Closed andk closed 10 years ago
I've filtered column three for values < 0.75
s/0.75/-0.75/
Thanks! Uhh... wow! Yes, this is very interesting and useful.
I was hoping they'd mostly be the fault of broken Test::* dependencies. Nope... they all seem to be a bit different. Poop.
MediaWiki-Bot-v3.4.2
is almost certainly not failing because of TB2. It is almost certainly failing because I have bugs in my test suite and/or the live site we test against has gone funky again.
Suggestion/request: Rerun it with our CPAN IDs on it. Make it easy for authors to find their names and problems without having to scan the listing of 100.
I agree with petdance - the CPAN IDs would be useful.
URI::Escape::XS appears to be a false positive. I can't get it to fail on 1.5.0a2.
UNIVERSAL::DOES has been fixed by 78f0c977ca3b79a7a33592557b77b4f14ca61f32
The following list is compiled from results of the regression testing done on analysis.cpantesters.org. In column 2 there is the R^2 denoting the quality of the finding. The closer it is to 1 the less disturbing influences have been identified, and the more likely it is that the measured effect is relevant. In column 3 there is the Theta, the direction and amplitude of the influence. I've filtered column three for values < 0.75 and column 2 to values > 0.35. Arbitrary decisions. 103 lines seemed more than enough to me and indeed the last items in the list are already lots of influences mixed in.
Hope this is interesting to you.
Oh, and splitting into individual tickets is left as an exercise:)