- From: David Woolley <forums@david-woolley.me.uk>
- Date: Sun, 30 Nov 2014 10:55:31 +0000
- To: w3c-wai-ig@w3.org
On 30/11/14 03:23, Adam Cooper wrote: > The question is who chooses to disable or which user agents don't > support CSS? > > I’d be very interested to see some dependable statistics on this. > The main purpose of this test is check whether the pages are written in HTML, or just using the HTML vocabulary. From an accessibility point of view, and given that most authors only provide visual styling, the main benefit is for the blind, although more generally, proper HTML semantics mean that a browser can re-target the page for other disabilities (e.g. pick out key navigation features, for those confused by the complex visual structure, or insert large guard bands around links, for those with poor motor control - the lack of such tools probably indicates the lack of well written HTML, rather than the lack of a market). (It also makes mechanical processing easier, although, unlike the with the original concept, current authors have mixed feelings towards this; they want Google to find their page, but they don't want the information to be extractable.) Whilst I've mainly turned it off recently to get round "subscribe to us" popups, I have, in the past, turned it off to get round particularly badly styled pages, or pages that require a version of IE that is not supported by the version of Windows I was using. I do remember a case where I was frustrated that Android doesn't allow you to turn off CSS and the only way of making the page usable would have been to do so, although I forget which page that was. The proper work flow for communications, is to decide the logical structure of what you want to say, then flesh it out with words. You should be able to write the HMTL at this point. Next you think about layout and visual effects, and add the CSS. Finally, you might want to tweak the wording, or re-order the structure, to improve the final presentation. Most pages seem to be designed the opposite way round, possibly with the logical structure never being considered.
Received on Sunday, 30 November 2014 10:55:42 UTC