2008/08/04/#launchpad.txt

[00:12] <onox> I just registered a branch, but I can't check it out
[00:12] <onox> how can I init the branch op LP?
[00:17] <mwhudson> onox: bzr init lp:~whatever/whatever/whatever
[00:34] <mtaylor> hey... my karma points went down in one of my projects!
[00:34] <mtaylor> what's up with that?
[00:36] <jml> mtaylor: I believe that karma degrades over time.
[00:36] <mtaylor> interesting...
[00:36] <mtaylor> but that means I can't just sit on my ass ... that I have to keep working to maintain my top spot!
[00:36] <mtaylor> bother
[00:36] <mtaylor> :)
[00:36] <jml> mtaylor: I've just typed help.launchpad.net/Karma into my browser to see what happens
[00:37] <jelmer> there was a time when karma only increased I think
[00:37] <jml> https://help.launchpad.net/YourAccount/Karma?action=show&redirect=KarmaCalculation
[00:38] <mtaylor> well... it is the "right" thing to do...
[00:38] <jamesh> mtaylor: it also helps to diversify, and get karma in multiple categories
[00:38] <Nafallo> yea, and the Swedish upstream Gnome translator owned all Karma :-P
[00:39] <mtaylor> hahaha
[00:40] <mtaylor> "Bug Marked as Fixed" is a karma-able action, but "Specification marked as merged" doesn't seem to be...
[00:40] <mtaylor> oh well
[00:41]  * mwhudson sighs
[00:41] <mwhudson> so this import suggests that there's some value for svn:special cscvs doesn't support
[00:41] <mwhudson> http://launchpadlibrarian.net/16298055/ascend-trunk-log.txt
[00:41] <mwhudson> turns out svn doesn't support it either:
[00:41] <mwhudson> mwh@grond:ascend-trunk$ svn up -r 416
[00:41] <mwhudson> ...
[00:42] <mwhudson> A    pygtk/interface/install-sh
[00:42] <mwhudson> svn: In directory 'pygtk/interface'
[00:42] <mwhudson> svn: 'pygtk/interface/.svn/tmp/text-base/config.guess.svn-base' has unsupported special file type '#!'
[00:42] <persia> mtaylor: Also, karma is influenced by everything everyone else does, so you get more if you do different things than other people.
[00:42]  * mwhudson stabs
[00:42] <mtaylor> persia: wow. fancy
[00:44] <jelmer> mwhudson, it seem it's perfectly valid to set svn:special on a file without that file being special
[00:44] <jelmer> it'll treat it as a normal file unless it starts with "link "
[00:44] <jamesh> mwhudson: and it shows that the UnknownSpecialFileType constructor wasn't tested
[00:45] <mwhudson> jamesh: yes, that too
[00:45] <mwhudson> jelmer: it seems that the svn client blows chunks on this repo though
[00:45] <jamesh> jelmer: the svn client doesn't generally let you fiddle with svn:special directly
[00:46] <jamesh> setting it to undefined values doesn't sound perfectly valid
[00:46] <jelmer> jamesh: The remote API does though
[00:47] <mwhudson> oh, so it's just the svn _client_ that's terrible?
[00:47] <jamesh> jelmer: right, but you run the risk of confusing the svn client
[00:48] <jamesh> mwhudson: symlinks are a client side thingee
[00:48] <jamesh> so this probably means that a normal file has svn:special set
[00:48] <jamesh> when it shouldn't have it set
[00:49] <onox> does code commits increase my karma? :)
[00:50] <jamesh> onox: not yet
[00:50] <mwhudson> it seems to be set to '*' on a bunch of files
[00:50] <mwhudson> ah well, not going to worry about it
[00:56] <jamesh> mwhudson: right.  If svn:special is set (to anything, I think), then the file contents are treated specially
[00:56] <jamesh> e.g. if the content is "link $target", then the file is treated as a symlink
[00:57] <mwhudson> and if it's set to something else, the svn client craps itself?
[00:57] <jamesh> if you have a normal file that starts with "#! /bin/sh" and has svn:special set, then it'll complain that it doesn't understand "#!"
