So it would not be unreasonable to say that
bad a: arguments cause 32% of the failures -
bad ip4: causes about 26% -
bad includes cause about 23% -
bad mechanisms cause about 19% -
the rest is apparently insignificant
Interesting, the (relatively) "new" processing limits cause no
significant problems. The "bad includes" probably include bad
redirections, that's tough.
If that was a list with 6000 problems, was each record counted
once, or once per mail ? And what's the number of "no problem"
in relation to 6000 ? We probably need something like this for
a "lessons learned" + "implementation interoperability report"
(or whatever the name is) later when SPF is updated to PS. Bye