unless also view.override_title_breadcrumbs = True [23:24] <wgrant> Argh. So many unStormified queries. [23:24] <danilos> barry, hum, I am surprised... what happens with the overrides of the page title then? [23:24] * barry must remember to update the wiki page [23:24] <wgrant> Stormified ones would be easier to fix :( [23:24] <barry> danilos: by default, nothing. or maybe i misunderstand the question [23:24] <danilos> barry, ok, so it's kinda kludgy, but will work for me :) [23:25] <barry> danilos: yeah, salgado-afk and i thumb wrassled over it :) [23:26] <danilos> wgrant, sorry about that, but stormified queries are sometimes harder to optimize (i.e. postgres can choose a bad query plan based on *text matching*; I've seen that with things like "AND is_something" vs. "AND is_something IS TRUE", depending on what exact condition is in the index, i.e. how it's spelled out) [23:26] <danilos> barry, right, thanks for the tip, very good to know [23:26] <wgrant> danilos: Ew ew ew. [23:27] <danilos> barry, however, I somehow have a feeling LaunchpadForms should use the `label` for it [23:27] <danilos> wgrant, of course, most of the non-storm queries are just old, never migrated code, but there are a few queries like these [23:28] <barry> danilos: wiki page updated [23:28] <danilos> barry, anyway, that's just random musings, having something is a great improvement [23:28] <danilos> barry, thanks! [23:28] <barry> danilos: you might be ultimately be right. it would be easy to change and it would be nice to DRY. maybe see what salgado-afk thinks? [23:28] <barry> danilos: sure! [23:29] <danilos> barry, can you please post to the list as well so everyone else can start figuring breadcrumbs in before Friday? [23:29] <barry> danilos: +1 [23:29] <danilos> barry, sure, not sure if we can make it happen this week (i.e. I'd rather have something people can rely on than keep changing it this week :) [23:32] <barry> danilos: good point. we're cutting it to the wire, so let's keep what salgado-afk has [23:40] <thumper> barry: are you available for a quick call? [23:41] <barry> thumper: not right this sec, but could be in a bit [23:42] <thumper> barry: could you ping me when you have some time? [23:43] <wgrant> cprov: Since there are quite a few callsites, would it be a good idea to create a general solution and add a property on IArchive listing all grouped archives? For now that would only be useful for primary, but it would make queries much simpler and seems more future-proof. [23:45] <cprov> wgrant: archive.distribution.main_archives ? [23:45] <wgrant> cprov: No problems will occur if partner is included? [23:45] <wgrant> I guess not. [23:45] <thumper> mwhudson: got a few minutes? [23:45] <cprov> wgrant: no, there is no intersection [23:47] <mwhudson> thumper: sure [23:50] <jml> maxb, what's the story with karmic? [23:51] <barry> thumper: i'm going to grab some dinner and will ping you when i'm done [23:51] <barry> jml: karmic works great! you need to install python-boto by hand if you wanna do ec2 though [23:51] <maxb> as far as upgrading goes? Don't yet. Give it at least one more publisher cycle at a minimum, or if you want to be safe, wait until tomorrow [23:52] <barry> with that caveat [23:52] * barry -> dinner [23:52] <jml> maxb, can do. [23:52] <jml> maxb, thanks for the warning [23:57] <jml> any registry folks still around? [23:58] <mwhudson> jml: good morning [23:58] <jml> mwhudson, hello

