[00:36] <pipegeek> o.O  compiz, as configured in gutsy, breaks alt-shift-tab.  Worth a bug report?
[01:04] <bmk789> can someone tell me the best way to start getting into ubuntu development?
[01:05] <mjg59> bmk789: Developing under Ubuntu, or developing parts of the distribution?
[01:06] <bmk789> parts of the distro
[01:06] <mjg59> One of the easiest ways is to find fixes for bugs
[01:07] <mjg59> In the long run, you'll want to be able to upload things yourself, which will involve going through the MOTU process
[01:07] <bmk789> whats that like?
[01:07] <mjg59> https://wiki.ubuntu.com/MOTU has information on it
[01:07] <mjg59> The FAQ is probably a good starting point
[01:08] <mjg59> #ubuntu-motu may also be a good place to ask questions
[01:08] <bmk789> ok ill look into that, what language would be most beneficial to learn?
[01:10] <mjg59> Most of our code is in either C or Python
[01:11] <mjg59> Probably a larger body in C, but a moderate amount of the Ubuntu-specific code is in Python
[01:12] <bmk789> ok thanks for the help, ill try to grab some books on C
[01:32] <JohnKarahalis> Hello everyone, congratulations on release!
[01:35] <JohnKarahalis> is anybody here?
[01:36] <RAOF> JohnKarahalis: Yeah, just exhausted :P
[01:37] <StevenK> RAOF: And you didn't come last night. Hmph!
[01:37] <RAOF> StevenK: Yeah :(
[01:37] <ajmitch> neither did I, so it's ok :)
[01:37] <RAOF> ;)
[01:37] <StevenK> ajmitch: Yeah, but you have the excuse of flights. :-P
[01:38] <JohnKarahalis> If you don't mind me asking, is anybody in here a paid employee of Canonical?
[01:38] <ajmitch> there are rumoured to be some around
[01:38]  * ajmitch is not :)
[01:38] <JohnKarahalis> I'm looking to interview a professional software engineer as part of a school project. A short interview of just basic questions really.
[01:39] <tonyyarusso> I'd guess that there are people not employed by Canonical who are also software engineers too.
[01:40] <tonyyarusso> JohnKarahalis: (it's sleepy-time at Canonical right now)
[01:40] <JohnKarahalis> ooh haha right
[01:40] <JohnKarahalis> damn time zones
[01:40] <ajmitch> tonyyarusso: it is?
[01:40] <RAOF> People sometimes email such inquiries to the various mailing lists, but I'm not sure how often they get responded to.
[01:40] <slangasek> tonyyarusso: it's never sleepy-time at Canonical
[01:40] <JohnKarahalis> unfortunately, the project requires that the person be salaried
[01:40] <tonyyarusso> slangasek: well, close enough...
[01:40] <tonyyarusso> the sane ones sometimes sleep
[01:41] <slangasek> but "software engineer" isn't exactly my job description :)
[01:41] <JohnKarahalis> slangasek: do you work for canonical?
[01:41] <ajmitch> what is it, "release monkey"?
[01:41] <tonyyarusso> teehee
[01:41] <slangasek> JohnKarahalis: yes
[01:41] <Burgundavia> trained chimp?
[01:42] <tonyyarusso> slangasek: you should get some kind of cloak, btw
[01:42] <JohnKarahalis> slangasek: Awesome! So what is your job title?
[01:42] <slangasek> ajmitch, Burgundavia: wow, I can see y'all are grateful that Gutsy is out... :)
[01:42] <ajmitch> slangasek: sure am :)
[01:42]  * tonyyarusso wonders if there are any mirrors still functioning
[01:43] <JohnKarahalis> slangasek: Do you program, or are you involved in support and things like that?
[01:43] <slangasek> JohnKarahalis: Ubuntu Release Manager
[01:44] <slangasek> so I think they probably let me program some when we're not in a release crunch time, but I haven't had an opportunity to see this in practice yet. :)
[01:45] <ajmitch> I can't imagine that you'd be kept idle for a few months
[01:45] <slangasek> tonyyarusso: why?
[01:45] <tonyyarusso> slangasek: so we can tell who you're with ;)
[01:45] <JohnKarahalis> slangasek: Ahh I see, so not much coding for you I guess. How is it working for a FS company though (that question is just out of curiosity by the way, not for the interview :)
[01:45] <tonyyarusso> slangasek: b/c all the cool kids have one
[01:45] <bddebian> tonyyarusso: heh
[01:45] <slangasek> ajmitch: heh, certainly not; I just don't know that it would include much "software engineering"
[01:46] <slangasek> JohnKarahalis: well, it certainly involves some coding, both for keeping the release machinery running appropriately and for helping to fix bugs
[01:46] <bddebian> So if you are all bored, anyone want to help me with a problem with a package? :)
[01:47] <JohnKarahalis> slangasek: That's awesome. I hope to work for a company like canonical one day.
[01:47] <slangasek> as for "how is it" - it's a lot more fulfilling than some software work I've done in the past
[01:48] <slangasek> bddebian: what package?
[01:48] <JohnKarahalis> slangasek: That's exactly what I wanted to hear! It seems like it would be more fulfilling.
[01:48] <bddebian> You don't wannt know. :-)  gnome-breakout
[01:48] <slangasek> bddebian: ok, so what's the problem?
[01:49] <bddebian> I have fixed everything I wanted to except that the path for the levels isn't getting set.  LEVELDIR is supposed to be $(datadir)/gnome-breakout/levels and I have datadir=/usr/share/games but it's still getting /usr/share
[01:49] <JohnKarahalis> slangasek: Do you think you would be interested in conducting a short interview with me via email some time over the next 2-3 weeks? It would be a simple interview, with questions like "What is a typical work day like", "What languages and technologies do you use", "How big is the team you work in".  If you're too busy I completely understand, what with Gutsy and what not.
[01:51] <slangasek> JohnKarahalis: do you have an estimate of how long the questions take to answer?
[01:52] <slangasek> though, "what is a typical work day like" would be a hard one for me to answer, hum - you'd probably find other Canonical folks better able to answer that :)
[01:52] <CarlF1> should the alt installer be hitting archive.u.com even though I specified a host ?
[01:52] <JohnKarahalis> slangasek: I mean, I wouldn't expect that it would take you more than 10-20 minutes to answer the questions. All I basically have to do is write a 3 page paper on the interview, it wouldn't be hard to do that if I had some background information about yourself and some answers to a few additional questions.
[01:52] <CarlF1> it does hit the specified host, but not after a long delay given the stress on arc.u.c, sometimes causing an error
[01:53] <JohnKarahalis> slangasek: But if you can't that's fine. I've got fallback options (not as good as a Canonical employee, but....) :)
[01:53] <slangasek> JohnKarahalis: sure, I could swing 10-20 min for that; I do think you could probably catch a better-suited candidate if you hung around here for a while though, or tried back earlier in the day (US time)
[01:55] <JohnKarahalis> slangasek: Yeah, I think I'll try tomorrow morning also. But could I put you down on my list of possible candidates?
[01:55] <slangasek> sure
[01:55] <slangasek> slangasek@canonical.com
[01:55] <slangasek> bddebian: source package showing this behavior?
[01:56] <JohnKarahalis> slangasek: Wow, thank you very much! What is your name by the way, my prof wants to know the names of the people we are considering. If you'd rather I get that info by email I'll send one to you.
[01:56] <slangasek> CarlF1: depending on what you mean when you say you "specified a host", no, you shouldn't have to hit the central server with the installer; but for support please see #ubuntu as mentioned in the channel topic
[01:56] <bddebian> slangasek: Well the current package just leaves stuff in /usr/share.  I ended up having to autoreconf in order to get localedir to work.
[01:56] <slangasek> JohnKarahalis: Steve Langasek (in my IRC client info, fwiw)
[01:57] <JohnKarahalis> slangasek: Ahh, this is my first time with IRC. Thanks so much! For now I think I'm gonna upgrade to Gutsy. Thanks again, I hope to talk to you soon.
[01:57] <slangasek> cheers
[01:58] <slangasek> tonyyarusso: anyway, my /whois already tells you who I'm with, I'm with the IPv6 goons :)
[01:59] <TheMuso> Congratulations guys on another great release.
[02:00] <bddebian> Oh, we released? ;-)
[02:00] <TheMuso> heh
[02:01] <CarlF1> slangasek: on 'stuff like this' I can't trust what i get in #u for filing a bug report.  in this case, I have a feeling it is part of https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/117398
[02:02] <slangasek> bddebian: you need to give me a source package showing the current behavior that you're trying to fix; I'm not going to guess based on a separate version of the source packge
[02:02] <CarlF1> so if not here, and not #u, any other #chans?
[02:02] <bddebian> slangasek: http://www.bddebian.com/packages/debian/gnome-breakout/
[02:03] <slangasek> CarlF1: you're using preseeding?
[02:03] <CarlF1> yes
[02:03] <bddebian>  src/Makefile.am:           -DLEVELDIR=\"$(datadir)/gnome-breakout/levels\" \
[02:03] <slangasek> CarlF1: and you're going through a proxy?
[02:04] <CarlF1> yes
[02:04] <CarlF1> well, if apt-cacher is considered a proxy
[02:04] <bddebian> I wonder how many packages I can stack on mentors before I get a single response.. :-)
[02:05] <slangasek> CarlF1: usually apt-cacher is not normally run using a proxy configuration AFAIK
[02:07] <CarlF1> slangasek: that fits with how I had to set the preseed options
[02:07] <slangasek> CarlF1: honestly, I don't see how that's a bug; the syntax in that bug listed under "the only way I could get it to work were with these settings" look like the obvious and correct way /to/ get it to work
[02:09] <CarlF1> then maybe the bug is the description for http://packages.ubuntu.com/edgy/net/apt-cacher  "caching proxy system"
[02:10] <slangasek> CarlF1: it is a caching proxy; but it's not an http proxy, it's an apt proxy
[02:10] <slangasek> i.e., it only caches and proxies .deb packages, not generic web content
[02:11] <CarlF1> given is is meant to be used for that it seems odd that the installer settings  would be "weird"
[02:12] <slangasek> CarlF1: they're not weird to me, but I'm the wrong person to judge because I've been a debian-installer developer for years
[02:13] <slangasek> CarlF1: effectively, the reason you configure it that way is because you're specifying your local apt-cacher as your mirror, which is precisely how apt-cacher is designed to function
[02:14] <CarlF1> i say weird because it dose what a proxy does, but instead of the installer using the proxy settings, I have to munge the host/dir names
[02:14] <slangasek> yes, because it's not an http proxy
[02:14] <slangasek> so if the documentation doesn't make it clear that "proxy" in the alternate installer refers to an http proxy, I think that's the bug
[02:15] <CarlF1> that would be a start
[02:15] <CarlF1> i'm assuming apt-cacher is better than an generic http proxly, right?
[02:16] <slangasek> hmm, I don't know that I would say it's "better" - it's a different use case
[02:16] <CarlF1> better for caching .deb s
[02:16] <slangasek> ok, then yes :)
[02:17] <slangasek> but the alt installer supports http proxies because some sites *require* all web traffic to go through a proxy
[02:17] <CarlF1> i see what you are getting at
[02:20] <CarlF1> The installers proxy settings propose aren't there to hit a caching proxy for speed, they are there for when it is the only way.
[02:20] <bddebian> Hmm, seems Mr. Langasek did one of the last uploads of stk too :-)
[02:21] <slangasek> stk?
[02:22] <slangasek> oh, pff, some package I uploaded in 2005?
[02:22] <bddebian> Heh
[02:23] <slangasek> CarlF1: well, if your site has a generic caching http proxy already, using it for .debs is fine too even if not mandatory, and probably saves you the trouble of setting up a separate apt-cacher
[02:24] <CarlF1> slangasek: cept I have only apt-cacher, no http proxy :)
[02:25] <slangasek> sure
[02:25] <CarlF1> I am adding a comment to the that basicaly says "not a bug"
[02:27] <ubotu> Launchpad bug 117398 in debian-installer "netboot install proxy apt-cacher" [Undecided,New]
[02:27] <bddebian> Frick, I hate source packages with tarballs :-(
[02:34] <slangasek> bddebian: I'm out for dinner, but I'm leaving gnome-breakout building and will deliver a diagnosis when I get back
[02:34] <bddebian> slangasek: Great, thanks!
[02:34] <StevenK> Diagnosis: "It's broken"
[02:35] <bddebian> No, I "fixed" it :-)
[02:35] <bddebian> the build system is fairly broken though :-(
[02:45] <bddebian> OK, f**k stk
[02:50] <slangasek> bddebian: um, when I build that source package from scratch, I get the correct path in the binary
[02:51] <bddebian> slangasek: The path is correct, I fixed that.  When it runs, it looks for the levels in /usr/share/gnome-breakout/levels, not /usr/share/games/...
[02:52] <bddebian> I am thinking the locales dir may be wrong at runtime too but since I don't know other languages, I'm not sure how best to test that.
[03:06] <Burgundavia> slangasek: as per my comment a few hours ago, we had to hold back while Gutsy was sitll in development ;)
[03:16] <bddebian> slangasek: Any thoughts or have I lost your help? :-)
[04:00]  * bddebian takes that as he's lost the help
[04:04] <Kano> hi, when do you set the default to gutsy on packages.ubuntu.com?
[04:30] <jdong> is it true that update-manager will automatically remove non-official repositories and also roll back or otherwise deinstall them? Or is that a Kubuntu/Adept feature? Or is it not a feature at all?
[04:32] <ScottK> jdong: It's supposed to.  I've helped one person today that had to remove some of the Automatix repositories manually.
[04:33] <jdong> ScottK: does it just remove the repositories in sources.list, or does it have a mechanism to remove thigns in Local/Obsolete, and/or downgrade packages back to official versions?
[04:33] <ScottK> AFAIK it just removes the repos.
[04:33] <jdong> ok
[04:34] <jdong> so users with unofficial stuff installed, I will still have to tell them to remove them
[04:34] <ScottK> I know for this release Automatix is recommending removing everything installed by Automatix and Automatix before upgrading.
[04:34] <ScottK> Yes.
[04:34] <ScottK> Getdeb makes the same recommendation.
[04:34] <ScottK> For their stuff.
[04:35] <bddebian> ScottK: And why did you help them? ;-P
[04:35] <ajmitch> recommendations & reality never tend to match up
[04:35] <ScottK> bddebian: I don't think I did.
[04:35] <bddebian> :-)
[04:35] <ajmitch> since it requires the users to be able to isolate the set of packages installed from elsewhere
[04:36] <ScottK> I took stuff that we could legally do and stuffed it into the official repos giving them fewer excuses to exist.  If that counts as helping, guilty as charged.
[04:49] <bddebian> ScottK: beauty :-)
[04:50] <Burgundavia> automatix seems to be very interested in helping
[04:50] <Burgundavia> I wonder if they are seeing less installs which each release?
[04:51] <ScottK> Burgundavia: I think they've had a "management restructuring".
[04:51] <StevenK> Burgundavia: That's an interesting point - are Automatix seeing less people using them due to us making an effort with things like easy codec installation, along with things like mjg59's analysis.
[04:52] <ScottK> The guy that appears to be running the show now at least has interest in cooperating and shows some inclination to listen when being told they're being stupid.
[04:52] <ScottK> StevenK: I think that's true.
[04:52] <Burgundavia> I wonder where arnieboy went
[04:52] <pwnguin> he's still there
[04:52] <ScottK> I'd like to see mjg59 do a followup of their Gutsy release to see if their claims "we fixed all that" are accurate.
[04:53] <StevenK> ScottK: Mail the tech board, I suspect the answer is, "Let's check to see."
[04:53] <Burgundavia> "WARNING: Automatix for Ubuntu 7.10 Gutsy will NOT be upgradeable from previous versions. You must either do a clean install of Gutsy or run Automatix in Feisty and uninstall everything before upgrading to Gutsy." <-- http://www.getautomatix.com/wiki/index.php?title=Main_Page
[04:53] <ScottK> The code to sigkill dpkg was at least commented out when I looked.
[04:53] <pwnguin> im sure mjg has plenty of other things to do ;)
[04:54] <ScottK> StevenK: I was sort of hoping he'd read the scrollback and be inspired.
[04:55] <ScottK> welcome back milli.
[04:55] <StevenK> ScottK: Heh
[04:56]  * ScottK ponders the text of his secvpn removal request.  Good technical terms for "This package is evil and must be destroyed."
[04:57] <StevenK> ScottK: "This package eats babies."
[04:58] <ScottK> Sounds about right (and does it using sudo - this part true).
[04:58] <pwnguin> ScottK: just susbcribe a young priest and an old priest
[04:59] <tritium> Every time you install this package...Please, think of the babies.
[05:00]  * StevenK chuckles
[05:00] <StevenK> ScottK: Are there technical reasons for why it should be destroyed?
[05:01] <ScottK> Yeah.  I'm not sure I'm knowledgeable enough to enumerate them, but the bottom line appears to be that the current package is unworkable and insecure and would have to be almost totaly redone to have a hope of working safely.  That's without looking at any of the upstream code.
[05:02] <ajmitch> StevenK: example - the initscript wants to write into /etc/inittab for start/stop
[05:02] <StevenK> *EW*
[05:03] <TheMuso> yuck
[05:03] <TheMuso> yuck
[05:03] <pwnguin> what about people who dont care about security but have to work with employers who do?
[05:03] <TheMuso> ok espeak says that weirdly
[05:03] <StevenK> ScottK: Okay, I think the way forward is to say the bottom line bit in a bug report, and ask pitti to do a short security review - if he finds major flaws, kill it.
[05:03] <StevenK> TheMuso: How does it pronounce it?
[05:04] <TheMuso> StevenK: yook
[05:04] <ScottK> StevenK: That sound like an excellent idea.  Thanks.
[05:04] <pwnguin> gootsy
[05:04] <TheMuso> gutsy is spoken correctly.
[05:04] <ScottK> I already got an opinion from keescook on secvpn adding itself to sudoers.  He wasn't a fan of that approach.
[05:05] <StevenK> Double ew
[05:06] <ajmitch> hence the slight reservations of having the package being installable
[05:13] <ScottK> Right.  Once I saw that I quit trying to fix it.
[05:30] <slangasek> bddebian: that was while I was still on the way to dinner...
[05:30] <slangasek> bddebian: anyway, it's doing the right thing at runtime too for me
[05:30] <bddebian> slangasek: Yeah, just got that with TheMuso.  Sorry to waste your time too :-(
[06:55] <dholbach> good morning
[07:04] <myrttiubuntu> how #ubuntu I computer no working
[07:04] <myrttiubuntu> plz help :)
[07:04] <myrttiubuntu> op #ubuntu-ops no help rude they broke computer no #ubuntu
[07:05] <Mithrandir> myrttiubuntu: that's offtopic for this channel though, this is a development channel, not a support channel
[07:05] <myrttiubuntu> i be like them they kick and no connect #ubuntu
[07:05] <myrttiubuntu> i can not connect #ubuntu
[07:05] <OpenSorce> the following review of Kubuntu 7.10 will be released to the public within the next 3 weeks.....not trying to be flamebait....just notifying you guys....my boss is sending an email to you guys as we speak http://bigcatlinux.com/kufailure.html
[07:06] <myrttiubuntu> people #ubuntu-ops no connect said they rude
[07:06] <Karnaugh> myrttiubuntu: I can't understand you, and nicknames like "ubuntusucks" is unlikely to get you any help
[07:06] <Burgundavia> OpenSorce: your review is incorrect
[07:06] <Burgundavia> OpenSorce: compiz-fusion is for Ubuntu, not Kubuntu
[07:07] <OpenSorce> Burgundavia, thanks for pointing that out I'll have it corrected
[07:07] <myrttiubuntu> Karnaugh i same nick as op but no connect #ubuntu, ops rude
[07:07] <Burgundavia> OpenSorce: the other thing you might want to change is that we don';t install prop. drivers for good reason
[07:07] <OpenSorce> Burgundavia, whatever
[07:07] <OpenSorce> Burgundavia, it either wors or it doesn't
[07:07] <Mithrandir> myrttiubuntu: you are offtopic here, please either stop being offtopic or I'll have to ask you to leave.
[07:08] <OpenSorce> *works
[07:08] <myrttiubuntu> all ubountu are rude
[07:08] <Burgundavia> OpenSorce: it makes no sense to argue with you, suffice it to say, that isn't going to change
[07:08] <OpenSorce> Burgundavia, not my job to get into the how and the why
[07:08] <OpenSorce> Burgundavia, okay :-)
[07:09] <Burgundavia> there, in fact, possible legal issues with shipping binary drivers
[07:09] <Mithrandir> OpenSorce: you should point out the multimedia keys didn't work with your keyboard.  They work just fine on my keyboard.
[07:09] <OpenSorce> Burgundavia, Mandriva does it
[07:10] <Burgundavia> OpenSorce: legal opinions differ on this matter
[07:10] <OpenSorce> Mithrandir, it's all taken from a sampling of one machine
[07:10] <OpenSorce> Burgundavia, okay
[07:10] <Mithrandir> OpenSorce: yes, and you don't say that, nor do you say what kind of hardware you have.
[07:11] <OpenSorce> Mithrandir, that gets added prior to final publication
[07:11] <Mithrandir> OpenSorce: I hope you are filing bugs about all the problems you have encountered as well
[07:11] <OpenSorce> This is a pre-edit copy, btw
[07:11] <Burgundavia> OpenSorce: I suggest you read this: http://www.ubuntu.com/community/ubuntustory/philosophy
[07:12] <OpenSorce> and there is an error I gave too many stars for the keyboard it will be 4 in the final release
[07:12] <OpenSorce> Burgundavia, it doesn't matter.....I am rating it for new user suitability....that holds the distro to a higher standard
[07:12] <Burgundavia> OpenSorce: binary drivers are not needed for average users
[07:13] <OpenSorce> Burgundavia, okay....either way I don't want to "troll" in here....just notifying you guys of the article
[07:14] <Burgundavia> as for your wireless, what kind of driver do you have?'
[07:14] <OpenSorce> Burgundavia, you are welcome to join #bigcat if you want to discuss this further.....I don't want to clog the channel with off-topic stuff
[07:16] <holycow> lol
[07:16] <holycow> dumbass
[07:17] <Burgundavia> be nice, holycow
[07:18] <ubuntunonakedgir> how fix?
[07:19] <ubuntunonakedgir> no help no #ubuntu how fix broken?
[07:22] <pitti> Good morning
[07:22] <Burgundavia> morning pitti
[07:24] <StevenK> Morning pitti
[07:30] <pizzadude> does anybody know the power output of the PCI bus?
[07:31] <ScottK> Burgundavia: Did you look at the other content on the site that "Journalist" pointed to?
[07:34] <OpenSorce> Burgundavia, I'm sorry did you ask me about the chipset in my wifi adapter?
[07:34] <Hobbsee> bye now.
[07:34] <Hobbsee> bloody tor
[07:34] <OpenSorce> Burgundavia, it's acx....which it seemed to think was a wired device
[07:35] <ScottK> OpenSorce: If you are a journalist, why are you putting your reviews on a web site of another linux distribution?
