Re: [PATCH 0/9 v2] difftool: teach command to perform directory diffs

Thanks.

I do not know a difftool user, and I wasn't paying close attention to the discussion but I recall these points raised and I do not recall the resolutions:

 * In [1/9], use of pass_through would mean 'git difftool' must be fed the
   options for difftool and then options meant for underlying diff
   machinery.  Is this limitation still there?  If so, isn't this a
   regression?  Shouldn't it at least be advertised to the users a lot
   stronger in documentation?

 * In [4/9], you remove the .exe extension when running git, which was
   there since the beginning of difftool 5c38ea3 (contrib: add 'git
   difftool' for launching common merge tools, 2009-01-16).  I think it is
   safe but are Windows folks OK with this?

 * In [6/9], the exit code in the failure case seem to be modified from
   that of the underlying "git diff" to whatever croak gives us.  Is this
   a regression, or nobody cares error status from difftool?  I personally
   suspect it is the latter, but just double-checking if you have
   considered it.

 * In [7/9], difftool--helper declares SUBDIRECTORY_OK, but there doesn't
   seem to be any inclusion of git-sh-setup in this script, and the patch
   does not have any effort to prepend $prefix to paths relative to $cwd.
   What good does the variable do here?