[00:01] <mwhudson> time for lunch
[01:06] <thumper> mwhudson: my work at the moment reduces code.launchpad.dev/mint from 350 queries to 150
[01:06] <thumper> mwhudson: but that is because the development focus isn't worked out properly
[01:07] <mwhudson> thumper: progress!
[01:07] <mwhudson> thumper: or you mean, it's fast because it's wrong?
[01:07] <thumper> mwhudson: it is just the listing class trying to determin dev focus
[01:07] <thumper> mwhudson: right now it preserves the existing behaviour
[01:07] <thumper> mwhudson: without the extra queries
[01:08] <mwhudson> the existing behaviour being "lp:ubuntu/karmic/bob" rather than "lp:ubuntu/bob"
[01:08] <mwhudson> ?
[01:09] <thumper> yeah
[01:09] <thumper> that's the next bit to fix
[01:13] <mwhudson> i see
[01:18]  * mwhudson afk for a few minutes
[01:25] <thumper> mwhudson: got it from 350 down to 50
[01:25] <thumper> mwhudson: now just to make it right :)
[01:43] <mwhudson> thumper: cool
[01:43] <thumper> I think I have it
[01:43] <thumper> 59 queries
[01:43] <mwhudson> thumper: for how many branches?
[01:46] <thumper> 72 queries for 130 branches
[01:46] <thumper> and it gets a nice star :)
[01:47] <mwhudson> thumper: that sounds better than the product branch listing currently
[01:48] <thumper> mwhudson: none of them have revisions :)
[01:48] <mwhudson> thumper: ah
[01:48] <thumper> mwhudson: I have a feeling that is where some of the extra queries are getting soaked up
[01:48] <mwhudson> certainly could be
[01:48] <thumper> mwhudson: but one thing at a time
[01:48] <mwhudson> yep
[01:50] <thumper> mwhudson: this is almost ready to submit \o/
[02:14]  * thumper found missing test coverage
[02:55] <thumper> mwhudson: has local launchpad sucked arse since you upgraded to karmic?
[02:55] <thumper> mwhudson: I'm seeing massive cpu spikes
[02:55] <mwhudson> thumper: haven't noticed anything odd
[02:56] <thumper> :(
[02:59] <thumper> when do we get memcached?
[03:08] <mwhudson> !!??!?!?!
[03:09]  * mwhudson finds that the sftp oopses are a twisted.conch bug that _i fixed_ over a year ago
[03:10] <thumper> ?
[03:12] <mwhudson> i guess we haven't updated our twisted in at least that long
[03:14] <mwhudson> thumper: https://bugs.edge.launchpad.net/launchpad-code/+bug/376279
[03:14] <mup> Bug #376279: AssertionError: still have data in FSETSTAT:  <codehosting-ssh> <oops> <Launchpad Bazaar Integration:Triaged> <https://launchpad.net/bugs/376279>
[03:14] <thumper> mwhudson: twisted time
[03:14] <thumper> ?
[03:14] <thumper> mwhudson: is twisted handled through buildout?
[03:14] <mwhudson> thumper: no, it's still in sourcecode
[03:15] <thumper> mwhudson: should it be buildout?
[03:15] <mwhudson> thumper: probably
[03:15] <mwhudson> i don't know if it would work
[03:15] <thumper> hmm..
[03:15] <mwhudson> huh, Twisted 8.2.0 is in the download cache
[03:16] <mwhudson> but not in versions.cfg
[03:16] <thumper> perhaps jml started
[04:08] <thumper> mwhudson: the slowness was firebug
[04:08] <thumper> mwhudson: having too much enabled
[04:08] <mwhudson> thumper: ah
[04:09] <thumper> *sigh*
[04:09] <thumper> lazr.editor isn't working properly
[04:10]  * thumper goes hunting for the actual source
[04:10] <thumper> mwhudson: I have an amost working inline commit message editor
[04:10] <mwhudson> thumper: yay
[04:10] <thumper> using the same (ish) code as bug descriptions
[04:10] <thumper> mwhudson: 2 issues though
[04:10] <thumper> 1) accept_empty doesn't work
[04:11] <thumper> 2) needs id="edit-description" for all the styling
[04:14]  * thumper files bugs
[04:21]  * thumper finds it isn't lazr-js
[06:05]  * mwhudson eods
[08:32] <adeuring> good moring
[08:47] <al-maisan> moin adeuring
[08:47] <adeuring> hi al-maisan!
