The :: syntax is not part of this proposal. It’s mentioned as a possible future direction. Shall we save that discussion until it’s formally pitched?
2 Likes
Before we file bugs, I think we should probably establish what the expected behavior actually is. I approached this proposal from the perspective that there was inconsistency in the language with regard to control over overload/name disambiguation in source files. Whether the language should adhere to a broader philosophy that code generated for a source file should never depend on transitively imported modules seems like a much bigger question that's worth a discussion.
+1
Did you consider adding an option where imports that are missing under the new rules yield a warning with a fix it?
There will be a fix-it (this is covered in the Implications on adoption section of the proposal).
2 Likes
I agree that this is broader than this proposal and I don't want to side track this thread further. This proposal is fixing an important problem and shouldn't be held back by an answer to the larger question. I think we should break that out into a separate thread to talk about expectations.
3 Likes
This is a fantastic cleanup for the language that makes code more robust and dependencies more explicit. +1, no notes.
Doug
6 Likes
Thanks to all of you for participating in the review. SE-0444 has been accepted; please see the announcement thread for details.
5 Likes