaking?
[08:41] <sharms> not sure intended behavior is to take up 3/4 a gig of memory?
[08:41] <sharms> call me crazy..
[08:41] <Chipzz> sharms: you're crazy
[08:41] <psusi> do you have a half gig of video ram?  a leak means it grows over time
[08:41] <Chipzz> sharms: those numbers are hardly accurate
[08:42] <Mithrandir> sharms: "X leaks memory" would be a useless bug anyway, you need to spend time to track down what causes a leak, for instance.
[08:42] <sharms> Chipzz: what do you suggest I use?
[08:42] <Chipzz> errr
[08:42] <Chipzz> what are these tools called
[08:42] <Chipzz> there are definately tools that will give you a better idea
[08:42] <Chipzz> but relying on top is stupid
[08:43] <Mithrandir> valgrind with various skins, pmap.
[08:43] <Chipzz> and you're very likely to be plain incorrect when you rely on numbers from top to make such a statement ;P
[08:44] <Chipzz> which is not to say X doesn't leak memory, it may very well do so; but most likely not for the reasons you think it is leaking memory
[08:44] <sharms> then I should file a bug against top, since nowhere in the manpage does it say ignore information top provides?
[08:45] <robertj> Mithrandir: if I'm following the line of discussion properly, since openssl is customarily distributed with ubuntu as a dependency of -minimal it might fall under the system library exception
[08:45] <psusi> top is fine... it's just that most people panic when they see X using a lot of ram like you just did, when most of it is simply the video ram on the card
[08:45] <Chipzz> sharms: no you shouldn't
[08:45] <Chipzz> sharms: there is no bug
[08:45] <Chipzz> (most likely)
[08:45] <Chipzz> sharms: but for example open gnome system monitor and take a look there
[08:46] <Chipzz> it will give you very different figures
[08:46] <psusi> huh?  they should give the same numbers?
[08:46] <Chipzz> psusi: gnome system monitor is showing you different things, hence different numbers
[08:46] <sharms> gnome-system-monitor provides the same information as top.
[08:46] <Chipzz> psusi: gnome system monitor is showing you "Writable Memory", which *is* what you actually care about
[08:46] <psusi> aye
[08:47] <Chipzz> sharms: no it does not
[08:47] <sharms> Chipzz: how so? I am looking at it right now.
[08:47] <Chipzz> sharms: gnome system monitor gives you writable memory, while top gives you total memory
[08:47] <psusi> it says it is showing resident memory, which should be rss
[08:48] <sharms> I am looking at the RSS field in top and the memory field on gnome-system-monitor
[08:48] <Chipzz> RSS is 69MB here; writable memory is 59MB
[08:48] <psusi> they agree for me too
[08:49] <Mithrandir> robertj: and it could easily enough be argued that all packages in Ubuntu accompany each other.
[08:49] <psusi> that's fine chipz, but by default g-s-m doesn't show writable memory by default, it shows resident memory
[08:49] <psusi> which is the same as the res column in top
[08:50] <robertj> Mithrandir: although OpenSSL's faq seems to indicate that it doesn't support that reading
[08:51] <sharms> My writable memory is 749MB.
[08:51] <Chipzz> there were actually patches to gsm for "easier memory profiling"; for example, taking into account memory that different instances of the same library share etc
[08:51] <Mithrandir> robertj: doesn't support my reading or your reading?
[08:51] <Mithrandir> robertj: it is in any case irrelevant since they're not the ones who have software licenced under the GPL so they can hardly interpret the GPL on behalf of others.
[08:52] <robertj> Mithrandir: "On many systems including the major Linux and BSD distributions, yes (the GPL does not place restrictions on using libraries that are part of the normal operating system distribution)."
[08:52] <Chipzz> sharms: do you know what mmap'ing is for example?
[08:52] <psusi> Chipzz: howso?  pick one process to bill for the memory and don't count it in the others?
[08:53] <sharms> Chipzz: yes but I don't see how you are helping this situation
[08:53] <sharms> I would rather wait until someone with more xorg specific experience can clarify things
[08:54] <psusi> sharms: typically the problem is that the x server mmaps the video ram, so it looks like it is using a lot of memory, when most of it is on the video card, not main ram
[08:54] <Chipzz> sharms: the video memory is mmap'ed into the Xorg server for example
[08:54] <Chipzz> which is why it appears to consume so much memory
[08:55] <johanbr> sharms: http://lists.freedesktop.org/archives/xorg/2007-March/022219.html
[08:56] <Mithrandir> robertj: at least I, and I believe the other archive administrators agree with me, that we work along the stricter interpretation I showed above.
[08:57] <sharms> johanbr: I understand that, but this guys RSS is 56m, mine is 769m
[08:58] <sharms> and I can guarantee my card does not have 700 megs of memory
[09:01] <sharms> I will assume since I have another ati card in the system next to me with the same software versions that is not exhibiting this behavior that it is most likely the fglrx driver, since the other system is using the open source radeon driver
[09:01] <psusi> sharms: how much does it use after a reboot is the important question
[09:02] <psusi> as an aside, I am very pleased that the feisty livecd correctly supports 3d accel on my radeon x850 with the open source driver
[09:03] <psusi> mesa seems to be making progress
[09:04] <sharms> ok
[09:04] <sharms> X uses 26m RSS at start
[09:04] <sharms> can anyone repaste my lines from above with the last numbers?
[09:06] <psusi> oh wow....
[09:06] <psusi> sharms>	 4923 root 15 0 916m 768m  18m S 5 37.9 342:52.82 Xorg 
[09:06] <sharms> thanks
[09:07] <psusi> that's one hell of a leak
[09:07] <sharms> yeah I will see if i can reproduce it
[09:09] <deep> hi 
[09:10] <deep> I want to add a new package to Ubuntu
[09:10] <deep> wso2wsas is a java based middleware server 
[09:13] <deep> anybody out there . .?
[09:13] <ivoks> deep: #ubuntu-motu
[09:14] <deep> ivoks:thanks
[09:21] <blackskad> hi all, is there anyone here from the desktop team?
[09:28] <ProN00b> congrats to the devs, upgrade to fawn was really painless
[09:38] <Arby> oops
[09:38] <Arby> sorry wrong chan
[10:06] <psusi> was there a significant change to fonts between dapper and edgy?  like their format changed or something?