[00:57] <mwhudson> ah right
[00:57] <mwhudson> oh, right
[00:57] <mwhudson> i'd become confused, i think i thought that svn:special was set to 'link target'
[00:57] <jamesh> the value of svn:special is not important: whether it is set or not is.
[00:58] <mwhudson> ok
[00:59] <jamesh> "bzr annotate" will probably blame me for the cscvs bug ...
[01:00] <mwhudson> heh
[01:01] <mwhudson> that's ok, i can just blame robert for all cscvs problems if you'd rather
[01:19] <awmcclain> "Teams" have ppas in addition to individuals, correct?
[01:19] <beuno> awmcclain, yeap
[01:39] <beuno> mpt, w00t, you made slashdot!
[02:04] <thumper> beuno: link?
[02:06] <beuno> thumper, http://developers.slashdot.org/article.pl?sid=08/08/03/1847220
[02:10] <thumper> ta
=== yannick changed the topic of #launchpad to: https://launchpad.net/ | Next meeting, all welcome: Thu 7 Aug 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions and spam reports: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com | Code Scanner is back!
[03:10] <Peng_> You can probably take the Code Scanner thing out of the topic now..
=== jml changed the topic of #launchpad to: https://launchpad.net/ | Next meeting, all welcome: Thu 7 Aug 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions and spam reports: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com
[03:11] <Peng_> :)
[03:12] <jml> (and who says we don't listen to our users?)
=== abentley1 is now known as abentley
=== abentley1 is now known as abentley
=== abentley1 is now known as abentley
[05:59] <kiko> <kiko> okay, flying out to toronto, see you all in some 16h!
[06:14] <elamou> i made an account on launchpad, but i've given only my first name
[06:14] <elamou> can i change it somehow?
[06:15] <elamou> or do i have to make a new account?
=== stub1 is now known as stub
[06:23] <spm> elamou: on the right hand side of your front page should be a 'Change details' link. Change your name via that.
=== Ursinha is now known as Ursinha-zzz
[06:25] <elamou> ok. i was in 'translations.launchpad.net/~...' and i could find it.
[06:25] <elamou> thanx
[07:05] <coolbhavi> hello please look: https://answers.edge.launchpad.net/launchpad/+question/41081
=== coolbhavi is now known as pulsar_work
[07:23] <\sh> good morning
[07:24] <emgent> moin \sh :)
[07:27] <Hobbsee> pulsar_work: that was filed 23 hours ago, on a weekend, and you're *already* bugging people about it?
[07:29] <pulsar_work> Hobbsee, here on weekends too we work ... I didnt know that lp admins will be on leave
[07:31] <Hobbsee> pulsar_work: most places don't work 24/7.
[07:32] <Hobbsee> pulsar_work: and as a general note, i'll also warn you that MOTU stuff also takes a while, and bugging them won't get your stuff done any quicker.
[07:32] <pulsar_work> Hobbsee, ok sorry for that
[07:32] <Hobbsee> no problem.
[07:33] <thumper> pulsar_work: answered, no public key#
[07:33] <thumper> pulsar_work: Invalid signature for right2bhavi@gmail.com: Signature couldn't be verified: (7, 9, 'No public key')
[07:33] <pulsar_work> thumper, I do have 3 keys
[07:33] <thumper> pulsar_work: I see that
[07:34] <thumper> pulsar_work: I don't know why it is Invalid
[07:34] <thumper> pulsar_work: but that is the error
[07:34] <pulsar_work> ok i ll update my pgp keys and add another key
=== pulsar_work is now known as coolbhavi
[08:57] <lifeless> mwhudson: is loggerhead down ?
=== poolie_ is now known as poolie
[09:05] <mwhudson> lifeless: it's working for me
[09:05] <mwhudson> using a fairly silly amount of ram, to be sure
[09:06] <lifeless> ah yeah
[09:06] <lifeless> I had a error before
[09:28] <persia> Is there a known bug about untruncated changelogs (e.g. https://launchpad.net/ubuntu/breezy/+source/ardour/+changelog) ?
[10:04] <mpt> beuno, ah, I wondered if that might happen
=== doko_ is now known as doko
[10:17] <\sh> hmm..who can I bug about launchpadlib? :)
[10:17] <wgrant_> mpt: Well done!
