[1751] in Commercialization & Privatization of the Internet
Re: what is a high speed network
daemon@ATHENA.MIT.EDU (William Manning)
Mon Dec 16 13:59:09 1991
From: bmanning@is.rice.edu (William Manning)
To: CK.MAN@rxg.xerox.com
Date: Mon, 16 Dec 91 12:56:02 CST
Cc: com-priv@psi.com
In-Reply-To: <"16-Dec-91 13:35:05 +1".*.Charles_K._Kuhlman.MAN@RXG.Xerox.com>; from "CK.MAN@rxg.xerox.com" at Dec 16, 91 4:35 am
CK.MAN@rxg.xerox.com
>
> This is my last comment on the subject (thank goodness!)
>
You really don't think you'll get off that easy do you?
>
> A good deal of the negative mail I got refered to how proud folks are of their
> high speed links with other systems, and how dare I bring up the issue of the
> end user. NOT ONE NOTE REFERED TO HOW A SYSTEM WAS WORKING TO GET THE SPEED TO
> THE END USER. I think there is a problem here.
>
> The end user is the life blood and rationale for the existance of a network.
> Systems admininstration activities must take a back seat to the needs of the
> end user. Systems administrators are NEVER justified in saying `my users are a
> pain in the ---'.
>
Ok, I think I get the point. I would like some clarification on "END-USER".
Just who is this creature? Is it the human being that is restricted to
the information flow through the senses OR is it the machine handling lots'o'
tasks, RPC calls, cycle stealing and who knows what else, while waiting for
its puny human to type in the correct response?
Can we assume that the general case of the end-user today is on some local
area network that has some bandwidth associated with it and a lower speed
link to some wide area network?
Or are you refering to the single machine that is NOT connected to any network
other than the wide area service?
Bandwidth to the "end-user" or "end-station" discussions ought to clarify this
point. I assume that you (& I) are looking at the case of the traveling sales
drone with attached dialup service participating in a global video-conference
from the laptop, mean while closing a deal via EDI transactions, and writing
the GAN (Great American Novel) in yet a third operations space. Note that the
operative word is --dialup--. Or is the case of the, for example,
omnipresent ISDN ports on every desktop and payphone, with the cellular mesh
not far behind. How many ISDN channels do we need for each "end-user"?
Are there cost benefit tradoffs? Does every "end-user" REQUIRE the same bandwidth
to EVERY end-point in the net? I guess that it all depends on how the end-user
is defined. Any thoughts out there?
Regards,
Bill Manning
713-285-5415
R.U. (o-kome)
