On Wed, Nov 19, 2025 at 01:11:33PM -0600, Glenn Washburn wrote: > The two commit reverted in this series degrade the GRUB tests and must be > reverted. The goal of any testing system for a project should be to find > bugs. That goal is not served by reducing test coverage, which is what > both of those commits did. Just because some (influential) users have a > different objective, such as making the test harness not error so they > can check the "GRUB tests pass" box after building, does not mean that > GRUB should override the testing goal of the project. It should be fairly > obvious when reading the results of the tests that there were issues in > running some tests and why the issue occurred. The solution to this is > not to pretend there was no issue, and the testing system should not help > the user in pretending there was no issue (no matter how much the user > would like it). These two reverted commits do just that, among other issues. > Of course, if the test requirements are so onerous that no one runs the > tests, that's also an issue, and patches are welcome to improve this. > However, due to the nature of the GRUB project and thus the kind of testing > it needs, GRUB will always require a complex testing environment. And > test coverage should not be sacrificed to simplify this environment. > > I would also like to add that there is a history of certain large distros > submitting patches that are in the (percieved) interests of those distros > and not in the interests of the GRUB project or community as a whole. > In my estimation, mostly this has been due to ignorance as opposed to intent. > Regardless, I hope in the future, people working for these distros can take > off their distro hat and put on a GRUB hat before sending patches. Distros > contribute a lot of valueable work to GRUB and I would like to continue to > see high-quality patches that are in the interest of the project as a whole. > > Glenn > > Glenn Washburn (2): > Revert "tests: Skip tests if required tools are not available" > Revert "tests: Remove -w param from mkfs.hfsplus command"
Taking into account the cover letter and updated GRUB Developers Manual I think these two patches have to be reverted. So, Reviewed-by: Daniel Kiper <[email protected]>... Daniel _______________________________________________ Grub-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/grub-devel
