Re: Extract Git classes from git-svn (2/10) (was Re: Fix git-svn tests for SVN 1.7.5.)

From: Michael G Schwern <schwern@pobox.com>
Date: Jul 19, 2012, 00:11 UTC
Message-ID: <500750AA.3060802@pobox.com>
In-Reply-To: <20120718105835.GA32215@dcvr.yhbt.net>

On 2012.7.18 3:58 AM, Eric Wong wrote:

> I agree with everything Jonathan said (and thank him for taking
> the time to point you in the right direction).

Thanks, you guys have been very nice to my flailing and failing. I'm going to back off and send out a sort of overview email so we can figure out how best to chunk this up.

Show 11 quoted lines

>> +++ b/t/Git-SVN/00compile.t
> 
>> +use Test::More tests => 2;
> 
> I prefer not declaring test counts and using done_testing() instead.
> done_testing() is favorable to me in at least 2 ways:
> 
> * done_testing() closely matches the behavior of the existing
>   sh-based test suite in git (which calls test_done)
> 
> * maintaining test counts leads to unnecessary merge conflicts

Yes, I concur 100%.  So much that I went back in my time machine and added
done_testing() to Test::More!  Also I killed Hitler, so now WWII ends in 1945.
 Things seem to have turned out for the better.

I love it when people advocate my features back to me. :) I didn't use done_testing because I didn't know your stance on using non-5.8 core versions of modules.

> Skipping the tests on old versions of Test::More (< 0.88) is acceptable
> to me (especially since integration tests provide the real coverage
> already).

It is very easy to bundle an uninstalled copy of Test::More, probably easier than putting in the code necessary to check for it and skip it. A lot of Perl modules do it. The usual thing is to put it into t/lib/ and add "use lib 't/lib'" to the tests. I don't see any reason why that basic technique wouldn't work here, with some minor changes to match the Git test suite.

I can help you with that, but I'd like to get through this SVN 1.7 fix first.

-- 
Being faith-based doesn't trump reality.
	-- Bruce Sterling