Ólafur Páll Geirsson

Thanks for the offer @kubum I am not sure yet if this is a desirable change or not, the git repository grows a lot bigger by keeping the generated website...

I agree with you @xeno-by that this change does not carry it's weight. I propose we close the ticket as "wontfix".

I've updated the PR to only sort Java class signature declarations.

This remains relevant since expect suite still sorts java declarations by kind and metacp doesn't preserve order across kinds. I don't think this is blocking for v4 however so unassigning...

This opens up a more general question of what symbols we should persist and how the spec should document this. I personally care most about being able to lookup information...

I think the first problem is to solve is defining the `SYNTHETIC` property https://github.com/scalameta/scalameta/issues/1492 In the JVMS https://docs.oracle.com/javase/specs/jvms/se8/html/jvms-4.html#jvms-4.1 the definition for `ACC_SYNTHETIC` is > Declared synthetic; not present in the...

Related question from https://github.com/scalameta/scalameta/pull/1690/files#r202204747 Should SemanticDB payloads contain package symbols? If not, we should update the spec to explicitly state this. I vote for "leave it undefined", but don't have...

@xeno-by I like `GENERATED` and `INVISIBLE` with your suggested meanings. Am I right in understanding that `INVISIBLE` symbols will always have the `GENERATED` property set?

This is not a blocker for v4.0.0, we can still fix it in the v4.x series if a use-case arises.

I use neither cli, only the `symtab` module. Feel free to merge metacp with metai 👍