ietf-mxcomp
[Top] [All Lists]

Re: on per-user macros; and the IETF's role in a deployment campaign

2004-05-11 08:58:59

In 
<77D06015-A35F-11D8-BA73-000A95CA7FAE(_at_)dbc(_dot_)mtview(_dot_)ca(_dot_)us> 
Marshall Rose 
<mrose+internet(_dot_)ietf(_dot_)mxcomp(_at_)dbc(_dot_)mtview(_dot_)ca(_dot_)us>
 writes:

As a result, I think it would be better for us to either not touch the
problem domain addressed by SPF, or to create a standard that is
(largely) compatible with the SPF install base.


perhaps i wasn't in the room when the secret handshake was
disseminated, but i was unable to find reference to this in the
charter.

In case it isn't clear when I said "I think it would be...", this is
my opinion.


as sympathetic as i am to arguments concerning the installed base,
that is only one of many competing concerns for the group.

Yes, I definitely agree.  However, when two proposals largely address
the same problem area, with similar costs, the one with the
substantially larger install bases is very likely to win out.  As a
result, I think we need to develop something that is better than SPF
in some significant way, or create something largely compatible with
SPF, or stay moot on the subject.  I don't think just being a
standard-track RFC will count as being significantly better.

In many ways, SPF is hindered by its own install base also.  There are
lots of things I would change if I could.  For example, I think the
"include:" mechanism is badly named because it doesn't act at all as
if the other SPF record is included in the first.  A better name might
be "if-match:" or something.  However, I think it would be hard to
change this now, and it will only get harder as time goes on.


Looking back on the discussion of semantics and syntax, only Phillip
Hallam-Baker has proposed anything reasonably concrete that is
incompatible with SPF.  While I think some of the stuff he has
suggested has merit (e.g. open vs closed sets), I personally (my
opinion!) do not think this qualifies as being significantly better
than SPF.


We are close to being half way through the month that we are supposed
to resolve the issues of semantics and syntax.  If we are going to
develop a proposals that is both concrete enough to be completed in a
reasonable amount of time and significantly better than SPF, we better
start nailing down the details soon.



-wayne