ietf-dkim
[Top] [All Lists]

Re: [ietf-dkim] Srsly.

2008-01-23 18:19:45
On Wednesday 23 January 2008 19:39, Dave Crocker wrote:

J D Falk wrote:
Jon Callas wrote:
SSP is an important, valuable, *optional* part of the email
infrastructure.

This is a very important point.  When the draft says "MUST," an
experienced i-d reader will know that it actually means "x must do y in
order to comply with this specification."  That's not so obvious for
other humans, especially when pretty much all of the conversation on
this list also has the inherent assumption that SSP will be everywhere
for everybody.

There will be entirely valid use cases for which SSP will not be useful,
and may even be damaging.  Ellen's is one of these: she's probably going
to have to change her entire business model as SSP adoption grows.  In
our discussions -- especially with people who aren't fluent in the
ancient & dusty IETF vernacular -- we MUST remember that SSP will never
and can never be any stronger than a SHOULD.

JD,

1. Yes, folks often forget that the premise to a standard is "IF you
embrace this standard, THEN various normative assertions apply.  IF you do
not, then they do not."  So all the musting and shoulding are strictly in
the context of those who have chosen to adopt the specification.

2. Your second point says that there should be an 'applicability statement'
for SSP, to clarify the scenarios in which it makes sense to use and the
ones it does not.


I think he said there will be cases where it's not useful.  I don't seen 
anything in that statement where he says we should engineer out exactly what 
those cases are.

Scott K
_______________________________________________
NOTE WELL: This list operates according to 
http://mipassoc.org/dkim/ietf-list-rules.html