[00:01] jml, thumper, other canonicalers; have any branches to ec2test currently/soon? [00:01] not yet [00:01] mwhudson, sorry, no. [00:01] Nope. [00:01] slackers! [00:03] * mwhudson tests the first 'faster --headless' thing [00:07] rockstar: call? [00:07] thumper, sure. [00:07] guys is deryck there [00:08] dhillon-v10, he usually cuts out about 4 hours ago. [00:10] rockstar: alright, thanks [00:10] rockstar: I wanted to talk to him about some code we talked about in mail [00:11] dhillon-v10, the bugs folks all work European hours. [00:12] wow, detached in 8m31 [00:12] in other news: make build is really really slow [00:12] three minutes of that was waiting for the instance to boot [00:13] rockstar: ah, thanks... [00:14] jml: want to review a really simple ec2test branch? [00:14] mwhudson, yeah sure [00:14] or anyone else, it's really simple [00:16] dammit, missed the */5 [00:19] rockstar: http://pastebin.ubuntu.com/271268/ [00:21] jml: you have mail now [00:25] thumper, just type "debugger;" in a single line [00:26] mwhudson, thanks. [00:27] mwhudson, so, make check depends on make schema? [00:28] mwhudson, also, won't this break test options like -1? [00:32] jml: make check runs test_on_merge [00:33] jml: which runs "make -C database/schema test" [00:33] jml: which sets up enough database for the tests to run [00:33] mwhudson, what about 'make build'? [00:33] jml: check: depends on build: [00:33] mwhudson, cool. [00:33] jml: i'm not sure what you mean about "options like -1" ? [00:34] jml: the fact that they start with a "-" ? [00:34] mwhudson, well, they aren't VERBOSITY [00:34] jml: VERBOSITY is a lie in the makefile [00:34] check: clean build [00:34] # Run all tests. test_on_merge.py takes care of setting up the [00:34] # database. [00:34] ${PY} -t ./test_on_merge.py $(VERBOSITY) [00:35] mwhudson, so if you pass in the perfectly cromulent options -vv -1, test_options='-vv -1', and then the 'make check' command will be incorrect [00:35] mwhudson, I think you need to quote test_options when you append it to command in build_test_command. [00:36] jml: actually, i don't think so [00:36] mwhudson, why not? [00:37] jml: because of the amount of times a shell gets to see the arguments [00:37] jml: subprocess.call(["make", "check", "VERBOSITY=-a -b"]) [00:37] is the same as running [00:37] $ make check VERBOSITY='-a -b' [00:37] in your shell [00:38] hmm [00:38] ok [00:38] actually i suspect it makes very little difference in the end [00:39] * mwhudson runs ./utilities/ec2test.py -o '-v -v -v -t xx-codeimport' --headless to be sure [00:39] heh [00:40] mwhudson, approved. [00:40] jml: thanks [00:41] mwhudson, I'm going to reboot my computer while spiv powercycles the router (!!!!!) [00:41] mwhudson: you can combine those to -vvvt for conciseness :) [00:41] thumper: of course, it's less of a test that way though [00:41] are you testing the args? [00:41] * jml would like to have phone calls today... :) [00:41] * jml reboots [00:41] jml: ok (!) [00:41] jml: because laptop no worky? [00:41] thumper: what do you mean? [00:42] ./utilities/ec2test.py -o '-vvvt xx-codeimport' --headless [00:42] thumper: that's what i rand last time, that worked [00:56] thumper, no, because karmic has a new kernel [01:18] thumper, mwhudson: I think both of you owe me calls :) [01:20] jml: yeah, i think you're right [01:20] jml: i'm going to have lunch first though, is a call in about an hour ok? [01:20] mwhudson, that's fine [01:23] jml: I'll call you after mwhudson [01:27] thumper, cool. [02:18] jml: call now ish? [02:19] mwhudson, I'm on w/ flacoste now, sorry. [02:19] jml: ok, ping me when you're done [02:19] mwhudson, will do. [02:22] * mwhudson kicks EC2 in the head [02:22] Started in 11 minutes 25 seconds [02:40] mwhudson, that's roughly half the time cut off, ent? [02:40] rockstar: no, that was just the time it took the instance to start [02:40] it was the first time booting a new AMI though, that might have been why it was slow [02:41] rockstar: do you want to look at this ec2test branch? [02:52] mwhudson, sure. [02:52] mwhudson, I'm not sure I'm the best qualified to offer technical advice on ec2, but I can do my best. [02:53] rockstar: here's the diff http://pastebin.ubuntu.com/271309/ [02:55] Oh, I didn't know you could create an AMI with ec2test [02:57] mwhudson, why move the dirs instead of copying