> Subject: [RFC-PATCH 2/2] t9001: adding --quote-mail option test We write messages at imperative tone, hence s/adding/add/ Tom Russello <tom.russello@xxxxxxxxxxxxxxxx> writes: > From: Tom Russello <tom.russello@xxxxxxxxxxxxxxxxxxxxxxx> Please use the same identity for email and commit to avoid this line. > --- > > > diff --git a/t/t9001-send-email.sh b/t/t9001-send-email.sh No diffstat again? Splitting a patch series as "code first; tests after" is not a good idea IMHO. When questionning the behavior of To: Vs Cc: in the previous patch, I would have appreciated having tests in the same message, to check that the tested behavior was indeed the one I was reading in the code. OTOH, having one patch to introduce "--quote-email populates To: and Cc: headers", and then another one for "--quote-email quotes the message body" would make the review much easier. Oh, BTW, this obviously lacks documentation (Documentation/git-send-email.txt). And that's one reason why the diffstat is useful: one can reply "this lacks tests and doc" before even reviewing the patch. Regards, -- Matthieu Moy http://www-verimag.imag.fr/~moy/ -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html