[01:09] <selves> having trouble getting the midi preferences straight. the inputs and outputs are greyed out
[03:33] <michaelh1> Hey, the blueprint icons aren't showing up for me in Firefox 6 on lucid.  Chrome is fine.  For example: https://blueprints.launchpad.net/cortex-strings/+spec/add-to-newlib has no icons next to the state or users.
[03:34] <mwhudson> michaelh1: are you logged in?
[03:34] <micahg> wfm
[03:34] <wgrant> micahg: Are you using fglrx?
[03:34] <wgrant> Bah.
[03:34] <wgrant> michaelh1: ^^
[03:34] <micahg> wgrant: no, nvidia
[03:34] <wgrant> michaelh1: fglrx + firefox >=4 have a bug when rendering large images, like Launchpad's sprites.
[03:35] <michaelh1> wgrant: yes, yes I am.  Large images?  These are the tiny icons like ! and the user guy.
[03:35] <michaelh1> Are they slices of a bigger image?
[03:35] <wgrant> michaelh1: They are, or you'd be loading dozens of images per page load.
[03:36] <michaelh1> wgrant: interesting.  Any work around?
[03:37] <wgrant> michaelh1: Probably not unless you are AMD. https://bugzilla.mozilla.org/show_bug.cgi?id=646611
[03:37] <michaelh1> Heh.  One day I'll find a distro+kernel+graphics combo that works everywhere...
[04:08] <wgrant> michaelh1: Actually, looks like it's fixed in either fglrx 11.8 or Firefox 7, since I've been using Firefox on oneiric for two days now, and the icons exist.
[08:33] <mrevell> Hi
[09:06] <czajkowski> mrevell: morning
[09:08] <mrevell> hey czajkowski
[09:10] <czajkowski> mrevell: how are you this fine and sunny morning
[09:10] <mrevell> czajkowski, Hey! otp at the moment
[09:10] <mrevell> but very well, other than losing my glasses
[09:11] <czajkowski> OTP ?
[09:13] <mrevell> czajkowski, sorry, on the phone
[09:13] <czajkowski> np enjoy
[09:24] <htorque> hello everyone! i changed my mail address in launchpad (and removed the old one) but can only log in at ubuntu single sign-on with my old address - should i expect a delay or is there something wrong?
[09:25] <maxb> htorque: Ubuntu SSO has its own completely separate idea of your email addresses. You'll need to make the changes there too
[09:25] <maxb> This is not at all clear, unfortunately, especially when SSO can masquerade as part of the LP UI
[09:26] <htorque> it's not clear because the login mask told me i could use LP credentials... ;-)
[09:26] <wgrant> htorque: Well, you can't log into Launchpad with that new address either.
[09:27] <wgrant> Launchpad doesn't manage authentication itself. It delegates that to login.launchpad.net -- which, confusingly, is just a different theme for login.ubuntu.com.
[09:27] <htorque> wgrant: oh, true.
[09:27] <wgrant> Its account database is separate from Launchpad.
[09:28] <htorque> so what exactly did i achieve by changing the address in my launchpad user page?
[09:28] <czajkowski> wgrant: ye do like to complicate things at times :)
[09:28] <henninge> htorque: which address LP uses to contact you
[09:28] <wgrant> htorque: You changed where Launchpad will send notifications, and which addresses LP will recognise as being you.
[09:29] <wgrant> htorque: eg. when your address shows up in packages, bug emails, etc.
[09:29] <htorque> i see
[09:29] <wgrant> czajkowski: This is all pretty horrible, yes.
[09:29] <htorque> that's rather confusing tbh
[09:29] <wgrant> 18 months ago login.launchpad.net used the Launchpad account database. Then it was split out.
[09:29] <wgrant> And Launchpad was then going to fairly soon become a generic OpenID consumer, and login.launchpad.net would go away.
[09:29] <wgrant> But the second phase of the plan hasn't happened yet.
[09:30] <wgrant> So we remain stuck in the rather confusing and suboptimal situation you see here :)
[09:30] <cheater> hi
[09:30] <cheater> wgrant, are you from the LP team?
[09:30] <htorque> anyways, i think i managed to change the address now - thanks for your help :-)
[09:30] <wgrant> cheater: I am. There are quite a few of us here now.
