[Tinyos-host-mote-wg] [Tinyos-2.0wg] 07/20/2005 Telecon Notes

Kristin Wright l.kristin.wright at gmail.com
Wed Jul 20 11:37:59 PDT 2005


-------------- next part --------------

Present: Ben Greenstein (UCLA), Vlado Handziski (TU-Berlin), Jonathan Hui (UCB), 
	 Phil Levis (UCB), Martin Turon (Crossbow), Kristin Wright (UCB)

Notes from 07/20/2005 TinyOS 2.0 Working Group 

07/20/2005 Agenda
-----------------
- UARTs
- status

07/20/2005 Discussion Notes
----------------------------------------------------------

Pre2 freeze
------------
- phil: freeze dates: Friday midnight, California time
  - exception: lib/Serial
  - vlado: going without eyes radio (kevin going back to washington, taking
    a month break before, many at tu-berlin on vacation for the next month

UART
-----
phil: interface summary
  - will still send message_t, but data within might be any
    of a number of formats
  - to figure out where data starts (offsets could be different),
    use a simple protocol
    -> what type? type 4.
    -> what offset does type 4 start on? n
  - interfaces checked into lib/serial
- phil: UART architecture summary
  - top layer: data packet, aware of message_t
  - middle layer: actual protocol, ACK packets, ex; exposes a
    byte interface to data component above (does not see all
    protocol machinery underneath)
  - bottom layer: actual encoding which, in case of current stuff, control
    bytes, escape bytes, etc.
- means that protocol component sees only proper data bytes that
  have 
- vlado: makes perfect sense; seems very similar to problem they
  have with byte radio; has a payload and footer. underlying components
  do exactly the same that he's outlined; would be interesting 
  to compare details
- vlado will look at phil/ben; phil/ben to look at vlado's
  - perhaps it could be generalized; in the near future probably too
    much work to unify
- martin: is it still using the HPL interface?
  -phil: nope; there's a notion that there's a platform independent layer
    - pointed to interfaces
- martin: in 3-layer breakdown, where would xlation of legacy packets fit in?
  - phil: would exist above those layers; micaz would 'register' legacy
    packet so they would be dispatched to that code; what is sitting
    above that is something that translates it 
- vlado: do your interfaces include handling configuration, baud rates,etc. 
  or using existing HPL UART interfaces?
  - use the existing HPL
- martin has a few more questions that will be handled offline w/Phil





-------------- next part --------------
_______________________________________________
Tinyos-2.0wg mailing list
Tinyos-2.0wg at Mail.Millennium.Berkeley.EDU
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-2.0wg


More information about the Tinyos-host-mote-wg mailing list