Skip to content

Conversation

@chrstg
Copy link

@chrstg chrstg commented Feb 14, 2025

No description provided.

Copy link
Member

@iloveeclipse iloveeclipse left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please provide meaningful commit message.

See

After that, please amend your commit and force push to same branch.

Please note, that we are at 4.35 RC phase so no API changes are allowed for this release.

@github-actions
Copy link
Contributor

github-actions bot commented Feb 14, 2025

Test Results

   541 files  ±0     541 suites  ±0   30m 11s ⏱️ + 2m 44s
 4 364 tests ±0   4 349 ✅ ±0   15 💤 ±0  0 ❌ ±0 
16 596 runs  ±0  16 486 ✅ ±0  110 💤 ±0  0 ❌ ±0 

Results for commit f213a14. ± Comparison against base commit b29048a.

♻️ This comment has been updated with latest results.

@chrstg chrstg force-pushed the master branch 3 times, most recently from 78e1d5d to ae774d9 Compare February 14, 2025 15:12
@chrstg chrstg requested a review from iloveeclipse February 14, 2025 15:21
@chrstg
Copy link
Author

chrstg commented Feb 17, 2025

Please note, that we are at 4.35 RC phase so no API changes are allowed for this release.

So we'll have to wait for the next release.
Anything to do for me until then?

@iloveeclipse
Copy link
Member

Anything to do for me until then?

Yes. Currently you have 3 commits including one merge commit.
Merge commits are not allowed, and for such a small change there is no need for multiple commits either.

Please squash your changes to one commit, rebase it on top of master branch and force push to same branch on github.

@chrstg
Copy link
Author

chrstg commented Feb 17, 2025

Anything to do for me until then?

Yes. Currently you have 3 commits including one merge commit. Merge commits are not allowed, and for such a small change there is no need for multiple commits either.

Please squash your changes to one commit, rebase it on top of master branch and force push to same branch on github.

isn't it good practive to separate functional commit and version increase commit?

@iloveeclipse
Copy link
Member

isn't it good practive to separate functional commit and version increase commit?

Correct, however I can't distinguish your commits in github UI because they have exact same commit message.
So updated instructions were: get rid of merge commit, provide proper commit messages, rebase on master & force push.

@chrstg
Copy link
Author

chrstg commented Feb 17, 2025

isn't it good practive to separate functional commit and version increase commit?

Correct, however I can't distinguish your commits in github UI because they have exact same commit message. So updated instructions were: get rid of merge commit, provide proper commit messages, rebase on master & force push.

force push is rejected with message
! [remote rejected] master -> master (refusing to allow a Personal Access Token to create or update workflow .github/workflows/maven.yml without workflow scope)

Either way i didn't branch on my fork.
I guess it's better to start over from scratch with a new pull request?!

@iloveeclipse
Copy link
Member

I guess it's better to start over from scratch with a new pull request?!

Yes. Id you work on master, you shouldn't do this in the future.
See also https://github.com/eclipse-platform/.github/blob/main/CONTRIBUTING.md#creating-a-pull-request

@merks
Copy link
Contributor

merks commented Feb 17, 2025

What I sometimes do is locally rename the branch I'm using for the PR, check out master again, pull, create a branch with the name previously used for the PR, commit the changes to that branch, and then force push that. That updates the current PR with completely new clean content rather than close a PR and open a new PR...

@chrstg
Copy link
Author

chrstg commented Feb 17, 2025

What I sometimes do is locally rename the branch I'm using for the PR, check out master again, pull, create a branch with the name previously used for the PR, commit the changes to that branch, and then force push that. That updates the current PR with completely new clean content rather than close a PR and open a new PR...

Thanks for the hint!
.. i managed the rebase, but I'm still on my local master branch

@chrstg
Copy link
Author

chrstg commented Feb 17, 2025

the Jenkins build failes with
04:51:13 [ERROR] [API ERROR] File unknown at line 0: The field org.eclipse.swt.SWT.NO_SEARCH has been added to a class (location: unknown)
04:51:13 [ERROR] [API ERROR] File MANIFEST.MF at line 1: The major version should be incremented in version 3.130.0, since API breakage occurred since version 3.128.0 (location: /home/jenkins/agent/workspace/eclipse.platform.swt_PR-1829/eclipse.platform.swt/binaries/org.eclipse.swt.gtk.linux.ppc64le/META-INF/MANIFEST.MF)

Did i increase the versions in a wrong way?

@merks
Copy link
Contributor

merks commented Feb 17, 2025

The API tools are very conservative about what is a breaking change. Adding a public static field to a class is considered API-breaking and hence it wants a 4.0.0. But that's not really acceptable so the error needs to be suppressed. Normally that's done via the a quick-fix in the IDE; was the error not already evident in the IDE?

@chrstg chrstg force-pushed the master branch 4 times, most recently from 80cf69b to d6c80a4 Compare February 18, 2025 08:16
@chrstg
Copy link
Author

chrstg commented Feb 18, 2025

The API tools are very conservative about what is a breaking change. Adding a public static field to a class is considered API-breaking and hence it wants a 4.0.0. But that's not really acceptable so the error needs to be suppressed. Normally that's done via the a quick-fix in the IDE; was the error not already evident in the IDE?

