[SensorNetArch] current HotOS draft

[Cheng Tien Ee]

+ Section 2, 1st para: "... which are functional components, their 
mechanisms... " mechanisms seems to imply implementation. In first para 
perhaps mention of protocol can be included. I feel that implementation 
should not be part of the architecture, so I would leave out the last 
sentence.

+ Section 2, 2nd para: I would say that the architecture must define the 
interfaces, services, and can define protocols. Because one of the 
primary objectives is the interoperability of components in the stack.

+ Section 3: I would add a short para on why IP needs to define a packet 
format and not SP. The reason would be that IP is both the narrow waist 
and a network protocol that does routing and addressing etc, SP does 
only the first thing, so there's no need for the others. I do agree with 
Prabal that at least some mapping for demux needs to be specified.

+ Section 3, para that begins "For example, it is clear from...": I 
think "... it is part of the SP interface" is supposed to be " ... is 
part of the SP service"?

+ Section 4.2: I think just saying TDMA itself doesn't directly imply 
that physical broadcasting is not possible. I think that's the case if 
the MAC coordinates when a receiver should be awake and receiving, 
otherwise turning the radio off to conserve power. If the TDMA is 
concerned only with transmission (i.e. schedules when a node can speak), 
but all the nodes are still turned on all the time, then we still can 
have broadcasts.

Typos:
+ Section 2, 3rd para: "In constrast to the Internet architecture"
+ Section 4.1, 1st para: "variety of techical challenges"
+ Section 4.3, 1st para: "power management to moviate why"
+ Section 4.3, 2nd para: "Energy contraints are a defining"


Phil Levis wrote:

>We had a discussion over the use of the term "implementations." Everyone
>agreed it was the wrong term. After some debate, there was consensus
>that we should use the OSI term of "protocols." I will change the text
>to reflect this.
>
>The second half of 4.3 and 5 still need a rewrite. I am going to do this
>now.
>
>The current draft is enclosed. If you have comments, PLEASE SEND THEM TO
>THE LIST BY 2:30PM.
>
>
>Phil 
>
>------------------------------------------------------------------------
>
>_______________________________________________
>SensorNetArch mailing list
>SensorNetArch at Millennium.Berkeley.EDU
>https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/sensornetarch
>