[07:04] bac, btw, I see mention of outerHTML: I found out that it's not supported by firefox 3.* at least (just fyi) [10:40] gmb, hi, do you think you could take a look at https://code.launchpad.net/~danilo/launchpad/bug-770241/+merge/59338 for me? it's mostly boiler-plate for tests/new module or actual tests :) [10:41] (and there's no OCR yet) [11:08] danilos: Sure, I'll look now. [11:09] gmb, thanks [11:20] danilos: r=me. [11:37] danilos i removed the outerhtml earlier [11:38] gmb, thanks [11:38] bac, yeah, I know, just mentioning another reason why it's bad (I wanted to use it as well, it's so natural sometimes :) [11:38] gmb hear anything from deryck yet today? hope he is ok [11:40] bac: No, I've heard nothing since he said they were basically battening down the hatches. [11:41] Fingers crossed. [11:41] they are reporting up to 65 dead in alabama. the storms just went through here [11:42] bac: I know, I saw. I can't find any news on Deryck's town (Dadeville) and besides he's quite a way out from the town itself, so it's all pretty sketchy. [11:43] But I doubt he's got power atm, one way or t'other. [11:54] bac: Apparently deryck has said on Facebook that there's no devastation near him. I guess he's just w/out power right now. [12:13] thanks gmb [12:21] gmb, I wonder how can one get more than one user link on the subscribers portlet? the code seems to assume it can happen with dupe subscriptions, but I am unable to reproduce it [12:21] (more than one user link for the same user) [12:22] ah, it can happen with teams, see it now :) [12:22] * gmb deletes the sentence he was about to post :) [12:28] gmb: just texted deryck. says they are fine and power is back on but no internet. [12:28] bac: Cool. Thanks for the update. [13:10] hi folks. Please make sure you do your qa asap; our qa lanes on the kanban board are almost full. [13:10] (my remaining one still is not on staging. :-/ ) [13:26] bac benji danilos gmb, kanban now & call in 4 [13:28] danilos, I'm going to ask you to identify, either via a white list or a black list, bugs in your claimed area that you think others can tackle without interfering/conflicting with you. We're getting near the end of the list of the other bugs, I think [13:29] FWIW these are "medium" bugs that I think are worthing of consideration (these are on my own list to consider) [13:29] https://bugs.launchpad.net/launchpad/+bug/771284 [13:29] gary_poster, ok, I'm loading the list now [13:29] <_mup_> Bug #771284: The empty +subscriptions page is a dead end

< https://launchpad.net/bugs/771284 > [13:29] https://bugs.launchpad.net/launchpad/+bug/771247 [13:29] <_mup_> Bug #771247: "added or changed in any way" should include when comments are added

< https://launchpad.net/bugs/771247 > [13:29] https://bugs.launchpad.net/launchpad/+bug/768997 [13:29] <_mup_> Bug #768997: Batch multiple comments from the same person/same bug < https://launchpad.net/bugs/768997 > [13:29] https://bugs.launchpad.net/launchpad/+bug/771282 [13:29] <_mup_> Bug #771282: Team subscription description addresses "you" when describing the nature of the subscription

< https://launchpad.net/bugs/771282 > [13:29] Some of them are easy, some of them I think are important-ish, some of them are both [13:29] thanks danilos [13:31] bac, no Skype-y [13:31] gary_poster: try again? [13:47] bac: https://pastebin.canonical.com/46847/ === Ursinha-afk is now known as Ursinha [14:03] IT WORKS. [14:03] That is all. [14:07] gmb, you fixed it? [14:07] I forgot to mention that I have a bit of a crazy schedule today, and somewhat tomorrow [14:08] I will be out for another "how is the baby doing" starting today at 10:30 (about 1:20 from now) and then again tomorrow in the morning. I also have some stuff at the boys' school this afternoon. [14:11] gary_poster: I think so. Writing tests to confirm it. [14:12] yay! [14:22] benji, I was looking at bugs for QA to see if it indicated a problem in our machinery somewhere. https://bugs.launchpad.net/launchpad/+bug/771269 actually has not landed, from a review two days ago. Looking at other ones... [14:22] <_mup_> Bug #771269: Structural subscription description is hard to read

< https://launchpad.net/bugs/771269 > [14:23] same for https://bugs.launchpad.net/launchpad/+bug/771260 [14:23] <_mup_> Bug #771260: In the "bugs must match this filter" accordion we should state that tags are space separated, without having to click a help pop-up

< https://launchpad.net/bugs/771260 > [14:23] oh, heh, same branch [14:23] gary_poster: yep, none of mind in "landing" have landed [14:24] benji, do you know why? Anything I can help with? [14:24] all seem to have had odd test failures in PQM, I'm trying to replicate and/or reland name [14:24] ok cool [14:24] s/name/now/ [14:28] danilos, bac, benji, I apologize, but my crazy schedule today may make it difficult to have calls at our regularly scheduled times. danilos, would you like to reschedule for 30 minutes earlier (in about 30 minutes), or for tomorrow? I bet you'd prefer tomorrow. :-) [14:28] bac and benji...actually, if we keep the calls short, and maybe start a hair early, I actually think we can keep to our normal times. Does that sound ok? [14:28] +1 [14:28] gary_poster, good guess, yeah tomorrow'd work much better for me :) [14:28] gary_poster: yep [14:29] benji, bac, cool thanks. danilos, heh, ok, I'll change it on the calendar. :-) [14:29] gary_poster: Ah. Ahahaha. I've just had a thought about my "fix". It could actually introduce problems, in that stuff that's only available to Bug Supervisors (and protected by transitionTo{Status, Importance, Assignee, etc.}()) will become available - and unprotected - at bug creation time. [14:30] We can fix that by adding the security code in createBug(), maybe, but I'm not sure whether that won't just cause us other problems (where "other problems" is a bit of a hand wave at this point). [14:31] gmb, I think I see. Adding the security code doesn't sound too bad on the face of it. [14:31] Yeah. [14:31] Particularly if it is small [14:31] But maybe rather than just passing the stuff to createTask() we should create the task and then call transitionTo*() appropriately. [14:31] actually, thinking about it, why is that stuff done in the view in the first place, I wonder? [14:32] ... ah, well, because we didn't have an API when that was written, that's why. [14:32] maybe so [14:32] well, you would probably know :-) [14:33] gary_poster: That view has been around since year 0 and does some weird stuff. I suspect that the security code living in it is a throwback to "thou shalt not do security checks in the model" which was our pre-API rule, IIRC. [14:34] yeah, that makes sense. I like the rule in the abstract. I might argue that it is an indication of a wart in the API stuff. But anyway... :-) [14:39] Yeah. Let's fix the small things ;) [14:47] benji, what's the story for https://bugs.launchpad.net/launchpad/+bug/769151 ? It is marked as in-progress by you, but is not on the kanban board AFAICT, and you are working on other things. It also looks vaguely dupe-y--like maybe it is something danilos is working on? [14:47] <_mup_> Bug #769151: Creating a subscription when unmuting a bug doesn't. < https://launchpad.net/bugs/769151 > [14:47] looking [14:48] thanks [14:50] gary_poster, benji: I was just about to ask about that bug myself :) [14:50] :-) [14:50] gary_poster, benji: I'd say I've fixed that already, but I'd be happy to take it on my plate to double check [14:51] +1 to either closing it outright if it is a dupe, or doublechecking and closing it. [14:51] I'm pretty sure it is invalid. I think it's only a bug on one of my dev branches. [14:51] ok cool. um...benji, you feel comfortable closing it outright, or should someone doublecheck? [14:52] I've marked it invalid [14:52] awesome. thanks benji and danilos [14:52] I'm relatively certain it was bogus [14:54] benji, on the contrary, I am pretty certain there was a bug like that until recently, because the underlying code was not very good at figuring out what do you really want when you select a radio button :) [14:54] feel free to reopen and reassign it if you think it's valid [14:56] well, one person thinks it is invalid, the person thinks it was valid but is now fixed...I'm happy to say "Invalid now" and move on, myself :-) [14:56] and in other news, there are no high bugs for me to choose from, which is kinda cool. :-) [14:56] particularly if Huw gives us fixes/ideas for the two that he's blocking on [14:57] what's up with allenap and team-yellow? was he a mole trying to infiltrate us? [14:57] gary_poster, benji: yeah, it's "invalid", just trying to point out that benji has not had lost his mind when he filed it :) [14:58] danilos, heh cool :-) [14:59] bac, :-) I added him to ~yellow because he reported a potential problem in a bug report but I wanted to see if he just needed the feature flag. He verified that once he was in ~yellow the problem was not there, so then I removed him again. :-) [14:59] mm, two smileys in one line. That's a bit much even for me. [15:00] heh, er, ha [15:00] lol [15:07] gary_poster: i just noticed bug 745660 was forced from committed back to in progress by lifeless but i don't know why [15:07] <_mup_> Bug #745660: Attempt to subscribe team that is not associated to distribution fails < https://launchpad.net/bugs/745660 > [15:07] bac, huh, looking [15:07] gary_poster, re bug 770345, the brokenness mrevell mentions in there seems to be the fact that his previous subscription doesn't get restored, which I know you are unhappy with, but I ain't exactly sure what can we do without it without major changes [15:08] <_mup_> Bug #770345: When muting a bug your name disappears from the subscribers list but doesn't re-appear when unmuting

