- From: Gregory J. Rosmaita <oedipus@hicom.net>
- Date: Sat, 28 Jul 2007 12:55:25 -0400
- To: Lachlan Hunt <lachlan.hunt@lachy.id.au>, public-html@w3.org, wai-xtech@w3.org
aloha, lachlan! i'm not jason, nor do i pretend to be as sagacious as jason, but perhaps i can explain the thrust of jason's message (with apologies to jason, of course) i'm quite happy to hear that we are in violent agreement on so much, but i did want to address the question of implicit versus explicit bindings. the reason youtube works for you is that you can perceive the entire page, and, thus, can sub-consciously associate disperate bits of data into a comprehensive whole. for those who -- for whatever reason -- cannot, however, and explicit binding is necessary, for how else is the user's assisstive technology to communicate to the user that item x is related to item y unless there is an explicit programmatic binding between the two. a screen-reader actually has some of the same limitations as a blind person -- i can tell you that i am feeling a plastered wall, but unless someone or something alerts me to the fact that what i'm feeling is actually a fresco, i will never know that i have encountered anything other than a blank plastered wall. what makes screen readers work is the ability to associate programmatic bindings with objects, so as to communicate to the user the nature, purpose and function of that object. failing that, a screen reader is reduced to "screen scraping" to obtain the contents of a document, akin to a static snapshot, while the page may dynamically change in response to user action; that change will not be available to the user, because he or she is working off a snapshot of the page as it was when first loaded, hence the constant need to "refresh" the screen reader's snapshot of the document, in order for it (and by extention, the user) to ascertain what -- or even whether -- change in content or functionality has occurred. when i am directed to a youtube page by a friend, the only interaction i have with the page is that the video plays on-load; i cannot control, restart, slow down, or find the scattered meta-data on the page, which means that i can only rely on URIs provided to me by others in order to listen to youtube content. i believe that google is attempting to rectify this situation with their beta version of a new interface for youtube, but the fact remains that without an explicit binding of the metadata to the object it describes/annotates, a non-visual user has no means at his or her disposal to mentally perform such bindings by using the sighted person's gestalt view of the page. for a screen reader user (or a blind person in general) there is no such thing as a gestalt view, other than that constructed by their assistive technology, which cannot communicate a working knowledge of the document to the user without explicit bindings. which is why the FIELDSET, LEGEND, LABEL model is so important to HTML forms as they currently exist. one cannot expect, as stated in the HTML5 draft that assistive technologies can obtain implicit groupings by walking the DOM -- there must be some sort of binding between objects and their descriptors or equivalents... that the FIELDSET, LEGEND, LABEL model is not more widely used is due mostly to either ignorance on the part of the author, or a lack of imagination when it comes to applying style rules to FIELDSETs and LEGENDs. the bindings provided by FIELDSET LEGEND and LABEL are especially important in forms embedded in a TABLE for layout purposes -- when i encounter a table-ized form, i first must inspect the form using TABLE navigation mode, which does not communicate to me either form controls or their state; in order to use the form, one must enter into "forms mode", but in order to ascertain that the correct form field has been chosen, one must leave forms mode, and return to either TABLE navigation mode or inspect the table-ized form using a virtual cursor, which is not the most effective way to obtain meaningful information (the only analogy i can offer is that using a virtual cursor to inspect a document instance is very much like the experience of a blind person without a cane or guide dog "air swimming" the room in an attempt to figure out its dimensions, contents, etc.) programmatic bindings, therefore, are essential to anyone who cannot perceive the document instance as a single, integrated entity, which is a far larger user base than simply those who cannot see. i hope that this helps to explain why EXPLICIT programmatic bindings are necesssary; especially as the perception of implicit bindings is in the eye of the beholder, and, therefore, implicit associations are extremely untrustworthy, not just from the point of view of a person's physical capacity to perceive implicit associations, but for reasons of, yes, internationalization, for one man's convention is another man's cundundrum. gregory. --------------------------------------------------------------------- CARTESIAN, adj. Relating to Descartes, a famous philosopher, author of the celebrated dictum, Cogito ergo sum--whereby he was pleased to suppose he demonstrated the reality of human existence. The dictum might be improved, however, thus: Cogito cogito ergo cogito sum--"I think that I think, therefore I think that I am;" as close an approach to certainty as any philosopher has yet made. -- Ambrose Bierce, _The Devil's Dictionary_ --------------------------------------------------------------------- Gregory J. Rosmaita, oedipus@hicom.net Camera Obscura: http://www.hicom.net/~oedipus/index.html ---------------------------------------------------------------------
Received on Saturday, 28 July 2007 16:56:23 UTC