Folks,
JL> I checked around and there are indeed a lot of them. Any scheme that
JL> doesn't support wildcards will break a lot of existing mail systems
JL> and make it impractical to deploy useful new ones.
I'm glad this case analysis was posted and that a discussion thread has
ensued.
It is probably worth being clear about the role of such a discussion:
It is useful for assessing some of the near-term impact of work and
possible ways the work can operate.
However some of the postings in the thread seem to have a tone along
the lines of "if we can toss out all of the cases, then we do not need
to support wildcards". That line of thinking is folly.
Wildcards are part of the DNS design and service. They provide an
essential mechanism for handling scaling.
To design a new use of a service that breaks an underlying
characteristic of that service is a good way to fragment the service,
over time, and to fill its software with complexity due to special
cases. It is often, also, a good way to break the service more
substantially, because of unforeseen effects.
Large-scale systems that operate with a large, long-term installed base
of users have their own inertia. We need to respect the subtleties that
such an operation embodies.
Caution needs to balance the innovation.
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>