[02:42] wgrant: have you ever looked at having un-proxied objects by default in the harness? [02:42] presumably there's something that makes that impossible or difficult [02:50] hmm weird, have some doctests failing with proxied instance reprs instead of a diff, have you seen that before? [03:03] blr: What are the failures? [03:14] wgrant: very uninformative https://pastebin.canonical.com/132631/ [03:15] blr: Hum, that's certainly odd. [03:15] (also, you might want to merge devel -- +all-branches no longer exists) [03:15] probably a result of changing the menu context from a product.development_focus (branch) to a product [03:16] Oh [03:16] You have a print somewhere [03:16] wgrant: hmm ok, maybe it will go away :P [03:16] browser.open isn't meant to print anything. [03:16] You have some leftover debugging code. [03:16] ah [03:16] I do to, thank you :) [03:17] s/to/too/ [04:53] wgrant: small branch to review here: https://code.launchpad.net/~blr/launchpad/dont-fear-the-repr/+merge/261174 [04:54] also if you're able to revisit project setbranch at some point (monday is okay) [09:26] wgrant: https://code.qastaging.launchpad.net/germinate/+git shouldn't show lp:germinate at the bottom, should it? [09:27] It's not an "other repository". [09:28] cjwatson: Probably not, but it required a GitCollection enhancement. [09:28] Or perhaps "All repositories" if it's hard to exclude that. [09:28] It's hard-ish for bzr. [09:28] For git we can just say "NOT target_default" [09:29] Trying to avoid the hideousness that plagues +branches. [09:29] Sure [09:30] Luckily we don't need a whole lot of stupid joins for git, so it's doable. [11:36] wgrant: There are still several users of DAS.supports_virtualized in the tree [11:37] wgrant: Oh, never mind, that's not the actual DB column [11:39] cjwatson: Indeed, all a property now. I might go back and update them all at some point, but I don't see much need. [11:50] cjwatson: Thanks. [14:45] wgrant: I wonder if we should have some way to get to the default repository's edit views from Project:+git, since we're generally trying to present that as a superset [22:13] cjwatson: Right, I believe we want to include all of the portlets, given the approach we're taking. [22:13] cjwatson: But that gets a bit weird with subscribers and two privacy portlets. [22:13] So I haven't done it yet.