- From: Chris Brainerd <Chris.Brainerd@cds.hawaii.edu>
- Date: Thu, 7 Aug 2003 17:36:20 -1000
- To: "Kynn Bartlett" <kynn@idyllmtn.com>, "David MacDonald" <befree@magma.ca>
- Cc: "Ben Caldwell" <caldwell@trace.wisc.edu>, <michaelc@watchfire.com>, <w3c-wai-gl@w3.org>
May I provide an alternative to the views expressed in Kynn Bartlett's message regarding link text. Poor link text does not only affect JAWS users who may choose to view a list of links out of context, but also those with cognitive disabilities and all surfers. Studies by Jakob Nielsen and others show most people only 'scan' pages and do not read them literally. An important aspect of scanning is looking for the very visible hyperlinks. Linked text should provide some context without the surrounding text. On the other hand, structured markup is invisible to non-JAWS users and provides little benefit unless the surrounding elements are also organized in a logical, hierarchal fashion that supports visible scanning of headings. I've also seen instances of UL and DL used for layout purposes that are confusing when read with a screen reader because the information does not represent a list. Chris Brainerd Instructional Designer Real Choices ACCESS Center on Disability Studies University of Hawaii Chris.brainerd@cds.hawaii.edu 808-956-9356 -----Original Message----- From: Kynn Bartlett [mailto:kynn@idyllmtn.com] Sent: Wednesday, August 06, 2003 11:29 AM To: David MacDonald Cc: 'Ben Caldwell'; michaelc@watchfire.com; w3c-wai-gl@w3.org Subject: Re: Techniques revision - Meaningful link names On Wednesday, August 6, 2003, at 10:58 AM, David MacDonald wrote: > So I recommend that we not justify any use ambiguous link names. Kibitzing: Such a requirement basically says that all links must make sense when taken out of context -- this is a bad thing if we are also emphasizing the importance of structure in documents. If we say that document structure is important and you must use it, and then we say that ALSO the page must be usable WITHOUT structure, we are asking an awful lot. Something like this: <section> <h>Teach Yourself CSS in 24 Hours</h> <p>Blah blah blah</p> <p><a href="http://amazon.example.com/blahblah">Buy it!</a></p> </section> ...is not ambiguous. From context, it makes perfect sense what is being presented here. Likewise: <dl> <dt>Kynn Bartlett</dt> <dd><a href="mailto:kynn@idyllmtn.com">Email</a>, <a href="http://kynn.com/">Website</a></dd> <dt>David MacDonald</dt> <dd><a href="mailto:befree@magma.ca">Email</a> </dl> ...is not ambiguous either. In both cases, the context provides meaning to the link. Removing the context -- as assistive technologies have been doing -- is not a proper solution for accessibility's sake. In fact, it is a practice which is harmful to accessibility. Constructing an outline based on headings is a good thing. Constructing such an outline and then adding links in is also a good thing. Even providing a straight list of links is reasonable, although not particularly useful since context is removed. However, Web developers should not be required to support this third option -- a flat list of links titles -- because that is not how markup is meant to function. Sometimes, context can be removed. Other times it cannot. The heading tags (<h1> ... <h6> or <h>) are designed to be able to abstract document structure in such a manner. The <a> tag is most definitely not built this way, at least when it comes to href links. (A list of ANCHORS, formed with <a name>, on the other hand, is a good idea -- and such anchors should probably have title attributes attached. Sadly, though, this is not currently supported by user agents, is it?) I caution you to very seriously reconsider the notion of requiring pages to "make sense" with vital context -- including accessibility elements and attributes -- removed from the presentation (as with lists of links). Such a practice is ultimately harmful to the cause of accessibility because (a) it can result in poor UI design, (b) it can result in rejection of accessibility by informed Web developers, (c) it encourages a practice which goes against the idea of structured markup, and (d) it does not encourage the assistive technology vendors to improve their abilities to generate document abstracts. --Kynn -- Kynn Bartlett <kynn@idyllmtn.com> http://kynn.com Chief Technologist, Idyll Mountain http://idyllmtn.com Author, CSS in 24 Hours http://cssin24hours.com Inland Anti-Empire Blog http://blog.kynn.com/iae Shock & Awe Blog http://blog.kynn.com/shock
Received on Thursday, 7 August 2003 23:36:22 UTC