[07:36] <OpenSorce> ScottK, that's one of my personal websites....it's just where I uploaded for review
[07:36] <ScottK> OK.  It looks like that web site of a soon to be Linux distro.
[07:36] <OpenSorce> ScottK, it is.....I don't know if I'd use the term "soon" :-)
[07:37] <ScottK> Odd thing for a journalist doing independent distro reviews to have.
[07:37] <OpenSorce> ScottK, and it's probably a project I'll have to hand off to someone else since it may be a conflict of interest for me
[07:38] <OpenSorce> ScottK, I'm a geek...we make geeky things :-)
[07:38] <ajmitch> Hobbsee: you may want to try & get him out of other channels also
[07:38]  * Hobbsee watches her troll meter go off
[07:38] <ajmitch> :)
[07:38] <Hobbsee> ajmitch: i'm somewhat limited by the fact that we have no staffers around.
[07:38] <ajmitch> unfortunate
[07:38] <Hobbsee> and i dont have ops on every single ubuntu channel
[07:38] <ScottK> OpenSorce: It sounds like you've been working on it for some time.
[07:38] <Hobbsee> yeah well, it is australian day.
[07:39] <OpenSorce> ScottK, yeah....about oh...a week :-)
[07:39] <Mithrandir> ScottK,OpenSorce: please take this elsewhere, it's not related to development of Ubuntu.
[07:39] <OpenSorce> Mithrandir, I apologize
[07:39] <OpenSorce> ScottK, #bigcat is open
[07:39]  * ScottK will note that the domain was registered yesterday and shut up.
[07:40] <OpenSorce> Mithrandir, sorry I just came back to answer a question about my wifi chipset
[07:41] <Mithrandir> OpenSorce: sure, which is fine, but if (plural) you want to discuss further, please do so somewhere else.
[07:42] <OpenSorce> Mithrandir, of course :-)
[08:28] <Hobbsee> sladen: yes, but you miss nice weather, compared to au.
[08:28] <Hobbsee> perhaps you should move to AU, and all the problems will be solved!
[08:45] <SlimG2> Why Isn't the locales for Firefox and Thunderbird included in their main package?
[08:46] <StevenK> They should be included in their own locale packs, becuase they are enormous.
[08:46] <SlimG2> mkay
[09:01] <sladen> Hobbsee: I'd consider it, but the latency is just unbareable
[09:04] <pitti> SlimG2: the main reason is that they use a custom method for translation
[09:04] <pitti> SlimG2: the entire FOSS world uses gettext, except for Mozilla and OpenOffice products
[09:05] <pitti> SlimG2: we are aiming at eventually providing them in the normal language packs, but it'll still take a while (maybe we'll manage in hardy)
[09:06] <SlimG2> pitti: thanks for explaining
[09:42] <pitti> mvo: can I ask you again to update the status of bug 120957?
[09:42] <ubotu> Launchpad bug 120957 in update-manager "UpdateManager fails to fetch dist-upgrade tarball" [Undecided,New] https://launchpad.net/bugs/120957
[09:44] <mvo> pitti: this is fixed with 0.59.25 in feisty-updates
[09:45] <pitti> mvo: and in gutsy?
[09:45] <mvo> pitti: but let me read up on it again to see why it was reopend
[09:45] <mvo> pitti: I'm pretty sure it is fixed in gutsy too, but I will double check
[09:46] <pitti> mvo: thanks; just wanting to make sure that it doesn't fall off the table
[10:10] <gaspa> pitti: hi. I had reported two bug for uplash (and a wished feature), and subscribed you.
[10:10] <gaspa> so, in both reports there are my fixes. could you take a look if I didn't make something wrong?
[10:10] <dholbach> hey seb128
[10:11] <seb128> hello dholbach
[10:12] <pitti> gaspa: ah, thank you! Will have a look
[10:13] <gaspa> pitti: I attacched my bazaar branch, and i got a source package in ppa.
[10:13] <gaspa> so you can try it.
[10:13] <pitti> gaspa: hm, I didn't get bug mail yet
[10:13] <gaspa> strange...
[10:14] <gaspa> right, you're subscribed only for bug 152952
[10:15] <pitti> gaspa: what's the use case for 152952?
[10:16] <pitti> gaspa: we should generally avoid asking questions to the user unless it's absolutely necessary (such as entering a passphrase for encrypted partitions, etc.)
[10:16] <pitti> gaspa: did you have something particular in mind?
[10:17] <ubotu> Launchpad bug 152952 in usplash "input command with timeout parameter" [Undecided,New] https://launchpad.net/bugs/152952
[10:17] <gaspa> we use it. for our products. we don't want that user sees terminal
[10:18] <gaspa> and i think it could be useful for someone else... but this is only a whish. The other two are bug fix, and probably more important
[10:19] <pitti> gaspa: right, I'm not saying I don't like it (after all, it's just an optional feature); I was just interested in a concrete use case (maybe we want it for something by default as well?)
[10:20] <doko> pitti: please accept gcc-4.2
[10:21] <gaspa> pitti: we ask things like "push 'f' key  if you want to force a check" ... or something like that... in that case the system should not wait forever for a key but instead goes on and boot properly
[10:22] <pitti> gaspa: ah, I see
[10:23] <gaspa> ;)
[10:26] <dholbach> ogra: do you still make use of human-cursors-theme or do you also use dmz-cursor-theme?
[10:26] <gaspa> pitti: and for Bug #154234? I'm not sure of this. I filed it cause i wasn't able to build the package in ppa.
[10:26] <ubotu> Launchpad bug 154234 in usplash "build-depend on libgd-dev " [Undecided,New] https://launchpad.net/bugs/154234
[10:26] <ogra> dholbach, dmz iirc
[10:26] <dholbach> right, human-* is in universe
[10:26] <dholbach> I'll ask for its removal in hardy
[10:27] <pitti> ah, that's a bit special; it works, because libgd2-xpm-dev (main) Provides: libgd-dev
[10:28] <pitti> gaspa: http://codebrowse.launchpad.net/~gaspa/usplash/usplash-addson/revision/172 looks fine, I'll apply that
[10:29] <gaspa> pitti, not at all
[10:29] <pitti> gaspa: thank you for your fixes! what's the other bug#? I still didn't get mail
[10:29] <gaspa> i'll write: "libgd2-xpm-dev | libgd2-noxpm-dev"
[10:29] <gaspa> i think it's better
[10:30] <pitti> gaspa: "libgd2-xpm-dev | libgd-dev" is best IMHO
[10:30] <pitti> gaspa: make use of the virtual package, but prefer a concrete one
[10:31] <gaspa> oh, right.
[10:31] <gaspa> if you want, i'll push it right now.
[10:31] <gaspa> (now you're subscribed to all three reports)
[10:32] <pitti> doko: done
[10:38] <pitti> gaspa: what's the third one?
[10:39] <gaspa> ... mmm.. which are the first two? :-P :-P
[10:39] <pitti> gaspa: #154234 and #152952, as above
[10:39] <gaspa> all are: 152933, 152952, 154234
[10:39] <gaspa> ok, 152933 then
[10:40] <pitti> gaspa: splendid! thanks
[10:40]  * pitti hands gaspa the "official usplash maintainer" badge
[10:40] <gaspa> lol
[10:40] <cjwatson> somebody's gotta be ;-)
[10:41] <pitti> ah, I'm TLA now :/
[10:41] <pitti> TIL even
[10:41] <gaspa> pitti: i'm not even a motu. ;)
[10:41] <pitti> gaspa: ^ that's "touched-it-last", and denotes the poor soul who gets all the bugs henceforth :)
[10:43] <cjwatson> dholbach: my ubuntu-main-sponsors membership is about to expire; care to renew it?
[10:45] <dholbach> cjwatson: sure
[10:46] <gaspa> pitti, but what should happen on launchpad? now it should appear something?
[10:47] <pitti> gaspa: I assigned the bugs to me and marked them 'in progress', so that I'll get to them soon
[10:47] <cjwatson> dholbach: thanks
[10:54] <Label> hi
[10:54] <Label> I've one question
[10:54] <Label> http://ubuntuforums.org/showthread.php?t=580557
[10:54] <Label> Is there any issue with fonts?
[11:29]  * ogra laughs 
[11:29] <ogra> debootstrapping stuff under /media makes the /proc and /sys mounts of debotsrap how up on the desktop :)
[11:29] <ogra> *show
[11:32] <cjwatson> I think virtual filesystems should be blacklisted from showing up on the desktop
[11:32] <liw> except for fuse ones, perhaps
[11:32] <ogra> yeah
[11:32] <cjwatson> right, but things like procfs, sysfs
[11:32] <ogra> but its funny it doesnt show u in mount output
[11:32] <ogra> *up
[11:32] <cjwatson> ogra: it'll be in /proc/mounts
[11:33] <ogra> might be (already wiped it)
[11:33] <cjwatson> ogra: mount reads from /etc/mtab and if you mount something in a chroot then naturally that isn't recorded in the base system's /etc/mtab
[11:33] <cjwatson> I do wish they'd fix /proc/mounts to expose the loop device information
[11:34] <Mithrandir> apart from possible politics, that should be easy enough, shouldn't it?
[11:35] <cjwatson> well, it's been outstanding for years
[11:35] <cjwatson> I haven't looked, but guess there's some compatibility implication
[11:36] <cjwatson> anyway, I'm not well-enough-informed so shall go back to prodding debootstrap
[11:37] <cjwatson> (wow, hoary debootstrap installed so much crap)
[11:39]  * Hobbsee curses the lack of decent scripting for irssi.  or the script that she's currently lacking.  either way.
[11:40] <Spads> When will someone write a decent irssi-killer in python
[11:43] <Hobbsee> Spads: i dont know.  so far, it seems that konversation is the irc client that sucks the least - and that's scary
[11:43] <Spads> it has a curses mode?
[11:44] <Hobbsee> nope
[11:44]  * Spads puts it on the "sucks most" pile
[11:44] <Hobbsee> but having to call up chanserv by hand is a ruddy pain in the neck.
[11:44] <Hobbsee> particuarly when on multiple channels,
[11:45] <soren> Hobbsee: http://www.pthree.org/2007/07/11/irssi-chanserv-and-nickserv-helper-aliases/ ?
[11:45] <Spads> nickserv is easy.  if you put your nickserv password in as the server password in your config, it gets passed along to nickserv
[11:46] <Spads> and that method is resistant to juping
[11:46] <sladen> Spads: didn't know that, thanks
[11:46] <Hobbsee> Spads: oh, nickserv is easy, yes.
[11:46] <Hobbsee> Spads: it's the problem of having to op for every channel.  but there is a script, and i'll get my hands on it when the guy comes back
[11:46] <Spads> sladen: it works more reliably on other networks, but it does work very often here
[11:50] <ogra> mvo_, would it be possible to make CDs that are used with g-a-i not open a nutilus window by default ?
[12:10] <Hobbsee> sabdfl: ping
[12:15] <mvo_> ogra: yeah, that is currently not ideal. is there a bug about it already?
[12:17] <ogra> mvo_, i dont think so ...
[12:50] <fatal__> in case anyone is interested, it seems like the ubuntu iproute package fixes a problem which already has a fix in the imported debian package.... I don't know if this "double-fixing" might cause problems.... I'm updating the debian package to use the ubuntu fix, and dropping the debian/patches/ip_address_flush_loop.dpatch .... (I guess you'll get this automagically on some future sync, but thought I'd notify you guys now anyway..)
[12:50] <fatal__> hth, hand.
