Daily Updates

09/17/2002

Today was another productive day.  Professor Looft will be happy to know
that our day started at 6 AM, when we left to drop Ethan off at the
airport.

We met with Mike again today to go over more task to action sequence
mapping.  In addition to the specification for "Map the Sky" that we
created yesterday, we now have action sequences for long-range
communications, solar sail actuations, and gathering data.

We also discussed how the streams between the low-level and the simulation
would function.  Command ids, sequence numbers, and the concept of an
"immediate mode" simulation were all discussed.  Mike is going to look
into feasible ways to handle command/response synchronization.

After specifying the "Map the Sky" task yesterday, we realized that our
representation of action sequences would have to change.  Instead of
storing a static list of actions for each task, we decided storing
self-executing action sequence classes would be wiser.  This method will
allow each action sequence to handle things such as looping.  Also,
parameter population will occur within the action sequence, making that
process a lot easier.  We updated the code base to include such a feature.

The Agents Technology Group had their weekly meeting again today.  The
remaining ANTS guys were on hand to give a status report.

We ended the day with writing.  The methodology and design chapters were
updated.  We'll continue to update these tomorrow morning, and present
them to Professor Brown when he arrives.

That's all for now.

ANTS2