Distributed Neuron in Croquet? Questions.

> Hello,
>
> An introduction on my part would be the easiest way to start.  My  
> name is Zachary Tong and I'm an undergraduate biology major, ex-
> computer-science major.  In my spare time I have been developing a  
> biologically realistic, distributed brain evolution project.  A  
> fairly complex neural network controls a relatively simple "body",  
> which interacts in a simulated environment.
>
> Its come to my attention recently that intelligence is as much a  
> product of environment and interaction with other organisms as it  
> is biologically encoded.  This means I need an environment that can  
> hold multiple organisms and allow them to interact in a semi-
> sophisticated manner.
>
> I was originally going to create this simulation environment  
> myself.  This not only adds a lot of work to my plate but will also  
> be much less flexible than desired.  I stumbled on Second Life,  
> which caught my eye.  It (and other virtual worlds) could provide  
> the environment simulation I need.  After looking into it, however,  
> I feel it would be more trouble than its worth, trying to hack my  
> project around their system.
>
> Croquet looks like a much more appealing option.  I had some  
> questions regarding feasibility however.  I realize Croquet is  
> young but these are the basic, fundamental requirements I would have:
>
> -Is it a persistent world?  Can I set up a zone that is dedicated  
> to the project and will persist even when people aren't there?  
> Since Croquet is P2P, where would this reside?  Would it be hosted  
> on my server (I'm assuming this is the answer)?

>
>
> Thanks for the help.  Croquet looks exceptional and the open  
> framework is a good thing to see after looking into SL.  In case  
> anyone is interested, here is my blog on my project:
>
> http://distributedneuron.net/blog/category/dneuron/
> http://distributedneuron.net/blog/
>
> Thanks again,
> -Zachary Tong

> Thanks for the fast replies and honest answers!  Its very much appreciated.
>
>> If P2P multi-user interaction isn't really a
>> requirement for your project, then you might want to look elsewhere.
>
> Multi-user interaction is a definite requirement.  P2P is not ideal as a
> classical server-client setup would also suffice, but it is also not a
> negative.
>
> My project is distributed in nature.  Classical brain simulation is
> often done on a supercomputer or cluster because of the computational
> requirements.  I've opted to evolve smaller "brains" and distribute the
> computation to volunteer clients.  Each client runs on a volunteer's
> ordinary PC.  Each client runs a small neural network and controls a
> single organism in the environment.  The environment should contain the
> organisms of other clients so they can interact.  Interaction is simple
> - moving, eating, "sensing" each other (vision, smell, etc), reproducing.
>
> I'm much more interested in the multi-user environment of Croquet than
> the 3D capabilities.  Infact, my client most likely won't even show the
> 3D environment, it eats up valuable processing time.  All I need is a
> place that can hold organisms, allowing them to interact in a spatial
> manner.
>
>
>> 2) You can just make sure that there is someone running the simulation.
>
> After thinking about it, this will happen automatically.  The simulation
> is a product of clients participating, guaranteeing someone will always
> be inside.
>
>
>> Yes. But don't assume it's slow. For many activities, it isn't.
>> Profile and know.
>>
>> However, remember the definition of the core model.  The simulation
>> state must be replicated on each participant, such that a snapshot gets
>> everything and such that all computation provides the same answers.
>> Getting C++ to do that is going to be very hard.  Very hard.
>
> I think I'm still ok, after reading about the Core model.  I can
> effectively treat my neural network code as a blackbox - feeding it
> input and receiving output.  This output (move, eat, etc) is the
> "behavior" of the client and what will be sent to other clients.  The
> only computation that needs to be done after receiving a snapshot is
> updated behavioral messages.  Just like real life, organisms only need
> to know superficial information about others such as their position and
> appearance.
>
> Can Squeak interface with DLLs?  I could separate my neural network code
> into a DLL.  Squeak handles all network and environment operations,
> feeding the neural network DLL inputs (spatial information, vital
> statistics, etc).  The DLL returns behavioral outputs which Squeak sends
> to other clients.
>
> Does this make sense?  Is this possible?  The biggest reason I don't
> want to port my code is A) performance worries which may or may not be
> an issue and B) its a ton of code I'd have to convert.  I just don't
> have time to rewrite 4 months of work :)
>
>
> Unrelated, but the router timestamps may work in my favor.  I've been
> struggling with a way to keep various clients working at the same
> speed.  Your computer may be much faster than mine, meaning your
> organism can update its behavior more often than mine.  This could
> impart an advantage (such as responding to threats quicker) that is
> independent of the genome, allowing unfit genomes to survive longer than
> the should.  By controlling the timestamps I could ensure that all
> clients have equal opportunity to play on the same speed.
>
> Thanks again for the help!
> -Zach
>
>
>
>
>