Re: [PATCH 1/2] t3200-branch: test setting branch as own upstream

[Jeff King <peff@peff.net>]

On Fri, Feb 28, 2014 at 04:28:38PM +0900, Brian Gesiak wrote:

> I would be in favor of using test_i18ngrep, but it seems like this
> test file overwhelmingly uses test_(i18n)cmp, even when inspecting
> stderr output.

We generally prefer "cmp" checks to "grep" checks, because they are more
rigorous. However, when testing human-readable output which may change,
sometimes being too specific can simply make the tests brittle and
annoying. Using a forgiving regex that matches keywords can be helpful.
So there's definitely some room for judgement.

I think what you posted as v2 looks OK.

> Making double-sure that all tests pass when run with "sh -x" seems
> like a larger endeavor.
> 
> Of course, I'd be happy to submit several patches if there's support
> for such a change. But as Peff points out it will be a large diff.

Yeah, I don't think it's worth the effort.

If you feel like continuing on this series, converting the warning()
into a die() would be a much more productive use of time (but if you
don't, I do not see any reason not to take the patches you've posted).

-Peff