On Tue, Feb 20, 2001 at 10:41:27PM +0000, Piers Cawley wrote:
> Surely the todo stuff should always run the test and complain if it
> succeeds. That's why there's a distinction between 'todo' and 'skip'.
Yes, that's the way it will work.
I think there's some confusion with todo() because I keep describing
it in terms of my version of ok(), not Test.pm's. I pass in an
expression which evaluates to true or false:
ok( 1 + 1 == 2 );
whereas Test.pm likes to use a got/expect style (yes, I know there's a
special case for a single argument ok(), but its impossible to wedge a
test name into that and has no future).
ok( 1 + 1, 2 );
todo()'s interface will simply mirror the local version of ok() (with
extra bits for a reason and optional eval).
I'm doing alot of naval gazing at the moment trying to figure out how
best to reconcile these two interface as well as add in names for
tests. Test::ok()'s interface is currently streched to the limit
trying to adapt to all sorts of styles and situations, I don't know if
I can wedge anything more into it. I may have to branch off a new
testing module and leave Test fundamentally unchanged.
--
Michael G Schwern <[EMAIL PROTECTED]> http://www.pobox.com/~schwern/
Perl6 Quality Assurance <[EMAIL PROTECTED]> Kwalitee Is Job One
