- From: Al Gilman <asgilman@iamdigex.net>
- Date: Wed, 19 Feb 2003 11:03:26 -0500
- To: Charles McCathieNevile <charles@sidar.org>, Dean Jackson <dean@w3.org>
- Cc: wai-xtech@w3.org
At 03:33 AM 2003-02-19, Charles McCathieNevile wrote: >My personal opinion, which I think matches PF's consensus, is that some >events in DOM2 such as onmouse* are indeed device specific. There are ways >of working around this - the current situation that included some >non-specific events alongside some specific ones doesn't strike me as a >great outcome :( > >There is discussion of this more and more frequently, and I believe that >the Batik SVG viewer has implemented a fairly device-independent set of >event triggers tied to particular user agent inputs. I hope to find time >to follow this up further, but it seems never to exist. This is a good point to work on getting into the document more clearly. Some applications, such as a query of a GIS for the nearest gas station, are intrinsically point-specific. Others just use a point to accomplish the isolation of an object which matches many points and has an articulable identity and functions independent of which point was used to indicate the object. The present state of the art is that much of the semantic information that an Assistive Technology needs to re-bind the interaction using other devices is buried in the script of the handler in a way which is not practical for the Assistive Technology to extract. The HTML WG has taken on a goal of creating an "XML Handlers" notational facility that is a complement to what exists in "Events in XML" that will get more of the critical information exposed in a way that ATs could use it. Yes, we would like to see migration that would put some of the current events in dis-use or else re-build their definition and implementation in a way that is more 'repurposable.' Al >cheers > >Chaals > >On Wednesday, Feb 19, 2003, at 12:04 Australia/Melbourne, Dean Jackson wrote: > >>* 3.4 Use a device-independent interaction and events model / module. >>Your example technique suggests using Activate rather than >>Click (I assume). I guess this means that you think the DOM events >>are device specific? (Note: I'm not arguing, I just want to know:) >-- >Charles McCathieNevile charles@sidar.org >Fundación SIDAR http://www.sidar.org
Received on Wednesday, 19 February 2003 11:03:21 UTC