Hello!

I would like to hear clarification on what happened in MARCHETYPES, as from
everywhere I look into this it confuses me.

I had opened
https://issues.apache.org/jira/browse/MARCHETYPES-76
https://issues.apache.org/jira/browse/MARCHETYPES-77

and https://github.com/apache/maven-archetypes/pull/18 fixing MARCHETYPES-76

It was dangling for some time, then I poked this list and it was imediately
looked into. TY!

The https://github.com/apache/maven-archetypes/pull/18  was reviewed, all
parts of it were discussed  and it was merged.
The only week point of it was that it mixed also one refactoring in
separate commit and I did not know it will be sqashed with merge. Knowing
that, it would be separate PR.
Simialrly it included two commits for readability - one with siomgleline
bump of maven compiler and second- much bigger - adapting the tests. I
think this two commits should in one PR and shold not be squashed. If they
will be squashed, its ok.

Today I had opened https://github.com/apache/maven-archetypes/pull/20 with
corresponding https://issues.apache.org/jira/browse/MARCHETYPES-78 It was
closed without reasoning - it did not even wait for reply to the single
question it posted.
In addition it was "surpassed" by
https://github.com/apache/maven-archetypes/pull/22 whcih was reverting the
PR#18  without justification and was including also PR#22.
Where it also included great

+    with:
+      jdk-matrix: '[ "8", "11", "17", "21" ]'

It did not justified anything else.
Later - after many confusing comments which proved a bit of
misunderstanding from apache side, the PR#18, which was approved by two
reviews as by policy, was reverted witout any explaantion or discussion
(???!????).

Note, this is no blame, I'm jsut confused by the events and I do not
understand why it have happened. And tbh, it is pretty demotivating.

So I'm asking, why the person who did it, did it, and also am asking what
is now expected form me.
If the reverted commit was so wrong, why it was let in?
If there is comment "ok all reverted. feel free to reopen PR with correct
changes"  what is wanted fro me, and why the person who acted literally
waste time of all others?

My understending of what is needed/wanted:
 - submit
https://github.com/apache/maven-archetypes/pull/22/commits/9eea2ff0ffb5f1a01b5ff6eeb2722ee71ac8f73f
(the jdk-matrix: '[ "8", "11", "17", "21" ]') as first and separate PR
without its own jira
 -- note that the requester will be not satisfied, as the integration tests
are passing on jdk 21. Fact that theirs output is not usable byt he jdk21
is not tested.
 -- also: no, I'm now not motivated to enhance the suite to fail properly
for all affected components
- submit
https://github.com/apache/maven-archetypes/pull/18/commits/9d59f3188638086501fa4a40b141360eabdc1a48
(the reffactoring) as separte changeset without its specific jira
- submit the
https://github.com/apache/maven-archetypes/pull/18/commits/22520884a7d0bbc8cf1dd248cf2873cff9ffc26f
(fix for MARCHETYPES-76) commit with the test adaptation of
https://github.com/apache/maven-archetypes/pull/18/commits/a7904177028ee6c3694856f8893434de77adc3cc
in one PR under  MARCHETYPES-76
- leave https://github.com/apache/maven-archetypes/pull/20 as it will be
looked into as resolving MARCHETYPES-78 with work in former PR#18 taken in
account.

As this - after all that waiting, reviewing, aprooving and reverting is a
bit reedundant work, i want to be sure that we agree on what everybody
wishes - as all is explained in MARCHETYPES-78 and MARCHETYPES-76 (+the
https://github.com/apache/maven-archetypes/pull/18 and paritally also
https://github.com/apache/maven-archetypes/pull/20 and
https://github.com/apache/maven-archetypes/pull/22)

Maybe one more missing expalnatin from me: I had first acted on
maven-archetype-quickstart because I ws dirrectly afected by it, it had a
bigger changes due to tests, and   I was not sure how t will be accepted by
wider community. I always had an intention to fix all others once I will
check how things go. And I'm really not sure why the fix for and new issue,
was a cause to revert quite good and approved change - by actually one of
the reviwers.

J.

Reply via email to