[13:36] <cjwatson> wgrant: Looks like you need some more preloading on SPRB
[13:37] <wgrant> cjwatson: Already pushed and awaiting more failures :)
[13:39] <cjwatson> Ah :)
[13:40] <wgrant> Though, shockingly, it seems like that's the extent of them.
[14:29] <cjwatson> wgrant: And now it's my turn, I see
[14:29] <cjwatson> Also WTF, I ran those tests, I'm nearly sure
[14:30] <cjwatson> I mean, I ran -t mail
[14:30] <cjwatson> oh, I had a layer setup failure I didn't notice :(
[14:30] <wgrant> Heh
[14:57] <wgrant> Oh, that testfix wasn't nearly as bad as I expected.
[15:00] <cjwatson> Amazing how much test failure verbiage can be caused by a small amount.
[16:58] <cjwatson> Hm, snap-listings is harder than I thought.  Branch.recipes etc. can be fairly naïve, but they get away with that because recipes don't support private branches.  I need something a bit more like Branch.getMergeProposals.
[17:01] <cjwatson> Well, maybe Branch.snaps is OK, but at any rate Person.snaps (or equivalent) has to limit to snap sources that you can see.
[17:02] <cjwatson> At least we don't have to deal with intersections between privacy of multiple branches.
[17:02] <cjwatson> BranchCollection.getMergeProposals /o\