them? [02:57] rockstar: faster [02:57] mwhudson, ah, okay. [02:58] rockstar: that's the whole point of this branch, make ec2test --headless let go faster by updating the download-cache rather than checking it out from scratch [02:58] mwhudson, are there any tests for this? I remember you remarking that there were no tests before, but I'm wondering if there are plans to change this. [02:59] mwhudson, yeah, I deduced it must have to do with making everything faster. [02:59] How does --extra-update-image-command play into making everything faster? [02:59] rockstar: no tests [02:59] oh right [02:59] that doesn't [03:00] ./utilities/ec2test.py --update-image launchpad-ec2test16 --extra-update-image-command 'bzr branch --standalone lp:lp-source-dependencies /var/launchpad/download-cache' [03:00] mwhudson, are you going to remove it though? It seems like it might make be of use. *shrug* [03:00] Ah, okay. So you aren't going to remove it. [03:01] ^ i ran that to make an image with download-cache already on it [03:01] mwhudson, r=me [03:01] rockstar: thanks [03:26] thumper, rockstar, jml: still no branches to ec2test? [03:27] mwhudson, not here. [03:34] mwhudson, no :( [03:34] mwhudson, I haven't actually done any coding today. [03:43] jml, who are you, and what have you done with the real jml? [03:48] he got promoted [03:48] Is all that going to become clear to us at some point? [03:48] There seem to have been lots of changes. [03:53] yeah, we ought to send an email out, I reckon [03:56] quick storm question [03:56] whats the default cache for a store? [04:31] thumper, hi [04:31] thumper, call now? [04:39] I just came across this bug: [04:39] https://bugs.edge.launchpad.net/malone/+bug/276789 [04:39] Bug #276789: inconsistent results for IBugTarget.searchTasks(has_patch=...) [04:40] abel has all but written a unit test as the bug report [04:40] sounds like a perfect bug report then? :-) [04:40] spm, hell yes [04:40] might be a good patch for a new developer to work on. [04:41] * spm ponders if this might be a good way to take advantage of the open source nature of LP and fix some of the most niggly losa impacting bugs :-P [04:42] spm, you mean, write bug reports that include unit tests? [04:42] I actually (jokingly) meant write code [04:43] but given most of my coding experience is assembler, cobol, perl and C... you won't like the patches [04:44] spm, heh [04:44] jml: I'm around now [04:44] thumper, shall we? [04:44] mwhudson, hi [04:44] jml: tho... I could perhaps draw on some php experience - would that help? I know you're a php coder by preference. ;-) [04:45] :( [04:45] rockstar: hello [04:45] spm: the last time I touched PHP, the tip of my finger turned into pure energy. [04:45] mwhudson, is there ever a case where an import branch would not have a code_import? [04:45] spm, it took a month for my eyebrows to grow back [04:45] spm: unittests in assembler would be awesome [04:46] rockstar: certainly shouldn't be [04:46] rockstar: there's no constraint in the db for that though [04:46] mwhudson, I just see a lot of unnecessary checking in the tal for the presence of code_import. [04:47] mwhudson: that implies a level of understanding of WTF the code is actually doing that, imho and experience, is missing from every assembler level program ever written. "keep randomly changing op codes, it's bound to do what we want eventually" [04:47] It certainly makes the page look funny if it doesn't exist, but I think that we have bigger problems should that occur. [04:47] rockstar: branch-index.pt is a little.... organic, i think you could say [04:47] mwhudson, it's not anymore. I fixededed it. [04:47] rockstar: this sounds like one of the smelly bits [04:47] mwhudson, great. Thanks. [04:47] rockstar: np [04:57] mwhudson, I don't see a function in the factory that allows me to create an import branch and a code import together. Is there one? [04:58] rockstar: