DatePicker UI #1039

[gnl42]

Got tired of seeing the DatePicker (and friends) truncated in Windows:

as well as in the JiraConnector in Windows:

Linux/MacOs was better but also had some issues:

With these changes it looks like:

and

and

The DatePicker popup was cleaned up a bit as well from:

to

For some reason that I don't understand, the Mylyn-Runtime.launch version differs slightly in appearance from when Mylyn is installed in an Eclipse instance.

[github-actions]

Test Results

 1 626 files  ± 0   1 626 suites  ±0   1h 30m 4s ⏱️ + 6m 35s
 5 327 tests  - 27   5 241 ✅  - 32   79 💤  - 2  3 ❌ +3   4 🔥 + 4 
13 470 runs   -  6  13 214 ✅  - 27  235 💤 ±0  9 ❌ +9  12 🔥 +12 

For more details on these failures and errors, see this check.

Results for commit 9c3c465. ± Comparison against base commit ee5fcfc.

This pull request removes 31 and adds 4 tests. Note that renamed tests count towards both.

AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testAbortBuild
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetJobDisabled
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetJobGit
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetJobs
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetJobsWithWhitespaces
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetManyJobs
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testGetNestedJobs
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testRunBuildFailing
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testRunBuildSucceeding
AllJenkinsTests JenkinsClientTest "JENKINS 2.506" ‑ testRunNestedJob
…

AllJenkinsTests JenkinsClientTest ‑ Unknown test
AllJenkinsTests JenkinsIntegrationTest ‑ Unknown test
AllJenkinsTests JenkinsUrlUtilTest ‑ Unknown test
org.eclipse.mylyn.jenkins.tests.AllJenkinsTests ‑ Unknown test

♻️ This comment has been updated with latest results.

[ruspl-afed]

Although added methods looks useful, I would avoid introducing new API during this change just to not mix two different discussions.

[gnl42]

Deleting it and then re-creating it doesn't sound very appealing.

Any other options?

[gnl42]

@ruspl-afed what would you like me to do?

[ruspl-afed]

Sorry, I missed your reply here. However, I do not understand it.
My point was to not introduce new API here, and focus on the UI itself.
And this doesn't mean deletion and re-creation, it means moving new type to an internal package for further API-related discussion.

[gnl42]

I think it should be ok now

[ruspl-afed]

Now we have two FontUtils types. I suppose we need inly one, located at internal.commons.ui

[gnl42]

Sigh, the refactor move only committed the new location, leaving the original location (and references) in place.

I've been seeing this recently, I do a commit only to have some of the changed files suddenly appear as unstaged afterwards.
Both in Windows and Ubuntu

[ruspl-afed]

The only thing to do here is to squash to single commit

[gnl42]

Done

[gnl42]

Wondering if we should treat fixture discovery failure as a "disabled/skipped" test?