[00:25] thumper, you around? [00:27] or mwhudson? [00:27] jml: hello [00:28] mwhudson, I'd like to talk about expanding the membership of the ~vcs-imports team [00:28] jml: ok [00:28] mwhudson, probably at least to include ~canonical-bazaar [00:28] jml: to "trusted community types" or ... ? [00:28] ah right [00:28] well that would make sense, yes [00:28] mwhudson, and maybe trusted folk interested in ubuntu-distributed-development [00:29] mwhudson, we could do that on a fairly ad hoc basis I think. [00:29] yeah [00:29] so, +1 [00:29] what is there to talk about? [00:29] i guess we need to consider the dangers of a member going rogue [00:29] mwhudson, that was it. I just wanted to confirm. :) [00:29] mwhudson, also, I don't have permissions to add members to ~vcs-imports [00:30] thumper is the owner these days, i think? [00:31] oh no, it's lifeless still [00:31] mwhudson, also, what do you think about putting ~canonical-bazaar into ~bazaar-experts? [00:33] jml: i thought you wanted to delete ~bazaar-experts :) [00:33] mwhudson, I do. [00:33] mwhudson, but while it's there... :) [00:36] lifeless, can you please add ~canonical-bazaar to ~vcs-imports? [00:37] mwhudson, is the import approval howto on the public wiki? [00:38] mwhudson, ahh, I see that it isn't. I can move it to the dev wiki and edit the CHR-specific bits if you think that'd be OK. [00:42] jml: sounds like a good idea [00:42] mwhudson, ok. [00:43] mwhudson, I guess the biggest risk in expanding the ~vcs-imports group is that it gives write access to a lot of branches. [00:53] * jml is offline for a while [01:22] mwhudson: is that an ack for jml's request? [01:23] lifeless: yes [01:23] mwhudson: thumper is an admin [01:23] jml: done [01:34] we are we adding? [01:40] thumper: scroll up a half page [01:41] thumper: how about we change the vcs-import team owner to be canonical-bazaar [01:41] * thumper looks [01:44] lifeless: you are the team owner at the moment [01:44] I know [01:44] I don't know if we can change the team owner [01:44] I can't [01:45] I can [01:45] its conceptual agreeement to changing it that is needed. [01:45] [only team owners & admins can change team owners] [01:57] lifeless: sure, why not [01:57] lifeless: we should be able to trust the members of canonical-bazaar [01:59] thats my thinking :) [01:59] done [02:58] wow the import fascist is sure annoying right now [02:58] yeay [02:58] h [02:58] damn [02:59] WTF are we supposed to do with questions that obviously are nothing to do with launchpad? [02:59] thumper: i assign them to ubuntu [03:04] * wgrant looks with confusion at the response to https://answers.edge.launchpad.net/launchpad/+question/96334 [03:05] wgrant: yay [03:12] ahh. structural subscription. [03:12] Yeah. [03:12] I have to wonder how so many people manage to do that. [03:18] hi [03:18] https://edge.launchpad.net/bzr/+download seems to be persistently (over hours) timing out [03:19] poolie: Presumably very related to your bug #499351... [03:19] Bug #499351: product downloads page causes browser hang/unhappiness [03:25] yes [03:40] i mean they're both basically reflections of the same problem, being that too much stuff is packed onto this page [04:10] dumb arse ec2 land command [04:10] yes there are multiple proposals, but one is already merged [04:11] I obviously mean the other one [04:15] * mwhudson eods [04:29] mwhudson, I've got a branch open to fix it [04:29] mwhudson, the import fascist stuff, that is. Maybe I'll fix it at the airport :) [04:29] jml: as in, up for review? [04:30] mwhudson, no, on my laptop [04:30] jml: i see [04:31] mwhudson, I saw some people complain about the warnings on IRC & email so I thought reducing the number of warnings might be a fun light-weight hacking job [04:31] jml: yeah, i had vaguely similar thoughts [04:31] jml: when are you next in an airport? [04:32] mwhudson, Sunday, I think. [04:32] mwhudson, also, I need to fix all the failures in the 'use-testtools' branch. [04:33] jml: there were failures? how boring [04:33] mwhudson, eh [04:33] mwhudson, I meant to say "heh" [04:33] mwhudson, testtools