MOO-cows Mailing List Archive
[Prev][Next][Index][Thread]

Re: Webbed MOOs etc...

At 08:26 PM 5/20/96 PDT, Will wrote:
> The Java client system is built on top of 
>the WWW system, loading pages for rooms and players/etc into frames as 
>needed. 
>   The Java system was designed with the idea of user-extendability from the
>ground up; ie, users can write their own applets to replace the basic ones we
>have programmed. Users can put applets "in" rooms to which "java-logged-in"
>players can connect to (we currently only have one example of this). Your
>entire interface can be revamped by you, as a programmer (not $wiz), if you
>put in the time to develop it.
>   Another design note: users logged in through the Java system and users 
>on regular telnet sessions are able to communicate/interact. The network 
>protocol and server are designed to allow simultaneous Java & telnet 
>logins. Java users and telnet users should be able to interact pretty 
>seamlessly. Only basic move/look/talk commands are implemented in the 
>applets so far, but the framework is there for further expansion.
>
>The server is at http://eth082.dorms.calpoly.edu:8080/
>or telnet to port 7777.
>Telnet and Java guest logins are allowed (but limited in number).

  Yup. Nicely done.  These folks are ahead of the curve.

It seems to take a while for people to realize that MOOs aren't "text-based
virtual reality servers" at all. They're virtual reality servers that
historically have a text interface, since that was the best technology at
the time.  I like to think about opening various windows into the world that
the MOO presents: telnet windows, Java windows (these use telnet protocols
generally, too), HTTP windows (for you web-heads), VRML windows (because
reality isn't flat), and who knows what else tomorrow.  What you can see and
do depends on the window you are working through, but it's all the same world.
  Interestingly, the VRML 2.0/Moving Worlds proposal implements behaviors
for VRML objects by associating Java applets with them.  This is already
functioning rather nicely (if a bit slowly) in the Live3D VRML plug-in for
Netscape Navigator (available for PC through Netscape's plug-in page, and
Mac through http://www.paperinc.com/l3dmac.html).  In the BioWeb interface
for MOOs, MOO object have a web document associated with them, a plain text
description, and also a VRML description.  What you see depends on the sort
of window you look into the MOO with.  You can embed java applets in the web
page associated with objects.
  Some people envision the future of net-based VR as a bunch of
VRML-associated applets all talking with each other and no central server.
I suspect the advantage of a central server (like the LambdaMOO server) for
coordinating and distributing information among people sharing a common
world is too great to give up soon.  It may be that the way to do this
right, maybe alluded to by the Cal Poly folks, is that each virtual object
has text, web and VRML descriptions, and also a Java applet that implements
its behavior, though the default applet associated with an object would
simply present available commands in some generic way, and pass
user-selected commands through to the MOO to the usual command parsing
mechanism.  The MOO code would perform those tasks relevent for all types of
connections, and for coordinating the VR world, while the applet itself
might handle some functions only relevent to people using fancier windows
into the MOO than a plain telnet one.
  Exciting possibilities, I think.
                                     Eric
Diversity University Java system:  http://www.du.org/dumoo/autoconnect_form.html
Non-java anonymous browsing: http://moo.du.org:8888/anonview/
  (Note that the Mac implementation of Java for Netscape Navigator does NOT
support telnet applets yet; only the PC and Unix versions do. The applet
runs but then can't connect. Damn.)
Home | Subject Index | Thread Index