In first place i didn't set up API tools / baseline correctly. Now it showed up and offered the quick fixes.

@chrstg
Copy link
Author

chrstg commented Feb 18, 2025

btw, the binary fragments require host bundle-version="[3.128.." . couldn't it be increased to [3.129 ?

@merks
Copy link
Contributor

merks commented Feb 18, 2025

Yes, I think these much match. Note that you will need to redo all this at the start of the next release cycle, so I'm not sure you want to bang your head on this now...

@merks
Copy link
Contributor

merks commented Feb 18, 2025

In first place i didn't set up API tools / baseline correctly. Now it showed up and offered the quick fixes.

It's probably better to use the setup which does all the necessary and desired things:

https://github.com/eclipse-platform/eclipse.platform.swt?tab=readme-ov-file#contributing-to-swt

E.g., it will update the target platform and the API baseline when the time comes for that...

@chrstg chrstg force-pushed the master branch 2 times, most recently from 081d132 to d7d4436 Compare February 18, 2025 08:39
@chrstg
Copy link
Author

chrstg commented Feb 18, 2025

Yes, I think these much match. Note that you will need to redo all this at the start of the next release cycle, so I'm not sure you want to bang your head on this now...

i was afraid of that :D but next time it'll go in a minute
parts of the .api_filters can be removed then

@chrstg chrstg force-pushed the master branch 4 times, most recently from b19cb9c to 32f3104 Compare March 19, 2025 11:59
@chrstg
Copy link
Author

chrstg commented Mar 19, 2025

Hi @iloveeclipse
can we get this one through the gate.

if ((getShell ().style & SWT.ON_TOP) != 0) {
GTK.gtk_tree_view_set_search_column (handle, -1);
}
if ((style & SWT.NO_SEARCH) != 0) {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we combine this "if" with the previous one, in a style consistent with the Tree / Table code below and extract the condition into searchEnabled() method?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

sure, did it.
(even when imho i think that the implementation of the no_search() method is ugly, too many negations)

@iloveeclipse
Copy link
Member

iloveeclipse commented Mar 20, 2025

can we get this one through the gate.

See my comment in code above, beside this it looks good & ready to merge. If you change the code, please first rebase on master, then amend commit & force push.

@chrstg chrstg force-pushed the master branch 5 times, most recently from 890323b to b30e803 Compare March 21, 2025 14:46
@chrstg
Copy link
Author

chrstg commented Mar 21, 2025

the build fails on PLATFORM = 'gtk.linux.riscv64'
java.nio.file.FileSystemException: /home/jenkins/workspace/eclipse.platform.swt_PR-1829: Read-only file system

i assume it's not from my changes ?!

@iloveeclipse
Copy link
Member

i assume it's not from my changes ?!

For sure not. It's Friday, both hard- and software is already in the weekend mode :-)
I've restarted the build.

@chrst4
Copy link
Contributor

chrst4 commented Mar 24, 2025

i assume it's not from my changes ?!

For sure not. It's Friday, both hard- and software is already in the weekend mode :-) I've restarted the build.

something seems broken on the linux-risc build. All pull requests seem to fail there.

@iloveeclipse iloveeclipse mentioned this pull request Mar 24, 2025
iloveeclipse added a commit to iloveeclipse/eclipse.platform.swt that referenced this pull request Mar 24, 2025
Just a test if we can rebuild natives on gtk.linux.riscv64

See eclipse-platform#1829
@iloveeclipse
Copy link
Member

something seems broken on the linux-risc build. All pull requests seem to fail there.

I've just triggered #1931 to have a proof.

tables eclipse-platform#1830

add SWT.NO_SEARCH flag
used in Table, Tree and List
to prevent the interactive search (popup) in a GTK environment when set
to true.
@iloveeclipse
Copy link
Member

I've just triggered #1931 to have a proof.

At least that one didn't wanted to rebuild binaries and failed for a different reason, but it was on top of master (yours is not). Let rebase this PR.

@iloveeclipse
Copy link
Member

Let rebase this PR.

... and finally we have a green state!

Copy link
Member

@iloveeclipse iloveeclipse left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I will handle search_enabled vs searchEnabled inconsistent naming in a follow up PR.

@iloveeclipse iloveeclipse merged commit ccbd39c into eclipse-platform:master Mar 24, 2025
14 checks passed
iloveeclipse added a commit to iloveeclipse/eclipse.platform.swt that referenced this pull request Mar 24, 2025
Follow up on
eclipse-platform#1829
review. Renamed the `search_enabled` method to have same name /
visibility as on Tree/Table classes.
iloveeclipse added a commit that referenced this pull request Mar 24, 2025
Follow up on
#1829
review. Renamed the `search_enabled` method to have same name /
visibility as on Tree/Table classes.
@chrstg
Copy link
Author

chrstg commented Mar 24, 2025

thanks a lot. Great!

Michael5601 pushed a commit to Michael5601/eclipse.platform.swt that referenced this pull request Mar 27, 2025
Follow up on
eclipse-platform#1829
review. Renamed the `search_enabled` method to have same name /
visibility as on Tree/Table classes.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

7 participants