Skip to content

Conversation

@yurikpanic
Copy link

@yurikpanic yurikpanic commented Aug 1, 2022

Fixes summonInline behavior in #15788

The fix is perhaps of a "hacky" kind. But it fixes the original issue described in #15788 and does not break any existing test.
Perhaps we need more generic way to distinguish error cases that mean "given instance not found" from "it was found but we have some error that should be reported to the user".

@odersky
Copy link
Contributor

odersky commented Aug 4, 2022

This should be for @nicolasstucki (or anybody else who volunteers to cover this) to review. But I am pretty sure
that a compiletime.error should mean that the candidate fails to typecheck, and I don't like the exceptions in the code.

The fact that this PR does not break any tests means nothing. Error diagnostics is severly under-tested.

@odersky odersky assigned nicolasstucki and unassigned odersky Aug 4, 2022
@nicolasstucki nicolasstucki self-requested a review December 23, 2022 09:24
@nicolasstucki nicolasstucki removed their assignment Apr 12, 2024
@som-snytt
Copy link
Contributor

Superseded by #22561 which tries to have a smaller footprint and also follows in the footsteps of a similar effort to smuggle out macro errors.

@som-snytt som-snytt closed this Feb 9, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants