On a capability-structured universe

I've occasionally pondered the idea of simulating a capability- 
structured universe; it would be essentially a capability operating  
system on a finite machine, but with a design such that the amount of  
computation occurring is constant ("conserved"); i.e. there are no  
errors such that programs stop and are not replaced by something else.

It occurred to me that the latter portion of Greg Egan's _Schild's  
Ladder_ depicts something similar to how such a universe (containing  
complex structures) might end up, or at least, similarly different  
from the universe we have. Well, so far as I recall the details; it's  
been a year or two since I read it.

-- 
Kevin Reid                                  <http://switchb.org/kpreid/>