"wayne" == wayne <wayne(_at_)midwestcs(_dot_)com> writes:
Personally, I would *really* like to see this changed to use the
following ABNF:
version = "spf2." ver-minor "/" ver-scope *( "," ver-scope )
ver-minor = 1*DIGIT
ver-scope = "pra" / name
Where name is already defined as:
name = alpha *( alpha / digit / "-" / "_" / "." )
I like this proposal; I think it is more flexible than what we
currently have in the drafts.
I think I'd prefer ver-token to ver-scope -- at this point there's no
need to suggest a particular use for the extention mechanism, even by
implication, and there may be other uses we haven't yet though of that
will make us regret the ver-scope name.
The idea would be that only records with 'pra' present as a ver-token
would be considered by the Sender ID draft. Yes, a future standard
can always stop Sender ID from interpreting the record by changing the
version string incompatibly (eg by bumping the major version) but this
approach gives more flexibility in terms of the way that the Sender ID
can be extended...
-roy