[ NNSquad ] Re: Comments by American Consumer Institute

[Richard Bennett <richard@bennett.com>]

The devil's in the details, Barry. Everybody agrees that some sort of 
response to congestion is necessary, because everybody (except Bob 
Frankston) recognizes that congestion is inevitable. But there are 
several responses that might be considered reasonable:

1. Drop packets willy-nilly
2. Drop packets from heavy users more than light users
3. Drop packets from heavy streams more that light streams
4. Drop packets from heavy applications like file transfer more than 
light applications like web VoIP.
5. Variations on above that stop streams altogether.
6. Variations on above that delay packets rather than drop them.
7. Other stuff that I'm not remembering at the moment.

I suspect that readers may find some of these policies "unreasonable" 
but I can assure you a case can be made for each of them based on 
maximum utility.

It certainly is true that some NN advocates argue that congestion would 
never happen if ISPs would get bigger pipes, and there are others who 
argue that bandwidth hogging never happens. And there are those who have 
argued multiple positions, such as Free Press (at one time, they said 
all streams should be degraded equally, and later said heavy users 
should be degraded more than light users; take your pick.)

It's not the job of the NN Squad to determine which of the above 
policies are virtuous and which aren't; rather, it's to see which ISPs 
are conforming their actual traffic management policies to the policies 
stated in their terms of use.

At least, that's the way I see it.

RB

  [ In my view, it is completely reasonable for us to discuss the
    pros and cons of these various policies and how they interact
    with the broad scope of traffic management operations.  That's
    part of analyzing the overall situation in a reasonable way
    that will be most useful for the range of interested
    observers.  "Raw data" presented without some form of reduction
    and comparative analysis is of only limited value to most
    persons, especially since most ISPs' terms of use are written in
    such a general fashion as to make point-by-point conformation
    analysis largely impossible based on raw data alone.

       -- Lauren Weinstein
          NNSquad Moderator ]


Barry Gold wrote:
> Brett Glass wrote:
> > ...This ConsumerGram provides a set of facts on which there seems to 
> > be no reasonable basis for disagreement, shows how these facts are 
> > important in resolving the network management debate,  reviews the 
> > main elements of the positions of the contending parties, and offers 
> > a consumer welfare perspective on the issues and their possible 
> > resolution. 
>
> I believe the usual term for these "facts on which there seems to be 
> no reasonable basis for disagreement" is "spin".
>
> A cursory examination of their argument shows the use of a strawman: 
> the assertion that NN advocates want everybody to be able to use 
> unlimited amounts of network capacity.
>
> In fact, people on this list have repeatedly said (and I believe the 
> law is clear in allowing) an ISP is permitted to discriminate among 
> customers based on quantity of usage.  They can charge by the packet, 
> by the byte, they can give you up to X MB per month at a fixed rate; 
> if you use more than X they can (a) charge you for the overage (as 
> many cell phone plans do), (b) cut you off (as would be the case with 
> prepaid cell plans), (c) turn down the speed on your cable modem, 
> allowing you to continue accessing the net but with significantly 
> slower down- and uploads.
>
> Also note the use of the term "reasonable network management 
> practices," which Robb Topolski just recently tore to bits.  I think 
> the term for this is "circular reasoning" or "begging the question" -- 
> in it's original meaning (assuming what you want to prove as a given 
> in your "proof").
>
> Either that or it's intentional, which would fit under the "big lie" 
> (a propaganda technique, not just a fallacy).
>
> Also note the term "bandwidth hogs," a pejorative which suggests that 
> those who oppose the writer's proposition are bad guys who should 
> therefore be ignored (argumentum ad hominem).
>
> A quick check of the ACI website doesn't reveal a lopsided 
> institutional bias, but the author of this paper is definitely using 
> "spin" techniques to make his case.  I don't know if this indicates 
> that he is unable to think more clearly, or if he knows his argument 
> won't stand up under rigorous analysis.
>
> I suspect the latter.

--
Richard Bennett