On Wed, Sep 3, 2008 at 1:25 PM, chromatic <[EMAIL PROTECTED]> wrote:
> Is the purpose of CPAN Testers to report whether a distribution will install
> on a given machine or whether its tests pass on a given platform?

Not exactly and not quite.

The purpose is to report whether a distribution can successfully
navigate a configure, build and test cycle.

We care about this because we presume that if it can do so, it can be installed.

As I suggested in my long essay, the problem in the CPANPLUS/Build.PL
case for an author is that knowing that CPANPLUS is broken is
essentially useless.  And hearing about it again and again on every
new release of a distribution is pointless and annoying -- making the
author less inclined to read the next FAIL report.

"Screw it -- it's probably just CPANPLUS again"

So that is why I think we have to find a way to prevent these kinds of
reports or let authors filter these out if they find themselves
afflicted by it.

-- David

Reply via email to