sam boyer

IMO, we need to let go of the notion that this is a simple case. Schematizing playlists on their own is one thing, and it is indeed decently simple, because...

> The meaningful bits that changed (id>uid) happened with no notice at 9.0 Missed this part of your comment when i first read it, sorry - if that's the case,...

(That said, if the goal is rigor, then IMO we need to embrace the notion that, at each schema version transition, we make a deliberate decision about what to do...

OK, after a fair bit of thinking and discussion with @sakjur, i've removed the commits that introduce the second schema and lenses, and am gonna leave this PR back to...

i think this one works best if we aim for a synchronized type that accurately expresses a schema we consider to be valid _today_, then defer any additional schemas, migration/lenses/etc....

closing in favor of #55622

@ryantxu this is the PR to actually make the generated types usable, FYI. When we talked about this a couple weeks ago, i indicated that i thought there was a...

Small update - talked with @joshhunt this morning, and i think we have a workable plan for how to make a "veneer" actually work - really, what to export from...

Update here - i thought i'd have this squared away earlier this week, but i ended up diving fairly deep on cuetsy to fix some outstanding issues. Most notably, we're...

Some of the dashboard schema rejiggering to create discrete TS types seems to have caused some test failures against devenv. i don't think any of those are meaningful, and i'll...