On 19 Dec 2007, at 02:05, chromatic wrote:
Sure - but I'd have expected that to be perceived as a specific
problem in an otherwise valuable system. It's not a rational reason
to
right off automated testing as a whole surely?
That depends on the ratio of useless to useful results.
Presumably the false negative rate achieved by the best modules is a
measure of how noisy the smoking system is. Given that the cleanest
modules regularly get a <1% FAIL rate over many tens of reports it's
not a huge reach to suggest that any module should be able to get
close to that. So I'd expect an author who was seeing a lot of
failures to look around on CPAN a little and observe that other
people's tests are doing better than theirs and then maybe wonder
whether it's their code that is at fault.
And I have to say being admonished by an auto-responder always stirs
my piss a bit :)
Agreed, if that means on your side of the pond what I think it means
on this
side of the pond.
"slightly irritate"
--
Andy Armstrong, Hexten
