[Top] [All Lists]

Re: SNPP - A Clarification

1993-09-03 13:56:42
In terms of simplicity: SNPP is no less complicated than SMTP.  Further,
in terms of clients, there are dozens of different SMTP clients, all of
which hide the protocol from the user.

Yes, but none are specifically designed with paging in mind, are they?

In terms of store-and-forward: if you think paging networks are
reliable, you probably haven't used one recently.  In the SF area, pages
take on the average of 5 minutes to be delivered, and one out of every
three of my pages gets dropped.  Because paging networks are simplex,
there is simply no way to ensure reliable behavior.  The best you can do
is report that IXO said it accepted the page.  Beyond that, anything
else is a fiction.

Actually, it depends on how you implement your paging network.  I am
working with a major nationwide paging firm.  If your pages are taking
5 minutes on average to be delivered, you should probably check them
out.  Numeric pages average 16 seconds, nationwide, and alpha pages are
sent within 1 minute.  All systems are monitored at the transmitter, and
should there be a problem, pages can be held and retransmitted.

Still, just because there may be problems with one part of a system, isn't
particularly good justification for breaking other parts.  I can at least
guarantee that information reaches the IXO.  I cannot justify degrading
reliability by rationalizing that paging delivery is unreliable.  Too many
people do that with too many things as is.

So, I think we are back to: why do we need another protocol (and
associated infrastructure), when we already have a widespread SMTP

Because I cannot see how SMTP is going to address all of these concerns.  If
I am missing something, perhaps you can help clarify the missed point.

Thanks for your input.


<Prev in Thread] Current Thread [Next in Thread>