I didn't explain myself very well. The intent is that both manycast and
pool use similar algorithms, but each is separate and implements the
data gathering scheme in different ways. I fully understand multicast is
not ubiquitous, not even outside our department here, and the only wide
area experiment where it definitely was ubiquitous was the now retired
CAIRN network once funded by DARPA. Once upon a time long ago we had a
private T1 tail circuit to CAIRN and did all kinds of interesting
multicast stuff with real time video, audio and whiteboard before
anybody even heard of it. Interestingly, some of my students work for
Comcast and they tell me multicast is alive and well within their
network, just not available for customer use.
So, the intent is that the pool concept and manycasdt concept should
learn from each other.
Brad Knowles wrote: