- From: Colin Clark <colin.clark@utoronto.ca>
- Date: Mon, 8 Dec 2008 16:14:20 -0500
- To: wai-xtech@w3.org
- Cc: Richard Schwerdtfeger <schwer@us.ibm.com>, Joseph Scheuhammer <clown@utoronto.ca>, oliver.keim@sap.com, stefan.schnabel@sap.com
Hi all, I'm the technical lead for the Fluid Project, an open source community of interaction designers and UI developers who do a lot of DHTML accessibility work. I've been following the ongoing thread about keystrokes for doing drag and drop interactions with great interest. Over the past year or so, we've done a lot of user testing and research about drag and drop interactions. As we delved more deeply into the problem of keyboard navigation, one of the things that became clear was how much context matters when coming up with the appropriate keystrokes and interactions. Based on our testing experience, we found that it's risky to think about drag and drop as a homogenous use case. There are lots of different types of activities that are powered by drag and drop: * Sorting items in lists, grids, or layouts * Moving or copying items within a tree or folder hierarchy * Creating file shortcuts or aliases * Opening files within a specific application * Referencing external files, text, or other media within an application * And so on... Not all of these activities are appropriate for the familiar cut/copy/ paste idiom. For example, prioritizing items in your "to do" list by sorting them in order of importance. You're not copying an element, you're simply moving its position relative to the others. Neither are you cutting an item from the list, since again, you're just changing position. Similarly, many of the other use cases listed above don't naturally correspond to the copy/paste metaphor. Copy and paste are probably the two most familiar actions in all of computing. By overloading them with different actions and semantics, there's a real risk that you will detrimentally affect the consistency of your user interfaces and confuse users in the process. As far as I can tell, the goal of the Style Guide group is to come up with a "generic" set of keystrokes for all different flavours of drag and drop behaviour. Assuming this is the case, I'd recommend that you avoid overloading the copy/paste semantics by using Ctrl-C and Ctrl-V as keystrokes for drag and drop. On the other hand, a more fruitful approach might be to try to break up the problem into specific contexts, such as sorting, cut/copy/ paste, shortcuts, and so on. Then it would be easier to come up with standards and best practices that are suited to the task at hand, along with shortcuts that match the user's expectations. In the end, this means you can design optimized keystrokes that will help the user accomplish their work faster. Our own Joseph Scheuhammer has articulated an excellent and in-depth rationale for this point. Aside from that, I'd be happy to lend a hand with exploring unique keystrokes for different use cases. I'm also happy to share advice and help from Fluid's interaction designers who have considered this issue in more depth than I have. I hope this helps, Colin --- Colin Clark Technical Lead, Fluid Project Adaptive Technology Resource Centre, University of Toronto http://fluidproject.org
Received on Tuesday, 9 December 2008 08:09:21 UTC