=== wgrant_ is now known as wgrant
[10:41] <thumper> \sh: you can bug one or more of flacoste, mars, salgado, barry, sinzui and I'm sure there are several others :)
[10:42] <\sh> thumper: thx :)
[11:13] <ignas> cprov: hi
[11:20] <gnomefreak> maybe its me but the java applet doesnt really work all that good, cant read it uploading screenshot
[11:21] <gnomefreak> http://farm4.static.flickr.com/3004/2730919869_06618ed9f9.jpg?v=0  is the applet screenshot
[11:22] <wgrant> gnomefreak: How's this Launchpad-related?
[11:22] <gnomefreak> wgrant: its LPs java applet
[11:22] <gnomefreak> oh damn
[11:22] <gnomefreak> never mind its freenodes
[11:22] <wgrant> Yes...
[11:50] <thorwil> found a typo on https://launchpad.net/+tour/branch-hosting-tracking: "And because they never touch your mailine they don’t need to ask for commit access." mailine -> mainline
[11:51] <theseinfeld> question, in the edge.launchpad.net where do you get the help anymore?
[11:52] <james_w> Hi all, I'm looking for an estimate of the timeframe when the API moves out of beta. I'm packaging launchpadlib, and want to have an idea of how much hassle there will be if a version makes it in to the stable release.
[11:52] <james_w> Are we talking Intrepid timeframe, launchpad 3.0, or beyond?
[11:53] <thumper> james_w: the only person I feel who can reliably answer that is kiko (who is travelling right now)
[11:53] <persia> I'm curious as to how stable the library code is?  Does it need more testing before it becomes a package?
[11:54] <james_w> thumper: cool, I'll ask him next time he's around, thanks.
[11:54] <theseinfeld> cprov. you might have the answer, where is the help hoover in the new launchpad? or is there any?
[11:55] <james_w> persia: thekorn is working on integrating it in to python-launchpadbugs, so we could let all of the scripts that use that try the library to see how buggy it may be.
[11:55] <theseinfeld> the so called InlineHelp...
[11:55] <theseinfeld> ?
[11:55]  * persia worries about aggressive adoption in a development release without some confirmation that stability can be achieved by FeatureFreeze (28th August)
[11:56] <james_w> I'm also after an impression of how much we are likely to see the headaches we have with python-launchpadbugs when launchpad changes
[11:57] <persia> I'll admit I'm not very happy with that one either, as the version released in Ubuntu frequently doesn't match the current version (and so may not work).
[11:57] <james_w> persia: true, though if everything goes through py-lp-bugs then we can at least switch back to using the html parsing code
[11:57] <persia> james_w: Well, I guess, but...
[11:58] <james_w> though the APIs at least stand a better chance of providing backwards compatibility
[11:59] <persia> james_w: If someone says that, I'll agree with you.  If nobody says that, I claim it's unsubstantiated hearsay :)
[12:00] <persia> (with "someone" defined as "someone who is in a position to speak authoritatively for the LP API", and probably is kiko :)
[12:10] <\sh> persia: if you want to be productive, please beg for "remove ez_setup" for wadllib and launchpadlib ,-)
[12:11] <persia> \sh: My philosophy about ez_setup is that if upstreams wish to make their code dependent on an unreliable resource, they may, but I will patch it out if I find it in Ubuntu.
[12:12] <persia> From an LP perspective, if they want LP tools to be available cross-distro, it may be easiest to use ez_setup, and let all the distros patch it out.
=== al-maisan is now known as al-maisan-lunch
=== al-maisan-lunch is now known as al-maisan
[12:20] <theseinfeld> so, in the launchpad, where do you get the help on? is there a link? hoover? or am I missing something?
[12:20] <andrea-bs> theseinfeld: help.launchpad.net ?
[12:21] <theseinfeld> yes, andreas-bs, i know about that, but there is no link to the specific components help like it used to be? is this a feature or a bug?
[12:22] <theseinfeld> we used to have it if you were hoovering, for example if you were in ppa, you could get help about ppa by just clicking something...
[12:22] <andrea-bs> theseinfeld: it is a feature: now there is the inline help
[12:22] <wgrant> It's not entirely replicated, though.
