Closed iloveeclipse closed 7 months ago
Few examples that I've checked show only expected changes.
Created PR's with touched bundles.
PS after all merged, we should also merge https://github.com/eclipse-jdt/eclipse.jdt.debug/pull/429 I will be back in an hour from now to check the state.
Triggered new build: https://ci.eclipse.org/releng/job/Builds/job/I-build-4.32/70/
Thanks, @iloveeclipse !
I checked a few samples of the artifact comparison, which all looked good to me.
In fact all I could see was the addition of some InnerClass attributes which are a side effect of adding a Class constant for an inner class to the constant pool. It seems the comparator doesn't report any changes just in the constant pool. This makes a lot of sense, since the constant pool is only an auxiliary structure.
I am sure @iloveeclipse is fully aware of such bytecode topics, so just mentioning :)
@stephan-herrmann : seeing you there at unusual daytime, couldn't resist to take the opportunity to ask if you missed this comment regarding failed tests? :-)
Nice teamwork. 🥇
@stephan-herrmann : seeing you there at unusual daytime, couldn't resist to take the opportunity to ask if you missed this comment regarding failed tests? :-)
I took note. But then I was sure nobody would be overly worried if a test using lambda fails at 1.7 :) at least not for a couple of days.
nobody would be overly worried if a test using lambda fails at 1.7 :) at least not for a couple of days.
This is not correct assumption, I monitor SDK builds state and every new test fail means extra time I have to spend to check if that is new regression or not.
nobody would be overly worried if a test using lambda fails at 1.7 :) at least not for a couple of days.
This is not correct assumption, I monitor SDK builds state and every new test fail means extra time I have to spend to check if that is new regression or not.
sorry. The main goal is of course to avoid the first failure in the first place, which in this case means: locally run new tests at all compliance levels. Will try to remember.
https://download.eclipse.org/eclipse/downloads/drops4/I20240416-0300/ is stable, closing this.
See
This is anticipated effect from https://github.com/eclipse-jdt/eclipse.jdt.core/pull/2341 (need to be double checked of course)