i think ICodeImportSet.new creates the branch along with the code import [04:58] rockstar: i may be misremembering... [04:59] mwhudson, ah, it does. [05:01] So what's the point of makeImportBranch ? [05:02] rockstar: argl [05:02] rockstar: i guess it's a way of constructing test data that is somewhat inconsistent with that found in real life :/ [05:02] mwhudson, yes, yes it is. [05:02] rockstar: grep for it? [05:02] i bet it's not called all that often [05:03] mwhudson, yea, I think I'm going to file a bug on it as well. [05:03] It should go away. [05:04] mwhudson, it was probably just a bi-product of when we broke up the makeBranch method to make all sorts of branches. [05:05] rockstar: yeah, indeed [05:06] rockstar: or redefine it to "return self.makeCodeImport().import_branch" i guess [05:07] mwhudson, yeah, something like that, although having two methods doing similar things seems a little odd to me. [05:48] thumper, hi [05:49] mwhudson, still no branches :P [05:49] jml: i got stub to test one [05:50] rockstar: hey [05:51] thumper, so it looks like branch titles have crept back in by way of the page headings. [05:51] :/ [05:51] rockstar: let's talk about this tomorrow in more detail [05:51] rockstar: I might get up early even :) [06:10] Oh great, looks like ec2test.py is broken on karmic. WTH [06:10] I thought jml fixed that last week. [06:10] How's it dieing? [06:11] Tells me my python-boto is not supported. I pastebin. [06:11] That sounds out of date. [06:11] If you remove that check, it works. [06:11] But I thought that check was gone in devle. [06:11] wgrant, ah, yeah, now that I think about it, it probably is... [06:12] rockstar: It was removed in devel r9357. [06:12] wgrant, yeah, looks like I was still using the lp-dev-utils version [06:12] rockstar: Ah. [06:14] yay [06:26] Is there a design somewhere for these new breadcrumbs/titles? [06:26] I wonder if some of this breakage is intentional. [06:28] wgrant: yes, there is [06:28] wgrant: it's on some wiki page somewhere, aiui [06:30] my brain is turning into fudge, time to stop for today [06:30] mwhudson: The only relevant one I know of is w.c.c/IForgetWhat/UI/Navigation [06:30] Could be there. Who knows. [06:30] i certainly don;t [06:31] OK. [06:41] wgrant, well, I think there's been so much churn that any wiki page you find will probably be out of date. [06:42] * rockstar also has a brain of fudge, needs to go eat dinner, and have his 1230 nap. [06:46] rockstar: ironic... now you're a slave of the clock because of your time-saver. ;-) [06:47] jtv, less is more, or something like that? [06:47] jtv, I've actually found ways to fudge the clock around get-togethers and such, because I can't always use my friend's bed for a nap. [06:47] rockstar: I was never really comfortable with that expression, from a formal standpoint :-) [06:48] jtv, :) [06:49] rockstar: heh, it hadn't occurred to me... You can't always expect to be at home at just the right time. [06:50] jtv, yeah, which sometimes leads to awkward situations. [06:50] I'm actually not even on Uberman anymore. I still have 4 hours of straight sleep starting at 4 AM - this is for my wife's sanity more than mine. :) [06:51] But that still means I'm only getting ~5 hours of sleep/day [06:51] Uberman? Strange name... No umlaut? [06:51] wow, that's gotta save you some tmie [06:51] time [06:52] jtv, uberman is the hardcore polyphasic, I'm on a more hardcore "everyman" schedule right now. [06:52] * rockstar goes to find viddles [07:03] I'm really not a fan of four thousand line LP Perl scripts that diverged from an unknown upstream version more than four years ago, with little useful VCS history. [07:57] wgrant, which ones are they? [07:57] jml: Just one four-thousand line one -- sbuild. [07:58] A nice fragile untested critical piece of lp-buildd. [07:58] :( [08:01] BjornT, hello [08:03] hi jml [08:04] BjornT, ready for a call? [08:05] jml: yes, in a minute. i'll get my laptop [08:05] BjornT, ok. [08:07] jml: i'm ready. skype? my username is bjorn.tillenius [08:07] BjornT, yes, skype. I've sent you an invite [08:28] good morning [08:29] Hi adeuring :) [08:29] hi noodles775! === abentley1 is now known as abentley [09:21] Morning [09:45] mrevell, hi [09:45] hi jml [09:46] noodles775, hi [09:46] cprov, hi [09:46] hi beuno [09:46] beuno: I'm guessing (or hoping) that he's asleep. [09:47] noodles775, isn't he in the netherlands? [09:47] beuno: nope. [09:47] hm, ok. I guess he moves around a lot. [09:48] noodles775, would you happen to know if it's easy (and cheap!) to get a count of the number of source packages in Launchpad? [09:48] beuno: it sounds like something easy for anyone with db access to do. Or you mean to display on a page? [09:49] noodles775, display on the homepage [09:49] One would first have to work out what is meant by 'source packages'. [09:49] it already hurts [09:49] wgrant, beuno: yes, I'm assuming you just mean all source packages. [09:49] noodles775, lets say yes [09:49] noodles775: That's unlikely. [09:49] well [09:49] As that wouldn't count PPAs. [09:50] Published SPRs, maybe? [09:50] ok, I can see how this gets complicated [09:50] which is why I left it out in my initial work [09:50] which is why I left it out in my initial work [09:50] er [09:50] so [09:50] Maybe all SPRs, in fac.t [09:50] I want to show some stats on soyuz on the home page === al-maisan_ is now known as al-maisan [09:50] beuno: There are already some stats at https://launchpad.net/ubuntu/+ppas. [09:51] aha! [09:51] that is fantastic [09:51] thank you wgrant, that takes me half way there [09:51] as for Ubuntu packages? [09:52] beuno: You probably want an all-time count, not just currently published stuff? [09:52] beuno, hi [09:52] wgrant, well, I don't know if it's as interesting to do an all time count [09:52] jml, hello hello [09:52] beuno: All the rest on the home page seem to be, but OK. [09:53] wgrant, true [09:53] would this include all the versions? [09:53] Yes. [09:53] it sounds like it would be hundreds of thousands [09:54] So, it's easy and pretty cheap to get the total number of almost-current (the number is updated by a cronjob; no idea how often that runs) sources across all of LP. [09:55] wgrant, any hints on where to look for that number [09:56] beuno: SELECT SUM(sources_cached) FROM Archive; There's no method for that yet -- only one that restricts to PPAs of a particular distribution. [09:57] wgrant, thanks [09:59] It would be nice to have realtime stats, but I don't know how expensive that would be. [09:59] noodles775, any idea why no H1? https://edge.launchpad.net/ubuntu/+ppas [09:59] beuno: yes, the branch barry landed assumes that view's implement the label property - that one doesn't yet. [10:00] noodles775, cool [10:00] s/view's/views gar. [10:07] beuno: I'm just looking at bug 429353 - why do we want the 'Enable edge redirect' there in the site_message, if (1) it'll only be seen if people are on edge anyway (note: on the homepage at launchpad.dev there is an option to re-enable the redirection) [10:07] Bug #429353: Move edge message to the footer, allow people to stop redirection [10:07] noodles775, one sec [10:07] otp [10:30] noodles775, back [10:30] right, this is only for people on edge [10:31] beuno: I've got two other questions, I added them to the bug. [10:31] noodles775, I'll reply there then [10:31] (well, one other question, one clarification). [10:31] beuno: great, thanks! === mrevell changed the topic of #launchpad-dev to: This is Launchpad Development Channel | Week 3 [10:43] I think it's pretty important for the message to be obvious, at least for dogfood/staging. [10:43] Not so much for edge. [10:44] wgrant, isn't the background obvious enough? [10:44] I agree though, edge is my target [10:44] beuno: Mmm, possibly. [10:46] My