RE: Navigational features

Bruce writes that the term "programmatically determined" is very
difficult to understand, and that having to refer repeatedly to the
Glossary interrupts the flow of reading the Guidelines:
 
<blockquote>
And it is a fine definition.  That is *not* the problem.

[Gregg writes:]
> It is a rather important definition and a term specific to this work.

[and Bruce resumes:]
Understood.  The problem is that the term is used so frequently and so
particularly that even after repeated trips to the glossary the phrase
interrupts naturally reading of success criteria.  I honestly believe
that replacing "programmatically determined" with "marklar" would be an
improvement -- readers would not be fooled into thinking they understood
what was meant.

 
</blockquote>
 
For what it's worth, the term "programmatically determined" appears in
the following 10 success criteria:
 
- 1.3.1 (perceivable structures within the content can be
programmatically determined)
-1.3.2 When information is conveyed  by color, the color can be
programmatically determined or the information is also conveyed by
another means that does not depend on the user's ability to
differentiate colors.
-1.3.3 Information that is conveyed by variations in presentation of
text is also conveyed in text or the variations in presentation of text
can be programmatically determined
- 1.3.5 When content is arranged in a sequence that affects its meaning,
that sequence can be programmatically determined.
2.4.1 Navigational features within the content can be programmatically
determined.
- 3.1.1 The primary natural language or languages of the delivery ynit
can be programmatically determined.
-3.1.2 - The natural language of each foreign  passage or phrase in the
content can be programmatically determined.
- 4.2.3 The role, state, and value can be programmatically determined
for every user interface component of the web content that accepts input
from the user
- 4.2.4 The label of each user interface control that accepts input from
the user can be programmatically determined and is explicitly associated
with the
control. 
- 4.2.6 Changes to content, structure, selection, focus, attributes,
values, state, and relationships can be programmatically determined. 

In 7 of the 10 success criteria listed above, "programmatically
determined" comes at the end of the sentence. In these cases, the phrase
"programmatically determined" *could* be replaced by the definition
without any other change to the success criterion. This would add a
total of 91 words to the Guidelines.
In the other three success criteria listed above, the phrase
"programmatically determined" occurs in the interior of the sentence. We
would probably have to rewrite the sentence in order to use the
definition instead of the phrase "programmatically determined." 
 
This would be easy for 4.2.4 (we could change it to read: 
- 4.2.4 The label of each user interface control that accepts input from
the user is explicitly associated with the
control and can be programmatically determined.  This would allow
replacement of "programmatically determined" with the definition)
 
It would be a little trickier for the other two.
 
Bruce, would this sort of edit address your concerns? There *would* be a
tradeoff-- the Guidelines document would become somewhat longer (though
the additional words would be diispersed through the document rather
than being concentrated in one place).
 
Thanks.
John
 
Dr. John M. Slatin, Director 
Accessibility Institute
University of Texas at Austin 
FAC 248C 
1 University Station G9600 
Austin, TX 78712 
ph 512-495-4288, fax 512-495-4524 
email jslatin@mail.utexas.edu 
Web  <http://www.ital.utexas.edu/>
http://www.utexas.edu/research/accessibility 

-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On
Behalf Of Gregg Vanderheiden
Sent: Sunday, December 04, 2005 2:02 PM
To: 'Bailey, Bruce'; 'Guide Lines list'
Subject: RE: Navigational features



Hi Bruce

 

Thanks for trying on this one.  I would really like to find another
term.  The ones you suggest though have the same problem.  They all
sound like phrases that one can guess the meaning of (and one would be
wrong).  In fact those would all lead a person to the problem Jim
Thatcher was citing.  

 

So let's see.   We need to bring some of the essential parts of the
definition up into the phrase so that the phrase is more self
explanatory.   OR we need to coin a term that people won't guess the
meaning of  (as you suggested with your  "marklar" - though that seems
to indicate markup in some way).    

 

AT determinable?  (but we don't just mean AT.  And that definition begs
one to define AT exactly)

 

User Agent Determinable (this is better - and since AT is a user agent.
Is it too narrow?   Does it actually get to what we mean?   If we don't
mean user agent determinable, what do we think is going to be doing the
determination?    

 

This one still has the problem that people will try to interpret it
without reading the definition and will exclude AT from the User Agent
Determinable.   Or they could say "well of course the user agent can
determine it or they couldn't render it" .    Actually that criticism
doesn't make sense since the user agent only needs to determine what it
should be rendered like to render it - it doesn't need to know structure
etc.    So may be it does work. 

 

Sorry for stream of consciousness - but was trying to walk Bruce's idea
since we are getting so much confusion on this and "programmatically
determined" is rather opaque to most readers.  

 

If we pluralize Agent to indicate that it might be (and needs to be)
more than just one agent (so AT is in there too) we get

"User Agents Determinable"     

It sounds enough like what it means.   It is just awkward enough to get
the reader to read the definition - and to remember AT and not just the
'browser' user agent.

 

 

What do you think?

 

Got other, better ideas?     Programmatically determined is a bit.....
well....   non-optimal as a term. 

 


Gregg

 -- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
Professor - Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 

 

 


  _____  


From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On
Behalf Of Bailey, Bruce
Sent: Sunday, December 04, 2005 10:32 AM
To: Gregg Vanderheiden; Guide Lines list
Subject: RE: Navigational features

> Can you be a little more specific?

The current subtly of the discussion about heading leads me to conclude
that the meaning behind "programmatically determined" is more fluid than
is being acknowledged.  But I will put that issue aside for the moment.

> "Programmatically determined" is defined in the glossary.
> If the definition isn't clear we need to fix it. 

And it is a fine definition.  That is *not* the problem.

> It is a rather important definition and a term specific to this work.

Understood.  The problem is that the term is used so frequently and so
particularly that even after repeated trips to the glossary the phrase
interrupts naturally reading of success criteria.  I honestly believe
that replacing "programmatically determined" with "marklar" would be an
improvement -- readers would not be fooled into thinking they understood
what was meant.

It is not about writing a good definition.  It is about finding a better
term.  I have to believe that there is a term of art that can be used
with the intended definition.  Machine processable?  Reducible to an
algorithm?  Wholly deterministic?  NP-complete?