W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 1999

Checkpoints 6.3-6.5, 9.3, 10.1-10.3

From: Wendy A Chisholm <chisholm@trace.wisc.edu>
Date: Fri, 02 Apr 1999 15:07:15 -0600
Message-Id: <Version.32.19990402142200.01e03d90@trace.wisc.edu>
To: w3c-wai-gl@w3.org
hello all - i'm back!  and amazed at how much work went into the guidelines
while I was gone.  wow!  thank you all.

I have six concerns about checkpoints 6.3-6.5, 9.3, 10.1-10.3.

1.
I believe the example for 6.3 focuses too much on NOSCRIPT and should
instead say, first and foremost, that the page must remain usable when
dynamic/interactive objects (scripts, applets, etc.) are not loaded (and
we're showing that it is possible to do so).  Some of what scripts create
is pure eye-candy which means that the page will still be usable without
the script loaded (rollovers, highlights, etc.)  Also, some of this
functionality can be provided by a style sheet. 

NOSCRIPT should only be used when the page is unusable without the script
and an alternative or description needs to be provided. Otherwise, we will
end up with pages with NOSCRIPT that provide useless information like "this
script creates image rollovers."  It is similar to providing null alt-text
for certain images.

Therefore, I would like to highlight that a. the page must be usable
without the script (i.e. don't use scripts to generate text, link to other
pages, etc.) then b. if the page is not usable without the script because
the script provides an important mechanism or presentation, use NOSCRIPT to
provide an alternative or a description.


2.
in checkpoint 6.4 - are we sure that user agents will provide
device-independent means to activate event handlers?  

Also, we say "ensure that event handlers are keyboard operable" - this is
device-dependent and we need to be device-independent.  In my experience
(using 4.0 browsers with win98), at this time to get the same effects with
mouseless navigation as with the mouse, both logical (onfocus/onBlur)
events and mouse specific (onMouseOver, onMouseOut) need to be implemented
for each object.  I would suggest making this checkpoint more general:
For scripts and applets, ensure that event handlers are device independent.

Similarly, checkpoint 9.3 suggests using onFocus, etc, and should say
something about using onFocus as well as onMouseOver or something.

The logical/device-independent dilemma is not only a user agent issue, but
a language one as well.  Early in the discussions of this topic an
"onActivate" event was proposed.  Therefore, I don't think it is a user
agent issue but an authoring good practice.  Even when user agents or the
languages have more logical events, it will still be a good authoring issue
to use the correct events.  therefore, I suggest dropping the "until user
agents" statement.

3.
checkpoint 6.5 - this needs to be narrowed in scope or the script portion
incorporated into 6.3.  

The script related ideas seem to fit in with 6.3 - if the page is unusable
without scripts, then provide an alternative - either a NOSCRIPT or a
simulation or something (see discussion on 1st point).  

However, I'm not sure what to do with the FRAMEs and image map piece of it.

This checkpoint was originally intended to convey the idea that if you use
a Java applet for a science experiment and someone can not access Java,
then they might be able to use a movie of the simulation, or a still image
and a text description, or just a text description.  


NOFRAMES, for legacy reasons, is still more of a P2, but this should be
qualified with an "until user agents" clause since more user agents and ATs
are able to navigate frames more effectively.  this is very much an interim
solution, since it is the old browsers that don't show anything for
framesets that need NOFRAMES the most.  Therefore, we might consider adding
it as a checkpoint in Guideline 10 - interim solutions.

4.
checkpoint 10.1 - "do not change the current window without informing the
user"  makes it sound like the author has to pop up a dialog window to let
the user know the window is changing.  In the dynamic pages I have recently
been testing, we use style sheets to hide and display content.  A screen
reader will read the text when it displayed, therefore we have effectively
notified the user by "causing" the screen reader to speak.  I would suggest
changing this to read, "do not change the current window unless the user is
able to determine the change."  However, this is dangerously vague.
Perhaps in this case, we have already satisfied the "until user agents"
clause of this checkpoint (as currently worded).  Also, this seems to fit
better with guideline 7 (user control of time-sensitive changes)

5.
checkpoint 10.2 - This seems that it ought to be a note for checkpoint
12.4.  I would suggest rewording, 12.4 to read, "Associate labels
explicitly with their controls.  Note.  when controls are not explicitly
associated, ensure that the label is properly positioned.  The label must
immediately, ....rest of 10.2..."

6.
checkpoint 10.3 - It has been mentioned before that we ought to move 10.3
to the table section.  I have often argued against it, since it is an
interim solution.  However, we have interim solutions all over now (with
the "until user agents" clause) and I think it fits best with tables.  Note
that the rationale for guideline 10 only mentions checkpoints 10.4 and 10.5
and based on my previous comments, we should do away with all but these
last two (and perhaps add one for NOFRAMES).

thoughts?
--wendy
Received on Friday, 2 April 1999 16:12:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:46:59 GMT