JLS discussions worth watching

[stephan-herrmann]

Here I will collect a few JDK tickets with recent updates, which might produce JLS changes:

Target 25:

• ✔️ https://bugs.openjdk.org/browse/JDK-8348901 "14.11.1: Allow 'null' case label for 'switch (null)'"
  • addresses my own question in https://mail.openjdk.org/pipermail/compiler-dev/2025-January/029306.html
  • resolution in 25: JLS has been adjusted to existing behavior of ecj & javac. No impl changes needed.
• ✔️ https://bugs.openjdk.org/browse/JDK-8351401 "Remove Uses of Boxed Primitive Constructors in JLS"
  • resolution in 25: new BoxType(n) has been replaced in JLS by BoxType.valueOf(n)
    • change in 5.11.1 (String Conversion) matches the existing status in ecj implementation
    • change in 8.1.2 only affects example code

No target currently:

Last checked for activity (negative): 2025-07-26

• https://bugs.openjdk.org/browse/JDK-8314380 "JLS 4. 10. 4. Least Upper Bound is a complete gibberish (full of bugs)"
• https://bugs.openjdk.org/browse/JDK-8314382 "Javac converts <? super FormalUpperBound> to <FormalUpperBound> against jls"
• https://bugs.openjdk.org/browse/JDK-8016196 "Inference: define supertype parameterization for wildcard-parameterized types"
  • this one has a long list of related tickets!
• https://bugs.openjdk.org/browse/JDK-7034922 "4.9: Clarify membership of intersection types for wildcard parameterizations"
• https://bugs.openjdk.org/browse/JDK-8347759 "Ambiguity about the downcast of short and byte in conditional expression"
• https://bugs.openjdk.org/browse/JDK-8328345 "5.2: Fix omission of widening-and-narrowing primitive conversion (byte to char/Character)"
• https://bugs.openjdk.org/browse/JDK-8016207 "Widening of capture vars occurs at unspecified times"
  • while originally blaming javac, it states that perhaps JLS should be adjusted to mirror valuable behavior of javac

We should periodically check if any of these yield any real changes.

[mpalat]

@stephan-herrmann - Good points. I have added this issue under the top level 25 issue for tracking.
Also added the general Top level query for language issues

[stephan-herrmann]

I have added this issue under the top level 25 issue for tracking.

That's fine.

OTOH, this ticket may well be running longer than BETA_JAVA25.

[mpalat]

OTOH, this ticket may well be running longer than BETA_JAVA25.

true. will keep note of this.