prepare to remove UninitializedError

[hrhino]

It doesn't have any Scala-specific semantics and is unlikely to be worth more than a custom exception type to anyone currently using it.

If there's some reason to keep it around that I'm missing, please let me know.

[sjrd]

Isn't that the exception thrown by the codegen for -Xcheck-init?

[hrhino]

that's what I thought at first, but that's actually UninitializedFieldError.

[NthPortal]

Is there a reason to remove it for 2.13 rather than 2.14? I feel like it's a bit late in the cycle, if it's not a foot cannon

[hrhino]

It doesn't need to go at all, in the objective sense.

However, it's in the scala namespace, and therefore imported by default in nearly every file (unless you use -Yno-imports or som's new thing in 2.13). It's also highly doubtful that it's useful, seeing as there's no way to even specify what is uninitialized.

All that said, I'm happy to rebase this on 2.13 and target the removal for 2.14. We could even not remove it at all.

[SethTisue]

I suggest that we deprecate in 2.12.7 and remove in 2.14.

[SethTisue]

note that it's not necessary for the deprecation message to actually say which version it will be removed in, we usually just say "in a future release" or just leave it unstated

if there is an informal intention to hold the removal until 2.14, that could go in a code comment to keep it from getting caught up in a sweep

[hrhino]

Alright, I can go with that. Updated.

[dwijnand]

/rebuild

[hrhino]

This conversation did not age well. Presumably "removed in dotty".

[som-snytt]

Maybe follow up with boopickle...