[ih] History of Network Virtual Terminal (NVT)

[Dave CROCKER]

On 1/22/2010 9:00 AM, John Day wrote:
> That was the brilliant insight, even more so than the NVT. Everyone else
> saw this problem as an assymmetric terminal to host protocol. What made
> Telnet unique and so useful was seeing it as symmetircal.


The 'symmetry' of having either side able to initiate an option assertion, 
including both simultaneously, was quite nice, especially since it eliminated 
timing sensitivity.

The larger construct that this was an entirely symmetrical connection between 
two terminals had the problem that /options/ didn't clarify the 'side' they were 
meant for.

While running a TCP/IP product team, we had a bid for a government contract that 
included a site visit to test our stuff.  We get to the demo of telnet and the 
gov't guy said he wanted to see echoing, or somesuch, /from the terminal to the 
host./

We explained why that made no sense.  He calmnly nodded and said he entirely 
understood the problem this requirement imposed.  He then pointed to the 
relevant specification text and it clearly said that this was all symmetrical 
and said his job was to verify conformance, not merely conformance of the stuff 
that made sense...

So, with everyone very clear that this was one of the more silly bits of code 
that the team would ever write, the lead programmer quickly added the code 
needed to make the terminal side process a DO echoing from the server.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net