A92: Project X [i3]
[Prev][Next][Index][Thread]
A92: Project X [i3]
Greetings everyone,
After some careful thought, I have decided to publicize my project. I don't
have any worry that someone may steal the idea. Considering that the work
already done on the project has almost put my in an asylum, and I am no
where near an ALPHA version. :)
Project X is almost a suitable title. Most of you Windows users are
familiar with Direct X. Direct X (now ver 5.0) is a hardware-abstraction
layer where programmers can issue calls and expect that a certain result
will happen on ANY hardware/software setup. If hardware exists for the
function call, Direct X uses it. If none exists, Direct X will emulate the
functionality. One subsection of Direct X is Direct 3D. And now for my
project...
**********************************************************
**********************************************************
*** [i3](TM) -- Insight 3D (C) 1998 HillVisions /\ ***
*** / \/\ ***
**********************************************************
**********************************************************
[i3] is the first three-dimensional graphics rendering subsystem targeted
for the TI-92 platform. Functionality will not be as involved as the
standard implementation of Direct 3D. Here is a list of features that
should be included in the final version:
* Dynamic View Window - The 10MHz processor in the TI-92 is
a problem when real-time graphics is expected. The
Dynamic View Window will update in a fashion that simulates
a real-time model. The window size, among other things,
can be changed on the fly. The 3D graphics rendering
subsystem uses a complex combination of the A-Buffer and
scan-line algorithms. A-Buffer is a hybrid form of the
standard Z-Buffer algorithm. Because of memory
constraints, a full version of the A-Buffer cannot be
implemented. The standard scan-line algorithm will be a
"stand-in" until the TI-92 Plus module becomes the optimal
platform.
* Aggressive Scene Management - I chose the word, aggressive,
as it is the best description of what the scene manager
does. [i3] will handle close to 90% of management for
those programmers who just want something quick. Expert
programmers may elect to use the following scene manager
for their needs. (See the note at the end.)
* Passive Scene Management - You should not infer that a
passive system is a weak system. On the other hand, the
passive nature of this manager will allow the expert
programmer close to 100% control over the scene. (See the
note at the end.)
* IPOM (Infinite-Precision Object Model) - Although infinite
is used in its loosest meaning, the heart of this object
model allows for precision that does not sacrifice memory.
Hopefully integrating with Hufflib, the object files will
be compressed and allow for larger scenes.
* Active Texture Model - The TI-92 does not have a display
that will allow for a lot of detail for textures. This
model will actually change the texture depending on the
other objects in the scene. [i3] will probably use the
7-shade grayscale library. The Active Texture Model will
not be an immediate feature as initial tests have shown
textures to be difficult to distinguish on the TI-92's
LCD screen. The intent for this model is to simulate a
complex texture using simpler, smaller textures. As more
objects are on the screen, it will essential that each
object be seen among the other objects.
* Expensive Atmospheric Model - Don't worry about paying for
anything. (See the note at the end.) Expensive, in this
context, refers to the sacrifice in real-time response.
This model will contain several features that just take
time to complete processing. It is my hope that each
feature will not dramatically affect the response. However
I cannot promise that these features will be fast. Some
features that are included in this model are listed below:
- Lighting (Simulated & Volumetric)
- Fog/Haze (Simulated & Volumetric)
- Atmospheric Attenuation (Simulated & Volumetric)
- . . .
I could keep listing, but I think most will get the idea.
Volumetric calculations are quite intensive. I hope to
implement a simulated version of the feature so that the
"effect" is there.
NOTE AT END:
There will most likely be two versions of [i3] released. The Aggressive
Scene Management version will be released to everyone and at no charge. The
Passive Scene Management only will be released on an individual basis and at
some cost. The Passive Scene Management requires some involved work with
the internals of the system. Don't expect the cost to exceed $5. I may
drop this charge, but I just want to let you know that one may exist. I
think that I deserve something as I spent the time on this project.
As most of you are thinking that I am insane and will never finish this,
rest assured that I have implemented most of [i3] in QuickBASIC on my
computer. I should have some Beta versions of [i3] ready for any
individuals who wish to try it out. I don't intend on writing games for
[i3]. I'm merely going to maintain the subsystem on which others may make
games.
I'll probably post an every-other-week update to the list so individuals who
are interested can be informed. Please don't pester me about Beta versions.
I will post something to the list when I have a Beta suitable for outside
use.
====
Aaron Hill (mozart@inlink.com)
IRC EF-NET: SeracOhw
Follow-Ups: