David M. Doolin
doolin at ce.berkeley.edu
Tue Jun 29 19:20:43 PDT 2004
Philip Levis wrote:
> On Tuesday, June 29, 2004, at 04:43 PM, David Gay wrote:
>> I think that Phil is arguing (and I would agree) that we could have
>> both if we only broke our monolithic components into smaller,
>> replaceable pieces. You can still have the monolithic component (it's
>> a configuration which assembles the smaller pieces), but you can
>> use/replace the smaller pieces too.
Agreed as well, exactly so.
But I think there is assumption here that the
application programmer has an intimate understanding of
the hardware details and system code encapsulted by all the
smaller bits and pieces, and that the application programmer
is expected to do this wiring into larger components as well.
If the application programmer is not familiar
with the hardware/system, how can this barrier be lowered?
This is perhaps a very tough question to answer. But
progress in this direction could help lower the
"PhD" factor required for deploying a TinyOS-based
More information about the Tinyos-devel