[12:23] <theseinfeld> it will be nice to add a button for more detailed help or popup window than the inline help
[12:24] <theseinfeld> for example, if i delete a package I get: Read the page help to understand how removals work.
[12:24] <theseinfeld> which there is no link whatsoever on the page, so you need to go to help.lp.net... PPA etc... that sucks!
[12:25] <bigjools> https://bugs.edge.launchpad.net/soyuz/+bug/195984
[12:25] <ubottu> Launchpad bug 195984 in soyuz ""Delete packages" page says "Read the page help"" [Undecided,Confirmed]
[12:29] <theseinfeld> added comments to th 195984 regarding this issue
[12:31] <bigjools> thanks for your comments
[12:31] <bigjools> this is causing some frustration, I'll see that it gets fixed soon
[12:49] <persia> bigjools: Re: bug #125987
[12:49] <ubottu> Launchpad bug 125987 in soyuz "Some uploads missing from +packages" [Medium,Confirmed] https://launchpad.net/bugs/125987
[12:49] <bigjools> persia: hello
[12:49] <persia> What is the use case served by presentation of partial information on /+packages?
[12:50] <persia> Personally, I'd be happy with only having the detail information available, but there may be some other use case.
[12:50] <bigjools> persia: I'm not sure I understand you
[12:51] <persia> To put that another way, I'd be happy not showing any information except on the detail info pages with everything (on a per type basis, or even on a per type & per-release basis)
[12:52] <persia> When I look at /+packages, I want to see the details, but I don't mind a little navigation.  On the other hand, partial information is useless to me.
[12:52] <persia> The question is: is partial information useful for someone else?
[12:53] <Hobbsee> persia: the use case for partial information is to show which packages have been uploaded most recently by the person, without making launchpad fall over.
[12:53] <persia> Hobbsee: Yes, but does anyone use that information for anything?
[12:53] <persia> (also, the page doesn't have that now due dropping multiple uploads of a package)
[12:53] <Hobbsee> persia: the reason it got cut from the full list, was because of the launchpad timeouts on those pages.  Thus, the launchpad way to remove the timeouts, was to remove the information used to generate them.
[12:53] <Hobbsee> persia: i'm just talking from the LP perspective.
[12:54] <persia> Hobbsee: I understand all that.  The question is whether it's worth bigjools trying to find a way to present some partial information without breaking things, or just build the batched complete detail pages and provide navigation.
[12:54] <Hobbsee> well, wsa the batched list still causing timeouts?
[12:54] <persia> Perhaps requiring the user to look at several different pages to see the information.
[12:55] <persia> The batch mode doesn't exist today.
[12:55] <Hobbsee> yes, but it used to
[12:55] <Hobbsee> i think, anyway.
[12:55] <wgrant> persia: It actually only shows published SPRs, so people who haven't uploaded since Edgy haven't uploaded anything.
[12:55] <wgrant> It was never batched AFAIK.
[12:55] <Hobbsee> they've done similar with the queues - cut down the maximum to some arbitary number, so the pages don't time out.
[12:56] <persia> wgrant: Well, people who uploaded in Dapper still get Dapper credit, but yes, that is another one of the ways in which it fails to be compete.
[12:56] <persia> s/compete/complete/
[12:56] <wgrant> I also wonder if it shows copies. Probably not.
[12:56] <persia> wgrant: Do you know of any use cases that benefit from partial displays?
[12:57] <wgrant> persia: No.
[12:57] <bigjools> Hobbsee: that is simply not true
[12:57]  * persia thinks that it may be easiest to skip having any default set of 60 or 150 or whatever packages, and just have the details available somewhere
[12:57] <persia> (perhaps with lots of different URLs if otherwise there are data collection issues)
[12:57] <Hobbsee> bigjools: which?
[12:58] <bigjools> persia: please add your comments to the bug, but a lot of people use that page for MOTU reviewing
[12:58] <bigjools> Hobbsee: the queue page
[12:58] <wgrant> bigjools: persia being one of the primary ones...
[12:58] <persia> bigjools: Yes, I'm one of the members of MOTU Council.  Partial information is completely useless for that purpose.
[12:58] <bigjools> heh ok
[12:58] <persia> I'm wondering what other use cases require presentation of that information.
[12:58] <bigjools> ok please write on the bug then
[12:59] <bigjools> persia: and please explain exactly what "partial" means to you
[12:59] <Hobbsee> bigjools: https://edge.launchpad.net/ubuntu/intrepid/+queue?batch=500 ?  Citation needed.
[12:59] <bigjools> Hobbsee: I worked very hard on making the queue page quicker, I did not reduce the batch size to do that
[12:59] <Hobbsee> that used to work most of the time, unless it timed out
[13:00] <persia> bigjools: partial is anything other than listing every upload for every package
[13:00] <Hobbsee> ahhh.  well done for helping it partially, then.
[13:00] <bigjools> persia: we can't do that without batching
[13:00] <persia> Also, your proposed solution looks like it works for me (as I said in my last comment).  I'm just wondering if you are trying to do more than you need.
[13:00] <persia> bigjools: Completely understood.  I'm not asking you to do that without batching.
[13:01] <wgrant> It used to work fine for everybody except one or two people.
[13:01] <bigjools> Hobbsee: it will time out still for batches with a lot of custom uploads, but that will be fixed soon also
[13:01] <persia> I don't even mind having to look up data by type and release target (so having lots of URLs).
[13:02] <bigjools> the batch size max of 300 is a Launchpad global
[13:02] <Hobbsee> why did it change?
[13:02] <bigjools> anyway guys I have a call right now, I'll catch up with you in a but
[13:02] <persia> On the other hand, I wonder if someone uses the data as presented: if not, it's worth looking into bug #43020 directly, and not bothering with a redesign of /+packages, instead just providing navigation.
[13:02] <bigjools> bit, too
[13:02] <ubottu> Launchpad bug 43020 in soyuz "Person +packages page should be batched (dup-of: 125987)" [Medium,Confirmed] https://launchpad.net/bugs/43020
[13:02] <ubottu> Launchpad bug 125987 in soyuz "Some uploads missing from +packages" [Medium,Confirmed] https://launchpad.net/bugs/125987
[13:29]  * persia is happy that the mooted 1024-byte limit on comments was never implemented
[13:30] <persia> bigjools: Thoughts summarised in the bug: please present to interested parties for review.  I'm still curious if you hear of any workflows beyond technical review of the people.
[13:30] <bigjools> persia: ok thanks for the input
[13:44] <\sh> guys, is there any way to see the applications which I gave auth to?
[13:46] <\sh> https://staging.launchpad.net/~shermann/+oauth-tokens that's it
[13:46] <\sh> thx \sh ,-)
[14:40] <sistpoty|work> hm... as I seem to have missed Rinchen again, any other lp admin around who could reset my password?
[14:56] <elamou> i cannot upload my program to launchpad, though i can download other people's software..
[14:56] <elamou> i generated a pair of rsa keys
[14:57] <elamou> i entered the public key at my account
[14:57] <elamou> bzr push –create-prefix sftp://bazaar.launchpad.net/~teamname/product/branchname
[14:57] <elamou> is that right?
[15:12] <LarstiQ> elamou: you're missing a hyphen on --create-prefix, but the most common problem is that either your supplied username to login is wrong, or your private key doesn't match the public you uploaded to launchpad. Can't say much more without an actual error.
[15:14] <elamou> bzr push sftp://bazaar.launchpad.net/~teamname/product/branchname   if i use this, where do I enter my username?
[15:14] <LarstiQ> elamou: sftp://lpuser@bazaar....
[15:14] <LarstiQ> elamou: or in ~/.ssh/config
[15:15] <LarstiQ> elamou: or with a sufficiently young bzr, 'bzr launchpad-login'
[15:15] <elamou> bzr launchpad-login -> No Launchpad user ID configured.
[15:16] <LarstiQ> elamou: eh, specifically, `bzr launchpad-login <yourlogin>`
[15:16] <LarstiQ> elamou: see `bzr help launchpad-login` for more info
[15:18] <elamou> and it will upload avery file in the . directory?
[15:22] <LarstiQ> elamou: if you push to there, and then from somewhere else branch it, you will get all the same commits as you had locally
[15:22] <LarstiQ> elamou: changes you don't commit don't end up there
[15:23] <elamou> LarstiQ,  "if you push to there, and then from somewhere else branch it" can you explain this plz?
