On 1 Sep 2013 03:22, "Antoine Pitrou" <solip...@pitrou.net> wrote: > > > Nick Coghlan <ncoghlan <at> gmail.com> writes: > > distutils itself is nigh impossible to work on, since it is > underspecified, and > > the fact it was added to the standard two years before unittest still shows in > > its test suite. > > I don't have any code coverage numbers, but distutils has a sizable test > suite (and has had since Tarek started working on it): > > $ ./python -m test -v test_distutils > [...] > ---------------------------------------------------------------------- > Ran 191 tests in 1.481s > > OK (skipped=14)
Agreed, Tarek did great work in creating a reasonable test suite, but the combinatorial explosion in possible build configurations is such that a rapid response cycle for issues is far more feasible than hoping to completely prevent regressions (at least, not until the PSF can afford the infrastructure to automatically do test installs of the whole of PyPI for new versions of the build tools across multiple platforms, which I don't see happening any time soon). We should check how well that suite runs with setuptools imported, though. Given the originally shared maintenance by Tarek, I believe it should be similar to the distribute (and hence now setuptools) test suite, but it's worth double checking that. Cheers, Nick. > > > Regards > > Antoine. > > > _______________________________________________ > Distutils-SIG maillist - Distutils-SIG@python.org > http://mail.python.org/mailman/listinfo/distutils-sig
_______________________________________________ Distutils-SIG maillist - Distutils-SIG@python.org http://mail.python.org/mailman/listinfo/distutils-sig
