Stamp code

[<aah@nofs.navy.mil>]

Tom mentioned that Chris or I would be good candidates
for fixing up the Stamp.  I wouldn't mind working on
the project, but I'm going to be gone all of October
and would rather work on the Linux code anyway.  However,
I've worked on PICs quite a bit, as well as dozens of
other micros for control purposes, and can offer a
couple of suggestions.
  Keeping ASCII mode is fine, though what I strongly
recommend is an even more ASCII mode than what Tom
is using.  I typically implement a crude command
parser with perhaps a two-letter command followed by
arguments.  This way you can hook a laptop up to the
Stamp and communicate in the field without any special
program other than Hyperterm or equivalent.  I'd
also suggest if the Stamp has sufficient resources to
have it do shutter timing for short exposures (say 10 seconds
or less) to relieve this function from the host computer.
The time delay for sending open/shut commands becomes
important as you go to shorter exposures.  If Tom is
concerned about data dropouts over the serial line, then
implement a checksum scheme on all commands (if the checksum
is there, use it; if not, ignore it, so you can use the
terminal/laptop for debug).
  The Stamp is the heart of the system, and it seems to
be working pretty well as-is.  So perhaps it is not worth
a lot of effort to recode.  However, recoding might not
be all that difficult since you have working examples of
all of the necessary functions.  It *will* be difficult if
you do not have a working system in front of you.
Arne