, listing <td>, comment <div>. [18:05] <sinzui> rockstar: so I think we want to change that test to verify that the div that represents the question (and maybe the other comments to be pt-BR [18:07] <sinzui> rockstar: I think we need to ensure that question-index.pt creates a title with the right lang, base-layout may need to change. [18:07] <rockstar> sinzui, okay. I won't worry about base-layout for now. [18:08] <sinzui> rockstar: right, file a bug. I knew about this when I created base-layout. I assumed I would be the person updating Answers [18:08] * sinzui simplified base layout to make the rest of the conversion simple. [18:09] <rockstar> sinzui, understandable. [18:11] <sinzui> rockstar: the 'dir' attr is important on the question in the listing table and question. I added an Arabic question to sample data so that you can verify the layout of the text [18:15] <abentley> rockstar: we are in testfix mode due to failure to initialize the DB. I want to submit an empty "testfix". Can I rs you? [18:16] <rockstar> abentley, I was just about to ask you if you could do a testfix. rs=me === salgado-lunch is now known as salgado [18:24] <rockstar> How do I clear the mailqueue so it stops trying to send mail? [18:24] <rockstar> It keeps hiding my logs and stuff because it's giving me failure messages in my terminal. [18:28] <intellectronica> rockstar: just get rid of /var/tmp/launchpad_mailqueue [18:36] <mrevell> night all! === gary_poster is now known as gary-lunch [18:48] <EdwinGrubbs> beuno-afk: ping === gary-lunch is now known as gary_poster [19:28] <barry> salgado: so i don't think my problem is that the sprints/+all link is being disabled. if i actually remove the 'condition="context/enabled"' from launchpad-inline-link.pt i /still/ don't get the sprints/+all link [19:33] <salgado> barry, looks like templates/navigationmenu-related-pages.pt doesn't render links that have linked=False [19:33] <salgado> barry, is this menu rendered using that template (i.e. view/++@related-pages) [19:34] <salgado> @@+related-pages, that is [19:34] <barry> salgado: it is, and i see that! [19:34] <barry> salgado: that doesn't seem right, does it? [19:34] <salgado> it doesn't. everywhere else we just omit the <a> tag for unlinked links [19:35] <barry> salgado: agreed. that should probably be done in launchpad-inline-link.pt which is what actually renders the link [19:36] <barry> salgado: yep, removing that tal:condition brings back the link [19:36] <salgado> barry, maybe we can even remove the conditional on link/enabled, if it's also done by launchpad-inline-link.pt [19:40] <barry> salgado: i think so. l-i-l.pt doesn't remove the href yet. i wonder how hard that will be [19:40] <barry> salgado: is it worth it? [19:41] <barry> salgado: disabling the href i mean [19:43] <salgado> barry, I guess you'll need to place the condition="link/linked" in the <a> itself and add another tag with condition="not: link/linked" to show the text without a link/icon [19:44] <barry> salgado: yep [19:44] <salgado> is that what you asked? === EdwinGrubbs is now known as Edwin-lunch [19:45] <barry> salgado: well, i'm actually asking whether it's worth disabling the link when you're on that link's page. personally, that behavior always bugs the hell out of me where ever i encounter it :) [19:45] <barry> salgado: i guess that's a question for beuno [19:45] <salgado> agreed [19:45] <salgado> I know he's planning to do that for breadcrumbs at least [19:45] <barry> or for beuno-afk even [19:46] <barry> salgado: so at least consistent annoyance is better? :) okay, ping beuno-afk for the pronouncement [19:51] <kfogel> danilos: thanks for your keeping an eye out on ubuntu-translators@ for replies in the "three wishes for Launchpad" thread [20:26] * rockstar lunches === Edwin-lunch is now known as EdwinGrubbs [20:36] <kiko> hey [20:36] <kiko> somebody needs to ping jamalta to go over a branch he has to fix bug 127489 [20:36] <mup> Bug #127489: Answers column is in the wrong order in person's "Most active in" table <trivial> <ui> <Launchpad Foundations:Triaged by jamalta> <https://launchpad.net/bugs/127489> [20:36] <kiko> https://code.edge.launchpad.net/~jamalta/launchpad/bug-127489/+merge/9564 [20:36] <kiko> any reviewers? === ursula_ is now known as Ursinha [21:16] <statik> hi launchpad-dev, i have a questions about stacking. i desperately need to push a 2a format branch to a project where the dev focus branch is in an older format. is there a way I can force a branch not to stack? [21:17] <statik> oops, wrong channel [21:17] <mwhudson> good morning [21:17] <statik> mwhudson, just the man who can save my life [21:19] <mwhudson> statik: bzr init lp:~whatever/whatever/whatever [21:19] <mwhudson> bzr reconfigure --unstacked lp:~whatever/whatever/whatever [21:19] <mwhudson> bzr upgrade --2a lp:~whatever/whatever/whatever [21:19] <mwhudson> bzr push -d ~/yourbranch lp:~whatever/whatever/whatever [21:20] <mwhudson> (there might be