> From the *end users* perspective all the *end user* does is polling
>
> There is no point in inventing artificial scenarios and impractical
> technicalities.

I thought we were having a technical discussion about ArgusConnect
and their technical vision for health messaging. I didn't realise that
how the "end users" thought the "magic" worked was really a
key design criteria.

> My point is that message *delivery* by email works. It is time proven, robust,
> scalable, available, and not a single good reason to needlessly replace it.
> It is a universally useful transport vehicle for whatever message format in
> unstable and unreliable networks

As much as I admire the engineering feat that is SMTP email it has
numerous reasons why it _may_ not be suitable for health care
messaging, and why its _possible_ that a better system could be
designed. Off the top of my head I'd go with
1) horrendously difficult to configure SMTP servers (sendmail.cf anyone?)
2) based on a store-and-forward premise that maybe isn't necessary
   anymore - sure the retry/resend mechanism is great, but are there
   really many emails that actually _relay_ through a completely
   unrelated host.
3) lots of legacy baggage (UUCP..)
4) weak delivery contract (certainly not a strong as
   dumping the message directly into a webservice that will immediately
   respond with an "accepted message" or "rejected")

Are you saying that the designers of SMTP wouldn't do anything
different if asked to design messaging in 2006? More specifically,
if you gave them the constraints that apply in health care, do
you think they would not design things differently.

Which is not to say that I disagree totally with you about
email messaging. You are right - it's proven, reliable, and
widely deployed. But it doesn't hurt to discuss
alternatives.

Andrew
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk

Reply via email to