Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!ames!rex!ginosko!uunet!philmtl!philabs!ccnysci!dan
From: dan@ccnysci.UUCP (Dan Schlitt)
Newsgroups: news.admin
Subject: Re: Warp speed Mr. Scott!
Message-ID: <3023@ccnysci.UUCP>
Date: 12 Sep 89 14:30:47 GMT
References: <1989Aug21.124002.11054@robohack.uucp> <1059@aurora.AthabascaU.CA> <1989Aug29.130347.10673@robohack.uucp> <3924@ditka.UUCP> <1989Sep6.004958.29593@robohack.uucp> <1090@aurora.AthabascaU.CA>
Reply-To: dan@ccnysci.UUCP (Dan Schlitt)
Organization: City College Of New York
Lines: 35

In article <1090@aurora.AthabascaU.CA> lyndon@auvax.uucp (Lyndon Nerenberg) writes:
>
>What I think we are all starting to agree on is that the labels currently
>in use don't reflect reality. My argument is that perhaps the idea of
>labels should be scrapped, and real numbers used instead. What these
>numbers represent is what must be decided. As I've said before, I think
>there are two likely candidates: the cost per kilobyte to send over the
>link, or the average delay to deliver a message over the hop (including
>time spent in the queue waiting for a poll).

It has been a while since I looked so my memory may be failing me, but
I thought those labels were numbers.  Like in
	#include	DEMAND	nnn
and that you could just use numbers if you like or adjust them by
adding or subtracting defined fiddle-factors.  Since any monotonic
sequence can be mapped into any other in a unique way, it seems that
the real argument is about the factors which should be used to
determine the relative weights for the links.  Two such factors are
mentioned above but there are others such as reliability of the link
which you might want to consider.  It seems that the previously agreed
upon factors are nolonger agreed to and that with the increase
rerouters and internet links they may not be relevant.

What is needed is an agreement about the factors.  The current small
integers which underlie the current system are quite adequate to carry
out the agreement once it is reached.  But remember that we are
talking about a uucp network --- there aren't any RFCs and there is no
cenralized authority to enforce decisions --- so any agreement will
continue to depend on the authors of the individual map entries just
like it does now.
-- 
Dan Schlitt                        Manager, Science Division Computer Facility
dan@sci.ccny.cuny.edu              City College of New York
dan@ccnysci.uucp                   New York, NY 10031
dan@ccnysci.bitnet                 (212)690-6868