[15:24] <elamou> i did a "bzr init" in the folder where my source code is
[15:24] <elamou> and, in that folder, i run the previous commands
[15:33] <LarstiQ> elamou: then nothing will happen
[15:33] <elamou> yeah. nothing happens
[15:33] <LarstiQ> elamou: you need to add files first, commit them, and then you have a revision you can push.
[15:34] <LarstiQ> elamou: http://doc.bazaar-vcs.org/bzr.dev/en/mini-tutorial/index.html
[15:43] <elamou> worked!! thanx LarstiQ
[15:43] <LarstiQ> elamou: glad to be of help.
[15:49] <sistpoty|work> Rinchen: just received your mail regarding resetting my lp password... so it seems that I used the wrong email address, right?
[15:50] <salgado> sistpoty|work, no, that shouldn't cause the application error you saw.  you didn't get an OOPS or anything like that, right?
[15:50] <sistpoty|work> salgado: no, just the error message
[15:50] <sistpoty|work> salgado: however I just tried with stefan.potyra@informatik.uni-erlangen.de and that worked...
[15:53] <sistpoty|work> salgado: want me to file a bug, now that I'm logged in again? :)
[15:54] <salgado> sistpoty|work, please do!  and sorry for the inconvenience
[15:54] <sistpoty|work> salgado: ok... and no problem... that way I had an excuse to not work on ubuntu and enjoy my weekend *g*
[15:55] <salgado> lol
[15:59] <sistpoty|work> salgado: bug #254672 filed
[15:59] <ubottu> Launchpad bug 254672 in launchpad "forgotten password does not work" [Undecided,New] https://launchpad.net/bugs/254672
[16:00] <salgado> sistpoty|work, thanks!  I just managed to reproduce it, so it should be fixed soon
[16:00] <sistpoty|work> salgado: cool, thanks a lot!
[16:06] <andrea-bs> Hello! Will Launchpad provide a PQM for hosted Bazaar branches, in future?
[16:08] <Rinchen> sistpoty|work, think so
[16:08] <sistpoty|work> Rinchen: situation already cleared ^^... thanks a lot
[16:08] <Rinchen> sistpoty|work, awesome
[16:10] <andrea-bs> Rinchen: the 1.99 milestone doesn't have a data target
=== ursula_ is now known as Ursinha
=== Rinchen is now known as Rinchen-sprint
=== salgado is now known as salgado-lunch
[17:45] <tacone> any way to get my hands on the new launchpad api ?
=== salgado-lunch is now known as salgado
=== matsubara is now known as matsubara-lunch
=== DasIch_ is now known as DasIch
=== abentley1 is now known as abentley
[19:14] <lbo> hi lp! I am new to lp and need a pointer to how to create a download zip file.
[19:17] <lbo> Anyone home?
[19:22] <nandersson> Are there plans to integrate the packaging (.deb's) in Launchpad?
[19:38] <tacone> nandersson: I guess that would be something quite hard to do :)
=== matsubara-lunch is now known as matsubara
[19:38] <nandersson> tacone, Well, I never assumed it would be easy :D
[19:39] <nandersson> tacone, I thought that there might be something similar to OpenSUSEs Build Service in the works
[19:39] <nandersson> i.e to connect the VCS to a Build Service
[19:39] <tacone> nandersson: that seems an entire different thing to me. am I wrong ?
[19:40] <nandersson> tacone, Well, it's an extension. VCS is for developing, Build Service is for packaging
[19:41] <tacone> launchpad has PPA'a
[19:41] <tacone> PPA's
[19:41] <nandersson> tacone, a Build Service would be the natural, next step
[19:41] <tacone> you send a package source and it builds it and insert it into your personal repository
[19:42] <tacone> but you have to package the software source yourself, and then upload it to launchpad.
[19:42] <tacone> is that what you were asking ?
[19:42] <LarstiQ> nandersson: what, exactly, does a Build Service do?