preference would be to have the edge warning right down the bottom, with the others moved to immediately below the watermark. That actually looks pretty good, IMO. [10:46] mrevell: Did you relaly mean to remove all that stuff from the /TOPIC? [10:47] soren: No, not at all. Thanks for pointing that out. === mrevell changed the topic of #launchpad-dev to: This is Launchpad Development Channel | Week 3 of 3.0 | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | 3.0 goal is to convert all templates: http://people.canonical.com/~beuno/conversions.html [10:47] mrevell: Sure. [10:48] soren, what are you doing here? [10:48] soren, are you going to start hacking on Launchpad? [10:48] wgrant, not sure how much freedom we have to diverge between the three [10:48] jml: It's called lurking :) [10:49] beuno: Three? Aren't there just two? [10:49] (but yes, that is a problem) [10:49] soren, but you've delurked now... does that mean you'll start sending in patches? [10:50] wgrant, dogfood :) [10:50] jml: If I had a clue where to start.. Maybe :) [10:51] jml: I haven't quite found my way around the code yet (not that I've tried all that hard). [10:51] beuno: Right, but staging and dogfood need the same kind of message. [10:51] yes [10:51] soren, there's a lot of it :) [10:52] soren, I'm more than happy to help if you have a specific thing you want to try [10:52] but maybe not tonight. I'm starving. [10:52] beuno: There's already edge stuff hardcoded elsewhere, so it's not a significant additional evil to add something in the footer. [10:54] wgrant, I'm fine with that, it's up to noodles775, who will need to pass the code review :) [10:54] wgrant: where's the other hardcoded edge stuff? [10:54] g'night all [10:54] soren, happy lurking :) [10:55] noodles775: The timeout template, the home page. [10:55] wgrant: currently it's just a config site_messages option, that is only used on edge, staging and dogfood. [10:55] jml: I'll keep that in mind. [10:55] jml: Good night :) [10:56] wgrant: ah yep, the is_edge, is_demo, is_lpnet etc. Great. [10:56] dear launchpad developers: please stop breaking the build [10:59] Morning, all. [11:24] hi, is there a way to get statistics (visitors, commits, downloads, etc) for our projects in LP? [11:29] kimus, you can get some of them [11:29] download stats, you've got in the download page [11:29] commits, in the code page, it gives you some of that [11:30] it's mostly scattered at the moment [11:30] beuno: so where can I put a feature request for this? [11:30] if you'd like something specific, feel free to file a bug, it will help us shape the information we present [11:30] kimus, https://bugs.launchpad.net/launchpad/+filebug [11:31] if it's something specific [11:31] if you feel it needs discussion [11:31] the mailing list is the best place [11:31] beuno, hi [11:31] deryck, hi! [11:31] first I will file the bug... then I can go to discussion if needed :-) [11:32] kimus, it's usually easier the other way around, but go for it! [11:32] kimus, mailing list at: https://edge.launchpad.net/~launchpad-users === stub1 is now known as stub [11:34] beuno: the "Subscribe to mailing list" link does not work. I will go to mailman interface to subscribe :-) [11:34] kimus, you need to join the team first [11:35] are you logged into LP? [11:35] I gave you an edge URL (sorry) [11:35] beuno: do I need to join the team to subscribe to ml? [11:35] kimus, yes [11:35] ok... [11:35] (we will change that in the future, curtesy of barry) [11:36] flacoste, hi === beuno is now known as beuno-lunch === beuno-lunch is now known as beuno [13:13] Hello. /me is around for 20mins or so, if anyone has thoughts on my py2.5-unittest-compatibility branch === matsubara-afk is now known as matsubara [13:25] flacoste, around yet? [13:26] deryck, has there been any decision on wether we're dropping mentoring? (please do!) [13:30] beuno: I have a