On 5/4/10 4:48 PM, John Levine wrote: > > And indeed, this might be the (or, at least, an) answer to the > >> concern (except of course for ADSP assertions made too broadly > >> because it can't cover this scenario. > > I have to say that ADSP is having just the pernicious effect that I > feared it would. Its actual utility is very, very narrow, people > want it to do other things, so they start insisting that the rest of > the world redesign itself to match their concept of ADSP. We seem to > have learned nothing from the experience with SPF.
ADSP offers limited utility as a transitional mechanism for stricter acceptance, and as an alternative to individual relationships with email providers throughout the world. Predictable cases of unintended message loss that ADSP might cause can be mitigated with a third-party authorization mechanism. The mechanism would allow domains a means to grant specific exceptions for the third-party services being used. Lessons have been learned. Unlike a comprehensive IP address list of authorized servers, as with SPF: A) a hash label mechanism scales to any number of third-party services within a single transaction. B) modified messages will not be confused with those directly from the sender domain. C) authorization is limited to specific domains, and not for any message handled by a server. -Doug _______________________________________________ NOTE WELL: This list operates according to http://mipassoc.org/dkim/ietf-list-rules.html
