Re: EARL properties details

At 10:51 AM 2001-03-15 +0100, Daniel Dardailler wrote:
>> I'm wondering: in the html example could we make the info in the comment 
>> field more machine readable?
>we could, we don't have to right now. we can add it later.
>> In fact, if we have a way to point into an invalid html document, do we 
>> even need the comment, since the expected result is pointing to it?
>the expected result is a pointer to the spec, while the comments is an 
>observation by the tool of a problem. 

AG::  There are two things going on here.

One is the potential to have further structure within the description of the
problem.  Between pointing into the text of the mis-parsing file and pointing
into the syntax, a syntax error could be pretty well fully formalized.  With
better language-description language we could go further to a complete formal
statement of the perceived discrepancy.

The other is the structure of EARL.  Syntax errors are but one example of what
EARL will capture. We need "what was wrong" to accept unstructured natural
language, too.  The trick is to structure the language with compartments, like
svg:desc that accept mixed content and we know how to interpret appropriate
structures when they appear therein.  Don't assume that the sub-parts are
one-to-one with fixed content models or you don't have enough flexibility to
deal with the mission statement.

We need EARL to be able to express the assertion that there is a problem,
together with as much of the steps needed to re-create the problem as we can
get, from someone who can only tell us that there is a problem, to the person
or work station where the origin of the problem can be ferreted out.  The
of detail should be variable within the overall functional flow of a
transaction description.


>> Len
>> At 04:49 PM 3/14/01 +0100, Daniel Dardailler wrote:
>> >3) a web page has an HTML syntax error
>> >
>> >on 2001/03/14, says that:
>> >  web-content at <>
>> >  fails
>> >    comments: syntax error, line12, missing ul
>> > 
>> >  according to <>
>> >  auto mode
>> >  purpose: checking html4 dtd content model
>> >  expected result: see
>> --
>> Leonard R. Kasday, Ph.D.
>> Institute on Disabilities/UAP and Dept. of Electrical Engineering at
>> University
>> (215) 204-2247 (voice)                 (800) 750-7428 (TTY)
>> <>        
>> Chair, W3C Web Accessibility Initiative Evaluation and Repair Tools Group
>> <>
>> The WAVE web page accessibility evaluation assistant: 

Received on Thursday, 15 March 2001 15:22:12 UTC