=== Ursinha is now known as Ursula === Ursula is now known as Ursinha === Ursinha is now known as Ursula === Ursula is now known as Ursinha === Ursinha is now known as Ursula [13:15] bac, benji, objectives today if possible (especially since I'm going on vacation) [13:16] also, is it safe to run lp2kanban? Should I just run it with [13:16] bin/py src/lp2kanban/bugs2cards.py -e gary.poster@gmail.com -p [13:16] or do I need to specify the board? [13:17] It looks like the ones I've done as test cases have not yet been synced [13:17] gary_poster: I did mine yesterday... hopefully correctly. [13:17] cool benji I'll look. thanks [13:18] benji, bac, oddly, I get [13:18] Traceback (most recent call last): [13:18] File "src/lp2kanban/bugs2cards.py", line 5, in [13:18] from argparse import ArgumentParser [13:18] ImportError: No module named argparse [13:19] gary_poster: did you run the buildout? that's a new dependency [13:19] I did, but maybe before I updated. lemme retry [13:20] if that doesn't work, perhaps it wasn't added to the install requires entry [13:20] benji, bin/buildout did nothing, and retrying failed again. We can look into it later [13:24] benji, updating setup and buildout made it work. I'l commit in a sec. Looks like it is updating ythe test board? [13:24] gary_poster: no idea, I haven't touched it lately [13:24] looks like -b yellow should work... [13:26] hi gary_poster [13:26] hey bac [13:26] got it to work [13:26] gary_poster: yes, yellowtest is the default. safer that way [13:26] agreed [13:26] gary_poster: and cards must be marked to get synced [13:26] but no cards are reported as updated (I mean, that's what the script reports). I'm seeing what the cnstraint is... [13:27] ah ok [13:27] title.startswith("sync:") or a tag in the description [13:27] coo; [13:27] cool, I see that [13:28] trying [13:28] bac benji danilos, call in 2 [13:28] ok [13:28] ack [13:29] gary_poster: for a second there I thought Julia had taken over your IRC client [13:30] bac, benji, ta-dar! :-) I synced my two bugs [13:30] yay! [13:30] ga ga goo goo [13:30] fwiw, does anyone find that leankitkanban is much less responsive lately [13:30] gary_poster: ooo, try again [13:30] maybe that was while the script was running :)) [13:30] heh [13:30] * benji wonders if it's onomatopoeia like a pirate day. [13:30] bac, running [13:31] I'll start the call when it is finished [13:31] note you have to reload the board [13:31] nice [13:31] it's done [13:31] no [13:31] it reload automatically [13:31] oh I see [13:31] I just wasn't patient enough the first time [13:31] sweet! [13:31] yeah, I got a notice how the board is updated [13:32] yeah, in my experience it behaves the same as when people update the board [13:32] gary_poster: how did you resolve the argparser issue? [13:32] bac, just updated setup.py and buildout.cfg. will commit [13:33] why did it work for me? is that required for things that are in the standard distribution? === Ursula is now known as Ursinha [13:39] bac: my psychic powers tell me that you're running Python 2.7 which has argparse and Gary is running 2.6 which doesn't [13:40] benji: hmm, i thought i'd used 2.6 [13:41] bac: that or you're using a python with argparse already installed [13:41] benji: i guess. i don't recall installing it. /me looks [13:42] gary_poster, lp:~danilo/launchpad/bug-772754-other-subscribers-remove-cruft [13:42] benji: i guess i'm wrong! it is in /usr/lib/pymodules so i guess that is something i installed? [13:43] I can't tell from here. ;) [13:43] * benji supresses a system-python-is-evil rant. [13:44] yep, i'd installed python-argparse at some point [13:46] hey benji i was thinking it might be a good idea to save off the initial state of the kanban board in case we mess it up in some way. i assume there is nothing in their API that would allow recovery. [13:47] yeah, I don't know of a way to set the entire board state (although, as you know, they let you retrieve it) [13:48] benji: i forgot to tell you that you and i swapped #launchpad help contact days this week! i covered for you