Roger Moser wrote:
it should say "Positional modifiers apply only to the
mechanism they immediately follow."
Yes, that would be clearer.
Better would be if the scope of a modifier is defined in
the description of that modifier. And not generally for all
present and future modifiers.
After seeing Wiliams's and Guy's variations of this idea I
still think that Mark's solution is the best. I'm a hardcore
KISS extremist.
you want to specify the scope after every mechanism if it is
not the default scope.
No, I don't believe in the scope idea. PRA is snake oil, it's
a broken predecessor of William's "eh", and "eh" is a marketing
gimmick for those who desperately want an anti-phishing option.
And v=spf1 has no scopes, so there it's op=pra,eh,dk,whatever
For v=spf2 the situation is similar: Whereever the MAIL FROM
passes, the 2822-From should also pass. Therefore only some
additional IPs for the 2822-From (resp. PRA) are "non-default".
Just add a sc=pra behind all corresponding mechanisms. Or use
an "include:_spf.pra.stuff.example sc=pra" if that's shorter.
It's really a spf2.0 resp. v=spf2 problem. If some relevant
scopes are identified where your notation works better we can
use it. But at the moment there are no other relevant scopes,
there's only the PRA rubbish. I don't believe in PRA, so why
waste more time with it than minimally necessary ? Bye, Frank