[om-list] simulating time/place and context

Luke Call lacall at onemodel.org
Sat May 26 09:59:41 EDT 2001


How about this for ideas on our model's handling of changes to data in 
time/place (lending itself to simulations or handling data that isn't 
true for all times/places), and context (such as what name someone would 
use for a thing, given time, place, or topic of discussion):

         1) simulating time/place changing could be done by the entities 
themselves knowing when & where they and/or their attributes are valid, 
by storing time/place information in the attribute data (like where Mark 
has some time info in the relational data model we talked about).
         Then, when the time/place changes in the way the system is 
viewing the data, the entities are notified by some event notification 
loop which sends them the change event or simply calls a time or place 
change method, and they respond by checking the internal info about what 
goes w/ what time/place and when the time/place requires it, switching 
to different "current" pointers for the values of their properties (or 
possibly instead returning different data for those values or displaying 
themselves/behaving differently, according their their new state). To 
optimize this loop a bit, we could have this loop or a separate loop 
track which objects have provided a datetime for the "next relevant 
change time", that indicates the next time interval on which the 
notified objects would care about a change, so that we only notify them 
if the time threshhold for a simulation update or has passed.
         So now, given the new time/place, the entities "are" different 
and their state and behavior reflect that. (I need to get a personal 
copy of design patterns so I can revisit the "state" pattern here and 
see if it's relevant--my usual one is at work. And I may be showing my 
ignorance about known best practices for writing simulations.)

         2) dealing with context could be done by having the database 
data simply reflect a new time/place to--so that the getter methods 
return the data that fits the situation. In this case the concept of 
"place" might have to include the idea of which person's mind we are in 
(a "context" or belief system). Mark I haven't thought about how this 
would fold with your schema entity yet.

Luke






More information about the om-list mailing list