Outside expert opinion on Web standards and WCAG (especially GL 1.3)

I wrote back to the same experts I had canvassed before:

<http://lists.w3.org/Archives/Public/w3c-wai-gl/2005AprJun/0145.html>

I caught them up on some of the Working Group's complaints about my 
proposed wording for 1.3 ("Ensure that information, functionality, 
and structure are separable from presentation"), and asked a few 
questions:

1. Does my original wording for an improved 1.3 need changes?

2. Do we need separate criteria for structure and semantics, or is 
the latter implied by the former?

3. Do we need a separate criterion for accessibility features?

4. Will "semantics" be a hard-to-understand term?

Their responses follow. The Working Group may view this as further 
expert opinion.


Gez Lemon:

>>QUESTIONS
>>
>>1. Do you have any improvements to any part of the wording?
>
>No :-)
>
>>2. Do you think we need separate criteria for structure and 
>>semantics, or is the latter implied by the former?
>
>I think your argument about a document being marked up purely as 
>divs and spans illustrates perfectly that the latter is not implied 
>by the former. A document being structurally correct doesn't imply 
>that the appropriate elements help convey the meaning for the 
>content.
>
>>3. Do we need a separate criterion for accessibility features?
>
>I'm not sure. I'm struggling to think of an accessibility feature 
>that would fall into the perceivable category, that wouldn't be 
>implied with a structurally sound and semantically correct document.
>
>>4. Do you believe that "semantics" will be a hard-to-understand 
>>term for developers? (Can you cite any definitions of the term in 
>>the Web-standards context, either in books or online?)
>
>The web development community use the term "semantics" a lot, so I 
>don't think the term will be difficult for them to understand.



Brothercake:

>>1. Do you have any improvements to any part of the wording?
>
>The intro to 1.3 begins "where permitted .." - but isn't that caveat 
>part of Level 1 success, and not acceptible to Level 2?
>
>Perhaps I'm misunderstanding - is "where permitted" already defined 
>to mean a specific set of criteria? It just reads to me as though 
>it's a get-out clause, as in "oh well, blah technology doesn't allow 
>this separation, so we don't have to consider 1.3 at all"
>
>I would prefer to see each guideline defined as the higest point - a 
>statement of ideal  - with the success levels being reduced 
>graduations up to that point.
>
>>2. Do you think we need separate criteria for structure and 
>>semantics, or is the latter implied by the former?
>
>This is tricky ... an HTML page is a structure and it should (but 
>might not) be semantically meaningful - you can't have unstructured 
>semantics, but you can have unsemantic structures; the latter isn't 
>so much implied by the former, as the former is implied by the 
>latter.
>
>So, maybe 1.2 might be better inverted - "semantic markup or coding 
>is used to define the structures to the extent possible for the 
>content"
>
>>3. Do we need a separate criterion for accessibility features?
>
>Definitely, yes - and your Flash example is perfect case; just 
>because a technology isn't perfecly accessible, that doesn't mean we 
>shouldn't try at all.  I suppose one could say "well don't use flash 
>then", but that's Level 2.
>
>>4. Do you believe that "semantics" will be a hard-to-understand 
>>term for developers? (Can you cite any definitions of the term in 
>>the Web-standards context, either in books or online?)
>
>Maybe at first, to people who are new to the subject, but no more so 
>than any other technical term; and you only have to go to the 
>dictionary for a meaningful understanding: 
><http://dictionary.reference.com/search?q=semantics>
>
>Those definitions are all easily reconciled to web development - 
>HTML elements are symbols; the semantics of the elements are what 
>those symbols mean - <cite> means a citation; <ul> means a list. 
>There you go.
>
>Now I won't say that's an _easy_ concept to grasp exactly, but it's 
>not particularly difficult either - it's just another thing to 
>learn, just as you learn what "encapsulation" and "degrading" means.
>
>I found a couple of refs:
>
>- <http://brainstormsandraves.com/articles/semantics/structure/> - 
>doesn't really define semantics per se, but explains the concept 
>practically by talking about the meaning and purpose of various HTML 
>elements
>
>- <http://www.codingforums.com/showthread.php?t=20105#post101559> - 
>me making the point in a forum post


Molly E. Holzschlag:

>>4. Do you believe that "semantics" will be a hard-to-understand 
>>term for developers? (Can you cite any definitions of the term in 
>>the Web-standards context, either in books or online?)
>
>I think that developers know the term, but don't necessarily make a 
>differentiation between semantics and structure. This is changing 
>though, as more and more standards advocates describe semantic 
>markup and structure more accurately to the public at large.
>
>I understand W3C concerns as to the use of the word semantic since 
>using it can conceivably confuse semantics in markup with the 
>Semantic Web. However, if explained well to developers, I trust 
>they're smart enough to understand the difference.
>
>In context here, it's accurate terminology in my opinion. I suppose 
>an alternative could be "meaningful markup." Not as attractive, but 
>also accurate.
>
>Regarding citations, Googling "semantic markup" gets me 681,000 
>results. I've personally used the term in at least two books, one of 
>them being _The Zen of CSS Design_. I've also spoken on clarifying 
>the meaning of this term numerous times.



-- 

     Joe Clark | joeclark@joeclark.org | <http://joeclark.org/access/>
     Author, _Building Accessible Websites_ | <http://joeclark.org/book/>
     Expect criticism if you top-post

Received on Monday, 23 May 2005 19:23:46 UTC