- From: Gregg Vanderheiden <gv@trace.wisc.edu>
- Date: Fri, 02 May 2003 15:17:58 -0500
- To: w3c-wai-gl@w3.org
- Message-id: <012201c310e7$ec5bb9f0$ac17a8c0@TOSHIBATABLET>
Web Content Accessibility Guidelines - Reformulation At the WCAG face-to-face at the CSUN conference, a quite different approach to web accessibility was raised with the group in connections with the discussions around conformance. This approach actually consisted of a number of different concepts. Discussions at the face-to-face and subsequently have embraced some, modified some and rejected or not adopted others. This document is an attempt to capture the current thoughts suggestions and issues of the working group members at this time. In order to not confuse the reader, the original proposals and forms are not presented. Instead, the description tries to focus just on the current thinking and areas of consensus or interest. It must be emphasized that the group is still thinking this through, so the description below is not a final, concise and fully integrated proposal or approach. It should give the reader a good idea, however, of the concepts being considered. Overview The revised approach to the guidelines is based on a number of observations and underlying principles. Observation #1: The overall checkpoints and success criteria can be divided into two general categories. A) Those guidelines that talk about adding supplemental (hidden) mark up, tagging or information which makes the page more accessible (e.g. Alt text, table/header mark up, making sure that any structure in the document is machine readable, etc.), but does not prescribe the default presentation of the page. B) Checkpoints and success criteria that make pages more accessible by prescribing how the default presentation of the page would be constructed or appear. (e.g., "add structure", "use or avoid specific language", "visually emphasized structure", supplement text with images, etc.) Observation #2: The recommendations also varied in terms of their general applicability. Some of the checkpoints and success criteria could be applied to all websites or content while others could not. Observation #3: Putting checkpoints and success criteria into the required category that could not be applied to all sites would mean that some sites/content could not make any accessibility claim. Observation #4: If the checkpoints and success criteria were included in the minimum which prescribed the language or appearance or presentation of pages, then it was likely that we would run into problems in ratification of the guidelines and even possibly challenges as restricting free speech. If we wanted to have the "minimum required" portion of the guidelines be consistent across implementers, agencies, countries, etc. Then, we would need to be quite careful about what we put into the minimum category. Observation #5: Different countries and agencies are interested in going to differing degrees of comprehensiveness with regard to their individual requirements. That is, different countries have shown an interest in wanting to create guidelines which had slightly different emphasis in terms of the overall set of guidelines they would require. Flexibility in determining what they wanted to have as a set of requirements was therefore seen as important. Observation #6 There is great concern among web developers that countries not each create different sets of guidelines which are either mutually exclusive or contradictory since they often have to create web sites that would be viewable from different countries and/or mirrored in different countries. Web developers also find it impossible to keep track of all the variations. However, if the only variation between entities was the items they chose from a standard set of guidelines, then at most, they would need to create pages which complied with the union or complete set of guidelines represented by the various jurisdictions they were interested in conforming with. They could just form a superset and build the web pages to that and they would comply with all of the various jurisdictions since they would all be proper subsets of the WCAG guidelines. Presumably, each of these different jurisdictions would use the minimum required set, plus some proper subset of the remaining WCAG ) checkpoints and success criteria. Observation #7: Once you got past the minimum, it was useful to have some other targets to shoot for. That is, it is useful to have some other group of guidelines beyond the minimum which a person could strive to "meet all of". This was likely to yield better conformance than simply having an open ended claim once you got beyond the minimum level. However, it might be possible to simply have something that would provide a measure beyond the minimum (e.g. compliance plus twelve) which might provide a similar motivation for people to keep trying to improve their sites. Observation #8: The use of server based services (e.g. transcoders and other services /processes that transform content) are a powerful new tool that can help to automatically convert less accessible content into more accessible content and convert special proprietary formats into general, standard and highly accessible formats. In writing the guidelines and the conformance, server side and "proxy server" or "third party server processed pages" should all be acknowledged and allowed as effective accessibility strategies for web content providers where these strategies are available at the same level as the default serving of pages (e.g. the user requiring these special services can get them at the same cost, with the same speed, and from the same locations as a user who is just accessing the "default" pages). Proposed Restructuring Approaches and Principles 1. A new conformance approach would be used that divides the guidelines and success criteria into 1) a minimum required set and 2) a group of additional guidelines and measures which would increase the accessibility beyond the minimum set. 2. The minimum set (Group A) would be restricted to those checkpoints that do not prescribe the language and presentation form of the default presentation of the page. 3. The body of additional checkpoints and success criteria may or may not be divided into additional groups. 4. At present, we see the "Extended Guidelines and Measures" as breaking down into two groups (group 2 and group 3) - one group (Group B) representing those things which do effect presentation, but which could be generally be applied to most or all sites and - a second group (Group C) which represented additional measures which could be taken, but which would not be practical or possible for all sites or which might require that the site content be rendered in multiple formats. 5. The last proposed set of descriptors for this three group model as of May 1st is: GROUP A - Those measures that can provide access without changing or constraining the presentation of the page - and that it was felt could be reasonably applied to all web content or sites. * These items address compatibility of the Web content with assistive technologies for all disabilities. * They also include accessibility that can be achieved using mass market web browsing technologies but that do not affect the default view of the content by all users. * These measures would constitute the core required set of checkpoints GROUP B. - Those measures that allow access beyond Group A but can be reasonably applied to all types of web content or sites. * These measures affect the presentation of the pages somewhat in order to make them more accessible. * They often allow access to some individuals without requiring any assistive technology. * These do not address all disabilities but allow many to access web content using mass market web browsing technologies alone. GROUP C. - Those measures that improve access, either directly or via assistive technology, beyond Group A but that cannot be applied reasonably to all web sites or content. * Some would require multiple presentations of the information or targeting of the web site to individuals with particular functional limitations. * Some would be an unreasonable amount of work to expect of all web content or sites. 6. Conformance with the WCAG guidelines would require the minimum or first group of items all be met. No conformance could be claimed without meeting all of these first items. 7. Entities wishing to create their own set of guidelines based on WCAG would have to include all of the minimum set or minimum group. They would be free to choose include and/or admit any other checkpoints or success criteria beyond, or from the rest of the WCAG guidelines as they saw fit. However, in order to avoid conflict, it would be extremely highly recommended that they not change any of the checkpoints or success criteria, but rather just adopt or exclude them. (If there were any items in the minimum set that particular entities had trouble with, it would be optimal if they could identify the items and the issues before adoption of the final guidelines so that the issues could be accommodated. 8. Page author/websites could claim conformance with the guidelines if any of the following were true: a. The default pages from their site met the guidelines. b. Alternate versions of the pages (which had the same content and functionality and were updated simultaneously with the default pages), were available from the same URI This could be done either automatically or from an accessible link on that same page. OR c. A link or mechanism at the same URI as the default page was available that routed the content through a (local or remote) processor that would change the content into a form which would conform with the WCAG guidelines (i.e. a page that was run through the transcoding server that was available for the same cost and the same hours and the same speed, etc. as the default page/content). Vendors may opt to either maintain and run such a transcoding/processing server as a part of their site or may contract with a third party to make such a server available which would quickly and efficiently process any pages from their site as required by users.
Received on Friday, 2 May 2003 16:18:18 UTC