Re: [colorforth] DARPA takes aim at IT sacred cows

["Samuel A. Falvo II" <kc5tja@xxxxxxxx>]

On Friday 12 March 2004 11:41 am, John Drake wrote:
> "The packet network paradigm probably needs to change," Gibson said.
> "I'm not advocating throwing out the Internet Protocol completely, but
> we must absolutely have some mechanism for assigning network
> capabilities to different users and that capability has to scale to
> large numbers of devices automatically. The commander wants to be able
> to send a message and have it delivered, completely, accurately and on
> time."

The problem is, packet-switched networks are the most reliable way to 
accomplish this goal.  Circuit-switched networks certainly do have their 
advantages, make no mistake.  But the problem with circuit switching is 
very simple: for every user who gains access to the network resources, 
there is one user who is denied (assuming full network capacity), which 
thoroughly defeats the purpose of DARPA's initiative.  Packet switching 
networks typically don't have this problem.

What needs to be discarded, plain and simple, is CSMA/CD (e.g., the 
Ethernet way of doing things).  CSMA is problematic no matter what, and 
CD doesn't do a thing to actively avoid collisions.  CSMA/CA is far 
superior in every respect, and even better still, is DAMA.  DAMA (and 
its very close relative, token ring [with or without early token 
release]) offer near 100% network utilization factors, while avoiding 
collisions completely.  DAMA has the advantage over token ring in that 
it can prioritize node traffic, and therefore, can implement quality of 
service rivaling any ATM network.  Token ring is somewhat easier to 
implement though.

--
Samuel A. Falvo II


---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com