Re: What's cooking in git.git (Jun 2010, #03; Fri, 18)

From: Brian Gernhardt <benji@silverinsanity.com>
Date: Jun 19, 2010, 04:27 UTC
Message-ID: <74E790B7-BDC9-479B-A7E7-259ACB7E60B9@silverinsanity.com>
In-Reply-To: <7v7hlvyi50.fsf@alter.siamese.dyndns.org>

On Jun 18, 2010, at 8:34 PM, Junio C Hamano wrote:

Show 7 quoted lines

> * ab/tap (2010-06-09) 4 commits
> - We use TAP so the Perl test can run without scaffolding
> - Skip tests in a way that makes sense under TAP
> - Merge branch 'jc/t9129-any-utf8' into ab/tap
> - Make test-lib.sh emit valid TAP format
> 
> I was not sure why TAP is worth the trouble, and I still am not sure.

Having looked it over, using prove (a TAP harness) would allow me to do things like save which tests fail to use either while trying to fix a bug or performing a bisection.

cd t/ prove --state=save t*.* cd .. git bisect start HEAD good git bisect run 'make && (cd t && prove --state=failed t*.*)'

Also, the fact that it makes the perl tests be more "normal" is definitely in its favor.

~~ Brian