[12:51] <fatal__> btw, the debian iproute package has moved to git .... http://git.debian.org/?p=collab-maint/pkg-iproute.git;a=summary
[12:57] <dholbach> MOTU meeting in #ubuntu-meeting
[13:03] <dholbach> cjwatson, Riddell, slangasek, Mithrandir: we're discussing universe and hardy schedule in the MOTU meeting in #ubuntu-meeting right now
[13:06] <liw> does Ubuntu have a document like the Debian Policy Manual? or a document describing differences from the Debian one?
[13:56] <pitti> cjwatson: Do you have some time today for looking into d-i/dapper to install lbm?
[14:19] <JohnKarahalis> Hello everyone
[14:29] <pitti> seb128: would you mind if we generally drop upstream changelogs from cdbs packages?
[14:30] <seb128> pitti: yes
[14:30] <pitti> seb128: they are often quite large, not terribly interesting to users, and easy to get for devs by apt-get source or looking at the upstream page
[14:30] <seb128> I find it really handy to have those available without having to getting megabytes of sources
[14:30] <seb128> hum
[14:31] <seb128> pitti: well if we really need extra space I guess it make sense
[14:31] <pitti> seb128: I'm currently working on a cdbs patch to symlink identical documentation files
[14:31] <pitti> seb128: in a generic fasion
[14:31] <pitti> fashion
[14:31] <liw> speaking as a sysadmin, I've often found it useful to read upstream changelogs when debugging things
[14:31] <seb128> we should consider using 2 CDs or DVD one day rather though
[14:31] <pitti> that shuold already help quite a lot, but if we would drop changelogs entirely, that would help even more
[14:32] <seb128> because we start doing a lot of compromise on what we can ship due to CD space which sucks
[14:32] <pitti> liw: I don't argue that they are useless, I'm just weighing CD space vs. the level of utility
[14:32] <pitti> and I'd rather drop those than eternally argue about which app to kick out next
[14:32] <seb128> well, I would prefer having extra translations rather than upstream changelog to be honest
[14:33] <pitti> that too
[14:33] <seb128> but I would like better having everything
[14:33] <seb128> is switching to an another format than 1 CD on the table?
[14:33] <pitti> the Debian changelog is more useful for me at least in many cases
[14:33] <liw> pitti, well, I'd be in favor of changelogs over, say, oo.o or firefox, but I'm cantankerous and contrary, and I know it's not going to happen
[14:33] <pitti> seb128: no
[14:33] <pitti> seb128: ATM, one CD is a design goal (and you don't believe how happy I am about it...)
[14:34] <pitti> well, it can always be argued otherwise, of course
[14:34] <liw> any chance of Ubuntu supporting usb sticks officially some day, by the way?
[14:34] <liw> (I'm all in favor of having at most one CD, personally :)
[14:34] <pitti> seb128: well, if you want to include the upstream changelog in your package, nobody will stop you, btw
[14:34] <pitti> seb128: it's just about not doing it any more by default
[14:34] <seb128> pitti: I like really having one CD, but if it means throwing changelog, translations, limiting the inclusion of some applications, etc...
[14:35] <pitti> well, let me first finish the symlinking
[14:35] <pitti> that's unanimous, I take it
[14:35] <seb128> pitti: I've not strong objection, I'm not sure how the userbase will react though
[14:37] <seb128> liw: that should be easy enough to do and would be nice for hardy
[14:37] <cjwatson> liw: we support installing from USB stick - what's missing?
[14:37] <seb128> liw: there was a post somewhere (planet or list?) this week about a script from redhat, somebody adapted it to Ubuntu and it seems to work fine
[14:38] <cjwatson> it's not particularly well advertised though
[14:38] <seb128> cjwatson: there is no obvious way to create the key for users
[14:38] <seb128> cjwatson: like a "click here to make an usb bootable key"
[14:38] <liw> cjwatson, well, I looked for that option when I installed my laptop after starting to work for Canonical, and I couldn't find any documented way of doing it
[14:39] <seb128> liw: there is https://help.ubuntu.com/community/Installation/FromUSBStick on the topic
[14:39] <cjwatson> seb128: hence "not particularly well advertised". I agree
[14:40] <seb128> cjwatson: right, I was already typing when you wrote that, I agree with you
[14:40] <pitti> pong
[14:40] <seb128> pitti: hi
[14:40] <pitti> hey seb128 :)
[14:40] <seb128> not sure why you pong-ed ;-)
[14:41] <pitti> uh?
[14:41] <pitti> I didn't
[14:41] <liw> pitti, multiple times on some channels
[14:41] <pitti> oh, argh
[14:41] <pitti> that must be that test autoresponse script going wild
[14:41] <pitti> sorry
[14:41] <dholbach> pitti: don't say "it was the cat" - nobody will believe you
[14:41] <StevenK> Hah
[14:42] <pitti> seb128: thanks, fixed
[14:42] <Hobbsee> pitti: you used /ame not /me ?
[14:42] <Hobbsee> or /a something
[14:42] <pitti> seb128: takes a little while to migrate to a new IRC client...
[14:43] <pitti> Hobbsee: I didn't use any command; it was just a test autoresponder which says "pong" when someone says "ping" to me
[14:43] <Hobbsee> pitti: ahhh
[14:43] <pitti> I meant to disable it, but forgot
[14:43] <StevenK> And it matches on "ping", not "pitti: ping"
[14:43] <pitti> right
[14:43] <pitti> when I actually tried it, it didn't respond to either
[14:43] <Spads> haha someone said typing
[14:44]  * pitti wonders why it suddenly started to work
[14:44] <pitti> *blush*
[14:44] <StevenK> Which is naughty, since words like 'typing' and 'zipping' ...
[14:44] <seb128> pitti: which one do you use now?
[14:44] <pitti> seb128: weechat
[14:44] <pitti> seb128: the one and only client I found which supports vertical splitting, and it has nice scripting capabilities, too
[14:44] <seb128> ah, you liked the screenshots with all the chans splitted on a same screen?
[14:45] <StevenK> So many terrible places we could take that name.
[14:45] <pitti> seb128: yeah, that was it; I have three vertical splits now
[14:45] <pitti> without the need to sort the xchat windows every day, alt+tabbing through them, and so on
[14:55] <Keybuk> cjwatson: do we want MoM to use unstable for 8.04 or testing?
[14:55] <cjwatson> I meant ubuntu-devel@lists :-)
[14:56] <dholbach> MOTU Q&A session in 2 minutes in #ubuntu-classroom
[14:59] <StevenK> Hrm. My poor newly upgraded Gutsy machine does not like LVM.
[15:00] <StevenK> Hrm. No, it's because it can't find the first hard disk.
[15:01] <StevenK> Fun message during boot like "ata3: failed to recover some devices, retrying in 5 secs"
[15:04] <StevenK> Looks like our good friend HPA is too blame.
[15:12] <soren> StevenK: Hardy amd64 sbuilder also has all uppercase dpkg output :)
[15:13] <StevenK> soren: Yay ....
[15:13] <StevenK> soren: Let's file a bug this time. Then mvo can fix it. :-)
[15:13] <soren> StevenK: So now we know it not the string "gutsy" that's messing it up. Big surprise.
[15:14] <soren> StevenK: Yeah, and then he'll say it's dpkg... and the iwj will say it's sbuild, and then elmo will shout at everyone for not getting anything done.
[15:14] <cjwatson> I'm inclined to suggest stracing the entire damn thing and look at what's fiddling with the tty.
[15:14] <cjwatson> you'll get a hell of a lot of output but it's faster than bouncing it back and forward between people who can't reproduce it
[15:14] <soren> cjwatson: If it was the tty that got told to spew only upper case it wouldn't be uppercase in the logs, would it?
[15:15] <cjwatson> soren: you can't tell the tty to spew upper case without it showing up in strace
[15:15] <StevenK> Grumble.
[15:15] <soren> Also, it's only dpkg's output. Not apt or the build.
[15:15] <cjwatson> we assume that it is some process under sbuild that's doing it
[15:15] <StevenK> soren: Shows in the builds logs for me.
[15:15] <cjwatson> therefore, strace -f sbuild
[15:15] <soren> StevenK: Yes, exactly.
[15:15] <StevenK> 2.6.22 doesn't boot, and 2.6.20 doesn't start X.
[15:16] <cjwatson> soren: oh, I misread - but even so, strace will show dpkg printing lower case and something else printing the same thing in upper case
[15:16] <cjwatson> so you'll be able to isolate from that what's at fault
[15:16] <cjwatson> or if it *does* show dpkg printing lower case, then you go and beat up some mad locale definition
[15:17] <cjwatson> s/lower/upper/
[15:17] <soren> cjwatson: But the logs I get in my e-mail is also upper case?
[15:17] <soren> cjwatson: strace and schroot is not an easy combo to work with, by the way.
[15:18] <cjwatson> due to set-id?
[15:18] <soren> I assume so, yes. I haven't really bothered to look into it just yet.
[15:18] <cjwatson> if that's it, make a private copy of strace and make it setuid root, then use that; see strace(1)
[15:18] <soren> If I just try now, I get: E: PAM error: Authentication service cannot retrieve authentication info
[15:19] <StevenK> Right. X with nv
[15:20] <StevenK> Now to figure out to disable HPA so 2.6.22 will actually boot
[15:22] <soren> cjwatson: Didn't help. I'll try attaching to the apt-get process inside schroot when it's running. That should work.
[15:22] <cjwatson> yeah
[15:23] <StevenK> soren: If strace is installed inside the chroot. It might be simpler to get schroot to spawn strace apt-get.
[15:24] <dholbach> can all teams please include their team reports on  http://wiki.ubuntu.com/TeamReports/October2007 ?
[15:26] <StevenK> Oh, grumble.
[15:26] <StevenK> Can you not disable HPA without recompiling the whole damn kernel?
[15:33] <dholbach> PriceChild, mathiaz, Burgundavia, Mithrandir, superm1, popey, bryce: if you teams have, can you please include their team reports on  http://wiki.ubuntu.com/TeamReports/October2007 ?
[15:33] <popey> wilco
[15:34] <popey> 22nd is the cut off isn't it?
[15:35] <PriceChild> gah IRC haven't had a meeting this month... will poke people about it and probably just add things we've decided outside meetings in -ops or ML for gutsy+
[15:35] <dholbach> popey: jono asked me to ask people to move their stuff :)
[15:35] <bddebian> Heya
[15:35] <mathiaz> dholbach: sure.
[15:36] <Hobbsee_> PriceChild: so stop being slack :)
[15:36]  * popey makes some stuff up
[15:41] <pitti> does anyone feel like eyeballing a piece of really horrible shell for me? (http://paste.ubuntu.com/1038/)
[15:41] <pitti> it's a cdbs patch to symlink identical documentation files to dependent packages
[15:45] <StevenK> pitti: My only thought is 'Ew' :-)
[15:46] <pitti> doko, seb128: ^ you maybe?
[15:46] <seb128> pitti: I'm looking at it
[15:46] <pitti> StevenK: alternative approaches appreciated; this is the kind of stuff I'd rather do in a 100-line Python script, but that's inaappropriate for a cdbs rule IMHO
[15:46] <StevenK> BenC: So, how can I disable HPA if I think it's the reason 2.6.22 decides to disable the SATA channel my hard disk is on...
[15:46] <doko> pitti: later, away now
[15:46] <seb128> pitti: I'm a bit unsure about the "iterate over Depends", it'll make the result Depends of the packages installed on the system
[15:47] <pitti> seb128: I tested it with tracker, which has a lot of libraries and packaes depending on them, etc.
[15:47] <pitti> seb128: well, it iterates over dependencies which are built by the same source package
[15:47] <StevenK> pitti: Of course, but at this point I'd rather get my desktop machine booting. :-)
[15:47] <seb128> pitti: right, but I mean what if nautilus and gedit have a common README.GNOME?
