[standards-jig] About outgoing packets.
ehz4 at yahoo.com
Sat Mar 15 02:02:44 UTC 2003
--- Peter Millard <me at pgmillard.com> wrote:
> Yes, so it keeps clients simple, and it allows all
> packets from "serverA" to
> "serverB" to be multiplexed over a single TCP
> socket. Otherwise, a single host
> would have to have N sockets for it's own clients,
> plus M sockets for all other
> clients, where M could potentially be huge. The
> current way, Q is the # of
> sockets for other servers, which will at least be
> several orders of magnitude
> smaller than M.
Thanks for the info. Actually, this leads into my
next question. If serverA and serverB are exchanging
all the messages, aren't they sending one message
after another? I'd imagine that in certain situations
a long queue could form. From what I understand, this
is one of the problems with SMTP.
The authentication issue of sending messages through
the server makes sense. I suppose if I don't need
authentication I could just act as a server, assuming
I'm that worried about the overhead of routing
Do you Yahoo!?
Yahoo! Web Hosting - establish your business online
More information about the Standards