some easier way, too early in the morning for that) [21:23] <statik> mwhudson, you are my hero [21:52] <gary_poster> maxb: ping [21:52] <maxb> pong [21:53] <gary_poster> maxb: hey. bug 413335 : do you want me to go upstream (I am upstream and can make releases)? or have you done it, or do you want to? [21:53] <mup> Bug #413335: Figure out what to do about zope.sendmail incompatibility with Python >= 2.5.1 <python-upgrade> <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/413335> [21:53] <maxb> gary_poster: Ah, right. The reason I've not going upstream with that is that I'm not sure what the correct action is [21:54] <rockstar> sinzui, I tire of answers tests. Can I delete the ones that are redundant? [21:55] <gary_poster> maxb: sure, understood, me either. we don't have to have an answer necessarily; we can brainstorm with upstream and see if someone has ideas. This is a problem for 2.6 as well? I'm not familiar with the threading changes. I saw your workaround. [21:55] <maxb> I don't think it's changed any further in 2.6 [21:56] <gary_poster> maxb: ok. btw, I landed the "all of zope from zope distributions" about 8 hours ago. I have another branch I hope to land tomorrow that brings us up to the most recent releases of our lazr.* stuff (and launchpadlib and so on) tomorrow--tests appear to all be passing. [21:56] <maxb> The problem is that in 2.5.0 and below you could use sys.exitfunc as a notification that the main/default thread had exited. In 2.5.1 and above, there is no such mechanism (other than the one that's private between the interpreter and the threading module, which I therefore monkeypatch) [21:57] <maxb> Neat! [21:57] <maxb> I was starting to teach myself about doctests and trying to understand REnormalizing [21:57] <maxb> doctests are great, except they suck at supporting variable output [21:58] <gary_poster> yeah [21:58] <gary_poster> I agree. others wouldn't go so far as "are great" ;-) [21:59] <gary_poster> RENormalizing helps, ellipses help. Maybe we'll switch to manuel later, which gives us a lot more flexibility http://packages.python.org/manuel/ [21:59] <rockstar> maxb, don't go into the light... :) [21:59] <gary_poster> lol [22:01] <maxb> I need to decide whether to shift my python2.5 efforts to be based on db-devel, then, I guess [22:01] <gary_poster> maxb: ok, for zope.sendmail, I'll write something upstream, and cc you. If you want to join in (and I hope you do) feel free to join the mailing list, or I'll forward notes you write. [22:02] <gary_poster> maxb: why db-devel? [22:02] <maxb> I thought your zbuildout branch was targetted at db-devel? [22:02] <gary_poster> no, devel [22:02] <maxb> oh, weird, the merge proposal said db-devel when I looked at it [22:02] <gary_poster> urg, I bet I screwed up. [22:03] <maxb> Yay, all the better [22:03] <gary_poster> heh [22:04] <maxb> Which is the mailing list in question? [22:04] <gary_poster> maxb: http://mail.zope.org/mailman/listinfo/zope-dev [22:05] <maxb> eek, all of zope on one list, that'll be interesting [22:05] <maxb> s/interesting/high-traffic/ [22:08] <gary_poster> heh [22:11] <maxb> Right, looks like I can probably mark lp:~maxb/zope3/launchpad-3.4-py2.5 as Abandoned at this point [22:12] <maxb> I'll let my desktop chew on the tests tonight [22:12] <thumper> danilos: a translation based breaking on lp builder [22:18] <barry> yowza! look all all the new log spew [22:18] * barry will bet gary_poster had something to do with that :) [22:19] <gary_poster> lol [22:19] <barry> gary_poster: so what does that buy us? === salgado is now known as salgado-afk [22:33] <gary_poster> barry: (sorry was on call) all of zope is now from distributions, rather than sourcecode. I have a branch that brings us to trunk (and distributions) for all of our own packages (lazr.* and friends) that I hope to get reveiewed and land tomorrow (I believe tests are now passing entirely). Then we need to upgrade to the zope versions that upstream have identified as 2.5/2.6 friendly and interoperating, and deal with the fal [22:33] <gary_poster> (which will probably be significant, we'll see). [22:33] <gary_poster> Then we fix the remaining launchpad bits, and switch to Py 2.5/2.6 immediately after the 3.0 release, is my intent. [22:34] * barry <3 gary_poster [22:34] <gary_poster> :-) heh [22:34] <barry> 2.6 please! [22:34] <gary_poster> +1 from me ;-) we'll see [22:35] <jml> hey guys [22:35] <gary_poster> hiya [22:35] <barry> gary_poster: well, we have to before karmic! [22:35] <barry> jml: hi! [22:35] * jml reads the good news in the backlog! [22:35] <gary_poster> actually, thanks to maxb, we don't: he's made the necessary debs already, AIUI from flacoste. [22:36] <gary_poster> but that doesn't mean we don't want to ;-) [22:36] <barry> gary_poster: right. is there any reason why we wouldn't switch to 2.6 if our test suite passes? [22:36] <gary_poster> barry: losa-type reasons [22:37] <maxb> ~launchpad/ppa is complete for karmic/2.4 other than launchpad-dependencies itself, which I'm waiting for someone to sync from my ppa [22:37] <gary_poster> (which is to say, none that I know of, but who knows) [22:38] <gary_poster> maxb: I thought I saw flacoste did that in the past 24 hours, but maybe I imagined it [22:38] <maxb> My instinct based on the kinds of changes is that any and all work done migrating 2.4->2.5 would also need to be done for 2.4->2.6, hence 2.5 as a comfortable stepping stone [22:38] <gary_poster> agree [22:39] <maxb> And I took a quick look at versions.cfg in devel just now - I think the versions there are clear for 2.5 already, though not for 2.6 [22:39] <gary_poster> you might be right [22:39] <gary_poster> i know 2.6 has some issues: zope.publisher has some problems that I need t record [22:40] <gary_poster> need to run [22:40] <gary_poster> bye all [22:40] <jml> gary_poster, ciao [22:40] <jml> dammit. [22:40] <jml> missed him by *that* much [22:45] <sinzui> rockstar: please delete redundant tests. I have deleted a lot in 2.2.8 [22:48] <rockstar> sinzui, roger roger. I've been doing my best. I'm surprised how many tests broke in various, and sundry, ways with just a few changes. [22:49] <barry> gary_poster: i guess i'm sayin': if we can/are going to move to 2.5 [22:49] <barry> and all our tests pass on 2.6 we should jump right to 2.6 and ignore [22:49] <barry> old pythons that aren't being maintained by upstream any more <wink> [22:49] <sinzui> rockstar: the tests verify the formatting instead of the content. I was young and naive. [22:50] <sinzui> rockstar: And I I wrote most of them again, I would be testing the view, not the template. [22:50] <sinzui> s/I/if/ [22:50] <rockstar> sinzui, yeah, had I known it was going to take me this long to fix 35 (!) failing tests, I would have just re-written them that way. [22:52] <sinzui> rockstar: I have written all view tests, then watch the story failures. I then add the fix to my view test and delete the page test. I do the 75% of the time. [22:53] <rockstar> sinzui, that's a good idea. I'll keep that in mind. === matsubara is now known as matsubara-afk [22:54] <jml> thumper, skype is still broken for me [22:54] <thumper> jml: :( [22:54] <thumper> ok [22:55] * jml restarts, just in case. [23:47] <maxb> Anyone have any thoughts on breaking compile up into compile-lp compile-mailman compile-lazr-js ? [23:48] <jml> maxb, on the surface, it seems like a good idea [23:55] <maxb> Hmm, I need a bit of buildout help... [23:55] <maxb> Error: There is a version conflict. [23:55] <maxb> We already have: lazr.uri 1.0 [23:55] <maxb> where do I start looking? [23:56] <wgrant> Is your sourcedeps branch up to date? [23:57] <maxb> the download-cache branch? [23:57] <maxb> yes [23:57] <wgrant> Hm. [23:58] <wgrant> Tried cleaning and rebuilding?

[00:01] rockstar: it is [00:01] sinzui: It looks like it should also be split into three branches: 1) Making all of the target interfaces descend from IStructuralSubscriptionTarget, rather than the model classes implements()ing it directly. This is because lazr.restful will only export one interface for each entry. [00:01] 2) URL/traversal for ISS [00:02] 3) The actual export. [00:02] thumper, awesome. [00:02] The first is 300ish lines, the second about 200. [00:02] wgrant: 1) agreed. This is a common practice to export [00:05] 2) barry had some trouble with that last month he landed a change to lazr.restful to make some kinds of object that are not traversable work. The behaviour depends on where the collection has an implied URL. [00:06] 2) barry's work may not be relevant, Do you imagine something like [00:06] /ubuntu/karmic/mozilla-firefox/+subscriptions [00:07] sinzui: I've used +subscription, rather than +subscriptions. There's inconsistency throughout LP as to whether the plural is used or not. [00:08] !?!?!?!??! wgrant: One of those mystery test failures of mine no longer fails (the package-diff.txt one) [00:08] maxb: Launchpad must like you now that you've added Karmic to the ~launchpad PPA. [00:08] heh [00:08] sinzui: /+subscription/ is the scheme. [00:09] that is good [00:09] wgrant: We have not moved structural subscriptions. were to you think they belong? lp/app or lp/registry? [00:10] sinzui: Registry. [00:10] good [00:10] The only other non-obvious changes are making ISST.removeBugSubscription take an unsubscribed_by, and implementing access control in that and addBugSubscription. [00:11] sinzui: Should I move them? [00:11] wgrant: you can if you like. There are tools to do it. I think exporting is more valuable than moving [00:12] sinzui: I'll just export for now, then. Thanks. [00:12] wgrant: I want to move it to registry and add answers and code. I think the move can wait since I am pushing to emphasis subscription and notification in 4.0 [00:13] maxb: What about the other mysteeeeerious failurs? [00:13] sinzui: XMPP! XMPP! [00:13] * sinzui high five's wgrant [00:14] Difficult, but rather useful