Hi, On Thu, 1 Jun 2017, Stefan Beller wrote: > On Thu, Jun 1, 2017 at 4:40 PM, Junio C Hamano <gitster@xxxxxxxxx> wrote: > > Johannes Schindelin <Johannes.Schindelin@xxxxxx> writes: > > > >> Also, about the commit IDs. As long as the tests are consistent (i.e. they > >> use test_commit rather than plain `git commit`, or at least call > >> `test_tick` beforehand), the commit IDs should actually be identical > >> between runs and not depend on the time of day or the date. > >> > >> The only exception to that rule is when some previous test cases call > >> `test_commit` but are guarded behind some prereq and may not be executed > >> at all. In that case, the precise commit IDs depend on the particular set > >> of active prereqs. > > > > Good observation. The tests written such a way may make later > > introduction of new hash function troublesome, though (we already > > have tons of them, and it is already a hassle just imagining that we > > will have to migrate them X-<). > > > > And what you gave below is an excellent suggestion to even solve > > that future headaches. > > We had a discussion off list how much of the test suite is in bad shape, > and "$ git grep ^index" points out a lot of places as well. Maybe we should call out a specific month (or even a longer period) during which we try to push toward that new hash function, and focus more on those tasks (and on critical bug fixes, if any) than anything else. I also wonder how we can attract (back) cryptographic talent to help us avoid repeating mistakes when picking a new hash algorithm. So far, the only undisputable expert opinion I read was from the Keccak team, and I did not have the impression that their opinion had any impact on the discussion. Needless to say: I think it should. Cryptography is hard. We proved it ;-) Ciao, Dscho