[e2e] Bandwidth Estimation workshop

[Michael B Greenwald]

   Thu, 02 Oct 2003 08:41:09 -0400
   "David P. Reed" <dpreed at reed.com>

   Look, guys.  I'm a pragmatic fellow, not a pedant.   However, my pragmatism 
   is focused these days on helping the public understand how to make policies 
   related to communications networks.
   
   The confusion between bandwidth and bandwidth adds a HUGE problem to this 
   space, not to mention the confusion between broadband and broadband, 
   narrowband and narrowband that result.
   
   The number of lawyers and economists at the FCC who understand this 
   distinction is tiny.   Congressional staff do not either.   

I believe you but I am surprised.  Is the distinction so hard to
explain?  If the discussion about units doesn't enlighten them, what
about the fact that often (all other things being equal) it's
desirable to minimize one and maximize the other?  (This won't explain
the difference, but it may make it easy for them to remember that
there are two distinct notions).
But enlightening congressional staff is probably off-topic for e2e so
I should return to the main point.

Having two separate terms for bandwidth and bandwidth will not aid in
explaining the difference in meanings.  It will only remind people
that there are two distinct meanings.  On the other hand, as I said,
(a) there are already many cues (e.g. units, min vs. max) to help people
remember that, and (b) having two terms is no guarantee that people
won't consider them synonyms (e.g. weight vs. mass)).

That said, I agree that a change that "merely" reminds us that two
meanings exist is valuable.  The main value of this reminder is to
people (including myself, mea culpa) who are occasionally guilty of
sloppy speech (i.e. using the term "bandwidth" in a context where the
listener/reader cannot disambiguate between the two meanings),
although we are perfectly aware of the distinction.  Two distinct
terms would force speakers/writers to be more precise in their speech
(but it will not help people who are imprecise in their thinking).

So while I agree with you in the substance of your argument (we would
benefit from having two distinct terms), I disagree about the degree
of importance this issue has.  I don't view the overloading of
"bandwidth" as an indictment of computer scientists or network
engineers as being guilty of sloppy thinking, and two distinct terms
may not help as much as you hope.  In other words, I don't think the
world (w.r.t. to "bandwidth" terminology) is so bad now, and I don't
think the world will be *that much* better if we change, so it seems
like an argument that reasonable people can agree to differ on, and
not something that marks computer scientists as voraciously ignorant.

							       Which leads to 
   a situation in which lobbyists for incumbents are makiing up "technical 
   facts" which have no bearing on reality.
   
   It may be impossible to stop the misuse of bandwidth as a term by the 
   jargonic digerati - I agree.  But context is not sufficient when you are 
   talking about systems-in-the-large.  Context is only sufficient for 
   disambiguation when focusing on microworlds where one deals with narrowly 
   abstracted problem spaces.