[19:43] <nandersson> LarstiQ, The Build Service takes the source code and package it in a .deb, .rpm or whatever. OpenSUSEs build service at https://build.opensuse.org/ has support for CVS and SVN
[19:44] <nandersson> LarstiQ, and it's pretty distro agnostic, they build for Ubuntu, Debian, Red Hat, Centos + their own platforms
[19:44] <nandersson> tacone, I'll look into PPA. Found a nice link - thanks :)
[19:45] <nandersson> LarstiQ, I think it will be of great help for ISV in order to package their code for multi platforms/distros
[19:46] <nandersson> LarstiQ, I've never packaged a package myself using Build Service, but it looks like they have a web frontend to the usual /debian directory.
[19:46] <tacone> nandersson: that can work but I am doubtful it will meet the ubuntu/debian quality requirements.
[19:47] <LarstiQ> nandersson: so, from an upstream release, it adds debian/ and then builds a .deb? As tacone said, I'm very doubtful the actual packaging would be solid.
[19:48] <LarstiQ> nandersson: for the last step, building a .deb from upstream source + packaging information, that is what PPAs do
[19:48] <nandersson> tacone, I guess it's all scripted. They resolve dependencies and everything. OpenSUSE is going to use this build service to build the entire next distribution OpenSUSE 11.1
[19:49] <tacone> nandersson: I shuoldn't say this, because I don't really know enough the other distribution, but I guess debian and ubuntu have higher quality requirements and guidelines for packaging software.
[19:50] <nandersson> In the build service, they start up a Xen-instance with the distro they are building for. The Build Service then checks all dependencies
[19:50] <LarstiQ> tacone, nandersson: if software is sufficiently standardized, this approach can work.
[19:50] <LarstiQ> ref python eggs, java/.net, etc.
[19:50] <nandersson> tacone, Well, if you have stuff that can't be automated (Which is a bad approach of course) I guess it'll be hard
[19:50] <tacone> LarstiQ: no doubt it can *work*. I doubt people would want it anyway
[19:51] <nandersson> The build service make sure that you don't break the repository - that everything works
[19:51] <nandersson> of course it doesn't solve bugs
[19:51] <LarstiQ> nandersson: well, a large part of packaging is integration
[19:51] <tacone> nandersson: launchpad *has* a build environment.
[19:51] <tacone> nandersson: to use launchpad ppa you actually send the package source, not a binary.
[19:52]  * tacone stops here, it's talking about things he's not enough familiar with :)
[19:53] <nandersson> tacone, Yes, but I guess you still have to do that work manually - I saw the movies with Daniel Holbach in the Youtube-channel UbuntuDeveloper
[19:53] <tacone> nandersson: btw, I find Suse's a great effort.
[19:53] <tacone> nandersson: with Suse's too you have to work manually. check their help pages.
[19:54] <nandersson> tacone, I have to look into that. I'm going to investigate how the PPA works as well :)
[19:54] <tacone> nandersson: http://en.opensuse.org/SUSE_Build_Tutorial
[19:54] <tacone> look at the code samples. does that look automatic to you? :-)
[19:55] <nandersson> tacone, hehe, well, not too automatic :D
[19:56] <cprov> nandersson: one can say there is a parellel between openSuse and PPAs
[19:57] <cprov> nandersson: openSuse requires you to create the SRPM spec file and then allow you to build the package in a controlled environment
[19:57] <nandersson> cprov, Yes, it looks like it. I'm going through the PPA Quick Start
[19:57] <cprov> nandersson: PPA requires you to create the debian source and then builds it automatically for you in a controlled env.
[19:58] <nandersson> cprov, Yes, they run every job in a virtual instance
[19:58] <cprov> both use Xen-based build environment
[19:59] <tacone> cprov: automated ways to create debs _do_ exist. check checkinstall ( https://help.ubuntu.com/community/CheckInstall ) for instance. But that's not what you need to prepare a real package. dependencies, copyrights, patches.. all require manual intervention. that's true for every distro.
[20:00] <cprov> tacone: I didn't say they don't exist.
[20:00] <tacone> sorry cprov, I was meant to target the message to nandersson
[20:00] <tacone> not you. mistake :)
[20:00] <nandersson> :)
[20:00] <cprov> tacone: np :)
[20:00] <nandersson> tacone, thanks for the link
[20:01] <tacone> at your service nandersson.
