Group Discussion Results

Three times 1 hour was devoted to group discussion. Results were presented by the group chairs to the whole workshop. (Usually 10 minutes + 1 slide per discussion group). Below are the (only slightly edited) transcripts (minus drawings) of those slides.

A-1: Information Representation

1. Social construction of information
2. Driving through different views
3. Top level goal and underlying knowledge
4. AI knowledge representation
5. Incorporation of multiple senses creatively
6. Basic interactions to enable social construction of representation
7. Virtual environments provide remote, persistent representation
8. Requirement for clarity and precision

A-2: User Interfaces and Clients

1. Separate model and representation (allows multi-modal interaction)
2. Allow world (server) to communicate to the client the actions it supports - client decides how to represent the actions
   • How to describe ?
   • MIME ? (or something similar) or moving objects into I/F ? JAVA? --> there are API issues
3. Need to specify common set of actions (client functionality) --> problem of different modalities
4. Client-server interaction
   • Content negotiation
   • Bandwidth <--> usability
5. Import/export of information to/from other systems
6. Client should be customizable by user
7. Server technology has impact on usability (LAG!!
8. Monolithic clients are bad, Clusters of separate functionality good (cyberdog, hotjava)

A-3: Integration of the 'real world'

1. VE should at least be as good as real world (RW), e.g. objects should at least have the same functionality as their RW counterparts (but could have more)
2. Integrate our everyday tools + RW services - electronic or real - don't reinvent but rather integrate common standards.
3. Social conventions:
   • Awareness:
     • What is going on in the VE, who is there and what are they doing
     • Focus of attention
     • How to represent this
   • Transfer of etiquette --> develop mechanisms
4. Underlying structure should be abstract, allows for multiple representations (text, 2D, 3D, audio) and multiple operations. Should be easy enough for users to develop.

B-1: Spatial Organization and Navigability

1. More spatial metaphors for the Web (in particular cues for paths)
2. Cues:
   • well-worn paths
   • indexed linking
   • guided tours
   • landmarks
3. Path:
   • looking back
   • looking forward
   • looking up in hierarchy
   • looking down in hierarchy
   • page author info
   • etc.
4. Search Engines:
   • predict [???]
   • direct
   • filter
5. 3-D Metaphors:
   • phone booth
   • catalogue
   • libraries
   • musea
   • entrance-hall
   • etc.
   • Is it a tool you carry with you, or is it a place ?
6. [lovely drawing skipped] More (forced) tagging of web-pages. An example is the living document system used for this workshop preparation.

B-2: Collaboration and CSCW in VEs

1. Sharability of things:
   • "work on ..."
   • "use of .."
   • between applications
2. Ergonomics:
   • as intuitive as possible
   • showing "the model" to the user
3. new framework for design --> build collaboration into the system
4. synchronous discussion tools
   • both for work and reflection on the work process
5. support for different roles
6. Need more research on how people work
7. Ownership: who can do what ?
8. Generally: more creativity needed (not necessarily rebuild the world)
9. VE vs. CSCW (Computer supported collaborative work): VEs include "being there, are multipurpose and have persistence.

B-3: Large Scale systems and distributed architectures

1. We should replace old-fashioned MOO architecture with distributed systems.
2. We should separate huge MOO DB into clients, many types of servers (routing, user id, location, ....)
3. Clients and Servers should have (partial) database replication
4. Database consistency is a quality-of-service parameter

C-1: Community (Users, Social Aspects, Evolution etc)

1. Why do we need them?
   • Humans are social animals --> belonging
   • Self reflection and identifying
   • Many can do more than one
   • Social lubrification
2. Communities through time:
   • Inherit [??] --> freedom of choice
   • Location --> spatial dilation
   • Homogeneity -> heterogeneity
   • Ideology -> common interest
3. Communities as problem solving tool ?
4. Positive:
   • quick action
   • powerful
   • easy to organize
5. Negative:
   • close off to new comers
   temporary existence
6. Tools + medium + context will result in different types of community
7. Ownership: social rules vs. "digital" [??]
8. Design: initiators, catalysts, splitting/merging

C-2: Integration of the WWW

1. What's the point ? C.f. aims of WWW: sharing, communication, hypertext, platform independence, collaboration, ....
   • Collaboration --> Annotation (GONE!) --> ???
   • See also: CSCW
   • Choice:
     • Add VE to WWW ?
     • ADD WWW to VE ? (better)
   • WWW + VE added (==> adding community to the Web)
     • Spatial metaphors
     • Enhance collaboration, interaction
     • Immersion
   • VE+WWW added
     • URLS: sharing (Masinter/Ostrom)
     • Access to lotsa stuff (other VE/WWW)
     • Multimedia
2. Are webbed VE's productive ?
   • VE buys you a lot (OODB, communication, user modeling, on-line experts)
   • Adaptive documentation (Jeni)
   • TaTTOO'95 exhibition (Chris)
3. Cyberspace desktops
   • Browsers are too BIG! do TOO much !
   • Integration of WWW/VE stuff into operating environment
   • Smaller units of functionality (API->Cyberdog, OpenDoc ..?)
   • [drawing skipped about document-centric not application-centric]

C-3: Objects in collaborative Work

1. Need right level of abstraction for the problem at hand - simple as possible (multiple levels are ok)
2. Standardized programming interfaces (at protocol level) for objects needed to build a VE.
3. Objects (at user level) do not have to look the same