[09:13] <mrevell> Morning!
[10:13] <jtv> mrevell: you may be interested to know that we have translation import from branches running on staging now.  At last people can do real testing there without admin help!
[10:14] <mrevell> jtv: Ah, I am interested! Would you like to write a blog post about it?
[10:14] <jtv> mrevell: yeah, I guess that's worth while
[10:41] <jml> hello
[10:42] <thumper> hi
[10:42] <thumper> does anyone know how many lines of code there is in launchpad at the moment?
[10:43] <jml> thumper, not off the top of my head.
[10:45] <jml> thumper, http://paste.ubuntu.com/298097/
[10:46] <thumper> jml: that includes the tests too right?
[10:46] <jml> thumper, yes.
[10:46] <thumper> jml: but is probably doesn't count the doc tests
[10:46] <jml> thumper, it's just the output of sloccount
[10:46]  * thumper nods
[10:59] <deryck> Morning, all.
[11:01] <jml> morning
[11:29] <jml> anyone got balsamiq running on Linux?
[11:29]  * jml is having some trouble.
[11:30] <intellectronica> jml: yeah, i do
[11:31] <intellectronica> jml: is it only balsamiq you've got a problem with. did you try other air apps?
[11:32] <jml> intellectronica, no, I haven't tried other Air apps.
[11:32] <jml> this is the first one :)
[11:34] <jml> intellectronica, does Air have a command line I can use to try to run balsamiq?
[11:34] <jml> as in, run the baslamiq installer (or whatever it is?)
[11:35] <intellectronica> jml: not that i know. i don't remember how i got it installed
[11:36] <intellectronica> jml: so, you can't even get it installed?
[11:37] <intellectronica> jml: did you follow http://www.balsamiq.com/products/mockups/help#linux ?
[11:39] <jml> intellectronica, I am following it now.
[11:39] <jml> intellectronica, Step 3 never happens.
[11:42] <intellectronica> jml: no idea. it "just worked" for me. you should try the balsamiq team - they're a pretty dedicated bunch
[11:43] <jml> hmm.
[11:43] <jml> intellectronica, are they on IRC?
[11:44] <intellectronica> jml: no, but they are on jabber and skype
[11:44] <jml> intellectronica, heh, ok. thanks.
[11:52] <jml> ahh, here we go, found the air app installer
[11:52] <jml> Error loading the runtime (libadobecertstore.so: cannot open shared object file: No such file or directory)
[11:55] <jml> intellectronica, are you 32 bit or 64 bit?
[11:55] <intellectronica> jml: 32
[11:57] <jml> intellectronica, that'll be it :)
[11:57] <intellectronica> orli? no air for linux 64-bit?
[11:58] <jml> intellectronica, complicated air for 64 bit
[11:59] <jml> intellectronica, I had to manually copy a library from /usr/lib to /usr/lib32
[11:59] <intellectronica> eugh
[11:59] <intellectronica> it's so hard to get excited about proprietary software, isn't it, even if you really try hard
[12:00] <bigjools> just the games :)
[12:01] <intellectronica> bigjools: games are not software
[12:01] <intellectronica> as in "candy isn't food"
[12:02] <bigjools> so games are what?
[12:05] <jml> candy
[12:05] <jml> executable polygonous candy
[13:21]  * jml lunches
[14:31] <sinzui> EdwinGrubbs: barry: bac: standup in 2 minutes
[14:57] <barry> reviewers, beuno, lurkers, developers -> #launchpad-meeting in 3m
[15:33] <EdwinGrubbs> james_w: ping
[15:34] <james_w> hi EdwinGrubbs
[15:35] <EdwinGrubbs> james_w: regarding bug 436507, would it be helpful to change the portlet name from "Releases in Ubuntu" to something else to indicate that you are in the distro context?
[15:35] <mup> Bug #436507: Distribution series source package page has no links to distribution source package page <post-3-ui-cleanup> <Launchpad Registry:Triaged by edwin-grubbs> <https://launchpad.net/bugs/436507>
[15:38] <james_w> EdwinGrubbs: I'm not sure I understand the question?
[15:39] <EdwinGrubbs> james_w: well, in the bug description, you mentioned that you can change to the distro context with the "Releases in Ubuntu" links, but that it was not obvious. Are those links good enough, but just confusingly named, or should there be a different link to get you to the right place?
[15:40] <james_w> well, I generally find myself dissatisfied with the distro/distroseries split in these pages. I'm not sure that just renaming that will help much.