on monday. you mind picking it up today? [13:49] bac: sure [13:50] I swear! Now the Gnome shell is crashing on me. Maybe it wan't Unity's fault. [13:56] gary_poster: after looking at bug 697735 for a little while, I've realized that either I don't think it should be considered a bug or I don't understand what it should do instead -- if you hand craft odd URLs you'll get wierd errors, what's wrong with that? [13:56] <_mup_> Bug #697735: ValueError raised submitting bogus query string < https://launchpad.net/bugs/697735 > [14:01] sorry, just saw that; looking [14:02] benji, from a high-level perspective, the problem is that OOPS reports are supposed to be actionable [14:02] benji, it's fine for the user to get an error [14:03] benji, but as you say, there's nothing for us to fix--so we should not get an OOPS [14:03] That's my understanding [14:03] makes sense, I was missing that angle [14:03] Of course, fixing it may be a little exciting :-/ [14:04] yeah, I'm starting to wonder if I made a bad pick. I'll timebox at least the first stab at fixing it to see where it goes. [15:02] anyone, what date did you choose for your allhands objectives' target dates? I have responded to all of them so I can't see them anymore :-P [15:03] Oh Apr 12 [15:03] nm [15:05] gary_poster, something random for me, mostly in March/April next year [15:05] :-) [15:05] cool [15:05] dragnob said in her email today april 12 [15:05] I dunno if that is real or not [15:06] but whatever, it's in the ballpark :-) [15:06] thanks [15:07] gary_poster, fwiw, all the MPs are up and linked to the bug, in order of how they are in my pipeline (and in order of their dependencies; -lp-names is the only one which is relatively independent) [15:08] gary_poster, -sections, -subscribers, -activity, and -loading should be most interesting [15:08] danilos, awesome. I got your last branch and I think it is built now; I'll look at the UX, and the diffs, soon. [15:09] gary_poster, fwiw, anything after -loading (inclusive) but before the last one (exclusive) will have two subscribers lists :) [15:09] danilos, heh, ok, good to know :-) [15:09] gary_poster, I wouldn't mind you reviewing a branch or two while I add tests for -actions :)) [15:10] danilos, ok cool. I should be able to start in about 10 min. [15:12] gary_poster, yeah, no worries, you don't have to review them (especially not all of them, that's why I split them up :) [15:12] gary_poster, I am also asking on #launchpad-dev, so if are starting a review, please claim it [15:12] ack [15:16] gary_poster: what is an "informational" oops? [15:16] an oops that's not an oops? :P [15:17] benji, yeah [15:17] we track them separately [15:18] we don't need to drive them to zero [15:18] gary_poster: so would inappropriate use of :int be an informational oops? [15:18] but it is convenient to use the OOPS machinery to collect the data [15:18] eh [15:18] no [15:18] not as far as I know [15:18] but we don't really want to collect anything [15:19] because we don't want to track itright [15:22] benji, do you know how to generate api docs from a lazr.restful app, like api.launchpad.net/+apidoc ? This is a question from ISD [15:23] gary_poster: not off the top of my head, but I'm sure I could figure it out quickly enough [15:24] benji, if that's ok, please do; my stack is getting too deep for last day before vacation :-) [15:24] heh, ok [15:37] gary_poster: there's no super-straightforward way, but it's not too bad. The best thing would be for them to look at utilities/create-lp-wadl-and-apidoc.py in LP. [15:38] benji, ok thanks [15:42] benji, I told Ricardo Kirkner, "pindonga," that he could ping you if he had questions. I suggest that the first time he asks you for help, you give it to him as soon as convenient; subsequent times give him focus no more than once or twice a day if possible, unless, in your judgement, you think you ought to give him more. Good relations are good, and they are doing important things. [15:43] k [16:02] gary_poster, fwiw, I found one of the problems "generic" webkit JS engine (eg. epiphany) borks