[15:48] <pitti> seb128: that isn't caught with this code
[15:48] <pitti> seb128: and I don't think it should be
[15:48] <seb128> pitti: ah, I didn't read the code careful enough then
[15:48] <pitti> seb128: so, what it does is:
[15:48] <seb128> I was having a first quick looks before trying to understand the detail of the regexp
[15:48] <pitti> seb128: it finds the intersection of dependencies and binaries from that source as candidates of packages to consider symlinking to
[15:49] <pitti> seb128: i. e. I cannot consider *all* binaries of the source as symlink targets
[15:49] <pitti> seb128: just those which the pacakge I'm editing depends on (or pre-depends)
[15:50] <seb128> pitti: well, what happens if let's say nautilus Depends on gnome-control-center and they have a README identic?
[15:50] <pitti> seb128: I never symlink copyright (Debian Policy), and only documentation files, since those are (1) the worst offenders, and (2) if it breaks, we don't ruin the entire system
[15:50] <pitti> seb128: line 12 in the pastebin will rule it out
[15:50] <pitti> seb128: since gnome-control-center is not a binary built by the nautilus source
[15:51] <seb128> doh
[15:51] <seb128> right, sorry, I misread this one
[15:51] <pitti> np, it's really horrible
[15:51] <pitti> seb128: such kind of questions is why I asked for a review ;)
[15:51] <seb128> the logic looks good to me
[15:51]  * pitti fixes the comment
[15:51] <pitti> # fix absolute symlinks created above to relative
[15:51] <pitti> that makes more sense ^
[15:52] <pitti> to "be" relative
[15:52] <seb128> yes, it's better
[15:52] <pitti> seb128: I'm reasonably confident that it is correct
[15:52] <pitti> it's not necessarily optimal
[15:52] <doko> pitti: why reimplement fdupes?
[15:53] <doko> pitti: there's no possibility to disable symlinking
[15:53] <pitti> doko: I found it doesn't make the code much simpler, and so I can avoid the dependency and depending on the output format
[15:53] <pitti> doko: ah, good point
[15:54] <doko> pitti: and it doesn't help for inter-package symlinking
[15:54] <infinity> pitti: What if you end up symlinking web docs, and the user's webserver doesn't follow links (for /usr/share/doc, or at all)?
[15:54] <infinity> pitti: Not that two packages should contain the same web docs ANYWAY, but being unable to disable to feature seems... Likely to cause a problem for someone.
[15:54] <pitti> line 10 > [ -n "$$CDBS_NO_DOC_SYMLINKING ] || ...
[15:54] <popey> dholbach: done
[15:54] <pitti> infinity: web docs should absolutely *not* be in /usr/share/doc IMHO
[15:55] <dholbach> popey: rock on
[15:55] <infinity> pitti: Ahh, I'm blind.
[15:55] <infinity> pitti: And, uhm, lots are.  Always have been.
[15:55] <pitti> infinity: that's why I confine it to /usr/share/doc
[15:55] <doko> pitti: and it doesn't help with the bloody gnome help files
[15:55] <infinity> pitti: We even used to serve /usr/share/doc from webservers by default for that reason.
[15:55] <pitti> doko: as I said, it's a first start, not complete yet
[15:55] <pitti> I rather do it in steps
[15:55] <pitti> doko: "inter-package symlinking"?
[15:56] <doko> pitti: I'm not sure if you find even one link ...
[15:56] <infinity> pitti: The obscurity > security camp is the only reason we don't serve /usr/share/doc anymore, not because docs aren't there (they are).
[15:56] <pitti> infinity: right, but if you explicitly server /usr/share/doc, this should be fine then?
[15:56] <pitti> doko: lots (such as changelog.Debian.gz, changelog.gz, NEWS.gz, etc.)
[15:57] <doko> pitti: cd debian/$(cdbs_curpkg)
[15:57] <infinity> pitti: The point is likely moot, as long as the feature can be disabled, since maintainers can just clean up breakage as they see fit (only have web docs in one package, which is correct anyway, or disable symlinks, or whatever)
[15:57] <doko> pitti: you can't find any duplicates you search for
[15:57] <pitti> infinity: ah, point; they shouldn't really be duplicated
[15:57] <pitti> doko: hm?
[15:58] <pitti> doko: ah, I see what you mean; $rootdir takes care that I will
[15:58] <infinity> pitti: My only practical concern is "what if people are already managing this differently, and changing CDBS breaks their carefully-mangles rules file to do something Very Bad"?
[15:58] <doko> pitti: ahh, ok.
[15:59] <infinity> s/mangles/mangled/
[15:59] <doko> infinity: they can disable it
[15:59] <pitti> infinity: do you see how it can break?
[15:59] <infinity> doko: Yes, but who reads every CDBS changelog and then checks if their package build changed?
[15:59] <pitti> infinity: if people already symlink it manually, that should be fine
[16:00] <pitti> since this is a general and by-default rule, it should be designed not to break anything, right
[16:00] <StevenK> Right. The find only looks for real files anyway
[16:00] <pitti> infinity: oh, I see that bug
[16:00] <doko> infinity: without it you'll never get people like our gnome maintainers to change their habits (although it's known ...)
[16:00] <pitti> infinity: I shouldn't use find -L
[16:00] <infinity> pitti: Say I have a setup that removes dupes forcefully and symlinks.  From the opposite package that you find first.  And I do it after this new code exeuctes.
[16:00] <infinity> pitti: And I end up with packages with no changelogs.
[16:00] <pitti> infinity: instead I shuold change the *target* test to -f || -L
[16:01] <infinity> StevenK: You're making the same assumption pitti is... That this code will never run *before* my hypothetical custom-mangling code.
[16:02] <StevenK> infinity: Ah ha
[16:02] <infinity> (So, you remove a dupe, add a a link, then I blindly "rm -f foo/changelog", which was the only "real" one left, because CDBS and I disagreed about which to keep)
[16:02] <pitti> http://paste.ubuntu.com/1040/
[16:02] <doko> infinity: well, it runs in dh_builddep ?
[16:02] <pitti> ^ new version with "disable" variable and symlink fix
[16:02] <pitti> infinity: no, since any code you run afterwards cannot change the .deb any more
[16:02] <mathiaz> dholbach: I've added the ServerTeam report to the Monthly report.
[16:02] <pitti> infinity: since that very rule calls dh_builddeb
[16:03] <dholbach> mathiaz: rock on
[16:03] <dholbach> :-)
[16:03] <pitti> infinity: and any code you run before is fine, since I only look for real files, not for symlinks
[16:03] <infinity> pitti: Kay, I admit to being mostly CDBS illiterate (by choice)... There's no way to modify the builddeb phase with rules?
[16:03] <pitti> infinity: there are quite a lot of hooks
[16:03] <infinity> pitti: If not, then my concern seems to be addressed.
[16:03] <infinity> pitti: If it can be hooked/altered, then it still stands.
[16:03] <pitti> infinity: but I deliberately put that code in between dh_gencontrol and dh_builddeb (which are in the same rule)
[16:04] <doko> pitti: where's the variable?
[16:04] <pitti> infinity: i. e. there is binary-predeb which you usually do to chmod files, remove files, etc.
[16:04] <pitti> doko: I suck, wrong paste
[16:05] <pitti> http://paste.ubuntu.com/1041/
[16:05] <pitti> try again
[16:05] <pitti> infinity: in the diff context you still see the dh_builddeb call at the very bottom
[16:05] <infinity> pitti: Ahh, so I do.
[16:05] <infinity> pitti: Alright, seems to address that concern, then.
[16:05] <pitti> infinity: I need to do it after dpkg-gencontrol to see all dependencies, and before dh_md5sums to avoid mangling it agai
[16:06] <pitti> W: tracker-utils: debian-changelog-file-is-a-symlink
[16:06] <pitti> bah, lintian
[16:06] <infinity> Does it have a direct dep?
[16:06] <infinity> lintian's not supposed to warn on linked-docs-with-direct-deps...
[16:06] <pitti> yes, it has (it doesn't consider any others)
[16:06] <infinity> Or maybe it's only smart enough to do the doc directory, not files.
[16:07] <pitti> it could be made more efficient by also considering transitive dependencies, but I guess that code will give us 90% of the effect already
[16:07] <infinity> Easy enough to fix lintian, your solution is policy-compliant.
[16:07] <infinity> Oh, wait.
[16:07] <infinity> Uhm.
[16:08] <doko> if [ ! -h debian/$$dep/usr/share/doc ] && [ -d debian/$$dep/usr/share/doc ]
[16:08] <infinity> You realise dpkg behaves pretty "special" when files change to links, right?
[16:08] <doko> pitti: ^^^
[16:08] <infinity> (Not as bad as when directories do...)
[16:08] <pitti> infinity: well, I do know that it is spethial when directories change to links; which is why I don't touch directories at all
[16:09] <pitti> doko: ah, can do
[16:09] <pitti> doko: hm, wait; why for the depending package? that shouldn't matter
[16:09] <infinity> pitti: For peace of mind, build on of these sets of packages, and upgrade one that got itself symlinks (but not the dependency)... And make sure dpkg didn't do anything weird.
[16:10] <pitti> doko: I need to add a check that $(cdbs_curpkg)/usr/share/doc isn't a symlink
[16:10] <doko> pitti: that as well
[16:10] <infinity> pitti: Weird things may include unpacking the links, then following them when "removing old files", and other really bizarre stuff.
[16:10] <doko> pitti: do you miss indirect depends?
[16:10] <pitti> doko: yes, I do, see above (first correctness, completeness later)
[16:13] <pitti> infinity: dpkg follows symlinked files? eww
[16:13] <infinity> pitti: It's had such bugs in the past.  I'd want to verify that it no longer does. :)
[16:13] <infinity> pitti: It's not SUPPOSED to.
[16:14] <pitti> infinity: ok; I could imagine that this would wreak havoc in interesting ways
[16:14] <StevenK> Oh, I forgot about that. Now that I've upgraded, tracker is taking over my machine.
[16:15] <Hobbsee> StevenK: it hasnt taken over mine yet.  i'm really wondering why.
[16:15] <pitti> infinity: at least I can dpkg -i the old and new package back and forth, and both symlinks and files work fine
[16:16] <infinity> pitti: Alright, good, good.
[16:17] <pitti> seb128, doko, infinity, StevenK: thanks for your review!
[16:17] <seb128> pitti: thanks for the work on that ;-)
[16:17]  * seb128 hugs pitti
[16:18] <StevenK> pitti: I don't see what I did, but sure. :-)
[16:18] <infinity> StevenK: Same thing I usually do.  Be grumpy while pitti makes random changes and blames his improved code quality on your grumpiness.
[16:19]  * StevenK grins
[16:19]  * pitti wonders why his shiny new tracker package still doesn't have any symlinks
[16:19] <Hobbsee> pitti: it has a crackmeter.  if the level is too high, no symlinks for you!
[16:20] <pitti> seb128: wouldn't it make -- kind of, just a little -- sense if tracker actually depended on any of the libtracker* stuff?
[16:20] <pitti> hm, I guess the daemon doesn't use any of its own client libs
[16:20] <StevenK> Nice how tracker-status doesn't actually work
[16:20] <pitti> although this proves that my code DTRT, it's still weird
[16:21] <jamiemcc> pitti: they are client side libs only
[16:21] <StevenK> So, if I start something that grabs a shedload of CPU, should tracker stop using CPU time and heel?
[16:21] <seb128> pitti: right, it doesn't use those libs
[16:21] <pitti> jamiemcc: ah, ok
[16:21] <Hobbsee> pitti: DTRT?
