Eval compile error in 2.12.0-RC1: Improvement or regression?

[ghost]

Compiling Eval.scala in 2.12-0-RC1 I get a compile error that is trivial to fix, but nevertheless represents a change as the code is fine in 2.10/2.1. So the question here is whether it is an improvement or a regression that we can highlight.

compile:

[error] core/src/main/scala/cats/Eval.scala:77: overriding value start in class Compute of type () => cats.Eval[this.Start];
[error]  value start has incompatible type
[error]           val start  = c.start
[error]               ^
[error] core/src/main/scala/cats/Eval.scala:78: overriding value run in class Compute of type this.Start => cats.Eval[B];
[error]  value run has incompatible type
[error]           val run = (s: c.Start) =>

Trivial fix:

@@ -74,8 +74,8 @@ sealed abstract class Eval[+A] extends Serializable { self =>
       case c: Eval.Compute[A] =>
         new Eval.Compute[B] {
           type Start = c.Start
-          val start = c.start
-          val run = (s: c.Start) =>
+          val start: () => Eval[Start] = c.start
+          val run: Start => Eval[B] = (s: c.Start) =>
             new Eval.Compute[B] {
               type Start = A
               val start = () => c.run(s)

[non]

At the least it seems like a change in inference -- either a a known-change in 2.12 or some kind of bug regression.

For our part I don't mind adding the type annotations (it probably makes the code easier to read) but it probably also makes sense reporting upstream to the Scala team. Do you want to do it? If not I'm happy to open one later on.

[ghost]

@non I very casually mentioned this on gitter scala/contributors. I'm happy to raise an issue on their JIRA later today. I'll post with link when done, so we can track

[ghost]

Done - https://issues.scala-lang.org/browse/SI-9931

[non]

Thanks!

[SethTisue]

is there a branch that's set up to use 2.12.0-RC1? I'd like to look at this for you, but I don't want to duplicate work you've already done, either

[ghost]

@SethTisue #1382 uses 2.12.0-RC1. If you compile this locally, please see the note in the PR re local scoverage

[SethTisue]

The discussion is on JIRA now, but I just wanted to say thanks for the report. It's enormously helpful to get reports like these.

[ghost]

@SethTisue and FTR for those that do not look into JIRA - Awesome work!!! Really, I mean it

[travisbrown]

This is changed and the issue noted in a comment in #1415, so I think we can close this issue once that's merged.

[SethTisue]

#1415 is merged