Miscellaneous ideas

[GoogleCodeExporter]

1. Allow for "per-agent" variables, which will vary initial values of 
turtles-own variables for each agent?  
But how would this interact with "CA" (clear-all) being called in setup?
And how would it interact with varying the number of agents initially placed in 
the model?  Then the size of the search space isn't constant... Ugh!

2. what about density distributions for the agent variables?

3. Better error reporting/handling: 
 - how to deal with NetLogo errors, or OutOfMemory errors?
 - out of memory?  cutoff after certain amount of time if not done?
 - add option for how to handle errors -- ignore/continue or stop dead?

Original issue reported on code.google.com by fsondahl@gmail.com on 7 Feb 2012 at 4:08