[16:21] <infinity> Hobbsee: Do The Right THing.
[16:21] <pitti> Hobbsee: "throws a gummy bear at Hobbsee"
[16:22] <Hobbsee> infinity: ahhh
[16:22] <Hobbsee> pitti: :P.  i dont eat lollies.
[16:22]  * Hobbsee throws the gummy bear to infinity
[16:22] <pitti> gummybear != lolly
[16:22] <infinity> Hobbsee: Neither do we, we just lob them at each other.
[16:22] <infinity> Hobbsee: You should have seen the laptop damage at UDU...
[16:23] <dholbach> mentos! :)
[16:23] <Hobbsee> infinity: i did :D
[16:23] <Hobbsee> infinity: for sevilla, anyway
[16:23] <Hobbsee> pitti: they class as a lolly in my book
[16:23] <Spads> MENTOS DO NOT BURN
[16:23] <Hobbsee> infinity: oh, udu.  right
[16:23] <infinity> Hobbsee: High velocity Mentos = Very Bad for laptop screens.
[16:23] <Hobbsee> Spads: really?
[16:23] <Hobbsee> infinity: hahhaa, yes
[16:23] <JohnKarahalis> Hey, everyone.
[16:23] <pitti> Hobbsee: hm, in my country, a lolly is a round piece of candy on a stick
[16:24] <infinity> Spads: Peeps do, though.
[16:24] <StevenK> infinity: Who coped a dented laptop screen?
[16:24] <infinity> pitti: That's a lolly-pop.  Australians, however, refer to anything vaguely candyish as a "lolly"... And I imagine that's of British origin.
[16:24] <soren> Would it be confusing if I started filing sync requests already?
[16:24] <Spads> Hobbsee: three independent reports here: http://www.pigdog.org/naked_splicer/html/tnipnaz.html
[16:24] <Hobbsee> infinity: excluding chocolate
[16:24] <dholbach> http://images.google.com/images?q=lolly - hm
[16:25] <infinity> StevenK: Daniel's took damage, and I think one of thom's many dents was from UDU.
[16:25] <infinity> StevenK: Others likely suffered too.
[16:25] <ScottK> soren: If you did, you wouldn't be the first.
[16:25] <StevenK> Neeat
[16:25] <pitti> soren: no, that's fine
[16:25] <pitti> soren: there are quite a couple of them already
[16:26]  * dholbach did today :)
[16:26]  * Hobbsee has a sudden, unpleasant thought
[16:26]  * ScottK did 9 or 10 yesterday.
[16:26] <JohnKarahalis> I was wondering, does anybody in here work for Canonical?
[16:26]  * pitti whistles innocently
[16:27]  * infinity points at someone else.
[16:27] <Hobbsee> no, no, anyone who works for canonical we took out and shot.
[16:27] <JohnKarahalis> no one? Is it still night time in Canonical HQ? I tried last night but not many were around then eihter.
[16:27] <Hobbsee> !hungover
[16:27] <Hobbsee> but now it's all the canonical employees, not just the release manager
[16:27] <JohnKarahalis> So are you all pretty much volunteers?
[16:28] <dholbach> !hangover
[16:28] <pitti> JohnKarahalis: quite a few of us are; why?
[16:28] <dholbach> hrm :)
[16:28] <Hobbsee> dholbach: bot is lagged, i think
[16:28] <JohnKarahalis> oh really
[16:28] <Hobbsee> (again)
[16:28] <JohnKarahalis> good
[16:28] <Hobbsee> !ping
[16:28] <pitti> ubotu: *shake*, wake up
[16:28] <agoliveira> At least the ones who matter :)
[16:28]  * agoliveira ducks quick!
[16:29] <infinity> agoliveira: Tsk.
[16:30] <JohnKarahalis> Well. I'm a first year SE major at RIT. As part of a project I am required to have a short interview with a professional Software Engineer (or computer scientist, you know, anything along those lines). Would anyone be interested in conducting a short interview with me. It wouldn't be long. Mainly questions like "What is a typical day at work like?", "How big is your project team", "What languages and technologies do you use?" Would anyone be
[16:31]  * Hobbsee beats agoliveira
[16:31] <Hobbsee> pitti: you'll make him fall over more, doing that :)
[16:31]  * agoliveira falls in love for Hobbsee
[16:32] <Hobbsee> oh dear.
[16:32]  * Hobbsee is very unlovable, she assures you.
[16:33] <soren> pitti: Ok, rock. I'll start filing them. Thanks.
[16:33] <agoliveira> Hobbsee: I'm married, I know how it goes :-D
[16:33] <Hobbsee> agoliveira: :P
[16:33] <Hobbsee> agoliveira: glad to hear it.  i'm not interested in ubuntu to meet a boyfriend, per se.
[16:34] <infinity> Hobbsee: Ouch, I'm crushed.
[16:34] <Hobbsee> infinity: "You Poor Bastard"
[16:34] <dholbach> where's the CoC police?
[16:34] <Hobbsee> dholbach: kickbanned.  i have ops in here.
[16:35] <agoliveira> JohnKarahalis, many people here can help but we all have some sort of different attributions. I, for example, work on the embedded project, doing mostly packaging for now and a bit of software development but if you think I can help, just call me in private
[16:35] <agoliveira> Hobbsee: Would be a very poor place to find one anyway ;)
[16:36] <JohnKarahalis> agoliveira: What do you mean by packaging?
[16:36] <agoliveira> JohnKarahalis: Create software packages in a format that a distro can understand, install, uninstall, etc. In our case, .deb. pcakages.
[16:36] <agoliveira> packages
[16:37] <JohnKarahalis> agoliveria: ahh right that's what I thought. Just wondering because my school has a degree in "Packaging Science" as in physical packages. :P
[16:38] <JohnKarahalis> agoliveria: Anyway, do you think you would want to participate. It wouldn't take more than 10-20 minutes to answer the questions probably, but I understand that you're all very busy so if you can't that's fine.
[16:39] <SlimG3> Is there any (plans for) logos representing the (current) ubuntu release codename? like a drake for dapper drake, this would be doable since all the codenames represents animals in different "moods"
[16:39] <ubotu> The Release Managers are currently hungover, and wont be releasing gutsy today.  No gutsy for you!  NOT YOURS!!!
[16:39] <Hobbsee> sheesh.  have some lag!
[16:39] <sladen> JohnKarahalis: I suspect lots of people will help;  might be better to ask via the mailing list though, and none of the people here are likely to be "typical" software people
[16:39] <agoliveira> JohnKarahalis: I would love to help. I just can't do this right now but if you poke me in about 1 hour, I think I'll be able to help.
[16:40]  * agoliveira thinks that ubotu snapped :)
[16:40] <JohnKarahalis> agoliveria: Awesome, thank you! I actually have a class in a little bit, but when would be another good time to contact you? I have about 2 weeks before this paper is due, so no rush.
[16:41] <persia> agoliveira: It's the bug count - ubotu keeps getting in trouble for flooding.
[16:41] <sladen> telflon flood-resistant upgrades for ubugtu
[16:42] <agoliveira> JohnKarahalis: So call me monday.
[16:42] <JohnKarahalis> Cool, what time? And IRC?
[16:42] <soren> pitti: Even though filing sync requests is ok, I don't suppose uploading new stuff is kosher at this point, is it?
[16:42] <sladen> Hobbsee:
[16:42] <cjwatson> sladen: depends how you count it. I certainly have been a traditional software engineer, though that's not so much what I'm doing at the moment
[16:42] <cjwatson> soren: you can upload, it just won't be accepted yet
[16:42] <pitti> soren: you can upload away, it'll land in unapproved
[16:42] <soren> cjwatson: Oh, ok.
[16:43] <agoliveira> JohnKarahalis: about this time is fine
[16:43] <SlimG3> Is there a dedicated channel for the ubuntu graphics/design/theme apartment?
[16:43] <soren> I for some reason assumed we were supposed to wait.
[16:43] <Hobbsee> sladen: ?
[16:43] <sladen> SlimG3: #ubuntu-art
[16:43] <sladen> Hobbsee: tab über completetion
[16:43] <agoliveira> JohnKarahalis: Just look for me on freenode and call me.
[16:43] <Hobbsee> sladen: ahhh
[16:43] <JohnKarahalis> agoliveria: Great! Thanks! And if you don't mind me asking, what is your official title?
[16:44] <SlimG3> thanks sladen
[16:44] <JohnKarahalis> agoliveria: And, sorry, but what do you mean by call? I'm new to IRC. Is a "call" like an IM?
[16:44] <sladen> Job title: Typical Software Programmer #437
[16:44] <agoliveira> JohnKarahalis: You know what? I don't have the slightest idea :-D
[16:44] <JohnKarahalis> agoliveria: haha that's fine, I have a general idea of your responsibilities, thats all my prof needs
[16:45] <agoliveira> JohnKarahalis: Yes, you can just open a private conversation like on IM.
[16:46] <JohnKarahalis> agoliveira: Awesome. Thanks so much for participating!
[16:46] <agoliveira> JohnKarahalis: My pleasure
[16:53] <pitti> doko: what did you mean with gnome help files?
[16:53] <pitti> doko: ah, symlinking identical pngs, and so on?
[16:54] <doko> pitti: yes. but IMO large -help guides should be splitted out anyway (like it's already done for gimp)
[16:54] <pitti> doko: right, but lots of small changes will help as well
[16:54] <doko> sure
[16:55] <pitti> doko, seb128: so I do the same "symlink to depending packages" trick for /usr/share/gnome/help?
[16:55] <seb128> pitti: /usr/share/gnome/help should not be duplicated
[16:56] <doko> seb128: did you change it?
[16:56] <seb128> pitti: do you have any example of where it happens?
[16:56] <seb128> doko: change what?
[16:56] <doko> seb128: gnome-games, totem,
[16:56] <seb128> doko: what is duplicated in totem?
[16:57] <doko> pitti: all the packages that were rejected before release
[16:58] <seb128> doko: and no, user help should not be splitted from the package
[16:58] <doko> seb128: then fix gimp
[16:58] <seb128> doko: what about gimp?
[16:58] <doko> seb128: it has splitted user help
[16:58] <doko> seb128: OOo as well ...
[16:58] <seb128> doko: that's a different source package which contain those
[16:59] <seb128> that's not part of the program
[16:59] <doko> seb128: how do you decide this?
[17:00] <seb128> doko: decide what?
[17:00] <doko> so the list is: gnome-applets gnome-games gnome-power-manager gnome-utils totem tomboy
[17:00] <doko> seb128: "that's not part of the program"
[17:01] <seb128> doko: apt-cache showsrc gimp-help
[17:01] <seb128> it's gimp-help
[17:01] <seb128> it's not from the gimp upstream tarball
[17:01] <seb128> I didn't decide it
[17:01] <seb128> upstream did
[17:01] <doko> seb128: sorry, that's a bogus argument; with the same argument I can add all the OOo help to OOo-common
[17:02] <seb128> doko: I think it makes no sense to remove the user documentation from programs
[17:02] <doko> pitti: you find the list of all current duplicates in the liveCD build log
[17:02] <seb128> most users expect F1 to work
[17:02] <seb128> and to have documentation
[17:03] <persia> Is it possible to do something at install time, with the right hints?  The packages will not get smaller, but the footprint will, and users will never not have the basic docs.
[17:03] <doko> seb128: yeah, but we do have to make compromises due to CD size. OOo opens a dialog and asks the user to install the help package. sure you could do better offering an option to start the installer/update manager
[17:04] <seb128> we surely don't want to start doing that for every desktop application
[17:04] <seb128> and I'm not sure we need that much space yet
[17:04] <seb128> let see how it goes without the duplication
[17:04] <doko> sure, removing gnome-games from the CD is an option as well
[17:05] <doko> seb128: we already know, we do have the numbers
[17:05] <seb128> doko: gutsy was not oversized and we will win extra space with the cdbs change, what do you need space for now?
