- From: Nick Kew <nick@webthing.com>
- Date: Tue, 30 Apr 2002 00:53:18 +0100 (BST)
- To: Phill Jenkins <pjenkins@us.ibm.com>
- cc: Access Systems <accessys@smart.net>, <w3c-wai-ig@w3.org>
On Mon, 29 Apr 2002, Phill Jenkins wrote: > > > ... but I mention the Linux to > >illustrate that any solutions MUST be OS neutral and not depend on > >proprietary software that is not provided free. > > I don't understand what free has to do meeting the 508 or W3C standards. Standards? I think you mean guidelines. WCAG makes some mention of using (open) standards in (implied) preference to (proprietary) hacks. Open standards are by definition OS-neutral. So I think it reasonable to infer that solutions MUST be OS-neutral. That doesn't imply that equivalent solutions MUST exist on different OSs, but it does imply there should be no barriers to their existing - such as proprietary, undocumented extensions to standard protocols. A position that states "there is a solution that works on some platform, THEREFORE all is OK" is inherently unsatisfactory, for all the usual reasons of lock-in. Moreover, a solution that requires an OS that is itself costly and will only run on expensive modern hardware is utterly inacessible to many, for obvious reasons. > By the way, since LYNX supports FRAMES, if authors put usable titles on > them, would that end the debate about frames? I can see several answers to that. Navigating framed sites with Lynx can be b***** annoying. Titles help, but a decent noframes (say, ten minutes thought on the part of the author) is much better. OTOH, many framed sites present bigger problems to graphical browsers, especially where they use those Pure Evil attributes that prevent the things being resized. OTOH, I happen to believe frames can be well-used, too. It's just a shame it happens so rarely. > Also is LYNX supported JavaScript, would that end the debate about > JavaScript? I take it you're not responsible for security on your corporate network? More generally, I'd say what Lynx or any other individual browser supports is of little relevance from the authoring end, unless an author is tempted to develop "for" it *at the expense of* standards compliance. -- Nick Kew Available for contract work - Programming, Unix, Networking, Markup, etc.
Received on Monday, 29 April 2002 19:53:30 UTC