memory (trying to find the mail) that there was pretty broad consensus on that on-list. [13:30] kfogel, on dropping it or on dragging it along? [13:31] beuno: dropping [13:31] coolio [13:31] thanks kfogel [13:31] beuno: wait, let me see if I can find [13:33] beuno: so far: https://lists.launchpad.net/launchpad-dev/msg00148.html, and then Martin Pool's response. But I think we had a more definite thread, let me see. [13:34] kfogel, don't worry [13:34] your answer is good enough [13:34] thank you [13:35] beuno: ok. Well, sorry -- I do recall an even more definite statement coming later, from multiple people like flacoste, etc. But my memory is like a... what's that thing called?... I can never remember... [13:36] heh [13:37] I feel like eating cheese now [13:37] kfogel, while you're here [13:37] any thoughts on the LP homepage? [13:37] I'm working on a mockup [13:37] beuno: hmmmm. let me take a look [13:38] and then I'm back to my vacations [13:38] so if there's anything you'd like to see in a mockup, now is the time [13:39] heh [13:39] * kfogel pondering [13:42] beuno: only thought I have is that it might be nice if the "What's New?" section were an RSS headline feed straight from the blog. But I'm sure that's been debated/considered before. [13:42] beuno: let me look at page stats for a sec [13:43] kfogel, having a feed from the blog would be orgasmic [13:43] beuno: I'm looking at https://help.launchpad.net/PageHits to see if it offers any clue as to what should be directly on the LP front page. [13:43] aha [13:43] I did not know about that page [13:44] WOW a *lot* of people visit /Legal [13:45] kfogel, I guess this page is a good one to extract a few things: https://help.launchpad.net/NewToLaunchpad [13:45] beuno: to me the big surprise is PPA [13:45] by far the biggest help destination [13:45] cprov: I noticed a problem this afternoon with primary ddeb uploads. Because they skip NEW, they all land in universe. [13:46] beuno: one complaint about https://lpstats.canonical.com/graphs/ is that it doesn't offer any comparisons between categories. A graph is about bugs, or it's about branches, say, but there aren't really any easy ways to get comparisons between bugs and branches. [13:46] wgrant: shh, they don't have overrides at all, do they ? [13:47] cprov: They don't, but it seems wrong that they all land in universe. [13:47] kfogel, agreed. You just need to put together some SQL for the LOSAs though, and they will get a graph [13:47] wgrant: yes, we have to fix it. [13:48] beuno: do you think a "report bug in Ubuntu" would be a good idea? Interestingly, Launchpad front page has no indication at all that the site is about building Ubuntu. Of course, the degree to which it is or should be Ubuntu-centric is a big question. [13:48] beuno: yup (re SQL) [13:48] beuno: you know there's a branch to propose merges to these days, right? [13:48] wgrant: they could inherit the source component (maybe) [13:48] cprov: I thought that. [13:48] can they be linked back to the regular deb in any way? [13:48] beuno: But maybe we just strip -dbgsym off, or get pkg-create-dbgsym to stick the original binary name in a field somewhere. [13:48] Er. [13:48] cprov: ^^ [13:48] mthaddon, I do not! how cool is that! [13:49] kfogel, I've brought that up in the past [13:49] kfogel, I think I'd like it, but it will also make Launchpad look even more Ubuntu-centric [13:49] beuno: there was an email to the list about it - https://code.edge.launchpad.net/tuolumne-lp-configs [13:49] bigjools: The names should all follow the pattern of $ORIGINALBINARY-dbgsym. I suppose it's reasonable to reject if something doesn't match up. [13:49] mthaddon, that is super great. Thanks for the tip. [13:50] wgrant: if dbg_name.replace('-dbgsym', '') == name, yes, that would be an option [13:51] cprov: btw, pitti ack'd the pkg-create-dbgsym changes and suggests that they be SRU'd. [13:51]