[MUD-Dev] Languages

[Jeff Kesselman]

At 08:03 PM 5/19/97 PST8PDT, you wrote:
>[Jeff K:]
>Yes, but with cable modems, a polling system is used, so there are no
>collisions and backoffs. 

No collision, but the bandwidth is still shared acorss ALL users in  agiven
period of time.  Its kind of like a data party-line.

> He
>seems quite confident that cable modems really are a workable solution.

Workable? Sure. But they've been touted as "dirt cheap T1" and they aren't,
not by a long shot.  As an alterantive to ISDN for those who DONT need to
send large amounts of data out, they may be okay.

>The big problem for them, and for anyone else trying to bring the internet
>into the home, is IP addresses.

Not a biggie. Theres INET-2 aroudn teh corner. meanwhile, why cant they use
dynamic IP like everyone else? When you turn it on, it requests an IP
number.  When you turn it off it frees it up.  or even better, ion activity
it gets a number (basicly like dial-onb-demand) and after a set period of
inactivity it relainquishes it.



>Yes, but I wouldn't count them out yet. The above-mentioned company, for
>example, is going at it the right way. They are stringing fibre all over
>the place, as fast as they possibly can. They are getting high-capacity
>routers and placing them throughout the cities, and they are actually
>trying to forecast expected usage and plan and build for it. If they had
>the physical plant in my area of the city (not expected until the fall),
>I'm sure it would have been unavoidable for me to have one! :-)

SOudsn quite different from American cable TV companies who are lookign to
leverage existing networks.  Even "fiber to the curb" is proving to be a
slow process of installation here.

JK