[15:40] <james_w> though I'm not sure how I would phrase a link to the distro context either
[15:41] <james_w> do you have some suggestions for alternative wording?
[15:44] <EdwinGrubbs> james_w: maybe, "Distribution source package for hotssh" or "Ubuntu source package for hotssh". What exactly is the difference between karmic's source package and ubuntu's source package? Are they shared? Do they have extra patches for each distro?
[15:45] <james_w> well, ubuntu's source package is an overview of what is going on in each release
[15:45] <james_w> the /karmic/ page shows you just the information for that release
[15:45] <james_w> compare https://edge.launchpad.net/ubuntu/karmic/+source/hotssh and https://edge.launchpad.net/ubuntu/+source/hotssh
[15:46] <james_w> I prefer to work in the latter, but LP likes to send me to the former
[15:47] <james_w> also compare https://edge.launchpad.net/ubuntu/karmic/+source/hotssh/0.2.6-2 and https://edge.launchpad.net/ubuntu/+source/hotssh/0.2.6-2
[15:48] <james_w> I can see why you might want the information for a single release, but LP linking to it tends to mean that I am on the page with less information
[15:49] <james_w> and I think the distinction will be lost on many. They won't realise the different contexts, and so won't know to switch between them. That's why I think renaming the heading wouldn't help that much.
[15:53] <EdwinGrubbs> james_w: it is interesting that /ubuntu/+source/hotssh has the heading "hotssh package in Ubuntu" instead of saying "source package"? Would it be better to combine the distro and the series pages, or am I missing something about their purpose?
[15:53] <james_w> I believe bigjools has even tried to do that
[15:54] <bigjools> I did
[15:54] <bigjools> I was on a road to Fail
[15:55] <bigjools> because it buggers up source package branches
[15:55] <jml> \o/
[15:56] <EdwinGrubbs> bigjools: is one of those pages a soyuz page and the other a registry page? The split confuses me.
[15:57] <bigjools> EdwinGrubbs: they're both registry
[15:57] <EdwinGrubbs> bigjools: oh, you have thwarted my attempt to make this someone else's problem.
[15:57] <bigjools> :)
[15:57] <bigjools> they used to be soyuz!
[16:00] <EdwinGrubbs> bigjools: do you think it would be less likely to fail if I just moved a bunch of the /ubunt/karmic/+source/hotssh info into the /ubuntu/+source/hotssh page, and the other page is only used when it absolutely has to?
[16:00] <bigjools> EdwinGrubbs: yeah that would work - be mindful of excessive queries and page load times though
[16:01] <bigjools> EdwinGrubbs: but I think james_w is more concerned about other pages that link to the DSSP page, they might be more useful linking to the DSP page
[16:01] <EdwinGrubbs> bigjools: the /ubuntu/+source/hotssh page already uses ajax to expand to show info, so the distroseries pages' info could just be dynamically loaded if they expand it.
[16:03] <EdwinGrubbs> bigjools: are there pages that should continue to point to DSSP no matter what, i.e., should I just convert links on a page by page basis or try to convert all the links from DSSP to DSP?
[16:03] <bigjools> EdwinGrubbs: we'd have to analyse each I think and see what's most useful
[16:04] <bigjools> because if we lose all the links then how do you get to the page? :)
[16:04] <EdwinGrubbs> bigjools: it all depends on whether that page has any info that isn't moved into the DSP page.
[16:04] <bigjools> EdwinGrubbs: yes, I ajaxified that page when I redesigned it, but we need to be careful not to overdo that
[16:05] <james_w> fwiw this would mostly satisfy me, I rarely go to the DSSP pages out of choice, and rarely need any information there. If LP stopped sending me to them I would stop getting frustrated that it apparently doesn't want to let me out of that context.
[16:05] <bigjools> EdwinGrubbs: the reason it's there is so that you can click on a tab and change facet to Code
[16:05] <bigjools> otherwise I'd be real happy to blow it away
[16:08] <EdwinGrubbs> james_w: which pages have links that send you to the DSSP pages most often?
[16:08] <james_w> let me think, there is one in particular
[16:11] <james_w> +publishinghistory does, and going to "Overview" from branches does
[16:11] <james_w> I understand why there
[16:11] <james_w> and I'm sure that the latter can be blamed on me
[16:16] <bigjools> ha, the latter is the reverse of my last justification to keep the page
[16:16] <bigjools> it's because it's in the same context
[16:16] <EdwinGrubbs> james_w: what is the full link for +publishinghistory?