on: "class" as a variable name, or more frequently, a function parameter :/ [16:04] danilos, interesting! [16:05] good to file away [16:05] and unfortunate :-/ [16:05] gary_poster, do you happen to know if that's a problem for Safari as well? I find it interesting because webkit JS should be identical in both, minus the version differences [16:06] I don't know off-hand danilos. I mosty use Chrom these days, which has a different JS engine AFAIK [16:06] gary_poster, yeah, it does [16:12] gary_poster, looking at the code, it seems your unlanded branch is the only one which tries to use it, though :) http://paste.ubuntu.com/623491/ if you want to apply it to your branch directly [16:13] danilos, cool, thanks, will do [16:14] gary_poster, ftr, it seems to be a problem with safari as well: http://blog.solution10.com/2010/02/use-of-class-in-javascript [16:16] cool [16:16] thanks [16:30] * danilos -> off, pushing branches up took longer than I expected, so still tests missing for the final branch [16:31] * danilos needs to add task cards to the kanban board as well... :/ [16:59] * gary_poster lunches [17:01] bac, I want to reorder the sections in danilos' final branch. I don't know if it affects yours. I think that "Maybe notified" should be last, as it is in the mockup (and I wonder if we should have a help icon with the explanation for what "Maybe notified" means). [17:01] I suspect that is in a later branch [17:02] but wanted to share now, in case it is in that one [17:02] I'll dig in more after lunch [18:24] ok [18:31] I'll just put the comment in the last MP and not worry about it, I think. [18:41] gary_poster: I have a pretty good solution to the bug we discussed earlier (bug 697735) but I wanted to get a consult to be sure it doesn't go too high on the hack-o-meter: https://pastebin.canonical.com/48394/ [18:41] <_mup_> Bug #697735: ValueError raised submitting bogus query string < https://launchpad.net/bugs/697735 > [18:41] k, looking [18:41] it scores a 3/10 on my unit here, but mine is known to be unreliable [18:41] :-) [18:46] benji, mm, BrowserRequest.__processItem is so tasty. [18:47] heh, how so? [18:48] sarcasm, just that it is so big and dense that copying and modifying is really unappealing, and then combine that with the __* munging and you have a really great reason to do the monkeypatch, which is a shame [18:48] benji, did you consider having a marker interface for the request instead? [18:49] we do that kind of thing a lot [18:49] oh, indeed [18:49] re. marker; well, it'd be a marker on the exception, right? That'd be fine with me. [18:50] benji, wouldn't it have to be on the request? [18:50] the reason I did it this way is because it mirrors the way we did something similar for the web service [18:50] since _isIgnoredException takes strtype [18:50] not the actual instance [18:50] it also (after I changed it) takes the exception instance [18:51] oh, duh, yeah [18:51] so yeah, on the exception sounds good [18:51] that way you could do it on the fly or as a general statement in code [18:52] ("this kind of exception is not oops worthy") [18:52] sounds good [18:52] cool [18:52] benji, the only thing that really irks me is the monkeypatch, and I see why you did it. Overriding __processItem does not seem like a nice option [18:53] well... [18:53] right, it seemed like the lesser of the two evils; the other option would be to modify zope.publisher to make this easier [18:57] I looked for the word "monkeypatch" in the code, and I only saw mailman do it, which is not regarded as one of our high marks :-) [18:59] benji, I'd get bac's opinion, which will be easy since he'll probably be your reviewer. I think I'd trust his opinion more than mine. The only other thought I had was that you could consider moving the monkeypatch to lib/lp_sitecustomize.py. It is where we do other similar sorts of things, and is where people will expect to find stuff like this AFAIK [19:00] benji, other than that, I think the interface-on-the-exception method of squelching OOPSes will be quite welcome, and should be announced. [19:00] k; I'll look in lib/lp_sitecustomize.py, and add a comment there that hopedully would have helped me find it when I was looking for a home for this thing [19:00] cool [19:00] k [19:01] * benji orders bright white, embossed announcement cards. [19:01] :-) [20:05] * benji is reminded by his phone to do CHR tasks. [20:12] hmm, we seem to be behind in project/license review (or I'm doing it wrong) [20:28] gary_poster: your lost soul is still lost: https://support.one.ubuntu.com/Ticket/Display.html?id=2433 What should I do for him? Ask a LOSA to change his account's email address? [20:31] benji, I was bad (and/or swamped in other things) today and didn't do CHR, and we are missing gmb, so you are getting a bunch. Sorry. :-( bac, is there any chance you could give some thoughts on help we could give to that guy--the oops that he mentions? [20:31] I *think* we just need to set a preferred email, but I'm not really sure, and going int the raw SQL feels a bit like diving into the deep end which might have some hungry sharks hanging around. [20:31] heh [20:32] So I'm hoping you have some insight on "tried and true" things to do to help that guy, bac [20:32] (from your registry days) [20:32] k, I'll set his email address; how do we know that we're not being soccially engineered into giving control of someone's account to someone else? === Ursula is now known as Ursinha-lunch [20:36] gary_poster: i'll look [20:37] benji, we contacted him at this email address [20:37] thanks bac [20:37] getting kids from school, biab [20:38] Hmm, I don't understand that. He said he doesn't have access to the account's current email address any more. [20:41] benji: i'm confused too [20:57] benji, bac, it may have been because it was compromised? The situation was that his account was sending spam [20:57] benji: so here is my take. the dude was suspended for spamming. it is unclear whether he confirmed that he resolved his spam problem, i.e. changed password, got rid of browser issue [20:58] i think when we reactivate his account he has to go to the 'forgot my password' dance to get reinstated [20:58] is that right, gary_poster? [20:58] (he said there was no browser issue) [20:58] if so, he's currently stuck as he won't be able to retrieve the link [20:58] forgot my password: well...we suspended his LP account [20:59] not his SSO acount AFAIK [20:59] gary_poster: then why can't he log in now? why is his account still showing as suspended (grey avatar)? [20:59] I don't know. I think it is because of an LP bug [20:59] if you look at the state in the administartion page [21:00] then I think you'll see that he is actuve [21:00] actve [21:00] active [21:00] (he said, typing slowly) [21:00] what I thought we needed [21:00] was to set a preferred email [21:01] because that's what the OOPS points to [21:01] but maybe logging in via SSO would work? [21:01] Should we tell him to make a new account, and then merge the new into the old? [21:02] if there's no accessible email associated with the old account, that won't work either [21:02] oh, i missed that there was an OOPS [21:02] no, a merge will not work [21:03] ok [21:03] gary_poster: what is the OOPS? [21:03] i don't see it [21:03] i don't think lifeless understands my silly attempts at humor. [21:04] bac, argh, sorry. This is not the most pitiful of souls, nor the one I was thinking about or talking about. I meant https://support.one.ubuntu.com/Ticket/Display.html?id=2120 [21:04] who has been unable to log in for two weeks now [21:04] aha! [21:04] bac, OOPS-1973H90 [21:05] bug 793670 [21:05] <_mup_> Bug #793670: User account missing preferred email after suspension/reactivation < https://launchpad.net/bugs/793670 > [21:05] gary_poster: what's the bug number for the redoing of the bug subscription overlay you were working on? [21:05] (I can't find it on the board.) [21:06] benji bug 772754 (Feature Work: (BIG BUGS)) [21:06] <_mup_> Bug #772754: After better-bug-notification changes, list of bug subscribers is confusing