and impressive. [00:20] would be awesome [00:20] wgrant: the mystery soyuz DONE != ACCEPTED one persists, unfortunately [00:20] maxb: Huh. [00:20] My thoughts exactly :-) [00:23] sinzui: How would I best test the traversability of ISS? There are no views for it and no webservice export in this branch. [00:25] canonical_url to verify you get the expected url [00:26] wgrant:I think you want to test Navigation [00:26] * sinzui think about how we test that [00:26] sinzui: Do I add the tests to canonical_url_examples.txt, or is there somewhere better? [00:27] I think the right place for this kind of test is a unit test in browser.test. [00:27] * sinzui looks for a traversal test [00:28] There's always notfound-traversal.txt! [00:28] * wgrant runs. [00:29] Bug c.l.browser.tests.test_branchtraversal looks like a reasonable example. [00:29] s/Bug/But/ [00:34] That looks like the best example. There is some traversal testing in potemplate-views.txt too [01:45] thumper, ping [01:45] jml: heading to class, back in a couple of hours [01:46] thumper, ok. please let me know when you return. [01:46] akc [01:47] thanks. [01:47] jml: you might enjoy https://code.edge.launchpad.net/~lifeless/bzr/test-speed/+merge/10633 [01:47] * jml looks [01:48] hey, result decorators [01:48] I know them [01:50] so this is a bit of a bug in the composite, that you can't cleanly intercept code around a test [01:51] I'm accumulating a list of such defects [01:53] cool [01:53] I think that's a good idea [01:53] 9 so far:) [01:53] lifeless, have you followed much the recent compatibility work in py.test? [01:53] not at the code level [01:54] I've only scanned the emails. [01:55] lifeless, that patch does look quite useful. [01:55] jml: yes :) [01:56] bzr's selftest selftests take 5 seconds now [01:56] I have plans to make that lower [02:00] Can somebody please ec2test and land https://code.edge.launchpad.net/~wgrant/launchpad/descend-from-istructuralsubscriptiontarget/+merge/10636? [02:01] (it's already approved) [02:05] wgrant, sure [02:05] wgrant, commit message? [02:06] mwhudson, ping [02:08] jml: Never been asked for one before! "Inherit interfaces from IStructuralSubscriptionTarget, rather than declaring implementation in model classes.", I suppose. [02:11] wgrant, thanks. I've kicked off the run, will let you know when it goes headless [02:12] jml: Thanks. Will I be emailed when it's done? [02:12] wgrant, yep [02:13] jml: Even better. [02:14] * wgrant finishes writing tests for the other two bits. [02:15] jml: hello [02:15] mwhudson, would you like to talk about things? [02:16] jml: sure, is in about 5 minutes time about right? [02:16] mwhudson, that'll be fine [02:18] Where were launchpad-dependencies packages kept between the time they were removed from dapper, and when they started being built in the ~launchpad PPA? [02:19] I've not had any response to my question about rescuing a branch, so maybe if those source packages are stored somewhere, fishing it out from there would be best [02:20] maxb, I can't remember, but I'm guessing on some Canonical-internal server [02:20] maxb, what's your 'rescuing a branch' question? [02:20] "[Launchpad-dev] A bzr branch for launchpad-dependencies?" [02:20] https://lists.launchpad.net/launchpad-dev/msg00562.html [02:21] maxb, ta [02:21] I could just ignore that period and import what's already public, but it seems silly not to ask first [02:22] They were on lpdebs.canonical.com, which is devpad. [02:23] There are versions available back to 2008 there, but probably no branch. [02:24] The versions in dapper actually had a branch in the source packages [02:30] jml: ready now? (that was a cook islands 5 minutes, btw) [02:30] mwhudson, sure [02:36] wgrant, it's gone headless. [02:37] jml: What does that mean? It started, checked out OK, and is now actually testing? [02:38] wgrant, exactly [02:38] wgrant, specifically, it's out of my hands now :) [02:38] jml: Thanks. [03:50] Agh! Test failures... *grumble* [03:54] jml: back [04:35] thumper, h [04:35] i [04:35] thumper, up for a quick call [04:35] jml: on with mwhudson right now [04:36] thumper, ok. please let me know when you're done. [04:36] ok [04:36] brb [05:05] jml: ping [05:06] thumper, hi [05:06] thumper, I'll put on my cloak and wizard hat [05:06] jml: ready when you are [05:06] * thumper doesn't get it [05:07] skype won't start :( === Ursinha is now known as Ursinha-afk [05:08] thumper, I'll have to reboot. please bear with me. [05:08] ok [05:13] :( [05:13] thumper: oh, i meant to ask, do we have any imminent sprints/other trips planned? [05:13] mwhudson: not yet [05:14] k [05:14] mwhudson: I'm thinking of one perhaps after the TL one [05:14] so not until at least October [05:14] ok [05:14] jml: still not working? [05:15] thumper, skype appears to be broken in karmic at present. [05:15] :( [05:15] jml: it was working yesterday [05:15] thumper, yesterday I hadn't updated in 2 weeks :) [05:15] :( [05:15] i'm busy on the 