On Thu, Jun 15, 2006 at 03:57:05PM +0100, Manuel López-Ibáñez wrote: > Mike Stump proposed "a batch tester that > would bootstrap and regression test on 2-5 platforms for all patch > submitters post approval but pre-checkin." My point is that a batch > tester post-approval is just wasting reviewers time, which seems to be > one of the reasons why it takes so much time to approve patches. If > there is a need for a batch tester, then it should be pre-review.
Agreed; it makes more sense not to waste reviewers' time with patches that fail the bootstrap/regression test. If we do have a batch tester, it makes sense to run it before asking for review. Otherwise, the reviewer will have to diagnose portability problems that could have been found by the batch tester. For a bug fix patch, if it hits a reviewer after going through a system that shows it works on our five most important platforms, in most cases the reviewer can just give a quick "OK" and we're done. The reviewer might ask for a rewrite if the fix is done in a kludgy way, or there's a coding style problem.