< https://launchpad.net/bugs/772754 > [21:06] doh! I've gotten so used to not looking in there that I totally missed it. [21:07] :-) [21:08] I'm tempted to be "conversational" when chaning bug statuses or marking them duplicates, i.e., adding a comment about what I'm doing and why, but that doesn't seem to be the culture so I'm abstaining [21:09] benji, dunno. I do it when I think it adds value [21:10] the only "value" I was feeling was along the lines of softening the "they just marked my bug as Low, they must hate me" reaction [21:10] I have about one hour before vacation...I could help with CHR or read about CoffeeScript. Which one sounds like more fun? Which one should I actually do? [21:10] Yeah [21:10] I've written those [21:11] I've become more...blunt lately with the new standards [21:11] So I have not [21:11] but do what feels right [21:13] gary_poster, benji: i have not idea what to do for amichai2 [21:13] gary_poster, benji: short of asking a LOSA to do a SQL mod to give him a preferred email address [21:13] gary_poster: I just finished CHR (whew, it took more than an hour), so I doubt there's anything to do [21:14] benji, I was pondering https://support.one.ubuntu.com/Ticket/Display.html?id=2514 [21:14] oh, feel free handle that one ;P [21:14] bac, that's what I was thinking. Would you feel comfortable assembling such SQL? [21:15] gary_poster: will it fulfill my SQL goal? :) [21:15] bac, lol [21:15] bac, it would help [21:16] sure, i'll see what i can do [21:16] cool [21:16] am i going to need a TL or higher to sign off on my losa request? [21:17] bac, yeah, I think TL can do it [21:19] benji, FWIW, I assigned it to Matthew Revell and sent a message to the requester and Matthew to that effect [21:20] cool [21:25] benji, fun: today Danilo alerted me that webkit falls over if you use "class" as an argument because it is a non-functioning reserved keyword in JS. All other browsers are fine with it. coffeescript compilation (and repl) clearly disallows. another win. [21:25] heh, cool [21:27] gary_poster: I assume that goes for all the other reserved, but unused words. There are lots of them. [21:27] benji, I'd hope so. dunno [21:27] I'd expect so tbh. this obviously has been around the block [21:32] gary_poster: it looks like there is enough info at https://wiki.canonical.com/InformationInfrastructure/ISD/Docs/SSO/ErrorSettingPreferredEmail for a losa to work with [21:33] bac, oh interesting [21:33] bac, +1 I guess...assuming he has an email at all, which I hope he does [21:33] I mean, registered in LP [21:37] benji: did you see statik talking about playing with coffeescript? [21:37] nope, I'll look in my logs [21:45] benji: it was only 15 minutes ago or so [21:45] gary_poster: is there any chance we messed up with that user? should he have gone from suspended to unactivated? [21:45] which would require following an activation link? [21:45] i can find no docs [21:46] bac, I went from suspended to activated, and that failed [21:46] then [21:46] sinzui said we'd had truble with that in the past [21:46] and suggested I go from deactivated to activated [21:46] so I did that [21:46] and that failed too [21:47] did you try unactivated? [22:03] gary_poster: the guy's email address was marked 2-validated. we forced it to preferred [22:03] https://launchpad.net/~amichai2 [22:03] looks much better [22:03] Go Banana Slugs! [22:12] yay bac! [22:12] thanks [22:13] gary_poster: the other gent was in the same boat. i think we've got a suspend/reanimate issue here [22:13] bac, huh [22:13] maybe make a comment on the related bug? [22:14] this guy: https://support.one.ubuntu.com/Ticket/Display.html?id=2433 [22:14] ok [22:14] yeah, I figured that was who you meant [22:14] if the bug is not critical it should be [22:14] I think I marked it critical [22:14] bug 793670 [22:14] <_mup_> Bug #793670: User account missing preferred email after suspension/reactivation < https://launchpad.net/bugs/793670 > [22:15] yeah [22:22] bye [22:22] tty; [22:23] l [22:23] need to get better at this typing while walking thing :-) === Ursinha-lunch is now known as Ursinha