[MUD-Dev] Re: Telnet Echo OFF/ON commands and other thingies..

Bruce Bruce
Thu Oct 1 21:24:45 New Zealand Daylight Time 1998


From: Chris Gray <cg at ami-cg.GraySage.Edmonton.AB.CA>:

>[Bruce Mitchener:]
>
> >To turn echo off: `[255, 251, 1, 0, 13]
> >To turn echo on:  `[255, 252, 1, 0, 13]
>
>I understand the first three bytes of these (IAC, WILL/WONT, ECHO), but
>what are the NUL and the CR for? Do some telnet clients require those?


Woops.  The evils of blindly copying something.  I'll pass this along to the
people that work on ColdCore.

>My testing (with a fairly small set of telnet clients) shows that
>some do not respond to this correctly. They signal back that they
>will start echoing again, but don't do so.


Can you tell me which clients had this issue?  I'd like to pass that along
also for test cases.

>Note also that if you try to do any telnet commands like this, you should
>be prepared to either pay attention to, or discard, the replies you
>should get back. Typically [255, 253, 1] / [255, 254, 1].


ColdCore doesn't implement a full telnet state machine.  There isn't
anything that would make it much different from normal beyond the attitude
that such a thing doesn't belong in the generic core since most text-mode
clients don't support much.

My opinion on this differs somewhat in that I'd rather see a server that
sent out all data encoded in some manner, and when people needed
telnet-style access, they'd connect to an auxillary server which rendered
the content from the encoded format into the traditional flat text format
and handled the basics of command parsing.  This server would handle the
telnet option negotiation.

Is this an appropriate place to carry on discussion of client capabilities
to match the new capabilities of servers?  Is there interest?  Erik Ostrom's
mention of TWin here didn't seem to gather a whole lot of interest.

- Bruce






More information about the MUD-Dev mailing list