Touch Interface Working Group Charter
PROPOSED: for review only

Scope

Web browsers and mobile devices are making increasing use of touch-sensitive inputs, such as with a screen, trackpad, or tablet interface, as the primary or supplementary interface for web applications. This enables web developers to build more intuitive and sophisticated applications that fit naturally with the device being used.

Touch interfaces have a long history going back to the 1960s, and faster, less expensive hardware and recent deployment on mobile devices have led to a proliferation of different approaches to software interface design. The aim of this group is to determine an appropriate set of functionality to standardize, and to define those features in way that may be deployed quickly, widely, and interoperably.

Touch events are simple, but may have characteristics different than mouse events, including pressure sensitivity. Pen tablets may add hovering and modal aspects to touch events. Multitouch events add yet another layer, allowing a user to use two or more fingers to accomplish a task, and may even permit multiple users to manipulate the user interface using the same input device, such as a screen input.

In addition to low-level direct input methods, the concept of representational events has emerged as a higher-level abstraction which may allow for device-independent inputs that map to touch inputs as well as more traditional mouse and keyboard inputs, allowing authors to express more intentionality and universal functionality.

There are four conceptual layers for touch and pen-tablet interactions: physical; gestural; representational; and intentional.

• physical: This is lowest level, and deals with the physical actions that a user takes when interacting with a device, such as placing one or more fingers on a screen and moving them, applying varying pressure on a tablet with a pen, or pressing a physical button.
• gestural: This is a middle layer between the physical and intentional, and describes specific mappings between the two; for example, a "pinch" gesture may represent the user placing two fingers on a screen and moving them together at the physical layer, and this may map to a 'zoom-in' event at the intentional layer. However, these gestures are normally specific to particular devices and platforms, and may not be suitable for standardization yet. Experimentation with different implementations and script libraries may yield more concrete goals for specific gesture events in the future.
• representational: This is the highest level of abstraction in the event model, and indicates the means by which the user is performing a task, such as zooming in, panning, navigating to the next page, activating a control, and so on.
• intentional: This is a level beyond the API itself, and indicates the action the user is trying to accomplish, such as viewing more or less detail (zooming in and out), viewing another part of the larger picture (panning), and so forth.

Finally, different devices may have associated metadata beyond the physical layer, such as modal information which should be available to content developers; for example, a tablet device may have multiple pens with different colors or brush styles associated with each. A method for the device to convey this to the application along with other event information is very desirable, and a generic metadata property should be defined to facilitate this in an extensible way that will allow for novel future device inputs.

The Touch Interface Working Group's deliverables include documenting appropriate use-cases and requirements, and a specification for events and APIs to enable low-level physical events and high-level representational events, which may describe some mapping between the two; for ease and clarity of authoring, the model should include a way to manage the complex relationship between mouse and low-level multi-touch events, and should include a generic metadata property for device-specific non-physical modalities. The Working Group also intends to include guidelines for providing alternative interaction methods for people with disabilities—specifically those disabilities involving motor skills and vision; these guidelines should include both traditional interaction methods such as mouse/keyboard, and interaction methods based on tactile feedback systems for touchscreens. These deliverables must apply to desktop and mobile browsers, and other non-browser environments where appropriate, and must be consistent with Web technologies designed in other working groups including WebApps, DAP, HTML, and SVG.

Out of Scope

In order to get the broadest possible participation and quickest standardization of the physical and representational events that are currently missing from the Web platform, the definition of specific gesture events are out of scope for the Touch Interaction Working Group at this time. A future charter may include gesture events as a deliverable, based on implementation and development experience.

Pen/stylus-tablet interactions have many different aspects; ink and handwriting APIs are related topics, but are outside the scope of this current charter. For digital ink representation, see the InkML specification.

Success Criteria

In order to advance to Proposed Recommendation, each specification is expected to have at least two independent implementations of each of feature defined in the specification.