< https://launchpad.net/bugs/770345 > [15:08] gary_poster, fwiw, the name does re-appear if one chooses "unmute and subscribe me" [15:08] it should be fix release [15:09] gary_poster: mrevell raises a similar issue in bug 770342 [15:09] <_mup_> Bug #770342: "Unmute" dialogue is headed "Subscribe to bug"

< https://launchpad.net/bugs/770342 > [15:09] bac, my guess is that he marked it in progress because of what wgrant said. I'll stare at it more in a sec; lemme read what danilos wrote... [15:12] danilos, gotcha, yeah. You are relatively sure that "unmute and subscribe me" already works (and was working when mrevell was looking?) [15:12] gary_poster, not sure if it was working when he was looking, I'd have to check with older devel for that [15:13] danilos, or production [15:13] lemme try there [15:14] gary_poster, doesn't seem to work on production [15:15] danilos, wfm on prod. at least on the path I tried. [15:15] gary_poster, ha, interesting :) ah, right, I've noticed troubles with epiphany, let me try another browser [15:15] gary_poster: So, looking good with my fix for bug 720147. I need to step out for a while now but I'm running it through EC2 to see if anything breaks and will come back later to make sure it gets merged (assuming no breakages). [15:16] gmb, fantastic :-) [15:16] gary_poster, yeah, wfm as well [15:18] danilos, ok. I'll mark this invalid and ramble on a bit in a comment about the situation [15:18] gary_poster, cool, thanks, I've just reassigned it to yellow [15:18] thanks danilos [15:21] gary_poster: regarding bug 770342 i can make some of the text changes mrevell requests, and they are improvements, but i think he's mainly annoyed at losing his subscription when muted. shall we just note and WONTFIX that aspect? [15:21] <_mup_> Bug #770342: "Unmute" dialogue is headed "Subscribe to bug"

