Markus Staab

the xdebug approach used here does only protected agains stack-frame endless loop (endless recursions). it does not protected against endless loops, which we also see when running infection on the...

Any opinions/feedback whether it is worth finishing this PR?

do we have some timing data at hand to detect which mutations ran into timeouts previously and auto-enable the xdebug flags for these? maybe via junit report files? ---- if...

Just leaving a note for myself: over the weekend I worked in PHPUnit itself and noticed that the PHPUnit result cache stores runtimes of test, which it uses to sort...

> not sure I got you. We use `junit` but it's produced by PHPUnit, while mutations timeouts - is completely on our, Infection side, so I don't think `junit` can...

Is missing a test

would it help if `$mixed - 1` would get a benevolent `(int|float)`? I think we did something similar recently

the error at https://github.com/pmmp/PocketMine-MP/blob/8dedbb747108c45b7534b8ac6f05e8464ee8eb22/src/crash/CrashDump.php#L229-L234 looks like it might be useful handle `IntegerRangeType` or `getFiniteTypes` to get arround the false positive ----- but maybe its just https://github.com/phpstan/phpstan-src/pull/3220#issuecomment-2220631819 ---- alternatively you could...

github actions run with 2 cores on the default instance. whether its a measurable difference need to be proved though.

@theofidry maybe phpcs should have some cpu core autodetection ;)