[17:05] <doko> seb128: translations
[17:06] <doko> seb128: java plugin for the browser
[17:06] <seb128> we should not start creating complication and make users run installers when clicking on an help menu only for the sake of having free CD space we don't use
[17:08] <doko> looking away from cd sizes is "creating complication" ...
[17:09] <seb128> ?
[17:09] <pitti> seb128, doko: so while the separation of help into extra packages is debatable and should be decided on a case-by-case basis, symlinking identical files should be ok, right?
[17:09] <pitti> doko: I'll test it with gnome-power-manager and gnome-utils
[17:09] <seb128> pitti: if that actually works yes
[17:09] <doko> pitti: yes, if you can turn it off; I think turning it on by default is the right thing to do
[17:09] <pitti> seb128: right, of course
[17:10] <seb128> pitti: dholbach tried on ubuntu-docs before gutsy no?
[17:10] <pitti> I don't know
[17:10] <doko> ubuntu-docs is fixed
[17:14] <pitti> mvo: can I ask you to verify bug 152113?
[17:14] <mvo> pitti: let me have a look
[17:14] <pitti> mvo: I'd like to put it into -updates today, so that we fix feisty->gutsy upgrades ASAP
[17:14] <mvo> pitti: oh, gutsy was missing?
[17:14] <pitti> mvo: the actual upgrade breakage is bug 116193 (fixed in the same upload)
[17:15] <mvo> pitti: ok, I check this out
[17:15] <pitti> mvo: yes, gutsy-proposed didn't have builds yesterday
[17:15] <pitti> mvo: thank you!!!
[17:21] <sladen> seb128: I would *love* if F1 didn't work.  It pops up everytime I try and aim for escape
[17:22] <sladen> separating docs out helps the embedded case
[17:24] <seb128> sladen: not really, or it needs to be made in a smart way, not by just not installing those and having a broken menu item
[17:26] <pitti> slangasek: swapping caps lock and escape works *wonders* for us vimers :)
[17:26] <pitti> (except that I always erroneously switch on caps lock when I sit on a different computer than mine)
[17:27] <ubotu> Launchpad bug 152113 in tzdata "Brazilian DST date change needs upgrade to 2007h" [High,In progress] https://launchpad.net/bugs/152113
[17:27] <ubotu> Launchpad bug 116193 in tzdata "error upgrading tzdata_2007e to tzdata_2007f" [Critical,Fix committed] https://launchpad.net/bugs/116193
[17:27] <slangasek> pitti: what does one switch tab with to avoid nick collisions? ;)
[17:28] <pitti> that is still a mystery to be solved
[17:50] <pitti> seb128: tested now; images in yelp do break with broken symlinks (so I rule out caching), and work again with fixed symlinks
[17:50] <RvGaTe> heya
[17:50] <pitti> seb128: that rule indeed buys us quite a lot of space without any noticeable behaviour change
[17:50] <RvGaTe> i thought i'd come in here to report that some hardware is not supported by default...
[17:52] <seb128> pitti: good ;-)
[17:53] <pitti> doko: I shamelessly stole the basic approach from your g-p-m upload :) (although I do not modify debian/<package>.links, I think this is too evil)
[17:54] <pitti> cjwatson: would you mind if I upload and accept cdbs? it's arch-all, toolchainish, and I'd like to have it before we start the rush of syncs and Gnome packages, for maximum effectiveness
[17:54] <RvGaTe> with this setup: motherboard, Abit IP35-E. And a hdd, Western Digital Caviar SE 200GB 7200rpm S-ATA..... is not detected the hdd... making it unable to install.... if you need any stats, just ask... (if you have a solution, #ubuntu ofc... :P)
[17:54] <cjwatson> pitti: not at all
[17:55] <pitti> seb128: that also means that we can probably drop most of the hacks from evolution et all
[17:55] <pitti> seb128: s/all/al/, d'oh my Latin
[17:55] <doko> pitti: all such packages like debhelper, devscripts should be uploaded now
[17:55] <pitti> right
[17:56] <seb128> pitti: good ;-)
[17:56] <seb128> doko: is hardy open yet?
[17:56] <cjwatson> https://wiki.ubuntu.com/NewReleaseCycleProcess
[17:56] <cjwatson> somebody should merge base-files
[17:56] <cjwatson> (and change it for hardy)
[17:57] <doko> ok, doing ...
[17:57] <cjwatson> the debootstrap change in question is syncable from unstable (or will be soon) whenever somebody feels the urge; but note that debootstrap-udeb has C code so it should wait until doko's happy with the compiler
[17:58] <doko> the buildds are still happy building it ... ;)
[17:58] <Hobbsee> doko: pedal faster...
[17:59] <cjwatson> doko: are you planning to upload glibc afterwards?
[17:59] <cjwatson> (still 2.6 rather than 2.7, I assume from our earlier discussions0
[17:59] <cjwatson> )
[18:00] <doko> [ Uploading job glibc_2.6.1-6ubuntu1_source
[18:00] <doko>  glibc_2.6.1-6ubuntu1.diff.gz 676.1 kB, ok (15 s, 45.08 kB/s)
[18:00] <doko>  glibc_2.6.1-6ubuntu1.dsc 2.2 kB, ok (0 s, 2.21 kB/s)
[18:00] <doko>  glibc_2.6.1-6ubuntu1_source.changes 11.4 kB, ok (1 s, 11.35 kB/s) ]
[18:01] <cjwatson> aha
[18:01] <doko> has a b-d on the new gcc, so accepting should be ok (plus gcc-snapshot)
[18:02]  * pitti pokes gutsy-proposed queue
[18:06] <pitti> seb128: so, as discussed, we'll push gnome* gutsy-proposed under a blanket approval and have it publicly tested extensively until after allhands, right?
[18:07] <seb128> pitti: yes please
[18:07] <doko> pretty please have a look at the pango changes first ...
[18:07] <seb128> pitti: I've only uploaded safe updates and read debdiff until now
[18:07] <pitti> seb128: you rock! :)
[18:07] <pitti> doko: yes, I'll process the entire queue now
[18:08] <doko> pitti: I mean *look* at the diff, and/or test OOo / firefox with the new version
[18:09] <pitti> asac, Riddell: please do follow the SRU rules for SRU bugs (subscribe ubuntu-sru, gutsy and hardy tasks, etc.); otherwise this will fall off my radar
[18:09] <seb128> pitti: the only one which has non trivial changes is gnome-games which should be ok, that's only games and upstream said the code was not used
[18:11] <pitti> cjwatson: aah, the traditional vim upload \o/ (not accepting yet, though)
[18:12] <cjwatson> heh, wasn't me this time :)
[18:13] <pitti> which reminds me that I need to merge mutt :)
[18:17] <slangasek> haha, best autoresponse evar on ubuntu-announce from a hotmailer: "You e-mailed <foo>@hotmail.com, this is no longer my e-mail address anymore<br><br>Since Microsoft is communist, it does not let me forward my e-mail, so i am stuck with this automated reply"
[18:17] <pitti> bdmurray: would you have some time to SRU-verify bug 152113? that'll include the fix for bug 116193 which breaks an awful lot of upgrades and thus I'd like to get it out ASAP
[18:17] <ubotu> Launchpad bug 152113 in tzdata "Brazilian DST date change needs upgrade to 2007h" [High,In progress] https://launchpad.net/bugs/152113
[18:17] <ubotu> Launchpad bug 116193 in tzdata "error upgrading tzdata_2007e to tzdata_2007f" [Critical,Fix committed] https://launchpad.net/bugs/116193
[18:18] <Spads> slangasek: haha yeah, I get those too
[18:18] <bdmurray> pitti: let me look at the report
[18:19] <sladen> slangasek: must admit, "communist" would not have been the first word that came to mind
[18:19] <pitti> doko: there is no pango in gutsy-proposed, BTW
[18:19] <slangasek> sladen: that's what's great about it :)
[18:20] <doko> pitti: and no glib?
[18:20] <pitti> doko: no
[18:20] <doko> fine =)
[18:22] <pitti> doko: was it meant to?
[18:22] <doko> I don't know
[18:23] <doko> pitti, cjwatson: http://people.ubuntu.com/~ubuntu-archive/queue/ doesn't show hardy yet
[18:27] <bdmurray> pitti: only gutsy needs verification?
[18:27] <pitti> bdmurray: yes, dapper/feisty/edgy are already verified by mvo and in -updates
[18:27] <pitti> bdmurray: mvo is on it, too, now (he just told me)
[18:28] <bdmurray> pitti: okay, I'll let him do it then. ;)
[18:29] <bdmurray> pitti: is -proposed available now?  I was looking at verifying the kopete crash with msn bug too
[18:29] <pitti> bdmurray: yes, it works; I'm just processing the queue, most stuff is flushed
[18:29] <pitti> bdmurray: will still take some 3 hours to build, though
[18:29] <pitti> (and get into the archive)
[18:31] <bdmurray> pitti: okay, it is still early here
[18:32] <mvo> pitti: verification daeon
[18:32] <mvo> done
[18:34] <bdmurray> mvo: I was looking at some update-manager bugs and noticed a couple with errors regarding medibuntu
[18:34] <cjwatson> doko: fixed
[18:35] <mvo> bdmurray: oh? can you give me a example please?
[18:37] <bdmurray> bug 153958
[18:37] <ubotu> Launchpad bug 153958 in update-manager "update-manager problem when upgrading from Feisty to Gutsy" [Undecided,Incomplete] https://launchpad.net/bugs/153958
[18:38] <doko> pitti: maybe accepting OOo at this point waiting for the bootstrap is not helpful
[18:39] <superm1> mvo, how come the extra repos aren't disabled before the package lists is updated to run update-manager's dist-upgrade?
[18:39] <superm1> i ran into something similar a few releases back, but never queried further into it
[18:39] <doko> trying to lower the build score by hand
[18:39] <superm1> mvo, it appears that they are disabled after the apt-get update
[18:40] <mvo> superm1: its important that it knows where it starts from, its important for the calculation of obsolete packages
[18:40] <superm1> mvo, but what about cases like this that the old repo is no longer valid?
[18:40] <mvo> superm1: it basicly compare the packages that are not downloadable anymore before and after the upgrade to figure out what is no longer needed
[18:40] <superm1> mvo, is there a better way to handle that
[18:41] <superm1> or not necessarily no longer valid, but perhaps not contactable at that time
[18:42] <superm1> perhaps catching the IOError, but still heading forward?
[18:50] <doko> cjwatson, pitti, seb128, Mithrandir: please approve gcc-4.2 in binary NEW
[18:58] <geser> does somebody know what happened to the accepted packages for gutsy-proposed?
[18:59] <geser> or where they are currently?
[19:17] <joshk> does any ubiquity-automation documentation exist, or am i going to have to trawl through the source
[19:17] <joshk> ?
[19:25] <bdmurray> I think there was a post to ubuntu-devel or ubuntu-devel-announce about it recently
[19:28] <evand> joshk: https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2007-October/002001.html
[19:46] <joshk> evand: cool! thanks.
[19:46] <joshk> so automatic-ubiquity will generate a preseed file after a normal attended install?
[19:48] <evand> no
[19:48] <joshk> oh, i get it
[19:48] <evand> you need to generate a preseed file yourself
[19:48] <evand> hit f6 at the CD bootloader (isolinux)
[19:48] <evand> and append url=http://yoursite/yourpreseedfile.seed
[19:49] <evand> the values in that file will get pulled into debconf
[19:49] <evand> which will be used by ubiquity
[19:49] <joshk> yeah, got it
[19:49] <evand> ok, great
[19:50] <evand> if you have any questions, feel free to ask them here (though #ubuntu-installer would be more appropriate) or drop me an email.
