There are a couple of things that I think must happen if any multitouch gesture “standard” is to be successful and see wide scale adoption:
1. Tolerance for undisciplined gestures
2. Concurrent gesturing and fluid gesture transitioning
3. Customization
4. Elegant gesture degradation
5. Interoperability
1. Having multiple gesture “techniques” available could satisfy everyone. Most users today have real trouble performing perfect pinch gestures on multitouch devices. In some cases they routinely trigger accidental results something we have been calling “knuckle drag”. You could argue that the multitouch device itself is too sensitive if it can detect touches with low of zero pressure but people generally expect touch screen to require effortless touch. One way to get around this is to have a single gesture with more than one avenues for point analysis. For example scaling an object can be performed with 2 touch points separating from each other but i should also be triggered when three four or five points separate from each other so that the user can decide their preferred gesture “technique”.
2. As with most functions people perform there is often a lack of clearly defined modes of operation. It is far more “natural” for users to fluidly move from one mode of operation to another in a dynamic and unprescribed fashion. One way that gestures must accommodate for this is to allow for fluid gesture transitions so that users can change from a rotation to a zoom gesture without having to removes touch points and without having to pause. Multitouch by its very nature should also promote explicit multitasking in the form of being able to perform to gestures at the same time with one or multiple hands on a single object or multiple objects. Even going as far a using using multiple applications, multiple people, multiple multitouch canvases all at the same time.
3. As people have pointed out in this thread there is no final standard in gestures yet. Gestures are still in the process of being adopted and even the names of identical gestures themselves are different on different devices, look at synaptics and apple. The more consensus there is on the “common” or “core gestures” the better not just across devices but also across applications. One way to negate this is to allow for programmability or customization of gestures on the OS level so that you can map your own gestures to specific functions in the app of your choice. Currently this has to be done explicitly but in the near future we should expect to be able to teach app using a system similar to predictive text for typing where you can dynamically train different apps and point gestures to functions on the fly. There may be some divergence from core “common gestures” but why not ? There needs to be common set that have a predictable outcome ultimately for universal accessibility but the rest should be what ever YOU want them to be.
4. With the release of flash 10.1 you can now build multitouch applications that can be embedded in a web browser (using swfs) and be an integral part of a website (without security issues). The question is often posed as to what people do when they don’t have a multitouch input device and they browse to these sites. The answer is elegant degradation. If you have a multitouch device flash can detect it through windows7 it will allow for gestural/multitouch interactions. If you just have a mouse then then you should have one of two choices: treat the site the same way you would any other (with single point interactions) or make sure that you can create multiple points of interaction using a mouse. This can be achieved by leveraging the keyboard and allowing users to “Ctrl” click to lay down another point to control with the mouse. Of course there is also there is the fact that no two multitouch devices are necessarily equal where some support 2 points while others like the 3M touch screen can support 10-20 points. Using the same idea of degradation mapping 3 point multitouch gestures can be predictably and consistency mapped to conceptually similar 2 point gestures so that the same end user function can be enjoyed using the same “mode” of interaction even if the device is dual touch only.
5. In many ways gestures should do the same thing on most devices no matter what the size or form factor. Of course that assumes that the devices support the same number of simultaneous touch points. Forcing the user to rotate suing a fundamentally different gesture on different devices will only confuse and obfuscate functionality especially when the two devices support true multitouch (10 points +). This can contradict with the gesture degradation but would only do so if dual touch devices do not transition quickly to incorporate true multitouch capabilities soon. Assuming dual touch devices do die soon then elegant degradation would only be used in the case where there is no multitouch device.
Of course this is all nice in theory but practically speaking one of the main hurdles I feel that stands in the way of ANY these points being successfully implemented in a widely adopted system is the current lack of structure regarding multitouch gesture categorization. I have been working on a system that takes some ideas from gesture & stroke libraries along with current popular multitouch gestures and mapped them into cognitively similar working groups in a effort to learn how to build on multitouch gesture libraries in a consistent manner that remains intuitive to the masses and yet still allows for simple extensions or fundamental innovations such as combining touch and fiducial or touch and pen interaction modes. There are allot of patterns in multitouch gesture analysis but I imagine it’s a matter of discovering what is fundamental and what is superfluous to multitouch aspects of the interactions in general.
