Jump to content

Dasher (software)

From Alfawiki
Adapted from the English Wikipedia article Dasher (software), available under CC BY-SA 4.0Updated 9 August 2025
Dasher
Company typeFree and open-source software
IndustryComputer accessibility, input methods
FounderDavid J. C. MacKay
Key people
David Ward
ProductsText input software
OwnerACE Centre
Websitewww.inference.org.uk/dasher/

Dasher is an input method and computer accessibility solution designed to facilitate text input without relying on a traditional keyboard. It operates through a pointing device, which may include a mouse, touch screen, or specialized equipment controlled by foot or head movements. This technology primarily assists users with disabilities or scenarios where using a keyboard is not feasible.

Distributed as free and open-source software under the GNU General Public License (GPL) version 2, Dasher is compatible with systems running GTK+. Supported platforms encompass Linux, BSD, macOS, Windows, Pocket PC, iOS, and Android.

The concept was originated by David J. C. MacKay and further developed by David Ward along with his team at Cambridge. Funding and backing have been provided by the Gatsby Charitable Foundation and the European Union's aegis-project. Presently, the ACE Centre manages the software's upkeep, with intentions to transition Dasher's licensing to the MIT License in the future.

Design

The user interacts by pointing to letters shown on the screen, while an underlying probabilistic model forecasts probable letter sequences. These predictions are emphasized visually to facilitate quicker and less strenuous text entry. The interface has been likened to an arcade-style game, featuring characters moving dynamically across the display for selection. Over time, the system refines its predictions by learning from frequent letter patterns, thereby optimizing its presentation.

Features

Dasher provides platform-independent data assets such as:

  • Alphabet specifications covering more than 150 languages
  • Customizable color schemes for letters
  • Language-specific training datasets for all supported languages

External links