"Roger Moser" proposed:
Chris Haynes wrote:
SPF should define (even at this late day) a new modifier for spf1: "pra"
with a single legal value "yes".
Instead of "pra=yes" I would add a scope modifier "sc=mfrom,pra" with
"sc=mfrom" as the default. The "sc" modifier may appear multiple times and
affects all terms except "all" on the right side in the same SPF record. The
"all" mechanism is independent of the scope.
For example if someone is doing SES, then the SPF record could be:
v=spf1 sc=mfrom,pra a mx sc=mfrom exists:%{l}._ses.example.com -all".
Roger
NO. I have the following objections to your proposal:
1) spf1 IS about mailfrom - always. We should not entertain the thought that
the record could be used by pra but not mfrom. We are _not_ offering a "United"
or "Unified" capability in spf1. We are offering the possibility of additional,
experimental uses. "Scope" is a concept we should reserve for any _later_ work.
2) We should not make it read as if mail-from testing within SPF is an option
like PRA testing. That would be bad politics!
3) Your multiple use of the "sc" modifier rule breaks the existing spf1
(implied) syntax. Modifiers are currently position-independent, so your example
could not work as you probably intend. For yours to work we would have to
re-define sf1 modifiers as being position-dependent, which would break existing
records.
I'm offering a tactical accommodation in spf1 which does no harm to existing SPF
records and tools.
You are proposing changes which, IMHO, perhaps belong within some future, united
protocol.
Chris Haynes