[16:17] <james_w> https://edge.launchpad.net/ubuntu/+source/kerneloops/+publishinghistory
[16:17] <bigjools> that's a soyuz page
[16:18] <james_w> it does it because it actually has lines for the publication of each version in each release it is in
[16:18] <bigjools> indeed
[16:18] <james_w> so going to the version+distroseries page is probably the right thing to do
[16:18] <bigjools> there's a separate publishing history for a series
[16:23] <EdwinGrubbs> james_w: am I missing something, or are all the links pointing to DSSP still need to do that? Would extra links help on those pages, or is the best thing just to put a very prominent link on the /ubuntu/karmic/+source/hotssh page pointing to /ubuntu/+source/hotssh as discussed before?
[16:23] <james_w> possibly
[16:24] <james_w> I imagine this is a symptom of a deeper issue, but I kept bumping in to it, and so wanted to report it
[16:24] <james_w> I think a link would be a good start, I just don't know how to phrase it
[16:25] <EdwinGrubbs> james_w: in some ways, it would be nice to put the link in the breadcrumbs below the <h1>, but I'm sure I would get complaints about that since it doesn't match the url hierarchy. The links could be called "All source packages in Ubuntu"
[16:26] <james_w> you mean "9base, aalib, ..."?
[16:26] <EdwinGrubbs> james_w: really, it should be "All source package in Ubuntu for hotssh"
[16:27] <james_w> that could work
[16:27] <EdwinGrubbs> james_w: do you think it would work to put that link to the right of the "View changelog" link?
[16:28] <james_w> I think that link can appear multiple times?
[16:29] <james_w> hmm, no, that seems odd
[16:29] <james_w> but in that case, yeah, I think that would be appropriate
[16:29] <james_w> something on the version pages would be good as well
[16:29] <james_w> they are even harder to escape from
[16:30] <james_w> oh, no, the breadcrumbs can take you to the non-version pages
[16:32] <EdwinGrubbs> james_w: would you also want a link on the version page that would take you to the DSP page in addition to the DSSP link in the breadcrumbs?
[16:33] <james_w> I would like it I think
[16:33] <james_w> but given that the other link would give a way out, it's not critical
[16:41] <bigjools> EdwinGrubbs, james_w: FWIW, earlier this year we tried to redesign the navigation for these pages, and we came up with something nice that then failed a bunch of simple use cases.  It's hard :/
[16:45] <james_w> yeah, I can appreciate that
[16:49] <bigjools> and there's a lot of resistance to change
[16:55] <rockstar> jelmer, ping
[17:26] <beuno-lunch> jml, can our call be after lunch?
[17:26] <jml> beuno-lunch, before or after the TL call?
[17:27] <beuno-lunch> jml, before
[17:27] <jml> beuno-lunch, sure
[17:27] <beuno-lunch> I'll be lunched in 30'
[17:27] <beuno-lunch> great
[17:53] <jml> james_w, can you do me a favour and pastebin the output of 'time ssh bazaar.launchpad.net'?
[18:03] <james_w> jml: ssh bazaar.launchpad.net 0.03s user 0.02s system 0% cpu 6.135 total
[18:03] <jml> james_w, thanks.
[18:04] <beuno> jml, ready when you are
[18:05] <jml> beuno, I don't have skype, what's the best way to do this?
[18:06] <beuno> jml, I can call you, you can call me
[18:58] <maxb> jml: Hi. You mentioned something about adding licensing to a sftp server a while back - wondering if you have time to say more?
[18:58] <jml> maxb, yes! I do!
[18:58] <jml> maxb, 'bzr branch lp:txsshserver'
[19:00] <maxb> \o/
[19:00] <maxb> thanks
[19:07] <jml> maxb, and also, https://bugs.edge.launchpad.net/txsshserver/+filebug
[19:07] <maxb> heh
[19:08] <jml> maxb, there was a long time between me making the branch and me open sourcing the branch
[19:09] <jml> maxb, so it's slightly abandonware, but I'm happy to hack more on it if someone finds it useful
[19:50] <maxb> What is that Launchpad page which tells you the definitions of all the kinds of image sizes LP uses? badge, mugshot .... etc.
[19:51] <salgado-sprint> +images, I think
[19:52] <beuno> +graphics maxb
[19:52] <salgado-sprint> or +graphics
[19:52] <maxb> aha!
