RE: Exporting the DOM

Ian's memo [1] is attached below with my comments labeled with "EH:"

[1] http://lists.w3.org/Archives/Public/w3c-wai-ua/2000JulSep/0366.html
> -----Original Message-----
> From: Ian Jacobs [mailto:ij@w3.org]
> Sent: Friday, September 08, 2000 12:12 AM
> To: Hansen, Eric
> Cc: UA List (E-mail)
> Subject: Re: Exporting the DOM
> 
> 
> "Hansen, Eric" wrote:
> 
> [snipped good discussion]
> 
> > Summary
> > 
> > I think that this [contradiction]
> > can be solved by making clear in the document that the
> > subject as a whole must export the DOM, but any individual 
> component may not
> > need to export the DOM. (There may need to be further refinements,
> > especially as they relate to information going to the user 
> interface that
> > does not come from the DOM.)
> 
> I agree with this. I would argue the following:
> 
> 1) The subject of a claim may have many components (in fact, is likely
>    to have more than one component).
> 
> 2) Some checkpoints may not be relevant to some of the components.
>    For instance, a component of the user agent that implements
>    a search algorithm but has no user interface and does not maintain
>    state in the form of a document object would not be affected by
>    user interface or DOM requirements. However, if that functionality
>    is available to the user, then it needs to be documented, so 
>    Guideline 11 applies.

EH: A statement like "Some checkpoints may not be relevant to some of the
components" begs the question, "Which checkpoints are not relevant to which
kinds of components?" You have elucidated above a principle that seems
relevant to the situation involving DOM requirements. It would be But is
that even really necessary? Instead of trying to define which ones are
relevant either piecemeal or even in bunches, is there an even simpler
approach. 

Perhaps you have already suggested the approach when refer, in this case to
functionality/content that is "available to the user". 

Perhaps the document should treat the subject of the claim as a "black box"
with inputs and outputs where we don't care _at all_ about the interaction
between components. Inputs to the Subject: (a) Web content from the Web
content author (including documenatation), (b) configuration settings and
control actions from the user, (c) program control data. Outputs from the
Subject: (a) Rendered content from author-supplied content and user-agent
produced content (notifications, alerts, repair info), (b) revised settings,
(c) work product (produced primarily through use of author-supplied
content), (d) history data, (e) trash (temporary files, etc.), (e) DOM
export, (f) other data export as required.

All output from the subject must be WCAG 1.0 conformant (at the AA-level),
unless other spec is applicable (e.g., DOM export) or unless it is a 'work
product' over which the user agent has no control. (Perhaps this ought to be
a checkpoint.)

I was going to say "If the foregoing is done properly then perhaps we don't
even need to specify which checkpoints a given component helps satisfy" but
I checked the 1 September UAAG document and there is no such requirement in
the claim so my comment is not necessary.

I suppose that the checkpoints need to be examined from the standpoint of
whether there is every any instance in which we must care about the
interaction between components.

> 
> 3) The subject as a whole must satisfy the requirements of the
>    document. The subject must export the DOM for access to the
>    document object, whichever component or components contribute
>    to that document object. 
> 
>  - Ian
> 
> 
> -- 
> Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
> Tel:                         +1 831 457-2842
> Cell:                        +1 917 450-8783
>

Received on Friday, 8 September 2000 10:46:41 UTC