forces you to upcall setUp and tearDown [04:33] mwhudson, because I was sick of stupid, hard-to-debug, order-dependent test failures. [04:34] ah right [04:42] mwhudson, also, there are some addCleanup tests that I forgot to delete [04:56] * thumper EODs === jtv is now known as jtv-afk === stub1 is now known as stub [08:23] good morning [09:10] Morning [09:13] Hi mrevell! [09:13] wgrant: hi, can you have a scan of the info I've added to bug 503440 and see if you can glean anything out of it? (it's the oopsing build you pointed out yesterday). [09:13] Bug #503440: building build without a bq.builder causing LocationError OOPS [09:25] wgrant: actually, the timing looks like it is actually the result of some manual SQL that was run at the time... I'll update the bug. [09:29] morning all [09:32] noodles775: Is it intentional that the logs in that paste are not in any kind of order? [09:34] wgrant: no, I just did grep 1376913 `find . -name 'buildd-manager.log.*'` > ~/build1376913.log [09:37] wgrant: new comment added with a most-likely reason. [09:40] noodles775: Ah, that's good. [09:40] So there's only one known bug remaining. [09:41] Seems so :) [09:41] Although nothing jumps out to me about that one. [09:42] Yeah, me either. And the fact that build 1376913 was dispatched without causing the issue so many times, and then fell over... [09:54] noodles775: So it is still happening, without bad SQL being invoked? [09:56] wgrant: There are two occurrences of 499421 that I'm aware of, the one linked above on 2009-12-21, and once during the break (on the 2010-01-01, as commented on the bug). [09:57] But we don't know the build for that second one yet (I'm just organising to have the losa query updated so that we get the build id when it occurs). [10:20] jml: just followed up to your db review... are you content with the answer? === jtv-afk is now known as jtv [10:50] hi adiroiban! [10:51] jtv: hi [10:53] adiroiban: sorry for running away from the discussion last night :) [10:53] jtv: don't worry. there is plenty of time to solve the problem this year [10:54] adiroiban: true... as years go, this one still has a lot of unused days [10:56] About the notion of an IProductTranslationPolicy: I was thinking that the TranslationPolicy object would give admin permissions to members of self.context.owner, self.context.translationgroup.owner if set, plus of course admins / rosettaadmins. That way, I thought you wouldn't need to have any specialized I{Product,Project,Distribution}TranslationPolicy. [10:56] But does that not work? [10:57] jtv: i don't know. I am still new to zope and LP code. [10:57] I don't know how to configure the view [10:58] in order to create vies for TranslationPolicy [10:58] but with different objects [10:59] I know that I can use an adapter to generate TranslationPolicy(IProduct) or TranslationPolicy(IDistribution) [10:59] adiroiban: you'd have one TranslationPolicy, and one view class, that wouldn't know or care what object it's operating on... but how to get from "product" to ITranslationPolicy(product)... that's a good question [11:00] I wonder if you can just register a view for interface ITranslationPolicy but class Product, and get what you want automatically. [11:00] This is a web framework after all; the whole idea is that you spend days to figure out how to get automatically in a few cryptic lines what you could have coded up in half an hour. Progress. :) === salgado-afk is now known as salgado [11:11] jtv: what do you mean by „register a view for interface ITranslationPolicy but class Product”? [11:11] something like this http://paste.ubuntu.com/352254/ [11:11] how whould to know know to create a ITranslationPolicy(Product) context ? [11:13] adiroiban: sorry, page. Register a page, with class=Product (and similar for Project & Distribution) and interface=ITranslationPolicy [11:14] bigjools, http://pastebin.ubuntu.com/352255/ [11:14] adiroiban: ah, no, it doesn't work like that :) [11:14] true [11:15] what you are sugesting is similar with my patestbin [11:17] I can think of having ITranslationPolicyProduct [11:17] and an adapter to build object for this interface === matsubara-afk is now known as matsubara [11:20] adiroiban: or maybe just having ITranslationsPolicy with a view, and adapters for IProduct etc. plus a single will create the pages you need [11:21] * adiroiban trying that code [11:26] jtv: where I should put this common view in browser/* ? [11:26] adiroiban: how about under translationgroup? [11:27] I think that's also, on the interfaces side, where the enumeration for open/structured/... lives [11:27] ok [11:28] but we will rename IHasTranslationsGroup to ITranslationPolicy , right? [11:29] adiroiban: yes === mrevell is now known as mrevell-lunch === mrevell-lunch is now known as mrevell [13:42] bac, hi :) I see you were approving a branch merge from jtv when you got this oops, am I right? https://lp-oops.canonical.com/oops.py/?oopsid=OOPS-1466EB1022 [13:43] Ursinha: i don't recall getting that oops [13:43] https://code.edge.launchpad.net/%7Ejtv/launchpad/validate-translations-file/+merge/16866/+index [13:43] the MP was claimed and done with no problem. hmmm. [13:43] bac, I saw that the link works fine, but that error seems odd [13:46] bac, I'll ask a code person, but thanks! [13:47] Ursinha: thanks for asking [13:47] bac, no problem! [13:53] hi Ursinha, did you ever find my oops yesterday? [13:55] james_w, I'm trying to find a message I sent to the channel yesterday [13:55] james_w, asking you if that was the right oops id, because I couldn't find it [13:57] oh, sorry I missed it [13:58] james_w, ah, no problem, my connection could be the culprit as well [13:58] james_w, anyway, did you copied/pasted that id? [13:58] yeah [14:01] james_w, ah, were you doing soyuz related stuff when that happened [14:01] ? [14:01] correct [14:01] with the API [14:02] I recall a bug that something in soyuz wasn't recording oopses [14:02] I'm trying to find it [14:03] james_w, bug 499423 [14:03] Bug #499423: No oopses recorded for upload processor [14:05] james_w, does that relate to what you're doing? [14:06] I don't think so [14:06] this was just an API call to get data from the db [14:07] OOPS-1466EC775 is another from around the same time [14:08] let me check [14:10] matsubara, the oops needs to be processed by the db oops tools to be found at lp-oops.canonical.com/oops.py/?oopsid=OOPS-1466EC775 ? [14:11] nope [14:11] Ursinha, ^ [14:11] james_w, both oopses were generated on 4th? [14:11] yeah [14:11] that's why then [14:12] only one is stored per-traceback per-day? [14:13] matsubara, is there a way to see the oopses besides that one? [14:13] james_w, we had an issue with oops-tools yesterday due to an enormous amount of oopses on staging [14:13] james_w, it couldn't process all of them in a reasonable time [14:14] ah, ok [14:14] Ursinha, search the logs on devpad [14:14] I'm pretty sure the other oops will be the same, so if the reason is known why it disappeared then I don't think we need to look much more [14:14] are the bugs associated with oopses done manually or automatically? [14:15] james_w, manually in a first place, we add the numbers as we're doing triage [14:15] because the bug linked to 1466EC775 doesn't sound right to me [14:15] james_w, and then when oops-tools processes the summaries the next time it considers the linking [14:16] Chex: ping [14:16] matsubara, mind explaining how it works in detail to james_w and why the linking isn't that precise? [14:18] matsubara, I'll have to run oops-tools-db to process all the 4th oopses, will suspend cron entries until it's not finished [14:20] james_w, we consider the exception type and exception value to generate a signature for an oops. we link a signature to a bug report manually. if later on an oops show up with the same signature it'll show up linked to the bug. problem with timeout errors is that the signature generated is misleading. [14:20] we need to consider other things (e.g. like the pageid) into the signature to make the signature -> bug link more reliable. [14:21] Ursinha, ok [14:21] makes sense, thanks === salgado is now known as salgado-lunch === fjlacoste is now known as flacoste [15:18] jelmer: ping [15:18] bigjools: pong [15:18] jelmer: how is that cheery pick branch going? [15:18] cherry! [15:20] Oh, right! Tests seem to pass so I'll pqm submit. [15:40] Chex: mthaddon: ping [15:40] sinzui: losa ping is a much better way to go [15:41] mthaddon: I certainly will use that in the future. [15:41] thx [15:41] mthaddon: Do you know why edge is not updating? [15:42] <^Willie^> hi there [15:43] <^Willie^> any idea where i must goto for freenx problems ? === matsubara is now known as matsubara-lunch === salgado-lunch is now known as salgado === beuno is now known as beuno-lunch === beuno-lunch is now known as beuno === matsubara-lunch is now known as matsubara === EdwinGrubbs is now known as Edwin-lunch === flacoste is now known as flacoste_lunch [18:14] night! === deryck is now known as deryck[lunch] === flacoste_lunch is now known as flacoste === Edwin-lunch is now known as EdwinGrubbs === matsubara is now known as matsubara-afk === deryck[lunch] is now known as deryck === noodles76 is now known as noodles775 [20:13] morning [20:57] thumper: around? [20:57] bigjools: aye [20:58] coolio, first, happy new year! [20:58] second, how's the build from recipe stuff coming along? [20:58] bigjools: skype? [20:58] bigjools: I could put off jml for a few minutes [20:58] bigjools: since it is your evening :) [20:59] I can't, my headphones are in my office outside :) [20:59] or you can phone my home number? [20:59] bigjools: it is coming along, mwhudson is working on it [21:00] bigjools: I can call your home :) [21:00] slower than i would like but oh well [22:05] spm: have a few minutes for me? [22:06] I'd like to ask about graphs [22:07] sinzui: you have a few minutes for a quick pre-imp call? [22:08] i do [22:12] sinzui: sorry. try again? [22:12] yes === Ursinha is now known as Ursinha-afk [22:42] yay for forgetting to plug my laptop back in, for power management being busticated so no software noticed and then for fsck [23:00] james_w: sure, shoot [23:00] spm: I'd like to graph some things about the bzr import service on jubany [23:01] one of them is an LP db query, but the rest of the info doesn't live in the LP db [23:01] what do I have to put in an RT request to allow such data to be graphed? [23:01] the 1st is easy - I'll show you where/how to set that up; the latter can be harder. [23:01] are names of scripts that spit out numbers sufficient? [23:02] start with as much detail as you can. we can work out who does what from there. [23:02] more or less. we prefer the results in cricket format; but can wrapper a converter if needed [23:03] james_w: eg None_localhost_disk_partition_pct_/:17@1226983004 <== name, sensible alpha format for ease of finding == win; value (17) and the timestamp [23:03] the data lives in sqlite now, which makes it much more accessible than before [23:03] ahh, very helpful [23:04] I'm not sure what that means, but hopefully cricket does :-) [23:11] spm: do you have a link on interpreting that format so that I can write a script to output it? [23:11] james_w: hrm. well the fields are: :@ [23:12] ah, easy enough [23:12] just trying to find an example script for you... [23:13] and can we don multiple lines of that? [23:13] do, I'm not suggesting you wear the output as an outlandish garment [23:14] hahaha [23:14] yes :-) [23:14] my home dir on chinstrap snmp_stats.tgz should give you a script you can point at any snmp service for the idea [23:16] thanks [23:49] * jml afk [23:53] anyone interested in LP apparently giving up in the middle of writing data back to the client? [23:53] I've caught it in the act with the API [23:54] :-| [23:54] it seems to happen frequently with the API, but I don't get it with the webapp that I have noticed [23:54] james_w: foundations would be [23:54] it may be an order of magnitude thing [23:54] james_w: file a bug [23:54] well, I mean I have the response in pdb [23:55] so if someone wanted me to get certain data this would be a great chance to do so [23:55] heh [23:55] umm... [23:55] flacoste: ping ^^^