[19:50] <joshk> hm, partman crsahed
[19:52] <evand> Yikes.  If this was using the automated installer, file a bug against ubiquity and attach /var/log/syslog and the preseed file you used.
[19:52] <evand> well, either way file a bug :)
[19:52] <joshk> i'm betting my preseed file isn't working, since i last used it with feisty
[19:53] <joshk> in short: ubiquity/components/partman.py:360, assertion failed
[19:54] <evand> hrm, you may be missing a question or two.  Check it against the preseed file I posted in that email to ubuntu-devel-discuss.
[19:54] <joshk> yeah
[19:54] <joshk> brb
[19:59] <cjwatson> though it's still a bug if it fails an assertion
[19:59] <cjwatson> joshk: hey, didn't you use to hack on partman? we could do with more people working on ubiquity's partman integration ... ;-)
[20:00] <cjwatson> basically it'll fail that assertion if it doesn't go through ok_handler with partman-auto/{init_,}automatically_partition as the current question
[20:00] <cjwatson> which may be a little too fragile ...
[20:01] <lamont> Do you want to continue [Y/n]?
[20:01] <lamont> E: Internal Error, Could not perform immediate configuration (2) on libc6
[20:01] <lamont> hrm.  I hate it when dist-upgrade says that
[20:02] <cjwatson> at some point I must grab Ian and find out what that means
[20:04] <elmo> it's an apt thing?
[20:04] <elmo> I thought
[20:04] <elmo> and had to do with not being able to break dependency loops (or whatever) for essential (or virtually essential like libc6) packages
[20:04] <cjwatson> hmm, you're right
[20:10]  * mvo_ hides
[20:13] <pitti> geser: gutsy-proposed binaries should be accepted automatically (unless they are new, of course, but that shouldn't happen)
[20:15] <pitti> mvo_: ah, there you are; my /query timed out on 'mvo' :)
[20:16] <[knap]> hello, any ubuntu audio team member?
[20:16] <Amaranth> mvo_: We missed a couple things for the blacklist
[20:16] <Amaranth> mvo_: And maybe we should use --indirect-rendering again for nvidia now that the driver is fixed
[20:16] <Amaranth> mvo_: So people don't get the memory leak thing as quickly (or at all depending on their hardware)
[20:17] <ajmitch> hi
[20:18] <pitti> mvo_: thank you for verifying
[20:18] <[knap]> or anyone involved with alsa
[20:20] <ajmitch> Amaranth: why would compiz refuse to use gtk-window-decorator? it looked weird upgrading compiz on my laptop (was a few weeks old) & the window decorations changing
[20:21] <Amaranth> ajmitch: If you have emerald installed it uses it
[20:21] <ajmitch> I can run gtk-window-decorator manually & it worked, I just never saw where it got set
[20:21] <ajmitch> interesting
[20:21] <ajmitch> is that optional, or it just happens?
[20:25] <pitti> doko: meh, lots of -dbg
[20:30] <[knap]> why isn't this bug fixed? https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/93859
[20:30] <ubotu> Launchpad bug 93859 in linux-source-2.6.20 "[Feisty] Very low volume on Toshiba satellite a100-155" [Medium,Confirmed]
[20:30] <[knap]> it is araound since 7.04
[20:31] <[knap]> still happens with 7.10
[20:31] <pitti> doko: done
[20:34] <LaserJock_> [knap]: well, the simple answer is because nobody's fixed it
[20:36] <[knap]> I would fix it myself if I had the knowledge, does this have to be fixed with the alsa guys or here in ubuntu?
[20:37] <evand> isn't model=auto the default for snd-hda-intel anyway?
[20:37] <[knap]> evand the line "options snd-hda-intel model=auto" doesn't exist in /etc/modprobe.d/alsa-base
[20:38] <evand> [knap]: my point is that I'm not sure it needs to be.
[20:38] <[knap]> ok, I don't know but it fixes the problem
[20:49] <[knap]> so is there anything i can do to get this fixed?
[20:50] <LaserJock_> [knap]: you could talk to kylem or TheMuso about it, I think they're in charge of the audio team
[20:51] <[knap]> LaserJock_ ok i will try that, thanks
[21:02] <Vlet> How much space does a full mirror occupy?
[21:03] <LaserJock_> of the archive of .isos
[21:03] <LaserJock_> *or .isos
[21:03] <Vlet> the archives
[21:04] <LaserJock_> hmm, I think it's something like 40GB for i386 and amd64
[21:04] <geser> Vlet: for one or all archs? for one release or all?
[21:06] <Vlet> geser: Well, depends on how much space each release occupies. I would like to at least get each arch for gutsy. I know many people here on campus running ubuntu, so I thought a local mirror would reduce load on other mirrors
[21:09] <geser> 40 GB for i386 and amd64 for gutsy is a good estimate
[21:09] <Vlet> thanks. I just wanted to make sure the machine I have had enough space before I started rsyncing
[21:16] <geser> Vlet: http://www.ubuntu.com/getubuntu/mirror
[21:16] <cjwatson> for reference, feisty+gutsy main+restricted amd64+i386+powerpc+source is:
[21:16] <cjwatson> 30599976        /mirror/ubuntu
[21:16] <cjwatson> of course that may be a totally useless number to you if you want to mirror universe+multiverse too
[21:17] <geser> Vlet: according to that page the complete archive is around 170 GB (I found a site mentioning it around 200 GB)
[21:19] <LaserJock_> hmm
[21:19] <LaserJock_> I should talk to the CS department at my uni about mirroring
[21:20] <LaserJock_> 170GB isn't really that much space wise
[21:20] <LaserJock_> bandwidth around campus at least wouldn't be much
[21:20] <ivoks_> 230GB is official mirror of all archs
[21:20] <ivoks_> but that will grow with new releases
[21:21] <ivoks_> hr.r.u.c is nearing 1TB of traffic :)
[21:22] <ivoks_> (in two days)
[21:24] <Vlet> I seeded 86 gigs of the iso today :)
[21:25] <LaserJock_> it'd be interesting to get bandwidth numbers for all the official mirrors for the first week of Gutsy release
[21:25] <cjwatson> cjwatson@rookery:~$ du -s /srv/archive.ubuntu.com/
[21:25] <cjwatson> 174628444       /srv/archive.ubuntu.com/
[21:25] <cjwatson> FWIW
[21:26] <LaserJock_> thanks
[21:26] <cjwatson> non-ports
[21:26] <LaserJock_> does that go back to breezy?
[21:27] <ivoks_> breezy wasn't first ubuntu :)
[21:28] <LaserJock_> ivoks_: I'm aware of that
[21:28] <cjwatson> LaserJock_: no, breezy isn't on archive.ubuntu.com any more
[21:28] <LaserJock_> right
[21:42] <Burgundavia> LaserJock_: what we really need is more crappy bandwidth country mirrors, but another mirror is always good, I suppose
[21:43] <LaserJock_> Burgundavia: well, it'd probably be a uni-only mirror I'd guess
[21:44] <LaserJock_> I don't seem like my uni cares much for Linux in general, let alone Ubuntu
[21:44] <LaserJock_> but I might convince them that providing a mirror for the uni would be chep
[21:47] <Burgundavia> it is a good way to get Ubuntu in
[21:48] <LaserJock_> I gotta get them to stop shutting down linux labs
[21:48] <LaserJock_> :(
[22:42] <hunger> I get a invalid signature from the ubuntu archive automatic signing key on gutsy-proposed.
[22:42] <gcarrillo> hi all
[22:42] <gcarrillo> does anybody know of an online browser for GNU source?
[22:42] <hunger> gcarrillo: ?
[22:44] <gcarrillo> does that mean the question was unclear or that you don't know :)
[22:45] <hunger> gcarrillo: I don't know what you want.
[22:46] <hunger> gcarrillo: A online browser as in some app to look up docs on the internet? firefox springs to mind here... ;-)
[22:46] <gcarrillo> sorry no
[22:46] <gcarrillo> http://cvs.opensolaris.org/source/
[22:46] <gcarrillo> for example
[22:46] <gcarrillo> thats the source browser for opensolaris
[22:47] <hunger> gcarrillo: I am not aware of something like that.
[22:47] <gcarrillo> i wonder if theres something similar for gnu
[22:47] <gcarrillo> ok
[22:47] <hunger> gcarrillo: Some projects offer it, but more on a per-project base.
[22:47] <gcarrillo> i wanted to look at the source for echo
[22:47] <hunger> gcarrillo: There is no central entity that can collect all the source and stuff:-(
[22:48] <gcarrillo> i thought gnu was the central entity for the userland stuff like that
[22:48] <gcarrillo> like command line tools and such
[22:48] <hunger> gcarrillo: "apt-get source bash" will get it for you.
[22:48] <gcarrillo> oh cool
[22:49] <gcarrillo> thanks ;)
[22:49] <hunger> gcarrillo: Nothing gnu related, just normal ubuntu/debian goodness. You can grab the sources of everything in the main/universe/multiverse repositories that way.
[22:51] <gcarrillo> ok so where in my fs will that source live?
[22:52] <gcarrillo> once i've apt-get'd it
[22:53] <geser> gcarrillo: in a dir below $PWD
[22:53] <gcarrillo> oh sorry
[22:54] <gcarrillo> dumb question
[22:58] <mdke> how does one get a package removed from the archive? just a regular bug report?
[22:59] <geser> mdke: yes, plus the usual sponsoring for non-MOTUs/non-core-devs
[22:59] <mdke> geser: so I just subscribe the sponsor group and ask them to remove the package? I thought it was much harder :)
[22:59] <slangasek> hmm? what sponsoring is involved in removing a package?
[23:00] <sistpoty> mdke: you should have a good reason though :P
[23:00] <mdke> sure
[23:01] <mdke> slangasek: I dunno, I was asking
[23:01] <slangasek> yes, I'm questioning geser's response
[23:01] <geser> slangasek: only the ACK from a MOTU/core-dev is the reporter is none before subscribing ubuntu-archive
[23:01] <slangasek> ah
[23:02] <mdke> right, I see
[23:02] <geser> MOTUs/core-devs can directly subscribe ubuntu-archive to the remove request/bug
[23:03] <mdke> ok, I'll do that. Thanks
[23:07] <mdke> actually, perhaps it's not necessary. The packages i had in mind (ubuntu-quickguide and ubuntu-faqguide) don't appear to be in the archive, they are only in my tab completion for apt-get install... perhaps a remnant of having a system upgraded over many releases
[23:09] <geser> packages.u.c doesn't know any of them (also for past releases)
[23:10]  * mdke nods
[23:10] <mdke> my bad
[23:10] <mdke> anyway, useful to know how to do it, sorry for the timewasting
[23:10] <geser> np
[23:49] <purpleposeidon> To the individual who thought that using UUIDs for all of the disks was a good idea: Please type=UUID=9dd73528-841d-4ced-a0fc-be130e88a5de  at least 7 times.
[23:49] <purpleposeidon> I'm sure you'll soon be almost as good as me at typing it.
[23:55] <sistpoty> for i in 1, 2, 3, 4, 5; do echo UUID=9dd73528-841d-4ced-a0fc-be130e88a5de; done #:P
[23:55] <sistpoty> (sorry, couldn't resist)
[23:59] <TheMuso> sistpoty: lol. The slight inconvenience of typing it pays off in the end.
[23:59] <TheMuso> Many a time a UUID instead of a device node has helped me considerably.