[MUD-Dev] 3D graphics (Was: The impact of the web on muds)

[Jon Leonard]

On Mon, Jan 26, 1998 at 09:41:03AM +0000, Mike Sellers wrote:
> At 03:24 PM 1/25/98 PST8PDT, Brandon J. Rickman wrote:
> >On Wed, 21 Jan 1998 08:29:17, Mike Sellers <mike at online-alchemy.com> wrote:
> >>At 10:12 PM 1/20/98 PST8PDT, Brandon J. Rickman wrote:
> >>True.  But those people are the capstone of the customer base, and to a
> >>large degree drive it.  Those who do not demand the fanciest graphics
> >>nevertheless *do* demand what were the fanciest graphics of no more than
> >>4-5 years ago.  IOW, the demand is not static for the bulk of the
> >>population, and it is the early-adopter/neophiles who drive this demand.  
> >
> >For the moment.  Andrew Grove (Intel) would have you believe this
> >demand will always exist.  I don't, but this is getting off topic.
> 
> For the moment!?  That 'moment' has lasted since at least the late 1960s,
> and has (in the various guises of Moore's Law) been remarkably consistent
> since then.  You may not agree with the way technology (in general,
> computers and grahics much more specifically) is affecting our society, but
> I think you'd be hard pressed to show that continual increasing and
> expanding expectations is some sort of "momentary" fad.  Of course, demand
> for better graphics may well level off in the future, but not in the next,
> say, 20 years, I think.

20 years is an interesting time to pick, in that it's probably about right.
If you extrapolate the Moore's Law growth of graphics hardware perforance
(1000x every 10 years) and intersect it with the maximum graphics hardware
that makes sense, they intersect in about 20 years.

The projected graphics hardware is enough to render triangles smaller than
the eye can resolve, fast enough to avoid flicker artifacts (100Hz or so).
At least, that's how I heard the argument from Michael Deering, who is a
graphics architect at Sun Microsystems.

I don't see leading-edge graphics as being particularly relevant to MUDs
though, at least in the next few years.  The dominant limit for graphics
is modem bandwidth.  In a changing environment, new model data needs to
be sent, a more severe limitation than graphics speed.

The alternative is to populate the world with pre-generated types of objects,
allowing better pictures, but equally allowing a less ambitious client that
runs on older hardware.  The graphics isn't the limit.

Jon Leonard