[09:31] <cheater> cool
[09:31] <czajkowski> wgrant: is that written sowmehere the confusion over the change of address being the SSO ?
[09:31] <cheater> i have found quite a few usability problems in LP that stop me from switching over from bitbucket. I've written down a user story. where could i submit it?
[09:31] <cheater> i was told launchpad-dev was a good place to ppost it
[09:31] <wgrant> cheater: A few of us have read it, but yeah, launchpad-dev is a good place to send it.
[09:32] <wgrant> cheater: I have a few comments :)
[09:32] <cheater> cool let me do it
[09:32] <wgrant> But I agree with most of your points.
[09:32] <cheater> cool
[09:32] <cheater> want to talk about it? or post comments on the page?
[09:32] <czajkowski> I think lp has some flaws but is a lot better than other tools. but like everything, things can always be improved upon
[09:33] <czajkowski> and with the likes of mrevell working on some usabilty stuff and changes coming, hopefully make it nicer. I do like the fact it's one of the easier teams to log a bug about and people will actually discuss them and take the feedback on board
[09:34] <wgrant> I'm not sure if mrevell's seen it yet. Probably not, given timezones.
[09:34] <wgrant> mrevell: Did http://cheater.posterous.com/launchpad-not-an-alternative-for-bitbucket find its way over to you?
[09:34] <henninge> czajkowski: thank you for the praise ;)
[09:34] <cheater> let me make a post at dev then
[09:34] <cheater> (the list seems fairly quiet)
[09:34] <wgrant> cheater: That's a good way to let lots of people see it.
[09:35] <czajkowski> henninge: hello there :) yer welcome. but it is to be fair. I've logged some issues, most have had some decent feedback on them and people do take the time to comment on them
[09:35] <wgrant> cheater: Lots of discussion happens on IRC, plus like half the team was on leave yesterday.
[09:35] <czajkowski> benefits everyone really
[09:39] <czajkowski> cheater: your page is interesting, but some of the stuff on the about a person page I guess is specific for Ubuntu people, which a lot of people on LP are
[09:39] <czajkowski> so signing the CoC and member since, and contact details are important to see
[09:41] <wgrant> cheater: Also, all that absence of information is only shown on *your* page. Not anyone else's.
[09:41] <wgrant> The person page wasn't designed as a dashboard, so it doesn't usefully behave as one.
[09:41] <czajkowski> and as a person on the membership board I do want to see when a person joined and signed the CoC tbh
[09:42] <cheater> OK. bear in mind that does not help me further because it is not important to me as a typical developer breadeater.
[09:44] <cheater> is it *really* that important to see this on the person's profile straight away though
[09:44] <cheater> why is it ever important when someone joined?
[09:48] <czajkowski> cheater: hmm I guess it has to be generic enough to cover all aspects and a lot of Ubuntu people login in there plus many others best case fits, which I guess in this case doesn't mean it ticks all the boxes for you though
[09:49] <cheater> i think this is some sort of schizophrenic situation where the same area tries to do something geared for the general audience yet tries to facilitate a very specific usecase very stongly
[09:49] <cheater> +r
[09:50] <wgrant> Mmm. It's a combination of that and being designed in 2005.
[09:52] <RAOF> I know jml wanted a developer dashboard where launchpad.net/~ is now, but there's always something more pressing to work on.
[09:55] <nigelb> heh, fond memories of this dicussion yesterday with wgrant :)
[09:56] <czajkowski> you are never ever going to please thousands of users, best one can hope for a a smooth user experience and easy to use, yes there may be areas that you won't use but thousands of others do use. Perhaps logging a bug if those fields are not used they are not displayed?
[09:56] <cheater> i think it's not the case of pleasing everyone at once
[09:57] <cheater> you can please different people in different ways
[09:57] <cheater> you can have two different kinds of user info page
[09:57] <nigelb> There's a lot of suckage of launchpad UI compared to other similar sites out there.
[09:57] <cheater> if what you have works for the ubuntu team that's great and i'm not saying throw it away
[09:57] <cheater> but make an alternative for people with my usecase
[09:58] <nigelb> github/bitbucket have a similar UI, and moving from that to Launchpad is a bit confusing.
[09:58] <nigelb> Launchpad is awesome for a sort of project where you expect it to be slightly more formal.
