Re: [PATCH] Several tests: cd inside subshell instead of around

From: Jonathan Nieder <jrnieder@gmail.com>
Date: Sep 7, 2010, 05:19 UTC
Message-ID: <20100907051922.GP1182@burratino>
In-Reply-To: <7vlj7ep2ch.fsf@alter.siamese.dyndns.org>

Junio C Hamano wrote:

> Jonathan Nieder <jrnieder@gmail.com> writes:

Show 6 quoted lines

>> I like it.  Affected test scripts:
>> ... (many) ...
>> and probably some others.
>
> For the record, I _do not_ like it at all.  Is it worth being that
> draconian and say each test must start at $TRASH?  What do we gain by it?

To clarify, what I meant was something like "I like it, but reality does not seem to permit it now".

Show 6 quoted lines

> After we vet and apply Jens and your "turn 'cd there && do this && cd ..'
> into a subshell" series, and the "modernize indentation style" series,
> perhaps we should stop and think?  In an ideal world, it would be really
> nice if each "test_expect_success/failure" in a single script were
> independent from each other (if we did so, GIT_SKIP_TESTS=4321.4 starts to
> be more useful), but realistically, I don't think it is worth the effort.

Ah, our emails crossed. I would rather that tests that affect the state be marked as such, nothing more. The "always return to $TRASH" rule is a lot stronger than that, unfortunately. What I really liked about your suggestion is that the "tests must start and end in $TRASH" rule could be checked with very little effort. Unfortunately, as the list of scripts illustrates, that rule doesn't describe what our current tests do.