[00:39] <wgrant> cjwatson: db=me
[00:39] <wgrant> er no, actually
[00:48] <cjwatson> Hm yes, I suppose it does
[00:50] <cjwatson> I wonder if we could contrive explicit tests for common kinds of FKs that need indexes
[00:50] <wgrant> cjwatson: Mostly for deletions, I mean, rather than anything relating to the new feature
[00:50] <cjwatson> Yeah
[00:50] <wgrant> And yes, a linter here would be good
[00:50] <cjwatson> Could just be in the test suite rather than linting the SQL as such - less convenient, but probably much easier to write, and would still run before stuff gets as far as an FDT
[00:52] <wgrant> Oh yeah, that's what I mean, sorry
[00:52] <wgrant> We already have a lot of the tools required
[00:52] <wgrant> Used by person merge and librarian GC
[00:53] <cjwatson> Right
[00:54] <cjwatson> Anyway, index pushed
=== juergh_ is now known as juergh
[21:46] <pappacena> For the folks working on OCI, I've added the screenshot of proposed UI for OCI recipe list on the doc (https://docs.google.com/document/d/1hwFnrmEKfg7rimkjj2TC9J5i5gdzUPMZJTUUTFGmlR4/edit#heading=h.bgo6hibhf4m). Let me know if it looks ok that way, so I can implement tests, preloads and performance checks...