[net2-wg] draft TEP 124

[Philip Levis]

On Feb 8, 2007, at 7:10 PM, Omprakash Gnawali wrote:

>
> I have committed some edits to TEP 124 based on our discussion from
> today's meeting. It should be OK for tomorrow's interop meeting with
> Sun. Let us know how it goes.

I sent them the version from yesterday, and am giving 15 or so slides  
on the protocol. I'll let them know that the document has been  
updated. I'll read through 124 today and give comments.

I was thinking that it might be good to revisit 118 and 119 (the  
programming interfaces to dissemination and collection). Getting them  
to community review would be a good step forward. David Gay has noted  
that there are some inconsistencies in power management which we  
should handle.

We might also want to visit the issue of how turning network  
protocols on/off affects the AM layer. It's clear that the 1.x  
approach of deep imperative operations (I turn on collection, I turn  
on the radio. I turn on dissemination, I turn on the radio. I turn  
off collection, I turn off the radio and dissemination is unhappy)  
isn't the right approach.

Right now you have to turn on each network layer and the data link  
layer separately. Way back when, OSKI proposed keeping a bitmask of  
services using an abstraction and only turning it off when they were  
all off. I'm not certain this is the right approach, but we should  
revisit the question in order to give some sort of meaningful on/off  
control abstraction.

Phil