RE: Email or Something Else

"David W Morgan" <[email protected]> Fri, 20 Feb 2004 20:49:35 -0000
Newsgroups gmane.mail.ng
Message-ID <000801c3f7f3$0c6ed6d0$6401a8c0@cyber>
Chris, I think that you may a point.  We certainly seem to be starting
to define an extensible service, with many conflicting requirements.

I had thought along the lines of a virtual data truck with many
"application" using the delivery services of a transport service.  Those
applications could be wide ranging but with a common require to transfer
data being user entities.

Kind regards
David

David Morgan
Strategic Systems

-----Original Message-----
From: [email protected] [mailto:[email protected]] On
Behalf Of Bonatti, Chris
Sent: 09 February 2004 12:10
To: [email protected]
Subject: Email or Something Else


I've been watching the postings here, and reading (as much as I
can stand ;-) all the diverse requirements.  The breadth of the
requirements expresed is really staggering.

- We want low throughput mobile devices, but we want to be able
to send large files.

- We want strong user authentication and tracing, but we want to
support anonymous traffic.

- We want simplicity, but we want flexibility and extensibility.

- We want backward compatibility, but we want a clean slate.

- We want instant delivery, but we want our messages saved till
we get back.

I find myself wondering if we aren't scoping requirements for
more than a single application.  Maybe we need a several
cooperating, or at least related, applications that together
cover the full spectrum of desired requirements.  Say we defined
several different "applications" with a compatible framework,
message formats, naming scheme and some means to aid MUAs in
determining which "mode" to use to for a particular message.
Perhaps a set of applications might include:

- A push-mode store-and-forward style transport;

- A push-the-heading and pull-the-body transport;

- A connectionless IM-style transport.

I'm not trying to move ahead into systemm design, per se.
However, I think the idea that we might be working on a SET OF
applications has considerable bearing on how we sort
requirements.

Thoughts, contrary or supportive, are of course welcome.

Best regards,
Chris