On Fri, 17 Sep 2004, Ken Williams wrote:

I think you have a couple misconceptions about cpan-testers. It's not some organization with an office and some testing machines, it's just regular people out there who have set up robots to test all the CPAN releases they see (or that interest them).

Actually, in the last couple of days I've gotten email from something that claims to be an automated testing service. (Sorry, but I deleted the email.)

Also, my hunch is that yes, Module::Install may be part of the culprit
here.  But without us seeing any of the output they sent you, it's, um,
hard to fix a bug if you can't reproduce it.

Touche'. :) In my own defense, the test report doesn't tell you what if anything the user did before running "make test". For example:

http://nntp.x.perl.org/group/perl.cpan.testers/144008

When every single test fails, there's a good chance that they didn't
install a required module. I guess I could create a special test case
(perhaps called "read_the_readme.t") that checks that prerequisites are
met.

----

Overall, I'm just frustrated. I really do work hard on getting good
releases, and I give very detailed instructions on (1) installing
prerequisites, and (2) giving me enough info to debug test failures.
But these folks seem to think that if it doesn't install automatically, it
must be a module bug.

Maybe I'll have Makefile.PL actually show them the README file and make
them confirm that they read it. ;)

David

_____________________________________________________________________
David Coppit                           [EMAIL PROTECTED]
The College of William and Mary        http://coppit.org/

"Government big enough to supply everything you need is big enough to take
everything you have ... The course of history shows that as a government
grows, liberty decreases." -- Thomas Jefferson

Reply via email to