Stephen Haberman
Stephen Haberman
Hi @ilikebasic , sorry for the late follow-up here -- honestly I am not an expert on `Any` encoding. You're right, it does seem `@type` is the official standard now......
This should be fixed in #1236 -- thanks @timostamm !
Hi @00KL ; historically ts-proto tries very hard to not generated unused code, so I sympathesize that this is annoying; it looks like this check is tripping us up: https://github.com/stephenh/ts-proto/blob/main/src/main.ts#L197...
This is done for recursive parents...
Another idea would be to use the `ts-patch` plugin to insta-watch (i.e. during a `tsc -w`) what the latest property -> return type mapping is (i.e. for #1205) and persist...
https://github.com/joist-orm/joist-orm/blob/main/packages/orm/src/changes.ts#L124
:tada: This issue has been resolved in version 1.232.0 :tada: The release is available on: - `v1.232.0` - [GitHub release](https://github.com/joist-orm/joist-orm/releases/tag/v1.232.0) Your **[semantic-release](https://github.com/semantic-release/semantic-release)** bot :package::rocket:
Should we enforce the `as: "a1"` matches the `a:1` of entities? When would we catch this? During `em.flush`, but how does a test "watch" `em.flush`? Maybe aliases are something that...
What about `a#1` vs. `a:1`? Which one should we use? If we use `a:1` it won't work until `em.flush`. If we "use sometimes `a#1` (pre-flush), sometimes `a:1` (post-flush)", then ids...
The "see through"-ness problem exists just the same for `hasOneDerived` / `hasManyDerived`, but yes. That said, I'm not sure "making the users 'just know' which of hasOneDerived / hasManyDerived /...