[Reality] Re: Current Task: OBJECT PROXIES.

[Jean-Paul Calderone]

On Thu, 4 May 2006 17:38:52 -0400, Calvin Spealman <ironfroggy at gmail.com> wrote:
>On 5/4/06, Jean-Paul Calderone <exarkun at divmod.com> wrote:
>>
>>I don't think IProxy.proxy needs to know who's asking.  The point
>>is that the proxy objects it returns have logic in them.  Currently
>>the illumination game system is extremely stupid: if a location is
>>lit, everyone can see things in it; otherwise no one can.  A more
>>advanced illumination game system would wrap objects and then provide
>>an alternate IConcept implementation.  If the light level is low but
>>non-zero, it can replace the thing's description with text like "a
>>dark oblong blob"; since conceptualization does take the observer
>>into account already, if it is dark but the observer has infravision,
>>a custom description that takes both of these facts into account can
>>be given to just that observer.
>>
>>Jean-Paul
>
>
>As things progress and more combinations of conditions, objects, and actors
>exist,
>how will you manage all these descriptions of the objects in so many
>environments
>or what kind of automation will come to describing them under the different
>conditions
>and to differently abled actors? That is, will you need to define that for
>some range of
>light in a normally-visioned person that "a dark oblong blob" is the
>description, or is
>that somehow deducable by the system?
>

It should be deducable by the system.  In all areas, Imaginary strives to avoid the combinatorial explosion of development/creative effort that often comes with even slightly complex systems.  This particular feature might be implemented by having the visibility system extract information from an interface which describes the size and shape of a Thing.  It can combine that information with the acuity of the observer and the particular conditions of the environment when deciding how far to degrade the description of the thing being observed.

Of course, that's still a complicated task, but it's one which can be improved independently of the rest of the system.  It's an example of what we call a "game system" - something between the core simulation engine and any application-level code.

Jean-Paul