On Tue, May 14 2019, Johannes Schindelin wrote:
> Hi Ævar, > > On Tue, 14 May 2019, Ævar Arnfjörð Bjarmason wrote: > >> On Tue, May 14 2019, Johannes Schindelin wrote: >> >> > What would you think about a mode where random test cases are skipped? >> > It would have to make sure to provide a way to recreate the problem, >> > e.g. giving a string that defines exactly which test cases were >> > skipped. >> > >> > I am *sure* that tons of test scripts would fail with that, and we >> > would probably have to special-case the `setup` "test cases", and we >> > would have to clean up quite a few scripts to *not* execute random >> > stuff outside of `test_expect_*`... >> >> I think it would be neat, but unrelated to and overkill for spotting the >> practical problem we have now, which is that we *know* we skip some of >> this now on some platforms/setups due to prereqs. > > I understand, but I am still worried that this is a lot of work for an > incomplete fix. > > For example, the t7600-merge.sh test script that set off this conversation > has two prereqs that are unmet on Windows: GPG and EXECKEEPSPID. On Azure > Pipelines' macOS agents, it is only GPG that is unmet. So switching off > all prereqs would not help macOS with e.g. a bug where the GPG test cases > are skipped but the EXECKEEPSPID test case is not. It won't catch such cases, but will catch cases where a later new test assumes that whatever the state of the test repo it gets is what's always going to be there. In practice I think that'll catch most such issues. The other GIT_TEST_* modes assume similar non-combinatorial explosion failure scenarios. I haven't gone back through the test suite's commit history to try to dig for other cases, so perhaps this mode is premature etc.
