Re: Event-Oriented Computing

[glen e. p. ropella]

At 09:16 AM 2/5/99 -0800, you wrote:
>GR> Why not?  We already allow system calls.  
>
>The difference is that, currently, the models are handcrafted to 
>Do The Right Thing.  In the "throw a snake into a fish tank just to see
>what it'll do" scenario, the experimenter doesn't have a understanding
>of the consequences of their actions.  If the foreign agent looks
>around for features to run, and stumbles across `unlink', `opendir'
>and `readdir', it isn't hard to imagine how files would start
>disappearing.

Understood.  OK, how about defining "system" calls on the Swarm
machinery that any agent can use?  Other than crashing the sim or
seg faulting, we're not looking at any real danger, there, right?
Of course, for it to be useful, we'd want some kind of insurance
that the sim wouldn't really crash, even if it just went into an 
infinite loop over a null op.

>The `read' end of serialization will generate a class on the fly if
>necessary, but it doesn't yet have any support for generating methods
>on the fly (other than inheriting them).  The method code has to
>come from somewhere.  The best we could do without an interpreter would
>be to provide method lookup by name.

That's fine, I think.  

>To me, constructing variant classes from existing methods isn't very
>interesting, although I could see how it could be used.  Ideally, it
>should be possible to have a simulation write new methods by dynamically
>generating code (and testing it in a sandbox).

Well, it's only interesting to the extent that you ball up your
behavior correctly.  One could imagine writing methods that 
wrap all the primitives in a language (like "+" and "=:") and 
then constructing *any* class that did anything in the same
way we currently program.  But, of course, that's ridiculous.
The point is that it can be made interesting if you do it right.

The other side of it is that doing such things with precompiled
methods isn't ever Right(tm), so I agree that it would be much
cooler to be able to generate code on the fly.  But, that's beyond
the scope of what I want to do, really.

Maybe I should just shut up and try to write an application
that has discovery machinery and then see how well it goes, eh? [grin]

glen
--
glen e. p. ropella                =><=          Hail Eris!
Home: http://www.trail.com/~gepr/home.html  (505) 424-0448
Work: http://www.swarm.com                  (505) 995-0818  


                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================