> Andrew writes:
> >We should be ready to build systems that will take advantage of the
> >systems to come.
>
> True, but you have to be careful -- there are real tradeoffs here.
> Products fail just as often because they assume that infrastructure
> will appear quickly when it doesn't, as they do because they are
> overly conservative and get overtaken by events. In general, you need
> to be careful, and realistic, about what's going to appear, and
> when. The sort of broadband transmission you're talking about *is*
> probably going to appear, but probably not tomorrow; the economics
> still don't quite work right. Probably a couple of years yet before
> it starts hitting average homes. Meantime, we *have* the Internet
> capabilities, imperfect though they are, and those can reach most
> motivated people (at least in moderately urban areas)...
>
> Basically, I just want to make the point that knowing that your system
> is ahead of its time is cold comfort when it fails. That's why I like
> to emphasize flexibility: build within today's constraints, but always
> keep an eye on where things are going, so you can adapt to improved
> conditions when they come down the pike, and switch over to them
> quickly...
The last paragraph, above, is actually what I was saying. I do not
suggest that we spend all our time building VR so it only runs on systems
we *think* will exist in the future. I am for building them in open ended
ways that will allow us to recycle much of the material when we build the
next system.
In general, however, the idea of being careful can sometimes be construed
as meaning that we should be conservative to the point that we do
nothing. This is not what I would suggest. The (real) world is full of
risk. We should embrace that, and be willing to accept some risk in order
to build new things which will reward our risk-taking.
Note: This points out another use of VR: Scenario Evaluation. We can
explore risky ideas in a cheap, controlled, risk-free environment.
There is a paradox which faces all new endeavors. That is: Should we wait
for the support before building the product, or will the support be built
because the product exists? It's the "Chicken and the Egg" problem. My
view is that we should take the risk and build the system. Waiting for
someone else to produce the support will simply wind up in a lot of
waiting, which gains us nothing. Waiting is for the damned.
Let's build VR worlds. If we have to, we can make enough money selling
them as games on CD-ROM until we get enough support to build a good
online system. Beyond that we can begin to include the rest of the dreams
we all have. One reasoned step at a time, but we have to keep taking
those steps, and we have to accept the inherent risk of each one.
--- Andrew C. Esh mailto:andrew_esh@cnt.com Computer Network Technology andrewes@mtn.org (finger for PGP key) 6500 Wedgwood Road 612.550.8000 (main) Maple Grove MN 55311 612.550.8229 (direct) <A HREF="http://www.mtn.org/~andrewes">ACE Home Page</A>