[09:58] <cheater> websites like github, bitbucket, ohloh, and a thousand others are proof that such a usecase not only exists but is very prominent
[09:58] <cheater> nigelb, i think ease of use has nothing to do with formality at all
[09:59] <nigelb> cheater: The thing is, LP is easy to use once you figure it out. There's not much help initially how to do stuff like create a project.
[09:59] <nigelb> I wish that bit were different.
[09:59] <cheater> nothing that you have to figure out is easy to use
[09:59] <nigelb> I disagree.
[09:59] <cheater> the rubik's cube is easy to use
[09:59] <cheater> you have no figuring out
[10:01] <nigelb> The UI workflow is different. You have to figure it out.  When you use another UX and try to look for the same, its not easy to figure out. That's where you get hit the first time you use LP.
[10:02] <nigelb> I use github and Launchpad.
[10:02] <cheater> i didn't have to figure out gmail when i first used it and it's a very complex program
[10:02] <nigelb> I know how different they are.
[10:02] <cheater> and it was so much different from previous web clients i have used
[10:04] <cheater> er mail clients
[10:07] <nigelb> Personally, I do agree that the UX could use a lot of rework. BUt, there exists some good things once you get used to it.
[10:07] <nigelb> (that's what I was trying to say, but failed at conveying)
[10:08] <cheater> i'm not saying there are no good things!
[10:08] <cheater> i'm just saying there are quite a few bad things.
[10:08] <nigelb> I agree. :)
[10:08] <cheater> :)
[10:08] <nigelb> Some bits make us all cry, yeah.
[10:09] <cheater> i think maybe it would be a good idea to separate the UX for the ubuntu team and see it as a different use case from the way that a day to day developer would use LP?
[10:10] <nigelb> With the little that I know about LP, I know that's going to be hard :)
[10:10] <nigelb> of course, someone can correct me :)
[10:10] <cheater> why would it be?
[10:11] <nigelb> A lot of things will get complicated
[10:11] <cheater> can you give me an example?
[10:12] <nigelb> Well, for one, what happens when someone uses both sides of it.
[10:13] <nigelb> Ubuntu developer as well as some other projects
[10:13] <cheater> nothing stopping him from using both
[10:14] <cheater> it's just a different view
[10:14] <cheater> but really! it's not only about things like that single user page
[10:14] <cheater> it's about two different pathways through the website
[10:14] <cheater> user stories, not just single web panels
[10:15] <cheater> but yes, if you're talking about the user are there could very well be a second user panel which shows the data in a similar way to github, bitbucket, ohloh, even facebook
[10:17] <cheater> but there are lots of other problems that can be solved without stomping on the ubuntu use case
[10:17] <cheater> sich as the activity log being fairly useless
[10:18] <cheater> or such as there being no facility to branch stuff via the web interface
[10:18] <wgrant> cheater: that's one thing I'm not sure about. Why would you want to branch stuff via the web interface?
[10:18] <cheater> or having to initialize the bzr repository from my pc with an emmpty push whereas lp could do that for me
[10:19] <wgrant> In the LP model, I just work out which project I want, run 'bzr branch lp:someproject', cd someproject, hack hack commit hack commit, 'bzr push lp:~wgrant/someproject/my-branch'
[10:19] <wgrant> Done.
[10:19] <wgrant> Because bzr doesn't need server-side branching.
[10:19] <wgrant> git does, because it doesn't have a workable implementation of branch stacking.
[10:19] <cheater> of course it doesn't *need* it
[10:20] <cheater> but it would be a good idea to have it
[10:20] <nigelb> what purpose does it serve.
[10:20] <cheater> the problem is that your workflow is something that needs figuring out
[10:20] <cheater> there's just too much figuring out to do with lp
[10:21] <wgrant> Well, we didn't have people trying to branch through the web UI much before GitHub convinced everyone that it was necessary.
[10:21] <cheater> it's like a car but there's no steering wheel and the pedals are replaced with a ddr mat
[10:21] <cheater> wgrant: i think you are missing the idea of branching via the web ui
[10:21] <cheater> it's not only a necessity because of the technicalities behind the git format
[10:22] <cheater> it's also a good idea because of several reasons
[10:22] <cheater> 1. launchpad uses some magical repository url strings that i can use in order to make things go into a new project. this is great for advanced users, useless for new users
[10:23] <cheater> 2. letting lp know where i branched something from can let it display an information about me that i have branched from that project
[10:23] <cheater> which creates a network of people
[10:23] <cheater> which is nice and allows you to do a lot.
