[darcs-users] Re: testing patches that change the test pref
mark at summersault.com
Thu Mar 24 01:09:14 UTC 2005
On 2005-03-23, Phil Frost <indigo at bitglue.com> wrote:
> No, because the changes I made to my testing framework and the change in
> the test pref are an atomic change. One is not valid without the other.
> I think the solution is that when testing a patch set, obtain the test
> command from the result of applying the patch set, not from the current
> (unpatched) repo.
[ Thinks more. ]
OK, I agree this seems reasonable enough.
Consider: If your record update affects test scripts, the 'test'
that's run to make sure the tests are good is testing the /new/ version
of your test, not trying to somehow test your new test against an
Thus, I think it could be considered consistent to update the
functionality as you suggest.
I'll file a bug report.
More information about the darcs-users