real-time control, proposed road map.

[Chris Albertson <chrisalbertson90278@yahoo.com>]

Michael, Rob and other TASS real-time control fans,

What do you all think of this set of tasks.  I think
it will get us from where we are to a state of the art
control system without to much pain and re-write.

1) Get "tait" Michael's TASS control program and
reorganize the code and Makefile so that the command
interpreter is separated from command execution.  All
command execution is done in a library (I mean a real
dynamic loadable library as in "libTASStait.so".)

2) Modify again so that it will read commands ONLY
from a socket. not from a file or a keyboard.

3) Write a very simple program that reads from a keyboard
or file and sends output to a socket.  This combined with
#2 above gets back the original functionality but also
allows this #3 program to be replaced at a later date
with something better. (see #7 below.)

4) Toss out Michael's command interpreter. (Sorry Michael)
and replace it with a Tck interpreter.  I would keep Michael's
commands and syntax except that now we get the addition of
a complete language.  This assumes Step #1 is done.

5) Modify program to make use of Rob's new kernel level
driver.  

6) Modify my Real time driver in the same way as #4 above.
(Sorry Chris)

7) Modify my real time driver to send commands to Michael's
driver via the socket interface (#2 above) and rip out
the code that
talks directly to the STAMP via the serial port.  My driver
then becomes a "high level layer" that translates "high
level commands" like 'Go to ra=12.45 dec=23.44' into a
series of low level motor control commands that Michael's
program accepts.  It also handles image display and tasks.
-- 

--
   Chris Albertson             
   chrisalbertson90278@yahoo.com
   Redondo Beach, California
   home: 310-376-1029
   cell: 310-990-7550