Let us start using VRML with standard Mosaic and a few CGI-Scripts

Karsten Wolf (Karsten.Wolf@erziehung.uni-giessen.de)
Thu, 18 Aug 1994 10:06:59 +0200 (MET)

Hi there,
my name is Karsten D. Wolf from the University of Giessen (European=20
Union). I am doing some research on selforganized learning and the use of=
WWW in education, interfacing MOOs to the WWW (with graphical interface).

I am a bit late in my time-schedule, therefore don=B4t expect me to take=20
any real action until December, but nethertheless I=B4d like to contribute=
some of my ideas to this effort.

When I am reading this group (I admit - because of the huge traffic -=20
sometimes I scan it more Data like - only without the processing power=20
8^) I came to the following conclusion:

VRML should make a mix out of WWW and MOOs with a graphical immersive (mayb=
audi?cal?tory?) interface a reality.

But what made WWW and MOOs so successfull? Among the addictive features=20
of the MOO, the info-value of the WWW, and the fun-factor of both there=20
is one key element:

The clients (and the servers, too) run on reasonable machines (with=20
reasonable machines I mean computers in the $1k - $5k range) o.k.,=20
therefore giving everybody with an internet access (or even - local -=20
without it) to tinker around.

And why is everybody tinkering around with html: because it is something=20
You can use in a try and error style. Edit a file, test it with a client,=
edit a file, ... Everybody with a ascii-editor can generate=20
hypermedia-documents (well, almost everybody, err, at least technical=20

To get anywhere in our effort, I think it is time to test various=20
proposals made here in the group. I therefore propose to install some=20
WWW-pages where we can test the different proposals. A www-page for a=20
vrml-proposal should have a structure like this (or like that, I am=20
really no control freak):

overview of the special vrml-proposal
tutorial to the special vrml-dialect
form-based input for vrml-code
form-based input for parameters

Then the server should process the code and deliver back a gif (or a=20
series of gifs or whatelse) as an imagemap or whatever controls Your=20
vrml-supports, and a Update-button. This doesnot solve the problem of=20
real-time-data which has to be solved to get exiting VRML, but one could=20
test various concepts.

Parameters (by form-input) like what computer you have to simulate=20
the scaling on the client should be processed by the server.

I believe that this would enhance the efforts, reveal problems in the=20
reality, and speed up the vrml-development.
Virtually Yours,

Karsten D. Wolf
Arbeits-, Berufs- und Wirtschaftpaedagogik
Karl-Gloeckner-Str.21 b
Justus-Liebig-Universitaet Giessen
D35394 Giessen
European Union

fon: (GERMANY)-641-702-5260
fax: (GERMANY)-641-702-5259
email: karsten.wolf@erziehung.uni-giessen.de
http:// expect our institute to be wired faster than 2400kps late Q1/95