Hi, On Dienstag, 17. März 2009, Lucas Nussbaum wrote: > That's why I'd like it to be confirmed that the rate of > false positives is more like 5% or 10% than like 90% or 95%, before a > link is added on the PTS. In my experience, it was more like 95%, > especially for upgrade tests. It seems you dont read what I write. Or you dont believe it. And only go with your old experiences. And you also dont seem to bother to check the current results. I've just checked _all_ failed logs on http://piuparts.debian.org/lenny2squeeze/fail.html and they are all real failures, with the exception that it's debatable whether modifying /etc/shadow is a bug or not. (Some people think added users should be removed, others not.) BTW, as you dont seem to bother to go to that website, for lenny2squeeze tests its currently 435 successful tests vs 14 failures. I've also checked all in http://piuparts.debian.org/squeeze/fail.html and those are real too. I havent checked all failures in sid (etoomany :) but here the ratio is also equally good: 4818 pass vs. failed-testing 175 (Sid is testing since a few days longer, thats why there are more tests being done.) regards, Holger, who also doesnt want to run around making noise about false positives...
Attachment:
signature.asc
Description: This is a digitally signed message part.