[10:23] <wgrant> Ah, more GitHub-centrism :P
[10:24] <wgrant> Projects in LP aren't restricted to a particular person.
[10:24] <cheater> i have never used github.
[10:24] <cheater> well maybe once 3 years ago to browse some code
[10:24] <wgrant> Well, bitbucket basically follows the GitHub model.
[10:24] <wgrant> Where only authorised people can have code in a project.
[10:25] <wgrant> Whereas in Launchpad, a project transcends people. Anyone can push up their own branches inside the project.
[10:25] <cheater> ok well then if projects do not belong to a particular person then replace "person" with "project" up there and you have something that applies to launchpad
[10:25] <wgrant> 20:23:29 < cheater> 2. letting lp know where i branched something from can let it display an information about me that i have branched from that project
[10:25] <wgrant> It's clear you're branching from that project, because you're pushing into it.
[10:25] <wgrant> Which means we're just down to docs, basically, which are pretty terrible, yes.
[10:25] <cheater> down to docs?
[10:26] <cheater> what does that mean?
[10:26] <wgrant> Your first point was that documentation/guidance is lacking.
[10:26] <cheater> no, my point was that figuring out needs to be done where it is not needed.
[10:27] <wgrant> Branching through the web UI is one way of providing guidance.
[10:27] <wgrant> There are others.
[10:27] <cheater> if you think people should have to use any documentation to make basic use a web service then that web service is bound to fail
[10:27] <cheater> yes inceed
[10:27] <cheater> indeed
[10:27] <wgrant> cheater: Documentation shouldn't be a separate thing.
[10:27] <wgrant> It has to be integrated.
[10:28] <cheater> i see what you mean there
[10:28] <wgrant> It's still documentation, even if not in the sense that people usually think of it.
[10:28] <cheater> yes, but "help files" don't work
[10:28] <cheater> that is what i meant
[10:28] <wgrant> Right.
[10:28] <wgrant> LP sadly has lots of those.
[10:28] <cheater> you're right, inline guides and tips are better than nothing
[10:28] <wgrant> And it is reasonably hopeless, because nobody ever looks at them.
[10:28] <wgrant> Because they shouldn't have to.
[10:28] <cheater> but they are not as good as a user interface that funnels me towards what i want to do
[10:29] <cheater> i think we are  in good agreement there
[10:29] <wgrant> Right, I count that as a user interface with documentation built in. Just not in the "oh, why don't I throw a 10-page wall of text at you, that will help" sense.
[10:30] <czajkowski> cheater: interesting post. not being smart now, but have you logged the bugs with those issues ,so they can either be marked wishlists or invlaid or someone may work on them if they were needed but not urgent. Discussion on irc is fine but rarely any decision will come from it, bar maybe some insight into it
[10:30] <cheater> no, i have considered that though
[10:30] <wgrant> czajkowski: I think the issues mostly require discussion more than bugs.
[10:30] <wgrant> They're overarching design issues.
[10:30] <cheater> i decided to post on the dev list first to figure out what the situation is
[10:31] <czajkowski> wgrant: yes but they could still be logged, that way feedback can be logged and read.
[10:31] <czajkowski> I'm odd I like it in a bug so I can reference it, I'm not likely to refence or search an irc log for a discussion, even though it's been interesting to read
[10:31] <cheater> i need to understand the issues more before i can make a useful report to any sort of tracker
[10:32] <cheater> i will take these irc discussions and put the relevant parts in each bit though
[10:50] <htorque> hi again! bug 833137 - can anyone in here maybe delete the last comment? i don't want my poor bug report to turn into a pointless flame war...
[10:53] <henninge> htorque: looking at it
[10:56] <henninge> htorque: it's not spam or insulting. It's somebody's opinion. Though it may be wrong, there is no reason to censor it.
[10:56] <henninge> Trolls are best ignored ... ;)
[10:56] <htorque> henninge: ok, thanks. though i keep an eye on it - such comments tend to attract flame wars. ;)
[10:57] <henninge> fair enough
[10:57] <wgrant> I'm surprised it's gone 14 hours without inspiring a flamefest.