3rd of october ( http://www.organicbeer.co.nz/events.html :) ) but i guess you wouldn't want it _straight_ after the TL sprint anyway [05:16] no [05:16] I'm flying then anyway [05:22] does karmic+1 have a 1) name or 2) a uds location/time yet? [05:23] 1) No. 2) Mid-November. [05:23] The exact answer to 2) is on the Karmic and Launchpad release schedules. [05:24] wgrant: do you know anything about a location ? [05:24] mwhudson: No. [05:24] They were very quiet about it this time. [05:24] We knew about Barcelona in Mountain View. [05:25] it'll be in the US somewhere [05:25] we haven't actually got a final location, which is why it hasn't been announced [05:25] Ahaha. [05:26] * wgrant checks and rechecks the clock. [05:27] * mwhudson is slightly confused by the dates on https://wiki.ubuntu.com/KarmicReleaseSchedule [05:27] || 30 || November 19th || || || Karmic+1 Developer Summit || [05:27] the 19th is a thursday [05:27] Right. [05:27] Ubuntu releases are on Thursdays. [05:27] ah [05:28] i guess the uds will (probably) be the week containing that thursday then [05:28] Yes. [05:28] mind you, no idea if i'd be going [05:28] mwhudson: want to? [05:28] thumper: depends on the dates :) [05:28] :) [05:44] What is happening with branch statuses? [05:44] so [05:44] It seems to me like that column in the listings would be much better filled with the MP status. [05:44] wgrant, not much. there's a general trend toward killing them. [05:45] * jml returns to his other point [05:45] well... [05:45] we tried [05:45] but people complained [05:45] I'd like a desktop application that: [05:45] a) registers me a launchpad account [05:45] b) sets up my GPG key [05:45] c) sets up an SSH key [05:46] and by 'desktop', I might actually mean 'command line bzr plugin' [05:46] but I might also mean 'GTK thing with fat buttons' [05:46] jml: what about a "quickly" application :) [05:46] I'm not sure, and I want to make this happen somehow... [05:47] thumper, not a bad idea [05:47] * wgrant would prefer 'bzr lp-register' [05:48] thumper, have we loaded 'Mature' with any behavioural consequences? [05:52] nope [05:52] the only consequence is marking branches merged in the scanner [05:53] and changing merged branches to development on new revisions [05:53] we don't change "special" branches, those linked to things [05:53] as in series [05:53] This could do with documentation... [05:53] but we don't have any special meaning with experimental or mature [05:53] The second bit in particular is unobvious. [05:54] wgrant: we could do with much more documentation on lots of areas [05:55] thumper: But there aren't too many areas where there is a script silently and secretly doing magic behind the scenes. [05:55] Magic that hides and unhides my objects. [05:55] wgrant, I agree. [05:55] it's kind of a misfeature in a way [05:55] Particularly as there's no email notification. [05:55] I mean, we shouldn't mark series branches as merged [05:55] jml: we don't [05:56] wgrant: there will be soon [05:56] but we probably shouldn't make linking a branch to a series the only way of stopping it from being automatically marked as merged [05:56] thumper: Ah, excellent. [05:56] thumper: I'm not even emailed when my merge proposal is automatically closed... [05:56] wgrant: that is the main email that will be happening soon [05:57] wgrant: we should also do the branch one [05:59] * thumper goes to chop food, bbl === abentley1 is now known as abentley [06:24] * mwhudson eods [06:31] mwhudson, ciao [06:33] jml: Did something go wrong, or is PQM just slow? [06:38] wgrant, not sure. [06:38] wgrant, pqm hasn't emailed me... [06:38] but it says it's not doing anything [06:39] Hmmm. [06:40] where do I file bugs about mailing lists? [06:40] bugs - launchpad [06:40] operation issues, answers [06:40] thanks. [06:40] launchpad-registry [06:50] * jml turns his attention back to pqm [06:51] wgrant, I don't know what went wrong. I'll submit directly to PQM and see what happens [06:52] jml: Thanks. [06:56] hmm. I can't figure out how to do this without fetching the branch to local disk [06:56] This often seems to be a problem. [06:57] Some manage to do it remotely. Some push it up under their own user on LP. It's all a mystery to me. [06:57] jml: patch time [06:57] the way others do it is to submit --dry run then manually create the job [07:07] ohhhhhhhh [07:08] the next time I need to do it I'll patch the submit tool to do it for me [07:08] wgrant, what happened was that I changed my local config to make 'stable' the submit branch (for more interesting diffing), but that makes ec2test & pqm-submit both try to land the change onto stable [07:08] wgrant, which doesn't work [07:15] jml: Aha, I see! [07:15] jml: Easy fix, then. [07:16] wgrant, right. your branch is going through pqm now. [07:16] jml: Thanks. [07:28] done. [07:28] (waaaay too slow) [07:31] Great. [07:34] Why are so many pagetests failing on devel? === jtv1 is now known as jtv [07:48] g'night