On 03/02/2017 07:13 AM, Duy Nguyen wrote: > On Wed, Mar 1, 2017 at 10:11 PM, Michael Haggerty <mhagger@xxxxxxxxxxxx> wrote: >> By trial and error, I found that the test succeeds if I comment out the >> "for_each_reflog()" test. By having that test write its results to >> `/tmp` where they won't be deleted, I found that the problem is that the >> `actual` results are not sorted correctly: >> >> refs/heads/new-master 0x0 >> refs/heads/master 0x0 >> HEAD 0x1 >> >> I don't know why it's so Heisenbergish. > > It happens consistently on my other laptop. And yes it looks like > sorting order problem, probably because of the underlying file system. > I did wonder about that at some point but never asked. We do not > guarantee any sorting order in the for-each api, do we? The for-each-ref iteration is defined to be sorted by refname, but it's true that the for-each-reflog order is undefined. So it was wrong for me to say that the results are not sorted correctly. The problem is that the actual output is ordered differently than the "expected" results, which the test incorrectly considers to be an error. Michael