FTFA:

> The issue was not present in the early access releases for macOS 14.4, so it was discovered only after Apple released the update.

There were some security fixes that were marked as under active exploitation in macOS 14.4, so my money is on those landing very steep close to the release date and this being fallout from that.

FTBR:

"It is also successful on the previous versions of Sonoma" (no mention of EA)

"It could not be reproduced on 23-ea+13, 22+36-2370" (EA versions of Java)

Right so it’s present in all shipping versions of Java back to at least 8, and the reproduction case works fine on Sonoma pre-14.4. Ergo, Apple changed something in 14.4 (a minor patch release), and it broke something in Java. I’m not sure what point you’re trying to make.

It’s possible that Java was doing something very weird, maybe undocumented/unsupported; but it’s not clear how this is anything other than Apple making a change (which I don’t see documented in their release notes) and breaking users.

Yes, pre 14.4 (14.3.1 to be exact) but NOT early access of macos. The whole thread and Oracle's blogpost is about that it "suddenly broke but it was still working in a preview release of macos".

Java DOES work with an EA version of Java according to the bugtrackers.

Oracle's PM doesn't know what's going on and is mixing things up.

Please read the actual bugreport instead of an article written my Oracle's PM who really does not want the crap on his account.

So I can release breaking changes to my API and expect consumers (who are software suppliers themselves) and it's their fault for not instantly updating their code? How is this not Apple when they were the ones that broke it?

Apple should not be at fault! As simple as that.

Did you read Oracles post?

"The issue was not present in the early access releases for macOS 14.4, so it was discovered only after Apple released the update."

JVM generates code on the fly, which normal applications are not allowed to do. So java by definition receives special treatment from the OS, which OSX seem to have broken now.

I did and also the bugreport.

https://bugs.java.com/bugdatabase/view_bug?bug_id=8327860

It says that "It is also successful on the previous versions of Sonoma", no mention of EA.

In the bug report it says it's actually not reproducable with an eary access version of Java "It could not be reproduced on 23-ea+13, 22+36-2370"

TBF I'd rather not take Oracle's Product Managment's word for granted.

It's not that Java has a special treatment, but all applications who "request it" are allowed (for example also JavaScript engines in browsers).

https://developer.apple.com/documentation/bundleresources/en...

There is no “special treatment for Java”. This is a well known technique that has been used for ages in the implementation of JITs. The behavior it relies on is specified by POSIX.

Definitely Apple.