On 8/10/07, Joe Swatosh <joe.swatosh@gmail.com> wrote:
> On 8/9/07, David James <james@cs.toronto.edu> wrote:
> > On 8/8/07, joeswatosh@tigris.org <joeswatosh@tigris.org> wrote:
> > > Update tests that test for ranges of revisions to expect what is returned.
> >
> > Why did the test output change here? Was the original expected output
> > wrong? Do you know what revisions changed the test output?
>
> I seem to remember discussion on the list about changing the
> representations of ranges. I haven't been able to keep up with the
> everything that has been changing and I didn't go through the history
> to figure out which revisions caused the these particular Ruby tests
> to start failing.
>
> So the short answer is I think the original tests were written before
> the merge and range apis had settled down.
>
> A quick search through pburba's commit messages suggests that it might
> be r25836?
>
> If it is unacceptable to make these changes without referring to the
> revision that caused the regression in the log message, I will revert
> until I have time to figure out (and be certain) what caused them to
> start failing.
Hi Joe,
No need to revert. Could you update the log message to reflect your
comments above, though? A simple note which says that the test output
changed some time during merge tracking development is probably fine.
You should also mention in your log message whether or not you double
checked that the new behavior is correct.
Cheers,
David
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Aug 11 07:22:17 2007