On Fri, Jun 11, 2004 at 01:26:36PM -0600, Wiggins d Anconia wrote:
> > On Fri, Jun 11, 2004 at 02:00:58PM -0500, Andy Lester wrote:
> > > Seems to me that if T:H is passed a test file, it's not unreasonable for
> > > it to expect at least one test.
> >
> > No, not unreasonable. But I think it's also not unreasonable to reserve
> > empty file to mean "no tests yet, ignore".
> >
>
> Couldn't an empty file indicate the inability for the installation
> process to write a file, which should be a failure case, and certainly
> one that needs to be caught?
Well, this failure should have been caught by the build system, but yes,
this is a possibility. If an empty file is too error prone, maybe we
could have a special output "no tests" that T::H would respect. Then
T::I would create a file containing
print "no tests\n";
This would be fine with me, though it has some feeling of overdesign.
:-)
> ok(1);
I prefer to eliminate extra noise. The situation I'm in is, I just
started using T::I, so only a few modules have any tests, and I would
see dozens of spurious "ok" lines for untested modules.
Andrew
