Re: input api naming conventions:

David Poehlman wrote:
> 
> I think we should make clear that we are using an api model rather than
> a physical device model here.

We are?

>  This would make our intent for the
> document more clear and free us from some of the ambiguity that results
> from the physical modality structure. 

But checkpoint 11.3 is about physical keys: it's a "one gesture"
requirement, not an API requirement. The API requirement to
use the keyboard API is checkpoint 6.7.

> I believe this is discussed in a
> response to us from brian on his last call issues?  It becomes important
> when cross input modality approaches are taken and when an input mode is
> indirectly used such as in the case of single switch mapped to keyboard
> api and mouse or morse code and voice input also fall into this class if
> I understand things and have read the document correctly.

We can discuss this further on the call today, but I believe
that the "single key" parts of 11.3 are not about APIs at all.

In fact, you might want to bind one physical key to something
that requires 15 individual commands to accomplish. I use
this in emacs all the time: keyboard macros.

I don't think it's important to accessibility to be able to
accomplish a task through one procedure call versus 15 procedure
calls. The API requirement is simply that you can operate the
user agent, not that you can do it with fewer procedure calls.

 - Ian

-- 
Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
Tel:                         +1 831 457-2842
Cell:                        +1 917 450-8783

Received on Thursday, 22 March 2001 13:16:51 UTC