Report coverage unconditionally on Travis

[greg0ire]

Q	A
Type	improvement
BC Break	no
Fixed issues	n/a

Summary

pcov is fast enough for this to be acceptable.

[greg0ire]

This raises the question about the Smoke Testing step… should I keep it and remove the sqlite job from test? That way if something is wrong with the unit tests we don't run a bazillion test jobs for nothing

EDIT: it's not really duplicate the runs are on different versions of PHP

[morozov]

This raises the question about the Smoke Testing step… should I keep it and remove the sqlite job from test? That way if something is wrong with the unit tests we don't run a bazillion test jobs for nothing

Definitely. That's the idea. If there's a unit test failure, you don't want to wait 10 minutes until the DB2 job completes. Ideally, I'd like to not run tests until all static analysis phases succeed (which was the case before we moved some jobs to Actions) but I'm not sure how this could be implemented.

[greg0ire]

I think this will be achieved by moving to GA, with job dependencies. BTW @bendavies , are you still planning to work on that? 🙏 🤞

[codecov-commenter]

Codecov Report

Merging #4053 into 2.10.x will increase coverage by 0.16%.
The diff coverage is n/a.

@@             Coverage Diff              @@
##             2.10.x    #4053      +/-   ##
============================================
+ Coverage     73.16%   73.32%   +0.16%     
  Complexity     5031     5031              
============================================
  Files           215      215              
  Lines         12825    12825              
============================================
+ Hits           9383     9404      +21     
+ Misses         3442     3421      -21     

Impacted Files	Coverage Δ	Complexity Δ
lib/Doctrine/DBAL/Configuration.php	84.84% <0.00%> (-9.10%)	17.00% <0.00%> (ø%)
...ib/Doctrine/DBAL/Driver/DrizzlePDOMySql/Driver.php	66.66% <0.00%> (-6.07%)	5.00% <0.00%> (ø%)
.../Doctrine/DBAL/Driver/AbstractPostgreSQLDriver.php	96.36% <0.00%> (-1.82%)	24.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Driver/PDOOracle/Driver.php	15.38% <0.00%> (-1.29%)	5.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Driver/PDOIbm/Driver.php	12.50% <0.00%> (-0.84%)	6.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Driver/SQLAnywhere/Driver.php	7.69% <0.00%> (-0.31%)	7.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Platforms/SQLServerPlatform.php	90.57% <0.00%> (-0.28%)	224.00% <0.00%> (ø%)
...ib/Doctrine/DBAL/Platforms/SQLAnywherePlatform.php	94.58% <0.00%> (+0.01%)	196.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Schema/Table.php	96.81% <0.00%> (+0.01%)	102.00% <0.00%> (ø%)
lib/Doctrine/DBAL/Platforms/DB2Platform.php	94.18% <0.00%> (+0.02%)	124.00% <0.00%> (ø%)
... and 11 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 87aad2f...6c83f5a. Read the comment docs.

[morozov]

Looks good 👍 Please squash.

[greg0ire]

Looks good +1 Please squash.

Should I really? I mean I deliberately didn't, since it does not have to do with running coverage unconditionally… using the latest stable could be something we would want to do or revert independently, or maybe there is a link I don't see?

[morozov]

Without the second commit, the first one fails on PHP 7.4.

[greg0ire]

Oh all right then, didn't notice.

[greg0ire]

Ok now continousphp is flaky too…

[bendavies]

I think this will be achieved by moving to GA, with job dependencies. BTW @bendavies , are you still planning to work on that?

I do have a branch that moves postgres and mysql tests to GitHub actions.
It needs work.
I can submit it I'd you'd like to work on it together.

[morozov]

While I like the idea as such, before we go full steam ahead, could we establish some expectations of the end result? E.g. Travis currently allows us to run 10 jobs in parallel (I believe it's a favor since the default for OSS is 5, may be outdated) and we execute ~40-50 jobs per build.

Will we be able to do the same on GitHub?

[bendavies]

While I like the idea as such, before we go full steam ahead, could we establish some expectations of the end result? E.g. Travis currently allows us to run 10 jobs in parallel (I believe it's a favor since the default for OSS is 5, may be outdated) and we execute ~40-50 jobs per build.

Will we be able to do the same on GitHub?

I think 20 as per https://help.github.com/en/actions/reference/workflow-syntax-for-github-actions#usage-limits

There might be other reasons that would prevent migrating.
Last time I checked, secrets were not shared with forks. Are any secrets used?

[morozov]

Last time I checked, secrets were not shared with forks. Are any secrets used?

I believe this is by design on any public CI. No secrets are currently used.