Re: Dynamic server ports

On 13 Dec 2001 thomas@inorbit.com wrote:

> Essentially, you connect to port 449 and send in a name from the service
> table such as "telnet" or "ddm" (case-sensitive and ASCII) and it
> returns a 32-bit binary port number for that service. To see it in
> action, just set up a comm trace and watch any PC connect. I have no
> idea how the same is handled on other platforms; it'd sure make a nice
> standard.

Well IMNSHO the use of port 449 for this stuff is stupid.  Want to connect
to the telnet server?  Use port 23.  Want to use other services?  Connect
to the appropriate port.  But wait, that is what 449 is supposed to be:
the appropriate port.  Why do we need a standard to tell us what is
already an existing standard?

Maybe you decide to move telnet off port 23 to help prevent abuse.  I can
just connect to 449 to find out where it went.  Or do a port scan.  So
what did we gain by using 449?  Nothing - just another firewall rule and
added complexity.

We already have a standard.  We don't need another one.

p.s.  I'm not upset with anyone (though it may sound that way).  I'm just
fed up with moronic Client Access that has this special need of port 449
when my other 5250 clients work just fine without it.  Kind of like the
Micorsoft paperclip:  fancy features I don't need or want and create
headaches.

James Rich
james@eaerich.com