-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Wayne Schlitt wrote:
Something that I've been pondering for quite a while is what should
happen if you have an SPF record with a redirect= modifier, but the
target domain is malformed, non-existant or has no SPF records?
I'm pretty sure the current language, and that of mengwong-spf-0[01],
calls for the result to be "None", since the redirect= modifier acts
pretty much as if the process is starting over.
Should redirect= be treated the same as include:?
No.
Should the result simply be Neutral rather than None?
No. Giving nominal results such as "Neutral" on "redirect=foo.bar" would
be _even_more_ unnatural than giving "None" on "include:foo.bar". We
should not try to anticipate the receiver's reaction to specific result
codes and then base the mechanisms' and modifiers' behavior on that. That
would be all backwards. *sigh*
"redirect=" should always work as if the SPF check started over. This
follows from the principle of least surprise, i.e. it is what the user
expects.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFCinLJwL7PKlBZWjsRAmdjAJ93W4rLQI3y3UBSeRYmEFTOycxgkgCfavfv
gt9j11FnFQ9+dOnTwvpj+cw=
=Z/NE
-----END PGP SIGNATURE-----