[19:52] <maxb> thanks
[20:00]  * jml is so gone.
[20:27] <mwhudson> good morning
[21:08] <thumper> morning
[21:18] <Ursinha> barry: hi :)
[21:18] <barry> Ursinha: hi!
[21:19] <Ursinha> barry: here I am bugging you again about maillinglists oopses :) have you seen this one -> https://lp-oops.canonical.com/oops.py/?oopsid=OOPS-1389XMLP50 ?
[21:20] <barry> Ursinha: nope, that's a new one
[21:20] <Ursinha> barry: do you mind to triage a bug when I create one?
[21:20] <Ursinha> which will be pretty soon :)
[21:21] <Ursinha> s/a bug/the bug/
[21:21] <barry> Ursinha: sure
[21:22] <Ursinha> barry: do you have a clue about that? I don't even know what to write in the bug summary :)
[21:24] <barry> Ursinha: just say "OOPS in MailingListAPIView.holdMessage() and reference the oops page.  please add a 'mailing-lists' tag too
[21:24] <Ursinha> sure, thanks a lot barry
[21:24] <barry> np!
[21:28] <Ursinha> barry: bug 457606
[21:28] <mup> Bug #457606: UnknownSender OOPS in MailingListAPIView.holdMessage() <mailing-lists> <oops> <Launchpad Registry:New> <https://launchpad.net/bugs/457606>
[21:31] <barry> Ursinha: got it, but i might not get to it until after our sprint ;)
[22:11] <Ursinha> gary-sprint: when you have a minute :) can you take a look at bug 51843, please? thanks!
[22:11] <gary-sprint> Ursinha: looking
[22:11] <mwhudson> Ursinha: hi
[22:12] <mwhudson> Ursinha: do you know why r9669 is on https://dev.launchpad.net/CodeTeamTestPlans/3.1.10 8 times?
[22:13] <Ursinha> mwhudson: this is something weird that happens sometimes, it seems the script isn't able to record the last revision recorded in the file and so it adds again in the next run
[22:13] <Ursinha> I wasn't able to reproduce the problem
[22:13] <mwhudson> ok
[22:15] <gary-sprint> Ursinha: reopened with explanation :-)
[22:15] <Ursinha> gary-sprint: thank you very much :)
[22:16] <gary-sprint> Ursinha: thank you!
[22:43] <Ursinha> bug 331990
[22:43] <mup> Bug #331990: The inline editor widget reports a JSON error when saving non-ASCII characters <javascript> <Launchpad Foundations:Fix Committed by leonardr> <lazr.restful:Fix Released by leonardr> <https://launchpad.net/bugs/331990>
[22:43] <wgrant> (continuing the discussion from #launchpad) Why wasn't that CPed weeks ago?
[22:44] <wgrant> The problem was rediscovered a couple of days after the release, and it is a really really bad bug.
[22:45] <Ursinha> wgrant: I took a note about it and will discuss on tomorrow's production meeting
[22:49] <jml> mwhudson, taking my product strategist hat off and elegantly throwing it onto the hatstand for a moment
[22:50] <jml> mwhudson, a twisted web thing for serving branches would be very cool.
[22:51] <mwhudson> jml: yeah
[22:51] <jml> mwhudson, and done right, it might help other areas of Launchpad that have similar "store lots of stuff and serve it with cool names" needs.
[22:51] <mwhudson> another thing would be something mod_python ish i guess
[22:51] <jml> mwhudson, particularly combined with the sftp server.
[22:52] <mwhudson> that might be less technologically exciting
[22:52] <mwhudson> jml: i guess so
[22:53] <mwhudson> abentley: your trick with pushing a 1.6 format branch worked, btw
[22:53] <mwhudson> thumper: i just spammed a review at you, it's very simple
[22:53] <thumper> ok
[22:53] <lifeless> jml: isn't that 'all of lp'  ? (store stuff, serve with names)
[23:00] <jml> lifeless, some more than others.
[23:01] <wgrant> PPAs... branches.. what else?
[23:57] <mwhudson> rockstar: https://bugs.edge.launchpad.net/launchpad-code/+bug/306290 is Fix Committed i think?
[23:57] <mup> Bug #306290: OOPS in code-import page when missing the correct CVSROOT  <oops> <Launchpad Bazaar Integration:In Progress by rockstar> <https://launchpad.net/bugs/306290>
[23:57] <rockstar> mwhudson, yes, thanksy.
[23:57] <rockstar> Er, thanks.