- From: Laura Carlson <laura.lee.carlson@gmail.com>
- Date: Wed, 23 May 2007 07:36:31 -0500
- To: www-html@w3.org, public-html@w3.org
- Cc: mleikam@csulb.edu
Another id/header example "in the wild": -- Michael Leikam <mleikam@csulb.edu> wrote: > here is one place where I'm using two complex tables with > id/headers: http://daf.csulb.edu/printshop/copying.html > > Frankly, they're a pain in the ass and I advise avoiding them wherever > possible. Besides, most tables benefit greatly in terms of usability > from being simplified. In the vast majority of cases, complex tables > are an anachronistic carryover from the print world where space is at > a premium and display functionality is absent. > > However, if colspan/rowspan are in the spec and supported by user > agents, people will build complex data tables and there must be some > way to make them accessible. Do those who oppose ids/headers have > any ideas on how a screen reader should parse an arbitrarily complex > table? If there's a better alternative out there, I'd like to hear > it. In many cases, it's possible to guess correctly with a > relatively simple heuristic, but I'd hope HTML5 is a bit more robust > than that. Laura
Received on Wednesday, 23 May 2007 12:37:21 UTC