=== micahg_ is now known as micahg [04:14] Oh, bah. Firefox, why are you hiding the URI schema for http:// URIs? :-( [04:35] because chrome did it and so its obviously cool. [04:35] nevermind that it completely messes things up [04:56] If Firefox is just going to copy Chrome, that's disappointing. [05:41] oh, I agree. [05:58] lifeless: <3 [05:58] thanks! [05:58] StevenK: Speaking of browsers, this might be entertaining - http://lh4.googleusercontent.com/-lwEG-pIhR3Q/TodEKkigwVI/AAAAAAAACgs/ARfogL6LqXM/s512/msft.jpg [06:02] nigelb: I saw that [06:05] :) === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [20:59] under one page of mediums left [21:38] lifeless: and hey, i just closed two of the bugs you reclassified [21:52] mwhudson: cool [21:53] three! [21:53] do I hear a four? ;) [21:53] going once [21:55] spring cleaning :) [21:58] sadly no more yet [21:59] I think I found the best spring cleaning material in tech-debt. [21:59] Although fixing one bug = at least 2 to 3 MPs :P [21:59] my current launchpad side project is made of tech-debt [21:59] nigelb: thats more like an attic :) - this is stuff that we have lying around uncategorised ;) [22:00] lifeless: heh [22:00] I can see the spiderwebs hang off those bugs :P [22:01] :) [23:33] wgrant: were you going to look at 863098? [23:33] Bug #863098 [23:33] <_mup_> Bug #863098: OOPS when retargeting a private bug to a context the user cannot see private bugs by default <404> < https://launchpad.net/bugs/863098 > [23:34] I wasn't. [23:34] ok. i can [23:34] hi wallyworld_, wgrant [23:34] Evening jelmer. [23:34] hello [23:34] wallyworld_: Thanks. [23:34] np. just wan't to make sure we weren't going to double up [23:35] i hate unity today. had to totally blow alway all my settings to get it to not lock up the desktop [23:37] wallyworld_: I suspect the solution will be to just subscribe the actor, as we do when making a bug private in the first place. [23:38] wgrant: that would work code wise. is that what we want to do policy wise? i guess so [23:39] + xdr: {use: 'flash'}, [23:39] or just set next_url to a different value [23:39] aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa [23:39] + add a notification [23:39] I wouldn't expect the actor to be subscribed; in fact it will be explicitly *against* one of curtis' rules to do that. [23:40] the rule being 'for "private projects" newly private-in-that-project bugs have no subscriptions other than the default set' [23:40] Private projects don't exist. [23:40] So rules around them aren't relevant yet. [23:41] c-i-p is a default-private-bugs project, where the reporter has a subscription regardless. [23:41] 'private bugs by default' seems to be your proxy for now [23:41] wgrant: the reporter is one of the default set of subscriptions [23:41] wgrant: the triager isn't [23:41] * wallyworld_ wonders whether to take the red pill or blue pill [23:42] hooray for not landing code until you have a use for it [23:42] lifeless: But the triager is the one reporting the bug in the "private project". [23:42] mwhudson: you referring to your mp from friday? [23:42] wallyworld_: yes [23:43] wgrant: ahhh, no I'm not :) [23:43] wgrant: meh, I can see arguments both ways. [23:43] mwhudson: do you have a specific use case for it now? [23:43] wgrant: whatever you guys decide, I think it needs documentation if it will increase the visibility of such bugs [23:44] wallyworld_: basically, it seems stuffing things in annotations is unpleasant, better to just have an IParticipationExtras interface that separately defines the things we care about [23:44] lifeless: I feel that all these changes are insane. [23:44] lifeless: Because they're making an existing confusing and exceptional privacy mechanism even more confusing and exceptional, in a gradual fashion. [23:44] lifeless: Rather than a one-off change to a sensible one. [23:44] wallyworld_: yeah, trying to (only) store the feature flag controller on the participation [23:45] wgrant: so the goal is a sensible one; I agree that gradual changes can be tricky. [23:45] wgrant: I wouldn't want to see massive code drops, but I could totally see stuff being feature flagged until all the bits are in place [23:45] mwhudson: +1 for that ff change [23:45] For stuff like this they're not tricky: they're dangerous. [23:45] lifeless: Right, that is what I intended from the start. But it seems everyone prefers gradual change :/ [23:45] Normally, sure. But not for something like this. [23:45] wgrant: talk to your team :) [23:46] wgrant: FWIW, to me, feature-flagged to ~launchpad isn't gradual change, because only we're affected, and we can deal. [23:47] wow, our lp-oops install has masssssive cruft in it [23:47] QUERY PLAN [23:47] ------------------------------------------------------------------------------------------------ [23:47] Seq Scan on oops_oops (cost=0.00..1991487.27 rows=2772 width=68) [23:48] wgrant: so, what's the consensus? subscribe or redirect with notification? [23:51] Not sure. [23:51] But looks like we need to roll back the latest devel rev. [23:51] It breaks lots of profiling tests, it seems. [23:51] (and introduces Flash, so getting rid of it is a bonus :)) [23:52] flash? [23:52] what for? [23:52] Bypassing cross-domain request restrictions. [23:52] I assume. [23:53] you mean someone wrote action script? [23:53] No. [23:53] YUI has some Flash embedded. [23:53] To use as an XHR transport. [23:53] To violate browser cross-domain request restrictions. [23:53] Because Flash is more holey. [23:54] i thought flash functionality was implemented using ActionAcript [23:54] ActionScript [23:54] This is a prebuilt SWF in YUI. [23:54] ok [23:54] yuck [23:54] Because fuck good practice, let's put proprietary compiled formats in our tarballs :) [23:54] the 503 page ? [23:54] lifeless: Yes.