[00:51] <cjwatson> wgrant: Thanks
[00:53]  * cjwatson requests a deployment
[12:35] <cjwatson> wgrant: Thanks for the reviews so far.  I obviously can't actually land git-recipe-model yet, since db-git-recipes needs to go in first and isn't approved.
[12:35] <cjwatson> So I'll just get the rest of the stack proposed today.
[12:35] <wgrant> Yep
[12:35] <wgrant> How much is left?
[12:36] <wgrant> I was waylaid by a ridiculous ISP transparent proxy bug today, so didn't get through them all.
[12:37] <cjwatson> I just proposed browser code to edit existing recipes.  After that I have ~1300 lines to sort out - I think one branch for a bit of BranchTarget/GitNamespace refinement, and one or two more for listing recipes and creating new ones.
[12:37] <wgrant> Oh, not so bad.
[12:38] <cjwatson> Would prefer not to land the last until the buildd stuff's in place, of course.
[12:38] <wgrant> Indeed.
[12:39] <cjwatson> I could add an FF; not sure it's worth it in this case.
[15:33] <cjwatson> wgrant: The recipe series is complete now.
[22:25] <wgrant> cjwatson: Re. Build-Depends-Arch, I'm not sure that we shouldn't use user_defined_fields.
[22:26] <wgrant> I'm not unhappy either way, but the specific fields exist because JSON didn't.
[22:27] <cjwatson> If you think that it's reasonable to special-case certain elements of user_defined_fields, I don't object to reworking things.
[22:28] <wgrant> I have no problem with only displaying things that are interesting to display.
[22:29] <wgrant> The only downside is the inconsistency.
[22:29] <wgrant> vs. the existing fields
[22:30] <cjwatson> Display, and the archivepublisher ordering tweak, though the latter's fairly insignificant.
[22:31] <wgrant> Oh they're not seriously sorted that way are they
[22:31] <cjwatson> It's manual
[22:31] <cjwatson> We could pick out B-[DC]-A from user_defined_fields
[22:32] <wgrant> I personally think it makes sense to do that.
[22:32] <wgrant> We don't win much by having dedicated columns.
[22:32] <wgrant> Unless you consider increased database patch numbers to be a win.
[22:32] <cjwatson> A little cumbersome in lp.archivepublisher.indices, but not horribly.
[22:32] <cjwatson> No, my recent MP history notwithstanding :-)
[22:32] <wgrant> Heh
[22:33] <cjwatson> I will note that upgrade.py breaks when we hit 100.
[22:33] <cjwatson> Might be worth a consolidation at some point ...
[22:33] <cjwatson> (Or fixing the glob)
[22:33] <wgrant> I think the last one was in like 2010.
[22:34] <wgrant> Could get stub to do a fresh dump once your latest patch is out.