Re: [PATCH 2/2] t3404: add a test for the --gpg-sign option

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Jul 2, 2016, 07:38 UTC
Message-ID: <alpine.DEB.2.20.1607020933080.12947@virtualbox>
In-Reply-To: <xmqqpoqxz6kp.fsf@gitster.mtv.corp.google.com>

Hi Junio,

On Fri, 1 Jul 2016, Junio C Hamano wrote:

Show 14 quoted lines

> Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:
> 
> >> that has a substring '-S"' in it to ensure that the codepath to
> >> parse --gpg-sign= on the command line of "rebase", and to the
> >> message we see here are working correctly, without actually checking
> >> if GPG is invoked at all, or if it is invoked the key given by the
> >> option is correctly passed to the invocation?
> >
> > Exactly. I want to test --gpg-sign even when there is no gpg executable
> > available.
> 
> The other side of that coine is that even when GPG is available, we
> do not see if it is invoked correctly at all.  That was what I found
> disturbing.

Okay, I see now.

However, is it not better to have even a limited test than none at all? Granted, we still would not know whether rebased commits would be signed properly. But if my trivial test case fails, we would still have a strong indicator that something is broken, with a very convenient way to debug it. And that is what a regression test suite is all about, isn't it?

Of course I agree that it would be very nice to have a test at a later stage that does exercise GPG if it is available. But would it really be so terrible to have a (simpler, not as complete) test that is exercised *also* when GPG is *not* available?

Ciao, Dscho