[e2e] Re: Cannara's views - ATM

[David P. Reed]

At 12:21 PM 4/17/01 -0400, Craig Partridge wrote:
>And one strong reason is unbundling.  If the goal is (a) to make it
>easy to track individual "wires" or "subcribers" and (b) do it in a way
>that doesn't force you to open up your management subsystem to others,
>ATM is very attractive.

The function is important, I agree, and ATM hardware was an off-the-shelf 
solution you can kludge to perform it (though why frame relay wasn't good 
enough is beyond me - maybe all those companies with ATM switches needed 
some market after the original one died?) So all those wonderful "features" 
in ATM turn out to be useless, it's just a way to provision a path for 
packets through the local access provider to some ISP's POP.  And why not 
throw PPPoET on top of it just to make it look like a dialup circuit so we 
don't give the users the idea that "always on" is possible?

Maybe the "next ATM" (aka mandated telco industry standard) entering the 
chute will adopt some version of the hourglass model, rather than the 
typical "kitchen sink" or "soup to nuts" model typified by ISDN, then ATM, 
and now 3GPP and Bluetooth.  Sure would save on industrywide NRE and 
ex-post-facto kludgery.




- David
--------------------------------------------
WWW Page: http://www.reed.com/dpr.html