On Wed, Jul 26, 2000 at 04:33:46PM -0600, Tom Christiansen wrote:
> >Are you speaking of the idea as a whole, or simply what tests we
> >should run on code examples?
>
> All the pod crud. It's bogus. Simple run each code segment
> through perl -wc. No more.
This means no "=for example" syntax? How are we supposed to detect
what is code and what is not? C<> is not consistantly used and is
often skipped for large code blocks.
Even if we were to get use of C<> consistent, we do not necessarily
want to run it all through a syntax check. The basic problem being
that many of the fragments will not stand alone and need some prefix
and postfix code to get them to run. As I mentioned, too many false
negatives will cause the test to rapidly be viewed as annoying and
bogus.
However, I am often wrong. Could you write a program to test the
feasibility of automating the testing without modifying the existing
POD? Maybe run it over a few of the existing perl man pages.
--
Michael G Schwern http://www.pobox.com/~schwern/ [EMAIL PROTECTED]
Just Another Stupid Consultant Perl6 Kwalitee Ashuranse
"None of our men are "experts."... because no one ever considers
himself expert if he really knows his job."
-- From Henry Ford Sr., "My Life and Work," p. 86 (1922):
