This document contains the Wonderland Roadmap. The dates included here are estimates of when we hope to have functionality complete. Please keep in mind while reviewing this roadmap that the core team of developers on this project are situated in a research lab without documentation, testing, Q/A and other resources typically associated with a product organization. This is a living document that will be updated periodically as our thinking evolves and as we develop a better understanding of how long it will take to implement different aspects of the system.

For a detailed set of release plans, please visit: http://wiki.java.net/bin/view/Javadesktop/ReleasePlans.

Release 0.4 - June 2008 - Feature Release based on 0.3

Starting at the beginning of 2008, two Wonderland efforts will be going on in parallel. We will be fleshing out platform features as well as developing new Wonderland applications. This work is being done based on the 0.3 code base. The items listed below should be completed by around June. This work, which is already in progress, is happening in parallel with the three major architectural changes described in the 0.5 release section below. As soon as the 0.5 work is far enough along, we will port all 0.3 and 0.4 features to the new architecture.

• General
  • Basic meeting recording
  • Preliminary 2D world builder
  • In-world object placement
  • User-defined placemarks
  • Enhanced web UI for managing Wonderland in hosted environments

• Audio Features
  • Enhanced telephony support (ability to carry out-worlders around world; dial-in)
  • New audio cell types - microphone, cone of silence
  • Personal volume control for all audio sources

• Applications
  • Video player
  • Video from web cams (standard view, video panorama)
  • PDF Viewer
  • VNC
  • Music in Wonderland

• Bug Fixes

Release 0.5 - Fall 2008 - Major Architectural Improvements

In parallel with the feature work going on in the 0.3 branch, three major architectural changes are in the works. The first is a re-design of all the major APIs. This should dramatically improve development going forward, although will require some porting of existing features in the short term. The second change is replacing Java3D with the more modern jMonkeyEngine (JME) scene graph. This change will also involve making some core changes to JME to better support Wonderland features. The change is needed for a number of reasons. There are features in JME that will substantially help to upgrade the look of Wonderland including render-to-texture and multi-pass. Additionally, there is an existing and growing tool chain around JME, including a Collada loader, which will be a huge benefit to Wonderland content developers. The third major architectural change is replacing the existing inflexible avatar system with a modern state-of-the-art avatar system that includes new bone models, advanced skinning features, and some inverse kinematics for avatar posing. The new system will substantially improve the expressiveness of the avatars and will support a wide range of animated avatar behaviors. This release involves major code changes to the system, so there is substantial risk that the date of this release will slip.

• Move from CVS to SVN source code control

• Redesign APIs
  • Draft specifications for new API design
  • Review period
  • Revise draft
  • Implement APIs
  • Port all existing features to new APIs
  • Improve developer documentation

• Migrate to JME
  • Evaluate JME and prioritize necessary enhancements
  • Make changes to core JME code
  • Remove Java3D and replace with JME
  • Port necessary features

• Design and build new avatar system
  • Create new avatar API
  • Avatar designs
  • Bone models
  • Skinning
  • Customizable avatar animations
  • Posing (inverse kinematics)
  • Avatar customization UI for end users

• New Module System
  • Plugable modules containing artwork and code
  • Deliver art assets via embedded HTTP server
  • Support for master and mirror artwork repositories
  • Package extensions (e.g. new cell types)

• New Features
  • HUD enhancements

Release 1.0 - June 2009 - Wonderland for Small and Medium-Sized Groups

This release will be the next major step beyond the Small Work Group release by providing more complete, stable implementations of the Wonderland core functionality. There are three target uses cases for the 1.0 release:

• Small work group meetings or classes
  In this type of gathering, a small group of 10 to 20 people work in a highly-interactive style. In the meeting or class, they share a number of applications, such as text editors, whiteboards and web browsers. Most participants gather in the virtual world, although a small number may participate by phone.

• Medium size all-hands style meetings or lecture classes
  A group of around 100 people meet and listen to a small number of presenters. The application sharing focuses on a small number of presentation tools such as slide shows or shared browsers. Presenters and some of the audience gather in the virtual world. Other attendees may choose to watch a web-cast of the meeting through their browser or dial in to the audio portion of the presentation.

• Interactive training
  A small class of 10 to 20 people are lead through a training program by an instructor. The instructor will use a combination of presentation tools and live applications to present the lesson. Students will use their own applications, as well as other techniques like role-playing, to interact with the lesson. All students will be in the virtual world, although they may review recordings of the class afterwards.

To support these use cases, the features planned for this release include:

• Usability
  • Improved camera control
  • Uniform mechanism for interacting with objects in the environment (e.g. HUD, selecting)
  • More natural mechanisms for working with applications in the world
  • Improved presence and messaging integrated into the client HUD

• Scalability
  • ~200 simultaneous users in a world with several rooms
  • Work with multi-node Darkstar

• Security and Access Control
  • Implement basic security on cells
  • Allow users to define security parameters for documents, applications and spaces in the world
  • Documents and applications can automatically inherit the security parameters of the space they are in

• Improved Application Sharing
  • Default share-aware apps (whiteboard, presentation viewer)
  • Server-based sharing of X11 and Windows applications
  • Firewall and NAT traversal for client-side shared apps

• Video, Audio and Telephony
  • Play recorded video in-world
  • Record and play back in-world interactions
  • Stream in-world live or recorded content to web
  • Private voice chat

• End user content creation
  • End user design and customization of basic building using 2D user interface
  • Placing 3D models in the world
  • Open animation file import

Release 1.1 and Beyond

The vision for Wonderland going forward is to make it the best possible environment for collaboration, be it for business, education, commerce, or entertainment. Our focus is on features that will enhance how people communicate, work together, learn together, and play together. Features that we will consider for future releases include:

Turning Wonderland into a primary work environment

We are aiming for an environment that runs applications well enough that users can comfortably do most of their real work within the virtual environment. The sort of features that would enable this include:

• Further application sharing improvements for desktop-level performance
• Scheme for allowing any application to run in-world with little or no effort

Flexible space

As a virtual environment, there is no need to stick closely to the physical metaphor. We want to take more advantage of the malleability of the environment to allow:

• Everyone to have a front-row seat
• Rearranging your in-world office to be near a different person or team room at any time
• Dynamically adding walls to a room as more space is needed
• Generating new spaces based on dynamic queries (eg, search for a music group and a room dedicated to that band is automatically created and populated with related songs, artwork, and other artifacts found on the web)

Share-aware applications

Sharing existing applications designed for single users is never going to be the ideal way to collaborate. The best possible user experience can only be accomplished when applications are built with multiple users in mind from the outset.

• create several more example share-aware 2D applications
• create several example share-aware 3D applications
• add API and toolkit features to make it easy for developers to build new share-aware applications

Avatar expressiveness based on live data capture

While some avatar expressiveness can be programmed, natural body language is an important step in dramatically increasing the emotional bandwidth provided by the virtual environment. We believe realistic avatar motions including head gestures, facial expressions, and hand gestures will work most effectively if these gestures are captured in real time from actual human motion.

• synchronize audio with avatar mouth movements
• video capture of head gestures (nodding and shaking) mapped to avatar movement
• video capture of facial expressions mapped to avatar facial animation
• video or sensor capture of hand and arm motions mapped to avatar movement

Federation

The Wonderland model is for individuals or organizations to be able to create worlds on their own servers, either inside or outside their organization's firewall. Like web servers today, we envision that Wonderland servers can be federated such that users can move easily from one organizations virtual world to another's in a way similar to moving from one web site to another.

• help develop and implement interoperability standards