[Jean Delvare] > My original patch had a test for the new --no-index option in > test/formats.test, which I can't see in CVS. Was in forgotten, or ruled > not interesting?
[Andreas Gruenbacher] > I didn't consider it important. Shall I add it nevertheless? Yes, please do. In my opinion, the whole point of a non-regression test suite is to test every individual feature so that we can hopefully notice when a change breaks any of them. In that respect, I don't think there are features which aren't important enough to deserve a test in the suite. If breaking a given feature isn't considered a problem, we probably would have to wonder why that feature was added in the first place. I really do like the test suite that comes with quilt. Not many projects include a non-regression test suite. As a developer, I find it very comfortable to be able to verify that my changes do not break anything, and I would encourage contributors to quilt to always add a test for every new feature or function they add, and to use the test suite themselves to ensure they have not broken anything. Including a test in submitted patches also helps understanding what the patch does, methinks. Thanks, -- Jean Delvare _______________________________________________ Quilt-dev mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/quilt-dev
