[Opensim-dev] Towards 0.5, what's left?

[Kyle Hamilton]

Yes.  You are correct.

The issue that exists is more one of "how does one think about the
problem?", though.

With scripts on primitives, you have an object-oriented approach (the
object contains the mechanism for interaction with the object).  With
the Region Module interface, you have a completely separate, much more
procedural or functional approach (depending on how it's implemented)
-- the objects are acted on by a mechanism extrinsic to the object,
the data that describe the object and how the sim and client and agent
all interact with it managed by a set of routines that reach their
noodly appendages into the action of the scene.

(er, sorry for the Flying Spaghetti Monster reference.)

For someone who's used to working with the object-oriented approach,
it's very difficult to think in a procedural context.  For someone
who's used to working in a procedural approach, it's very difficult to
think in an object-oriented context.  And for someone who's used to
working in LSL, C# is a whole new ball of wax.

I appreciate that everyone is looking at it from a "well, this isn't
done yet, but you can work around the lack of it by using this other
thing that can do the same types of things only differently" view, and
I appreciate the willingness to offer workarounds for incomplete
functionality.  I'm just not sure that the workaround can be
implemented by everyone. ;)

-Kyle H

On Dec 16, 2007 9:54 AM, Sean Dague <sean at dague.net> wrote:
>
> However, the point that Stefan was making is, given that you have full
> source access to the simulator, and the Region Module interface is
> reasonably well defined, you can do this at a sim layer right now,
> giving yourself more access to the environment in the process.