There are enough of us now (79 at the time of writing) to hopefully get
some interesting discussions going and since everyone is being quiet at the
moment I'll throw a topic into the virtual ring.
My question is "What features do we need in order to move from environments
which are merely multi-user to environments that actively support
collaboration?"
There are several examples of systems that are multi-user in some sense but
not collaborative. Two common examples are:
- multi-user databases. These often actively attempt to hide the presence
of other users by giving each user the impression that they have there
own private database [1].
- The web. The web provides multi-user access to information but cannot be
said to be collaborative. However, people have implemented collaborative
extensions to the web.
There are several examples of multi-user virtual environments and these all
support collaboration to varying degrees. At a minimum users need to be
aware of each other and almost all multi-user systems support this by
providing some form of "embodiment" which is visible to other users. Users
also need to be able to communicate and some systems provide textual
communication between users (MASSIVE, Alphaworld, MOOs/MUDs).
Experience with MASSIVE (developed by Chris Greenhalgh,
http://www.crg.cs.nott.ac.uk/~cmg/massive.html) suggests that for VR, in any
case, real-time audio communication between users is necessary for natural
communication to take place. Although text is used as a back up when audio
quality is insufficient.
And thats probably about as far as collaboration support goes in most
cases.
What else do we need? How about:
- the ability to pass information (perhaps embodied as documents in the
environment) between users - I'm not a "Mudder" but I believe that this
has been implemented in some MOOs
- Applications that actively support multiple users. Most current multi-user
VR applications that I'm aware of support multiple users by simply "not
caring" which user the input comes from. However, in certain cases
applications need to contain user related state. An example of this would
be the whiteboard application in DIVE 2.2.
- Asynchronous communication and communication "at a distance" (in the
environment, not geographically). Again I've seen this done in
MOOs. Tomas Axling at SICS (where I'm physically present at the moment)
is also doing some really cute stuff in VR using virtual telephones that
can contact agents who are not in your current field of view and which
may be some distance away.
Anything else? Comments anyone?
Dave
1) T. Rodden and J.A. Mariani and G. Blair, "Supporting Cooperative
Applications", International Journal of Computer Supported Cooperative Work
(CSCW), 1992, 1(1-2), Kluwer.
Dave Snowdon
Communications Research Group Tel: +44 (0)115 9514226
Department of Computer Science Fax: +44 (0)115 9514254
The University of Nottingham E-mail: dns@cs.nott.ac.uk
Nottingham NG7 2RD, UK <http://www.crg.cs.nott.ac.uk/~dns/dave.html>
"This mind intentionally left blank"