| === almaisan-away is now known as al-maisan | ||
| === al-maisan is now known as almaisan-away | ||
| cr3 | hi folks, is there a wiki page that describes the policy used by Launchpad for triaging bugs? for example, I heard you don't use the medium priority, is that documented somewhere? | 19:24 |
|---|---|---|
| cr3 | nevermind, I just found my answer in the aptly named page: https://dev.launchpad.net/BugTriage | 19:25 |
| lifeless | cr3: there is also | 20:04 |
| lifeless | cr3: https://dev.launchpad.net/BugTriage/Background | 20:04 |
| cr3 | lifeless: do you happen to know of tools to generate more elaborate reports of bugs for a project or a person? I know of jkakar's kanban that's pretty cool, but I'd like to have a look at other options | 20:05 |
| lifeless | theres the ubuntu qa tools | 20:06 |
| cr3 | lifeless: thanks, I'll pull the master branch and have a look | 20:06 |
| wgrant | wallyworld_: I wonder if you could just return the types of the accesspolicies that exist? | 23:33 |
| wgrant | I think that would work | 23:33 |
| wallyworld_ | perhaps. i'd rather filter by ones in use | 23:33 |
| wgrant | The access policies that exist are the access policies that are in use, ± a day | 23:35 |
| wallyworld_ | sure, we now have a garbo job of course. but a day is a long time | 23:36 |
| wgrant | I was thinking last night that we should probably prune on sharing_policy change, too | 23:36 |
| wgrant | So if you change it and they're unused then they evaporate | 23:37 |
| wgrant | immediately | 23:37 |
| wallyworld_ | that sounds best to me. i don't like the job method of cleaning these things up. fine for a safety net, but i think we can do better | 23:41 |
| wgrant | Well, the job is necessary unless we want to be firing off more jobs every time an information type changes | 23:42 |
| wgrant | But we should probably do it inline on sharing_policy change | 23:42 |
| wallyworld_ | that's what a meant - do it inline, not more jobs | 23:43 |
| wallyworld_ | if it can be done efficiently | 23:44 |
| wgrant | Even when an artifact's information_type changes? | 23:44 |
| wgrant | We *could* do that, but it's likely to cause some contention and isn't obviously a significant benefit | 23:44 |
| wallyworld_ | it's a benefit since it allows the in use policies to be maintained. but how much so, not sure | 23:46 |
| wallyworld_ | if the info type changes, so long as we ensure there's a policy if there were none previously | 23:46 |
| wgrant | You can't change the info type to something not allowed by the sharing policy, and changing the sharing policy ensures that there's an access policy, so that direction is not an issue | 23:47 |
| wgrant | It's just cleanup that needs consideration | 23:47 |
| wallyworld_ | it may ensure there's an access policy but the policy may be cleaned up before it can be used | 23:48 |
| wgrant | It won't be, since we don't clean up anything that's allowed by either sharing_policy. | 23:48 |
| wallyworld_ | that's good | 23:49 |
| wgrant | (modulo that bug last week) | 23:49 |
| wallyworld_ | ah right, yes | 23:49 |
| wgrant | So, I think it's probably sensible to just return the APs that exist. Separately we should factor most of the pruner into a model method, and call it from both the garbo job and setFooSharingPolicy | 23:51 |
| wgrant | Since your new method has to return all the stuff permitted by sharing policies, and all of the types involved in access policies, but the model ensures that everything permitted by a sharing policy has an access policy | 23:52 |
| wallyworld_ | the method already returns the permitted stuff, just is missing the current in use stuff | 23:54 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!