Update PCT

[timja]

No description provided.

[timja]

➜  plugin-compat-tester git:(162dffc) git bisect good
ae2dd7130a56f7298437cd0a72cd7d4a06eb3fe3 is the first bad commit
commit ae2dd7130a56f7298437cd0a72cd7d4a06eb3fe3
Author: imonteroperez <[email protected]>
Date:   Wed Sep 8 11:51:51 2021 +0200

    Bugfix on dealing with non-executed test phase due collateral issues

 .../org/jenkins/tools/test/PluginCompatTester.java | 13 ++++++----
 .../tools/test/util/ExecutedTestNamesDetails.java  | 29 ++++++++++++++++------
 .../tools/test/util/ExecutedTestNamesSolver.java   | 22 ++++++++--------
 .../jenkins/tools/test/PluginCompatTesterTest.java | 24 ++++++++++++++++++
 .../src/test/resources/bad-surefire-exclusion-list |  1 +
 5 files changed, 67 insertions(+), 22 deletions(-)
 create mode 100644 plugins-compat-tester/src/test/resources/bad-surefire-exclusion-list

[timja]

According to https://ci.jenkins.io/job/Tools/job/bom/job/PR-863/2/pipeline-console/?selected-node=332

The build is erroring when injected tests fail

[jetersen]

Perhaps the injected tests do not add value to PCT?

[timja]

Perhaps the injected tests do not add value to PCT?

Yeah possibly, we already ignore them in bom

bom/pct.sh

Line 41 in b9fa01a

if [ -f $t/test-classes/InjectedTest.class -a \! -f $t/surefire-reports/TEST-InjectedTest.xml ]

I'm reverting the PR affecting us in our fork of PCT anyway jenkinsci/plugin-compat-tester#343

Upgrading should be dealt with separately to fixing the jep-229 issue imo

[timja]

I manually deployed a SNAPSHOT because the folder is setup to merge with master bringing in changes that cause bom to fail.

I tried reverting the PR that caused issues but was having difficulty.

this could work in the meantime what do people think?

[jetersen]

Whatever works but could you get a link to a commit sha at least of what is in the snapshot? 🤔

Would be nice to finally land a proper PCT fix?

[timja]

done

[jglick]

Better to get jenkinsci/plugin-compat-tester#344 to pass tests—by ignoring stuff if necessary.

[timja]

ha, my last attempt of just reverting the PCT change that affected bom didn't go so well as there were other changes. Unless there's a simpler fix I'd probably just revert every PR in the last year =/.

pr-merge pain -.-, can't build an isolated PR

Not sure if I'll have the time to fix properly.

[jglick]

Ack. PCT needs a lot of work.

[jglick]

See #657. Ought to remain in draft until it is picking up an incremental build (if not an actual release).

[jglick]

jenkinsci/plugin-compat-tester#345 oof