[10:57] <wgrant> It is encouraging.
[10:58] <nigelb> Or depressing. No one cares anymore.
[11:23] <henninge> adeuring: Hi! ;-)
[11:23] <henninge> adeuring: I am off to lunch ...
[11:23] <adeuring> henninge: I'm afraid that I don't have much time today to work as a help contact :(
[11:24] <henninge> adeuring: np, I'll just leave my name in here, then.
[11:24] <adeuring> henninge-lunch: cool, thanks
[11:49] <tkamppeter> Can someone fix bug 689323, I get always a timeout error when trying to access it in LP.
[12:17] <henninge> tkamppeter: Hi! Are you saying the bug page is timing out for your?
[12:17] <henninge> s/your/you/
[12:18] <henninge> tkamppeter: it's not timing out for me, so if you have an OOPS-ID, that would be very helpful
[12:24] <tkamppeter> henninge: Error ID: OOPS-2068AW52
[12:34] <henninge> tkamppeter: I don't see a bug open for this. Can you please file a bug with as much information as possible, including the OOPS id?
[12:48] <lag> Are there any Launchpad/packaging gurus around that could help me out?
[12:48] <lag> I'm trying to upload a binary package to a PPA
[12:49] <lag> the *.tar.gz has the files in, but the *.deb does not
[12:49] <lag> Any ideas?
[12:53] <wgrant> lag: You can't upload .debs to Launchpad.
[12:54] <wgrant> lag: You must upload a source package and let Launchpad build it.
[12:54] <lag> wgrant: That's exactly what I did
[12:54] <lag> wgrant: And the resultant *.deb does not contain my files
[12:54] <wgrant> lag: Have you tried building it locally?
[12:54] <wgrant> It sounds like your debian/rules is incorrect.
[12:55] <lag> Hmm
[12:55] <lag> wgrant: I'll check
[12:57] <lag> wgrant: Should fdr binary work?
[12:58] <wgrant> lag: fdr binary?
[12:58] <lag> wgrant: How would one normally build a binary package manually?
[12:59] <wgrant> lag: Have you read the Ubuntu Packaging Guide?
[12:59] <wgrant> https://wiki.ubuntu.com/PackagingGuide
[13:00] <lag> wgrant: I haven't, but this should just work
[13:00] <wgrant> Why?
[13:01] <lag> wgrant: fdr binary == `fakeroot debian/rules binary`
[13:01] <wgrant> Oh.
[13:01] <lag> wgrant: Because it's worked before
[13:01] <wgrant> One would normally use dpkg-buildpackage -b
[13:01] <wgrant> Which is roughly a wrapper around that.
[13:01] <lag> dpkg-buildpackage -b == debuild -b No?
[13:02] <wgrant> debuild is a wrapper around dpkg-buildpackage is a wrapper around fakeroot debian/rules binary
[13:02] <lag> Got you
[13:02] <wgrant> So yes, debuild -b will work too.
[13:02] <lag> Right, when I use -b I get "No packages to build"
[13:02] <lag> But I didn't get this before
[13:02] <wgrant> Is your debian/control empty?
[13:02] <lag> All I've done is clone an existing package and fork it
[13:03]  * lag checks
[13:03] <wgrant> It needs at least two stanzas.
[13:03] <lag> No
[13:03] <wgrant> One for the source pacakge, and one for each binary package.
[13:04] <lag> wgrant: http://paste.ubuntu.com/677960/
[13:05] <wgrant> Oh, it's this sort of package.
[13:05] <wgrant> Are you sure it's real?
[13:05] <wgrant> And been built before?
[13:05] <lag> wgrant: What's the voodoo that tells the scripts to "place these binaries into the deb"
[13:05] <lag> The binaries are pre-built
[13:05] <lag> All LP has to do is slam them into a *.deb
[13:06] <wgrant> lag: debian/rules needs to put them there, generally use helper scripts like debhelper.
[13:06]  * lag is wondering how this worked before?
[13:06] <wgrant> Do you know that it has worked before?
[13:06] <wgrant> Just because the OEM has built it, doesn't mean they built it properly or that it worked :)
[13:07] <lag> I've used it may times
[13:08] <wgrant> Oh.
[13:08] <wgrant> You renamed the package.
