Yes, I guess it's possible to argue convincingly for each one of those
alternatives.
My personal preference for a default would be the second I think, but
that is mostly based on its similarity to the behaviour of software
I've used in the past. I tend not to like the first option as much
because
Michael Bedward a écrit :
> In my working copy of trunk I've added AndConst and OrConst to the
> coverage processing operations by some simple copying/hacking of
> Martin's code. Would it be useful to add this to svn ?
Probably, but there is an open question that I have not yet resolved (which is