Interesting article on ATM in Wired (of all places)

Since I ordinarily do nothing but bash the journal of hype and unreadable
blue-on-orange type, I figured I should express appreciation for Steve
Steinberg's "Netheads vs. Bellheads" article starting on page 145 of the
October issue. The ATM v. real networks battle is one that is becoming
increasingly important at gads of institutions, such as mine. In addition
to economic efficiency, there are civil liberties reasons to favor the
current packet-switched technology over circuit-switched bit-mangling. The
nexus on this technical front isn't as tight as it is on the encryption
front, but I'd urge people in positions of technical or budgetary
responsibility to line up with the good guys, i.e., us netheads. You don't
want an Internet controlled by ATM technology. It would make censorship,
wiretapping, and other forms of nastiness by government and other armed
thugs far too easy. I'm speaking only for myself, of course, and of course
there are technical reasons I'd prefer to implement gigabit Ethernet as
well. 

The leader "Why the net should grow up" in the October 19th issue of _The
Economist_, whose net reporting has improved markedly with the addition of
some new staff, may also be of interest. Some overlaps with the Wired
article, though the writers probably weren't technically savvy enough to
realize that they were talking about the same thing. More comments on that
when I find time. 

-rich

Rich Graves
>
>Since I ordinarily do nothing but bash the journal of hype and unreadable
>blue-on-orange type, I figured I should express appreciation for Steve
>Steinberg's "Netheads vs. Bellheads" article starting on page 145 of the
>October issue. The ATM v. real networks battle is one that is becoming
>increasingly important at gads of institutions, such as mine.

I liked this article too because it does anticipate something crucial
going on in the telecommunications/cyberspace arena.
however I think it's a huge disservice to cast the future of the internet in
terms of a black/white either/or situation of packet vs. 
circuit switching (you're already hot into the emotional rhetoric
by calling the latter "bit mangling").

who here believes that circuit switching is superior to 
packet switching or vice versa? it's a lame debate, like
arguing whether apples or oranges taste better. 

the future network is going to be a blending of both. it will be
a collision of the cultures mentioned in the article you cite.
it won't be a battle, it'll be a blending. both sides already
concede that the other side has advantages. already protocols
are being designed that incorporate both technologies. 

>I'd urge people in positions of technical or budgetary
>responsibility to line up with the good guys, i.e., us netheads.

give me a break. what you are essentially arguing about in the
two technologies is a timed, guaranteed bandwidth vs. packets that
arrive but without timing constraints. depending on the application,
either may be the appropriate choice. the network of the future will
support both. please, please do not try to polarize this simple
issue. there is absolutely no doubt that much technology, such
as audio/video broadcasting, needs "guaranteed bandwidth" as
offered by ATM. there is also no doubt that support thing kind
of guaranteed bandwidth is expensive and not superior for all
applications.

 You don't
>want an Internet controlled by ATM technology. It would make censorship,
>wiretapping, and other forms of nastiness by government and other armed
>thugs far too easy. I'm speaking only for myself, of course, and of course
>there are technical reasons I'd prefer to implement gigabit Ethernet as
>well. 

oh, brother. "circuit switching is the mark of the beast". I find
it rare to find technically adept people holding such irrational,
almost religious-like beliefs about technology. but it is highly
entertaining <g>

it never ceases to amaze me how much FUD is generated when a 
new technology is introduced. a bazillion people all cower
and scream, "gosh, could [x] mean the end of [y]?" apparently
this is far more fun than realizing, "gosh, [x] will find a 
niche ALONGSIDE [y], but neither will destroy the other".
this happened with Java and all the new languages, it happens
with the "network computer", etc. ad nauseaum. sigh!!
maybe it is a sign of a certain kind of intelligence that
can't simultaneously hold two ideas at the same time (like
walking and chewing gum? hee, hee)