< https://launchpad.net/bugs/770342 > [15:23] bac, agreed about your evaluation, and with your suggested resolution. [15:24] gary_poster: ok, updating the bug [15:24] * gmb -> afk for a while [15:25] thanks bac. Please also either mention that we can make the internal change, but that it will require significant restructuring that will cause at least a couple of weeks of extra time; or ping me when you have made changes to the bug, and I'll add that comment. [15:31] dr appt. biab [15:50] i'll be gone for a bit too [17:13] * danilos -> out, tty tomorrow [17:14] * gary_poster back [19:02] benji, I'm writing a bug report and am trying to remember a phrase that you like. It's something like "a destructive convenience" but that's not it...the idea is that something looks nice to use, but when you actually use it, it makes the situation worse, not better. The convenience leads you down a bad path. Does this ring any bells enough for you to remind me of the phrase? [19:02] gary_poster: attractive nuisance [19:02] benji, that's it! thanks. :-) [19:02] np [19:25] gary_poster: are we chatting nowish? [19:25] yes, bac. you available now? [19:25] y [19:27] k one more sec [19:27] no rush [19:28] bac, could you read http://pastebin.ubuntu.com/600387/ and then Skype me? [19:28] that's a draft of a bug report [19:28] Easier to read: http://pastebin.ubuntu.com/600387/plain/ [20:58] benji: could you review this wee branch? [20:58] https://code.launchpad.net/~bac/launchpad/bug-770342/+merge/59415 [20:58] sure [21:00] bac: it looks fine. I would have expected to see some changes to the call site(s) as well. [21:00] benji: yeah, me too. [21:01] heh [21:01] * bac looks again [21:02] benji: i think my call site was eaten by my recent merge from devel [21:03] hmm [21:03] * benji suspects gary is about to call him. [21:04] * benji tries to remember where he put his headphones. [21:04] * gary_poster realizes that it is later than he thought [21:04] benji ping me when you have headphones [21:04] or just call me [21:14] http://pastebin.ubuntu.com/600387/plain/ [21:15] gary_poster: When you've got a second I've got a slightly knotty problem that needs another pair of eyes. [21:15] benji: diff updated at https://code.launchpad.net/~bac/launchpad/bug-770342/+merge/59415 [21:17] gmb, ack. Is it re the comment you made recently to allenap? I will be gone in 4 minutes, so sending me an email might be best. I can look at it probably starting in 45 minutes [21:18] gary_poster: Okay, that works. I'll email you. [21:18] Thanks [21:18] cool, np [21:21] * benji looks at https://code.launchpad.net/~bac/launchpad/bug-770342/+merge/59415 [21:29] bac: sorry if I'm wrong, but I would have expected there to be more than one calls of create_new_subscription_overlay; there are two on devel at the moment [21:30] benji: yes, but the other isn't changed as it uses the default value [21:30] or do you find that icky(TM)? [21:30] ah, gotcha [21:31] given that there are only two calls, I personally would have done it without the default, but I'm not bent out of shape about it [21:31] originally i thought there were more call sites so i went the default value route to minimize disruption. now that there are only two it is a coin toss [21:31] benji: cool, i'll make the change [21:32] bac: you may also want to run lint, I think you have a long line [21:32] good catch [22:04] ok benji, updated [22:05] looking [22:06] bac: done === Ursinha is now known as Ursinha-afk [22:26] gmb, if you are still around, I don't understand why transactions don't save us in the scenario you mentioned. Isn't everything in a single transaction? [22:26] and also if you are still around, why are you still around? :-) [22:26] gary_poster: Hmm. Yes, it is... so maybe it's to do with the way I'm testing things. However... [22:27] (And I'll address your other point in a minute) [22:27] :-) [22:27] ... I've gone back to thinking that maybe we *shouldn't* be doing security stuff in createBug(). [22:27] We don't, AFAIK, let people call createBug() directly from the API (which would be our biggest security hole) [22:28] And we have checks and balances in place in the UI to prevent people who aren't allowed to set milestone, etc. from doing it (save the edge case I came up with). [22:29] So maybe what we should do here is JFDI, pass the parameters to IBugTaskSet.createTask() (without going through transitionTo*()) and fix anything *if* it breaks. [22:29] well, the edge case can't happen within transactions, if that helps anything [22:29] But maybe - coming back to your other point - I'm not able to see the wood for the trees right now. [22:29] I was trying to get this done before tomorrow, since I've got a four-day weekend, but maybe that's just not feasible. [22:29] within a transaction, Alice either will or will not have the permission [22:30] So you are out tomorrow and Monday, right? [22:30] Yes. [22:30] (and Danilo is out Monday and Tuesday) [22:30] ok [22:30] gary_poster: Actually, you're right. Transactions should save us the worry. [22:30] ... but in tests, we don't use them unelss we do it explicitly. [22:30] That explains my problem. [22:31] gmb, it sounds like this is at the "it's fixed except for the fact I'm really tired" :-) [22:31] stage [22:31] Fair point. [22:32] I'm happy to try and run with your branch tomorrow, if you think that would work [22:32] that's exactly what I was hoping you'd say. [22:32] Is there anything I need to know more in order to do that, do you think [22:32] heh, ok :-) [22:32] I think it needs one more test, but nothing big. I'll email you with details when I've committed and pushed what I've got. [22:33] awesome, gmb. Thanks. Finish up and quit soon. Thank you very much for working on this [22:33] np [22:50] gary_poster: Branch pushed (lp:~yellow/launchpad/filter-on-bug-creation-bug-720147), details emailed to you. [22:50] * gmb -> Very, very AFK. [22:51] awesome, thanks gmb [22:51] Have a good weekend everyone. [22:51] :-) bye, have a great weekend!