Merry Christmas!

Just before Christmas, I sent an email to the TinyOS 2.0 working group about the involvement of the TinyOS community in the Zigbee Alliance. For those of you that don't know, there's a new version of TinyOS under development that changes almost all of the system interfaces. Part of this effort is to define an architecture to build sensor networking protocols. Zigbee is an industry-led alliance to standardize networking for wireless sensors.

My proposal had just one point:

Zigbee should be the first network protocol released for TinyOS 2.0

I detailed a few reasons why:

1. Zigbee is a standard.
2. TinyOS has received lukewarm, at best, recognition in the commercial sector. Embracing a standard protocol suite will bring attention to the project and interest from commercial users.
3. A standard protocol suite is valuable not just for commercial use, but also for academics.
4. A TinyOS-Zigbee implementation opens the door to improvements that the TinyOS community may propose to the Zigbee 2.0 technical working group.
5. Zigbee will help frame the design of TinyOS 2.0's networking architecture.
6. Think of the opportunities! "TinyOS inside" stickers for lightbulbs, thermostats, forklifts, trucks, toasters, microwaves, etc!

As of the writing of this post, there has not been a single response to my proposal. As per most of my suggestions, this very important proposal for saving TinyOS has been ignored and discarded by the infamous TinyOS "czars".

Read the full text of my manifesto in the TinyOS 2.0 working group archives

There are about 500,000 hits for "TinyOS" on Google compared to the almost 2 million hits for "Zigbee", yet there are only 9,000 hits for the combined search terms "TinyOS" and "Zigbee".

Well being in an equally controversial mood consider the following points:

1. Why does Zigbee need TinyOS? Using your example of light bulbs, thermostats etc., if they can be controlled using simple assembly routines or a C library that Chipcon or whoever supplied with their kit, why use TinyOS?

I think a larger problem here is the lack of documentation and training for people wanting to become involved with TinyOS. How many people post messages on tinyos-help wanting simple programming help or non-academic literature about TinyOS? I know Moteiv run training courses and consult but in order for TinyOS to gain more traction there needs to be more options (or one very large Moteiv :) ).

2. How does the community resolve the IEEE standards process against the open source nature of TinyOS? I checked the Zigbee site and the only way you can get hold of the specifications is agreeing that any use of the spec for non-Zigbee members is for non-commercial purposes. How does the community expect a company to adopt Zigbee-TinyOS if they can't use it for their purposes?

Furthermore, if an open source version of Zigbee existed how can the community cover themselves against patent infringement cases?

3. Focusing research efforts on Zigbee so that improvements can be made to the standard is a good idea, but would anybody on the Zigbee 2.0 Technical Committee listen?

Just a random observation:

I attended some ZigBee Alliance presentations on sensor networks in Sweden and they had text from your SP-SenSys paper as one of the slides (quote-unquote style) ... apparently they thought that your comments about/against ZigBee were very aggressive

How come a certain soft-corner for ZigBee now? If the TinyOS 2.0 community should read one email from you in a year shouldnt it be about implementing SP in T2?

I don't mean to be controversial, and I am not an IP lawyer, but I wonder if your "ZigBee-on-TinyOS" proposal would infringe patents? I am assuming that there are patents that cover ZigBee.

Does anyone know for sure?

Another item: I am preparing a proposal to develop a "communication system" for mobile medical sensors that would alert health care personnel when the patient's vitals go south. We decided to base our approach on 802.15.4, and are evaluating ZigBee and TinyOS for the rest of the stack. We are currently leaning toward TinyOS because it seems to have advantages over ZigBee in this application where the "health care provider node" may be mobile.

I think it's great to go after the "building automation" applications that ZigBee is focused on, but don't forget there are applications where ZigBee may not work well. Seems to me that ZigBee should try to emulate TinyOS :) But I like your idea on the stickers!

Jay