- From: Livio Mondini <livio.mondini@gmail.com>
- Date: Wed, 26 Sep 2007 10:07:32 +0200
- To: "Benjamin Hawkes-Lewis" <bhawkeslewis@googlemail.com>
- Cc: "Al Gilman" <Alfred.S.Gilman@ieee.org>, "Karl Dubost" <karl@w3.org>, "www-html@w3.org" <www-html@w3.org>
2007/9/26, Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>: > 1. What problem you're trying to solve that affects authors or consumers > of HTML or XHTML? Ambiguity and complexity required at user agent and designer for describe and understand type of table (some attributes, class, role, ecc ecc) > 2. What possible solutions there might be? Declaring on dtd two type of table, you solve at radix all problems. One element for layout table, and a element for data table. This clear problems at all levels. > 3. Which solution would be best, in your view, and why? For me, the best way is two separate elements on dtd. > Because it's not at all clear from what you're saying. Yes, i know, but think on responses here: all claim css. Is not a problem of css. Is indifferent type of character, color, spacing, all typography and positioning ecc ecc in table itself. Try to think on UA point of view, or like a blind person. For example, wcag2 claim on eliminate summary attribute on layout table. Is an error for me, disambiguate table model and maybe allow designer decision on that. Now, UA like Jaws speak for all tables [table]. Is correct, html permit this. Maybe tomorrow Jaws or other parser read <layout_table> and speak nothing, or do some elaborations. Sorry for my poor english, i try to explain a complex problem, is necessary to refer at an abstract model, without css presentation. Dont think on table represented on monitor, the level that i try to discuss is more bottom, on markup parser. Livio
Received on Wednesday, 26 September 2007 08:07:51 UTC