Re: EARL: Simple Sample

At 04:48 PM 2001-02-14 -0500, Wendy A Chisholm wrote:
>I'd be happy to work on an example, but I don't know exactly what that 
>means.  What do you envision?  An evaluation report of site A written in 
>EARL?  For example, if I look at the output of WAVE on 
><http://www.w3.org/WAI/>http://www.w3.org/WAI/  I can see that WAVE asserts a
reading order of the 
>page.  WAVE also asserts that every image has alt-text and that none seem 
>to be suspicious.   Now I need to figure out how to state these assertions 
>using the syntax we've been discussing...
>
>First attempt will be to assert that I think the alt-text on the image that 
>linearizes the table is ok.  Using Sean's latest summary of a syntax [1] 
>combined with what we discussed on Monday [2] and Len's summary of that 
>discussion [3]...I have thrashed the syntax, but here's a stab...probably 
>in the dark. (I've combined my references in this paragraph with 
>prefixes...hope that isn't confusing)
>

Simple, RDF-free statements:

  Wendy .sez. (([path to] ALT in question) .is. appropriate)

.sez. and .is. are clearly going to be popular.  but let's assume
'appropriate'
is going to come back again.  Clearly we have .sez. is a relationship
between a
person or similar agent and an utterance.  The utterance in this case is
indeed
an assertion.  RDF says something about them.  Next we have .is. which is a
relationship that links entities and qualities.  At least one meaning of the
English 'is' is to do that.  RDF may elide .is.  There may be property syntax
in RDF so we can say the equivalent of "appropriate(ALT in question)" as an
assertion, a duple rather than a triple.

We could get more precise and technical, if somebody asks Wendy why she feels
this is appropriate.

  Wendy .sez. ((ALT in question) .satisfies. (good pattern [reference])

The "good pattern" is in a graph transcription of the normative documents
[WCAG, AERT, the World-Wide heap of prior manual evaluation results, etc.].
 It
would be yet another patch of graph where it talks about an image which plays
the role of a control, and (the ALT for the IMG and the action of the control)
enjoying an equivalence relationship.  But writing out the good pattern
requires blank stubs or placeholders for the image at least.  The model of
HTML
will if we encode it right get you a text string placeholder linked by an
attribute-relationship labeled ALT to the IMG placeholder if you have an IMG
placeholder already.

I guess that is the way I would approach examples.  Write down everything
evaluative you would say about some web page.  Convert the prose to
entity-relationship [instance] graphs.  The RDF or EARL is a transcription of
the graph form.  The trick is in understanding the culture involved in drawing
out literate graphs made up of well-modulated primitives that, like the Dublin
Core, encourage reuse.  That will take some false starts and some comparing
what it was you wanted to say about different pages.

There's an alternate approach, if the graphs don't feel natural.  Just write
out your reactions to the page as a bunch of English assertions.  Different
people should do this with different pages.  Then go through and identify
(magic marker) all the items you made statements about, and the properties or
relationships that you asserted about them.  Compare and contrast these
content
objects and properties or qualities that different people used in their
commentaries.  Encode this in a WordNet lattice.  There you have a vocabulary
for EARL 0.1.

Al

>[1] @prefix 
>s:<http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Feb/0060.html>htt
p://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Feb/0060.html
>[2]
<http://www.w3.org/WAI/ER/2001/02/12-minutes.html>http://www.w3.org/WAI/ER/2
001/02/12-minutes.html
>[3] @prefix 
>l:<http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Feb/0097.html>htt
p://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Feb/0097.html
>@prefix h:<http://www.w3.org/TR/xhtml1/>http://www.w3.org/TR/xhtml1/
>@prefix earl:<http://earl/>http://earl  #doesn't exist
>
>#use XPath to express which image we're making an assertion about
><h:img[@src="/Icons/detab"]> has <h:@alt> of "change column layout"
>
><earl:person> <#name> "wendy"
>
>"wendy" <earl:asserts> "the alt attribute of <h:img[@src="/Icons/detab"]> 
>has <l:quality>"
>
>#The assertion shouldn't be a string, but wasn't sure how to fit it into 
>the vocabulary or syntax.  All of these should be strung together into one 
>assertion, shouldn't they?
>
>Not sure what to do about reading order.  Sean, is this the type of 
>exercise you were hoping we would do?
>
>--wendy
>
>At 01:57 PM 2/12/01 , Sean B. Palmer wrote:
>>One of the problems with coming up with a simple EARL example is that EARL
>>is so encompassing and generic. If we had an example of EARL where a page
>>was being WCAG rated (i.e. rubber stamped), there is nothing in that that
>>can provide an example of testing an SVG graphic for accessibility or one
>>of the many thousands of things that EARL could be used for.
>>
>>Maybe we should all work on a few examples each, and then submit them to
>>the list? That way, we shall get many issues covered at once. If all
>>examples had some sort of summary of problems faced, etc., then that would
>>be even more sueful. If we can first come to some knid of agreement about
>>the vocabulary, and how we use it, then we can all see how many more terms
>>we needed to invent for our actual practical examples. Note that if you
>>invent a new property, it might be possible to define it as being related
>>to something that has already been defined in the EARL vocabulary... this
>>is the kind of modular thing that I keep ranting on about.
>>
>>Is this a good idea? I think pragmatism will warrant useful results for
>>EARL at this stage.
>>
>>--
>>Kindest Regards,
>>Sean B. Palmer
>>@prefix : <<http://webns.net/roughterms/>http://webns.net/roughterms/> .
>>[ :name "Sean B. Palmer" ] :hasHomepage
<<http://infomesh.net/sbp/>http://infomesh.net/sbp/> .
>
>--
>wendy a chisholm
>world wide web consortium
>web accessibility initiative
>madison, wi usa
>tel: +1 608 663 6346
>/--
>

Received on Wednesday, 14 February 2001 21:01:14 UTC