An OASIS technical committee for User Interface Markup Language (UIML) is being formed based upon a proposal from interested members. The purpose of the TC "is to develop a specification for an abstract meta-language that can provide a canonical XML representation of any user interface (UI); the language should be capable of specifying the requirements, design, and implementation of any UI. The committee will use the UIML version 3.0 specification created by Virginia Tech's Center for Human Computer Interaction, Harmonia, Inc., and other organizations on uiml.org as a starting point" for the TC work. According to the announcement, a "general motivation for a canonical UI representation language is to accelerate the development of tools for UI development. If practitioners from these fields build tools with UIML, then the tools can interoperate. Just as XML made toolbuilders more efficient (because tools built for XML work for any XML vocabulary), so can UIML make UI toolbuilders more efficient (because tools built for UIML work for any vocabulary representing any concrete UI implementation language). Thus the TC's work will serve to assemble the jigsaw puzzle pieces of UI and HCI technology that have been created."

Members proposing the OASIS TC include Didier Courtaud of Association Aristote and three individual members Marc Abrams (Virginia Polytechnic Institute and State University), Jim Helms (Harmonia), and Hedy Alban (Max Shevet Consulting).

From the TC description:

The TC differs in scope from other committees and working groups that address concrete UI implementation languages (e.g., assembly language, C, C++, CSS, Java, HTML, VoiceXML, WML, XForms, XHTML, XSL-FO). Instead, this TC focuses on an abstract language, UIML, which expresses UIs at a higher level than the concrete languages. UIML's goal is to subsume in expressive power all concrete languages, and to permit efficient mapping from UIML to any concrete language.

The TC has a secondary purpose -- to use UIML to bridge UI-related fields:

• Techniques from the Human Computer Interaction field. These include UI models, transformational techniques, techniques for computer automation of UIs, usability engineering, generalization of the Model-View-Controller, and new UI metaphors.
• UI designers that want to think in terms of domain-specific abstractions (e.g., navigation maneuver for an automobile UI, material path for a factory automation UI), rather than at a widget level.
• UI designers that want a way to capture author intents and map UIs to different devices to create accessible UIs.
• Techniques to internationalize UIs.
• Techniques that help integrate UIs with Web Services (e.g., WSDL).

Proposed Deliverables

• Assessment of how UIML fits [with] other committees and working groups addressing user interfaces (projected for completion 90 days after first meeting)
• List of open issues in the UIML 3.0 specification (projected for completion 120 days after first meeting)
• Revised specification based on issues list (projected for completion 240 [days] after first meeting)
• Implementations of the specification by different parties
• Compliance test suite
• Best implementation practices
• Revised Specification based on implementation experience