[13:08] <wgrant> But you didn't rename startupfiles.install
[13:08] <lag> wgrant: http://snapshots.linaro.org/11.05-daily/linaro-hwpacks/lt-snowball/20110812/0/images/hwpack/
[13:09] <lag> Take wpack_linaro-lt-snowball_20110812-0_armel_supported.tar.gz and open it
[13:09]  * lag checks
[13:09] <wgrant> I can see in startupfiles-v3 that there is still a startupfiles.install.
[13:09] <wgrant> But debian/control doesn't define a startupfiles binary, so debhelper won't do anything with it.
[13:11] <lag> wgrant: So startupfiles.install -> startupfiles-v2.install
[13:11] <wgrant> lag: Or v3 in the one I'm looking at, but yes.
[13:11] <lag> wgrant: Awesome - bear with
[13:12] <wgrant> lag: This is using fully automatic debhelper, so it's fairly opaque unless you read the docs.
[13:12] <wgrant> For specifics on *.install, man dh_install
[13:12] <lag> I'll look in a bit, I just want to get it working for now
[13:15] <lag> wgrant: Woot!
[13:19] <lag> wgrant: Re-deputting now
[13:28] <tkamppeter> henninge, bug 837391
[13:29] <henninge> tkamppeter: thanks
[15:30] <deryck> henninge, I can take IRC now.
[15:31] <henninge> deryck: thanks ;)
[15:45] <pmatulis> mneptok: o/
[15:56] <mneptok> pmatulis: ahoy!
[16:02] <mneptok> komputes: pmatulis is stalking me. stalk him for me.
[16:03] <mneptok> O:)
[16:22] <nigelb> Aww, no more Launchpad T-shirts?
[16:31] <komputes> mneptok: hehe, I can do that for you...
[16:43] <pmatulis> mneptok: you're looking very innocent there
[17:00] <Lekensteyn> Hi all, I've just uploaded some packages to a PPA. All of them are ready, but one is still in "publishing" state. Can I safely upload another package which depends on the packages, or will it use an older, published version?
[17:05] <maxb> Lekensteyn: Build-dependencies can only be satisfied by published versions
[17:08] <Lekensteyn> maxb: thank you, I'll wait then
[18:11] <deryck> abentley, pitching to you now.
[18:12] <abentley> deryck: roger
[19:35] <adam_g> hi.. ive got a stale LP account that i no longer use, do not know the passwd for and do not have access to the email associated with it.. anyone know how i'd go about deleting that? :P
[19:43] <kiko> adam_g, you have to open a question to get thar account merged with your active one
[19:49] <adam_g> kiko: ah ha, thanx
[22:50] <Sweetshark> Hi there, launchpad fails to set an existing branch as project branch with a timeout. Any hints?
[22:52] <Sweetshark> (https://launchpad.net/df-libreoffice/ and https://code.launchpad.net/~vcs-imports/libreoffice/core)
[22:56] <Sweetshark> oh, lots of joins, so I impolitely repost:
[22:56] <Sweetshark> Hi there, launchpad fails to set an existing branch as project branch with a timeout. Any hints?
[22:56] <Sweetshark> (https://launchpad.net/df-libreoffice/ and https://code.launchpad.net/~vcs-imports/libreoffice/core)
[22:58] <mwhudson> Sweetshark: looks like https://code.launchpad.net/~vcs-imports/libreoffice/core is making the scanner angry
[22:58] <mwhudson> so it's probably a db locking issue
[22:59] <Sweetshark> mwhudson: I wonder why?
[22:59] <mwhudson> i'm guessing the branch is rather large
[22:59] <Sweetshark> *cough* *cough* 700MB git repo import.
[22:59]  * Sweetshark tries to look innocent.
[22:59] <mwhudson> large in terms of revision count will be the issue here
[23:00] <Sweetshark> mwhudson: pretty much the whole history of OOo/LO since 2000. Is that enough
[23:00] <Sweetshark> ?
[23:00] <mwhudson> Sweetshark: i don't know, i'm just guessing
[23:02] <Sweetshark> ~292993 revisions/commits
[23:03] <mwhudson> huh, that's a lot, but not a mindboggling amount
[23:04] <Sweetshark> (and that has only the merge commits of the feature-branches from the CVS/SVN days -- those are huge then of course)
[23:09] <mwhudson> Sweetshark: can you try to link again?