[20:01] <nandersson> tacone, I found "Packaging Overview"
[20:02] <nandersson> I liked the videos with Daniel Holbach, but my reaction was that it looks quite manual - and a lot of that job could be done more easy through Launchpad - but I might be wrong
[20:02] <tacone> nandersson: beside automatic launchpadding everything, if you're interested in packages, I suggest you to check the #ubuntu-motu channel. there you can find more expert and critique answers to your questions.
[20:02] <nandersson> tacone, I'll have a check there! Thanks
[20:02] <tacone> nothing.
[20:39] <awmcclain> Hrm. This is frustrating. How do I see which team owns a project?
[20:41] <beuno> interesting...  I can't find it either
[20:41]  * beuno pokes mp
[20:41]  * beuno pokes mpt
[20:41] <beuno> ah, he's not here
[20:41] <beuno> awmcclain, can you file a bug for that?
[20:41] <salgado> beuno, awmcclain, there's a bug for that already
[20:41] <awmcclain> Hrm. Boo for that.
[20:41] <beuno> salgado, cool, thanks
[20:42] <awmcclain> Can I search teams?
[20:42] <salgado> awmcclain, launchpad.net/people
[20:45] <awmcclain> Great, thank you.
[20:46] <awmcclain> Does registering a hosted branch actually create the branch? Or do I need to branch from trunk, then push to the new branch?
[20:52] <thumper> registering a branch doesn't create it
[20:52] <thumper> but it does show you where to push :)
[20:54] <tacone> any news about the launchpad api ?
[20:55] <geser> tacone: besides https://help.launchpad.net/API/launchpadlib?
[20:55] <tacone> don't know what is that. is that the new api they were talking about ?
[20:58] <geser> tacone: if you mean the new API from the last LP announcement, than yes, that's the currently available documentation for it
[20:58] <tacone> ok
=== mcasadevall_ is now known as NCommander
[21:59] <mrevell> beuno: hey
[21:59] <beuno> mrevell, evening'
[21:59] <seb_kuzminsky> how do i remove a branch from my launchpad account?
[22:02] <tacone> seb_kuzminsky: you mark it as abandoned. it won't get deleted, but it won't show anymore in the list.
[22:02] <tacone> seb_kuzminsky: there's also a delete button. a small red circle
[22:02] <seb_kuzminsky> hm, how do i do that?
[22:03] <tacone> you can change the status of the branch in the details section of your branch
[22:03] <tacone> or you can press the deletion button on the branch main page.
[22:03] <salgado> seb_kuzminsky, there's a link with a '-' sign right beside the name, on a branch's page
[22:03] <seb_kuzminsky> oh i see, the little - by the name, thanks!
[22:04] <seb_kuzminsky> that removes my subscription ot the branch...  it goes away when there are no more subscribers?
[22:06]  * tacone doesn't know.
[22:06] <seb_kuzminsky> oh well, it went away from my list of branches so i'm happy, i guess ;-)
[22:06] <tacone> seb_kuzminsky: what's your problem ? you don't want to see the branch again or you have some privacy problem ? (password etc)
[22:06] <seb_kuzminsky> thanks!
[22:06] <tacone> oh, ok nice.
[22:07] <seb_kuzminsky> no privacy issue, it's just outlived its usefulness and i dont want to waste any resources
[22:07] <tacone> ok
[22:08] <seb_kuzminsky> thanks again, problem solved, bye :-)
=== Martinp24 is now known as Martinp23
=== oojah_ is now known as oojah
[23:09] <directhex> can someone be a sweetheart & remove a couple of files from my PPA, so i can finish setting up a package before going to bed?
=== mcasadevall_ is now known as NCommander
=== salgado is now known as salgado-afk
[23:38] <RainCT> directhex: "a couple of files" = "some packages"?
[23:38] <RainCT> directhex: if yes, there is a "delete packages" link
[23:38] <directhex> RainCT, i uploaded an incorrectly versioned package, and marked it for deletion. i suspect some race conditions came into play, and i now have a stray .dsc and i386 deb
[23:39] <RainCT> oh
[23:39] <directhex> whereas the lpia and amd64 debs are correct
[23:39] <RainCT> I don't know then, sorry
[23:40] <directhex> i could probably delete it again & wait a few hours. maybe.
[23:40]  * RainCT is off now, good night