Describing all functionality without using too much jargon in a paper is a challenge. Understanding the work of others is even a bigger challenge. We have witnessed some confusion about the Midas framework in some papers and would like to clarify them on this page.

Multi-user support

Midas inherently supports multiple users (in contrast to what Julia et al.[1] states). It is even one of the core features to support concurrent interaction. To better position the [1] paper, the authors can claim that Midas does not provide a lot of constructs to deal with conflict resolution besides:

  • Salience (priority) level per rule. However this is not frequently used due to the fact that gestures might not be triggered at the same time. This is mostly limited to offline (discrete) gestures or simple online (continuous) gestures.
  • Control facts can be used to enable or disable rules as used in various rule systems
GUI symbiosis

Spano et al.[2] argues that Midas is affected by `spaghetti code` because it integrates UI components within the gesture definitions. However this integration is fully optional but very beneficial to describe certain complex gestures. In the Midas paper we actually describe the spaghetti problem as a case where multiple gestures definitions are interwoven in order to provide a form of conflict resolution. These implementations can for instance be found in the Android framework.

[1] Julia, Carles F., Nicolas Earnshaw, and Sergi Jorda. “Gestureagents: An agent-based framework for concurrent multi-task multi-user interaction.” Proceedings of the 7th International Conference on Tangible, Embedded and Embodied Interaction. ACM, 2013.

[2] Spano, Lucio Davide, et al. “GestIT: a declarative and compositional framework for multiplatform gesture definition.” Proceedings of the 5th ACM SIGCHI symposium on Engineering interactive computing systems. ACM, 2013.

research/midas/clarifications.txt · Last modified: 2014/04/09 13:06 by lhoste
Trace: clarifications