ietf-mxcomp
[Top] [All Lists]

Re: MARID compatibility with SPF records

2004-07-18 12:07:22

Folks,

MO> At what point is it safe for receivers to assume that mybikeshop.com's 
...
MO> I do not know how big this problem is, but given that there are 10s of 

Networking standards that require "assumptions" do not work very well.
The point behind a standard is to specify things explicitly.

With respect to retaining syntax but changing semantics, here is a
simple example:

I offer service for "free".

Previously, everyone has used the word "free" to mean "there is no cost
to the consumer".

However for anyone offering services using the alice-in-wonderland
standard, the word "free" will mean that the consumer is obligated to
pay US$ 1,000.

I choose to conform to the alice-in-wonderland standard.

The question to the reader is:  How the consumer to know whether I
conform to this standard and, therefore, whether the consumer's use of
my service will cost them US$ 1,000?

Now apply the same question to new-vs-old semantics on a DNS record.


d/
--
 Dave Crocker <mailto:dcrocker(_at_)brandenburg(_dot_)com>
 Brandenburg InternetWorking <http://www.brandenburg.com>
 Sunnyvale, CA  USA <tel:+1.408.246.8253>, <fax:+1.866.358.5301>