-
Notifications
You must be signed in to change notification settings - Fork 50
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Prepared Eclipse GEF Classic for 3.23 Devlopements #627 #628
Conversation
Just out of interest: What version of the Eclipse IDE are you using? I always find it weird that a new baseline requires adjustments to the Manifest versions and older versions of the API tooling report false positives due to improperly handling e.g. the |
I'm ussing the Eclipse IDE for Committers:
I also don't know why this is necessary. I somehow inherited the process from my predecessor and as it worked I just followed it. Maybe I'm doing unnecessary or wrong things. |
At least for the WindowBuilder project, I only ever increase the feature versions, unless the API tooling complains for whatever reason. In the end, I don't mind if the Manifests are also updated, as long as it it for a reason that makes sense to me... |
@ptziegler I agree with you that my approach is strange. Changed it back and let's see how it turns out. |
The Eclipse Platform updated to Lucene 10, which requires at least Java 21... |
that means we also need to upgrade to java 21? |
That's the part I still don't understand... Technically, the Platform still supports Java 17 if you use Lucene 9.x. But Orbit already has Lucene 10.x, which is then referenced via the Eclipse Help feature in the Eclipse SDK feature... |
The help bundles have also changed their lower bound. There is no need for GEF to change bounds to use Java 21, but you do need your build to support it when using Eclipse 4.35. I think probably it's sufficent to change this in the Jenkinsfile and you might as well plan ahead for the promotion to require Java 21: |
As a workaround, this error is reduced to a warning if you include https://download.eclipse.org/tools/orbit/simrel/orbit-aggregation/2025-03 in the target platform. The 2025-03 staging repository doesn't contain org.apache.lucene.analysis-smartcn 9.12.1, which is why p2 can't use this Lucene version. i.e. something like this should work with both Java 17 and Java 21:
As long as we support Java 17, we should also build and test against it... Whether we require Java 21 is then something I'd like to discuss separately. |
For EMF I occasionally test it against older target platforms: https://ci.eclipse.org/emf/job/all-target-platforms/ I'm not 100% sure how Tycho launches tests but it seems to me that if a test should respect the bounds specified if they are specified using the JRE in the toolchain. |
Tycho is now happy and
|
I've seen that as well and while I still believe that the API tooling is wrong (as we haven't modified them), I don't think it makes sense to go bug hunting. So your initial PR likely contained everything that needed updating. |
Note that the "only the qualifier has checked" problem can be caused simple by using a new version of ecj. This happens in the platform builds all the time because they often update their ecj version from recent integration builds. |
- Update the version number of all features to 3.23.0 - Bump maintenance version number of all plugins - Update the target platform for the API baseline to GEF Classic 3.23.0 - Update the pom file target platform checking to GEF Classic 3.23.0 - Update the development target platform to Eclipse 2025-03 - Added changelog entry - Removed outdated api problem filters eclipse-gef#627
I know updated all versions that where complained by Tycho and kept all that where not complained. What do you think? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looking good. Apologies for this constant back-and-forth with what to update.
No need to apologize, I'm happy to better understand this and make it more correct. |
#627