On 5/10/13 5:29 PM, "Justin Mclean" <jus...@classsoftware.com> wrote:

> Hi,
> 
>> I see. But shouldn't we rely on the mustella console output?  That's what we
>> did when in SVN.
> At the time you run the test sure, but you may forget which one failed after
> fixing a bug or two and git status is a lot lot faster than rerunning the
> tests.
That presumes the test failed with a bitmap failure which isn't always the
case.  And each run cleans up bad.pngs from prior runs beforehand.

I suppose we could try taking failures.txt out of the gitignore and changing
mini_run to delete it if there are no failures, but even then, it will only
show you if your last run failed.
> 
> Given that no one is likely to check these files in and they indicate an issue
> than needs to be fixed I think it's safer to have have them in .gitignore.
You can remove them from gitignore if you want, but just realize that their
existence is not a foolproof way of knowing you have failing tests.
> 
> Thanks,
> Justin

-- 
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui

Reply via email to