Re: WD-WAI-USERAGENT-19991105 review

"Håkon Wium Lie" wrote:
> 
> Enclosed are my notes from reviewing WD-WAI-USERAGENT-19991105
> 
> The document is important. Hopefully it will be able to turn a
> negative trend -- the web is less accessible today than it was five
> years ago.
> 
> One general comment before proceeding to more specific issues: I think
> the document puts too much emphasis on UAs supporting various APIs
> relative to the UAs enhancing accessibility itself.
> 
> For example, section 5.3 states:
> 
>  > Provide programmatic read and write access to user agent
>  > functionalities and user interface controls. [priority 1]
> 
> and section 10.3 states:
> 
>  > Allow the user to change and control the input configuration. Users
>  > should be able to activate a functionality with a single-stroke (e.g.,
>  > single-key, single voice command, etc.). [Priority 2]
> 
> I would suggest reversing the priorties here -- it's more important
> that the UA supports (say) single-key functionality than that it
> supports an API for changing the UI.

Added as issue 155.
http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#155
 
>  > 5.6 Conform to W3C Document Object Model specifications and export
>  > interfaces defined by those specifications. [Priority 1]
> 
> Adding support for DOM effectively turns a browser into an editor.
> This is often beneficial, but in memory-constrained enviroments it is
> often impossible. By making this Priority 1, a whole segment of UAs
> will fail conformance and might therefore pay less attention to the
> Guidelines in general. I suggest changing it to Priority 2 and limit
> the requirement fo the read (i.e. not write) portions of the DOM.

Added as issue 156
http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#156
 
> I suggest changing all sections in Guidline 5 to reflect this.
> 
>  > 6.2 Conform to W3C specifications when they
>  > are appropriate for a task. [Priority 2]
>  > For instance, for markup, implement [HTML40] or [XML]. For style
>  > sheets, implement [CSS1], [CSS2], or XSL.
> 
> XSL is not a W3C Recommendation and although it may turn into one some
> day, it is bad practice to include forward references in a
> specification. 

We have an XSLT reference in our list of references, but not reference
to XSL. I'd recommend that we include a reference to XSL (not XSLT)
but clearly indicate that XSL style is only a Working Draft.

We could push the XSLT reference up next to XML...

What do you think?

 - Ian

-- 
Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
Tel/Fax:                     +1 212 684-1814
Cell:                        +1 917 450-8783

Received on Saturday, 4 December 1999 14:30:56 UTC