begin oe_protect.scr
thad01@xxxxxxxxxxxxxxxxxxxxxxxxx <thad01@xxxxxxxxxxxxxxxxxxxxxxxxx> espoused:
> The challenge with this course is that the university wants it to
> be available to students very early on without a huge pile of
> prerequisite courses... perhaps Programming I and II, Data Structures,
> etc, but not the higher level things you might expect like Computer
> Graphics, Simulations, or Artificial Intelligence. I think their
> purpose for the course is to increase enthusiasm for the CS program
> in general and stem some of the early attrition that the program
> currently suffers with.
>
It's several years since I've run up a real compiler in anger (I just
hack together occasional scripts now), but here's my take on what might
be interesting...
For real-time games, getting the physics right (ie., convincing) could
be interesting, as well as for multi-node real-time games, maintaining
synchronicity between nodes, handling network congestion/packet delays and
such like. Interface design decisions are also interesting. How about
sound? Creating a stereo sound-stage, ensuring that stage positioning
relates to display position of artefacts, sound level vs distance - how
about (something I've not seen much of) including the /speed/ of sound
in the physics, so that distant collisions are heard later than near ones?
If you want to get really creative, you could look at system latencies
involved in getting realistic lip-sync :-)
Usual trade-offs between sprite/AI processing time and
display/sound/networking. If you can get an artist involved, you
could talk a little about making it /look/ right? If not, the maths of
perspective drawing, single & multiple vanishing points and transforming
"3D" objects into perspective space is probably interesting too.
For non-real-time gaming, it's all about the maths of system modelling
of one kind or another. Perhaps you could pick an example and work
through it? Something like lincity (the original one, not the new one
which seemed to miss the point of the original entirely), or perhaps the
Java-based railway game I've seen around - set in South America. In the
main, though, this is about Simulations...
You could also look (briefly) at game markets and marketing? Currently,
most games come from three geographies - Japan, UK and US. The style
of games is different from each, and the market preferences differ
significantly; there are more team games from Japan and UK than US,
say, which prefers single-hero games. You could also even consider the
perennial games manufacturer problem, that being why don't girls like
computer games all that much?
My suggestion for an overall approach would be to pick a reasonably
simple game format (dealer's choice), and do a series of lectures which
go through the market choices, design decisions, structure and intended
implementation, choice of libraries, platforms, i/o, through to writing
the actual game!
I don't think that you could do this course without some aspects of CG,
Simulation, AI, Data Structures and basic programming coming up, /but/,
if it's a teaser for later more dry & abstract courses, it could provide
useful tasters for each of those areas.
Anyway, just a brain dump!
--
| Mark Kent -- mark at ellandroad dot demon dot co dot uk |
Experience is what causes a person to make new mistakes instead of old ones.
|
|