all [07:49] Night jml. [07:49] Thanks. [08:22] good morning [08:25] Is there an idiom for running the same test class over several target model classes? [08:26] In particular, I need to test a new traversal mixin for all the pillars and more. [08:26] Or can I just get away with testing it once? [08:30] wgrant: there's bzrlib.tests.multiply_tests [08:31] mwhudson: Is there? [08:31] * wgrant can't see it. [08:31] something like that [08:32] 'pydoc bzrlib.tests.multiply_tests' [08:33] mwhudson: Oh, in __init__.py, I see. [08:41] mwhudson: But do I need to do it? None of the code tested differs between the targets -- the only possible way it could fail is if somebody removes the mixin from one of its descendants. [08:45] mwhudson: And if I do need to, I probably can't use it straight from bzrlib, can I? [08:47] wgrant: i don't know about need to, but yes, you can just use it [08:47] for a testcase-style test, of course, it would be a bit awkward (and not really a good idea) for a doctest [08:48] mwhudson: I actually just found another related test which does the same thing for a doctest. [09:11] wgrant: another example (but of multiple implementations of an interface) is lib/lp/soyuz/tests/test_hasbuildrecords.py [09:16] noodles775: That makes four existing rather different ways of doing it, I think... [09:16] Morning [09:43] good day, im having problems setting up launchpad instance. I have following setup: nginx proxy (public-ip) - intranet server running apache/launchpad (private-ip) [09:44] for now LP works great if i run firefox on nginx-server and go to https://launchpad.dev [09:45] LP is running on intranet server on two ip-aliases .50 and .51, setup using HOWTO from https://dev.launchpad.net/Running/RemoteAccess [09:46] any pointers how to do some rewrite magic on public ip server so i could access intranet LP from outside [09:47] i guess i have to do some mod_rewrite magic to parse https://public-ip/launchpad url to be proxyed intranet server as target launchpad.dev? [09:48] as it seems that its not possible to access LP using just ip-addresses? [09:49] LP makes extensive use of vhosts. You are probably not going to have too much luck rewriting it from under a subdirectory. [09:52] ok, is there any pointers then how is could start progressing. I guess someone must tried using nginx as a frontend proxy for LP? [09:52] I doubt it. [09:53] I know of four proxied LP instances. They all use apache and pound, and are run by Canonical. [09:53] What are you trying to do? [09:55] basically im trying to setup system with single public ip and multiple intranet servers to serve different kind of stuff all proxied by nginx, for example i have 2 trac instances running on intra servers, and now tried to setup LP also to sit in intra [09:55] im not actually sure if my approach is compelety sane :D [09:55] What are your plans for the Launchpad instance once you get it working? [09:56] to be quite honest, dont actually know, im just man in the middle trying to get it "working" [09:56] That really affects how it should be set up. [09:56] uh, nice [09:57] this is kind of, "try to get it running" task [09:59] maybe i should then slow down and do littlebit of background investigation (rtfm that is) any pointers on documentation of dependecy between set up procedure and actual purpose how LP will be used [10:00] The only documentation is for setting up local development instances. [10:09] ok, i guess what we want is local development instance, but with remote connectivity on top [10:11] That's what /Running/RemoteAccess is for. [10:11] But proxying is way out of that page's scope. [10:12] neurocid: what are you going to be using it for? [10:17] bigjools: i guess most intereting parts are code hosting and ppa:s [10:17] neurocid: but what are you going to use it for? [10:17] umm, im sorry, what you mean [10:18] There are licensing restrictions on the default Launchpad images. You cannot use them on a non-development, non-testing instance. [10:18] neurocid: are you setting up a development host so you can improve the code, or are you running your own service? [10:19] running own service [10:20] neurocid: then you won't get much help here. You are also breaking the law if you don't replace all the trademarked graphics. [10:21] bigjools: ok, that is absolutely not my purpose. i really have to check the licensing terms [10:22] neurocid: Why do you want to run your own service, rather than using launchpad.net? [10:23] wgrant: well as i said im just "man in the middle" with task "see what LP can do/try to get it running" [10:24] neurocid: You must have some idea of what it's being tried for [10:24] If all you wanted was to see what it can do... well just wander over to http://launchpad.net [10:25] i guess purpose is to codehosting&trac integration [10:25] i.e. for use, not development of launchpad itself [10:26] I'd dearly like to run a couple of private instances myself, but the license says "NO!" [10:26] neurocid: one of the major features of Launchpad is that there's only one of it, which makes it much better at collaboration, which is one of its aims [10:26] There are two very different types of private instances. [10:26] One I consider acceptable, one less so. [10:27] But both are impossible without replacing the images. :/ [10:28] maxb: as i said, i wasnt aware of those licensing terms, and now that i am, i have to reconsider all of this [10:29] And there are so many images, that it's an effective prohibition [10:29] That is most likely the point. [10:29] Although it does also serve to protect the brand, I suppose. [10:34] ok, but i thank you all for kind answers, just re-read Getting page and actually saw the note about image/icons licensing [10:38] by no means any licence violation was not my purpose, this just plain ignorance, good that you set things straight [10:40] Hey [10:40] https://dev.launchpad.net/Getting recommends installing in a virtual environment [10:40] I have not stayed on top of virtualization stuff [10:41] what is the current latest, greatest, easiest way to set up a virtualized ubuntu guest on an ubuntu host? [10:41] ddaa: vmbuilder [10:41] cool, never heard of this one :) [10:42] It is awesome. [10:42] there's some good help pages [10:42] Takes a couple of minutes on a good host. [10:42] and apt-proxy :) [10:43] ddaa: Though vmbuilder only builds the VM image -- does your CPU support KVM? [10:43] and soren is really helpful! [10:43] wgrant: presumably, it's reasonably recent thinkpad [10:44] ddaa: I've not set it up in a virtual env [10:44] ddaa: Make sure it's enabled in the BIOS; even my three-month-old T400 came with it disabled. [10:44] ddaa: long time no talk :) [10:44] hello thumper [10:44] ddaa: how's tricks? [10:44] currently CTO of a business without income, funding, or a team [10:45] but working on learning my trade [10:45] ddaa: good luck [10:45] intending to draw inspiration from launchpad for Windmill integration [10:50] * bigjools repeats the good luck [10:50] thanks luck would be useful [10:51] but at the moment, we need funding more :) [10:52] Meeting investors in a couple of weeks for a few 1e5€ [10:52] I don't think it's at all necessary to install lp in a virtual environment. You just have to pay attention to what it's doing to your apache and really really heed the warnings attached to launchpad-database-setup [10:53] I was a lp coder for years. [10:53] I know the kind of stuff it does in the system. [10:53] then it's even less necessary to use a virtual environment :-) [10:54] I do have a dep on a specific apache config [10:54] I'd like to keep lp stuff in a box if I can. To save me some trouble. [10:55] Fair enough. I've not used the fancy virtualization stuff much, so I tend to regard it as overkill if there is a non-virtual simple solution [10:59] ddaa: I got it working in a vm [11:00] it's easy, the trick is to specify the right sources.list entries to vmbuilder, then you're good to go [11:00] bigjools: can you be more specific about "the trick"? [11:01] I initially ran it in a chroot, but once I worked out what evils it did I moved it out. [11:01] Morning, all. [11:02] Morning deryck. [11:12] yeah, the chroot approach gets a bit insane. Especially if you have a postgres inside and outside the chroot [11:12] I had a postgres outside the chroot, and inside another two apart from LP. [11:13] It's sometimes a tad confusing when things stop working. [11:14] * wgrant demonstrates a masterful command of understatement [11:14] heh [11:14] It's even more confusing when you install postgres in a chroot with another postgres already running in another. [11:14] Because it works, but creates the cluster on 5433 instead. === henninge_ is now known as henninge [11:58] danilos: How do I return an empty ResultSet? [11:58] henninge: use the condition along the lines of 1=0 or something [11:59] danilos: does that get optimized somewhere to not execute the query? [12:01] henninge: not sure, but postgres would be very quick with that query, so unless you have that done a big number of times (when you can use the same one), I'd just go with it [12:02] danilos: thanks [12:02] henninge: you can look through the Storm code to see if there's something for it, and if not, file a bug (and/or fix it in there :) [12:02] storm code .... [12:02] terra incognita ... [12:02] danilos: err, not now ;) [12:04] mrevell: ping [12:05] hello there danilos [12:05] mrevell: I need you to come up with nice one paragraph introduction for translation groups on translations.launchpad.net/+groups [12:05] * mrevell looks [12:05] henninge: there's an EmptyResultSet [12:05] and danilos, FYI ^ [12:06] danilos: To replace the existing two paragraphs, right? [12:06] mrevell: I've got http://pastebin.ubuntu.com/259233/ so far (mostly what it was) [12:06] * mrevell looks [12:06] bigjools: yep, seen it, thanks [12:06] bigjools: cool [12:06] thanks [12:06] mrevell: you'll be free to provide pop-up help later on, but I just want to get our 3.0 template migration off the zero now :) [12:07] :) [12:17] hi danilos! [12:17] danilos: I thought your migration branch was already landing yesterday... Did I get that wrong in my notes? [12:18] jtv: which one? that one was for 'Translator' forms, this one is for translation group listing, and I've got another one coming up for translation group overview (they have both gotten UI reviews yesterday, if that counts :) [12:19] danilos: there was some confusion in the UI call last night because no conversions showed up in the overview yet. [12:19] jtv: haven't you seen the total number of the templates go below 50? [12:20] danilos: I thought it was 51 last I checked, though I haven't made a point of remembering... [12:20] jtv: so, we've had 5 template removals (replaced by generic-edit.pt) so far, which is 10% of the numbers... my branch in review moves another one to generic-edit and migrates one template [12:20] jtv: we had 52 at least at one point [12:21] danilos: cool, thanks [12:21] * jtv scribbles [12:22] danilos: I thought you converted two TranslationGroup templates in the same branch, so I thought I'd recognize the removals as a piece of blue bar in the graph. [12:23] jtv: nope, that would have resulted in ~2000 line diff, which is why it's all split up [12:23] danilos: I don't know what's wrong with reviewers nowadays... they've gone so _soft_... [12:23] :-P [12:24] jtv: heh, right, I see you still push them with >1k branches ;) [12:24] *cough* *cough* [12:25] Around 600 I thought, "this is headed towards the branch limit, better split off what I have now." First question I get on -reviews is "can't you split it up?" [12:25] jtv: also, the simple insertion of transaction.commit() with every potemplate helped reduce the langpack exporter transactions; stub mentioned how after that is CPed (waiting now), we should again start killing all transactions that are longer than 3h [12:25] jtv: will that be a problem for export-to-branch? [12:26] danilos: not at all, we have a similar thing cherrypicked there. :) Thanks for adding them to langpack; I've been yearning to get that behind us. [12:26] jtv: ok, great to hear that [12:27] danilos: migration of +translations-to-review is being reviewed; it was very easy because the page was simple, and I hope it opened the door for the rest of our Person pages. [12:27] jtv: it might, depends on what you decided to do with navigation menu items [12:27] danilos: "related pages." [12:28] jtv: ok, we can do the same on all the rest person pages, except the main one :) [12:28] Looks a bit ugly, but consistent with what we have elsewhere. I also added the +translations-to-review page to the nav menu. [12:28] * henninge lunches [12:28] jtv: you can also move it to the right side menu [12:28] henninge: be quick [12:28] danilos: in this case I couldn't justify a sidebar. [12:28] danilos: just have to walk to the kitchen today [12:29] danilos: stop distracting henninge so he can eat :-P [12:30] Ooh. [12:30] * wgrant likes and loathes devel r9224. [12:31] wgrant: I am just scared ;) [12:32] danilos: Look ok? http://pastebin.ubuntu.com/259245/ Will make that translation group article live after lunch, which I'm taking now. === mrevell is now known as mrevellunch [12:33] mrevellunch: sounds good, thanks === abentley1 is now known as abentley === matsubara-afk is now known as matsubara === abentley1 is now known as abentley === carlos__ is now known as carlos === salgado-afk is now known as salgado === mrevellunch is now known as mrevell [13:39] cprov: I see the email discussion about that issue is rather dead, even though it's blocking uploads of various packages. [13:40] wgrant: we have to resurrect it, I guess [13:41] It's interesting that it's happening so frequently. [13:42] I suppose in most cases people have logged into Launchpad once, and done nothing at all, so their Person was removed a couple of weeks ago. [13:42] But that seems unlikely! [13:55] beuno, ping. === abentley1 is now known as abentley [14:06] salgado: hehe, I just renamed the breadcrumbbuilder to breadcrumb in my branch :) [14:06] salgado: and then I see the email ;) [14:06] heh [14:07] danilos, lp:~salgado/launchpad/real-breadcrumbs has a fix for the bug you reported and is on ec2 now. if all tests pass it'll reach PQM in 3h [14:08] salgado: cool, I'll try it out and see if it works, and just get my branch landed with no rootsite set knowing that once your branch lands it'll be fixed :) [14:13] danilos, I guess that means you won't be writing any tests for your changes, right? tsc, tsc [14:14] salgado: bread crumbs? nope, I am just lazy like that... fwiw, I have tests for canonical_urls, but I don't want to test that breadcrumbs are broken or not broken when they should be generically tested [14:14] deryck, hi [14:14] salgado: i.e. it'd be like testing that launchpadform works everywhere properly [14:15]