[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [jdt-dev] About Y-Builds (was Re: 4.35 Y-Build: Y20250111-1000 - BUILD FAILED)
|
Are you proposing our implementation should capture the full
history of every feature that once was a preview feature, perhaps
in several iterations? So when there is a change between previews
or between preview and final version, JDT should still mimic each
of the preliminary variants?
I thought this discussion is about simplifying development. The
way I understand this latest proposal it is the exact opposite.
On 20.01.25 11:44, Mickael Istria via
jdt-dev wrote:
>
Master could easily include support for Java 24; as
long as the file support Java 23, the Java 24 code
wouldn't harm; just as it won't harm after Java 24
is released and the BETA_JAVA24 branch is merged.
I request you to look at the code
around preview flags, for e.g.,
AST#isPreviewEnabled() . In master, preview can only
be enabled with Java 23, but in BETA_JAVA24, preview
can only be enabled with Java 24.
I see some cases where the condition is used with
`ast.apiLevel() == AST.JLS20 &&
ast.isPreviewEnabled()`. Isn't it a good thing we could try
to use more?
Note that the same "problem" exists with Javac: is you
used a Java N preview feature, the feature is standardized
in Java N+1 and you use Javac N+1, then the feature is
enabled in, independently of the -preview flag.
In practice, it's not really harming as usage of preview
features seems to be rarely reaching production code. At
least, not so many people complain about this possible
misalignment between the (older) spec and the (newer)
compiler behavior when it comes to toggling preview
features.
Similarly, JavaFeature that
specifies the @since version a preview feature is
supported, will change between master and BETA.
OK, so maybe we could consider to refine it, replacing
the boolean with some "previewFeatureIn" listing the related
JDK versions ?
_______________________________________________
jdt-dev mailing list
jdt-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jdt-dev