On 26/06/2011 12:00, Barbie wrote:
On Sun, Jun 26, 2011 at 10:56:37AM +0200, Lee Goddard wrote:
> What is the concern over anonymous data? That it is not really
> anonymous?
Its the fact that an author has no ability to contact the tester to
understand why a test run has produced a FAIL report. It then becomes
an irrelevent report if we cannot verify the validity of the report.
My point of view is that I would rather see more results, regardless,
but I accept your argument.
> How about an opt-out when CPAN is configured?
...There are
many companies that would see this as a security risk...
And On 26/06/2011 12:03, Gabor Szabo wrote:
If anything then an opt-in would be better - more acceptable - but
I'd rather go for education.
I've been programming Perl since 1997 or so, and although the old cpan
testers script has been replaced or augmented by a number of modules, I
still rarely have the time to set any of it up. So although, as a
short-term contractor, I work in lots of locations, various locales, and
on a variety of OS and hardware, I very rarely send in test reports like
I did when I was a university. Education and information is not the
issue in this case, it's just a matter of time - and operator laziness!
If there was an option, when configuring CPAN, to send test reports, I
would check it in 90% of cases - the 10% being the paranoid clients.
Most of my friends and colleagues feel the same, though I'm sure they'd
not be bothered to even express a concern or desire as I do here.
Is there a reason that `cpan` should not be extended to facilitate this?
Cheers,
Lee
