- From: Karl Groves <karlgroves@gmail.com>
- Date: Thu, 30 Jan 2014 14:12:08 -0500
- To: Shadi Abou-Zahra <shadi@w3.org>
- Cc: public-wai-evaltf@w3.org
- Message-ID: <CABScKPABaeniXEMjvE6nBScY5GXZJqy0+FcuJ3kUBBS91YpGFw@mail.gmail.com>
Shadi, The portion you cited is clear enough IMO, however the guidance I cited seems to be in direct conflict, especially as the words "full page" contains a link to normative WCAG documentation stating that partial pages cannot be considered in conformance claims. On Thu, Jan 30, 2014 at 12:35 PM, Shadi Abou-Zahra <shadi@w3.org> wrote: > Hi Karl, > > Thank you for your prompt feedback! > > There is a note in this section that reads: > [[ > While it is important to check the conformance of each web page and web > page state in the sample to each WCAG 2.0 Success Criterion, repeated > components such as the header, navigation bars, search form, and others > usually do not need to be re-evaluated on each occurrence. For example, if > the same header code is used on two web pages then it is evaluated on the > first only in most situations. > ]] > > A similar statement is earlier on in the introduction to Step 4. > > Maybe these need to be made more clear? Do you have some suggestions? > > Best, > Shadi > > > > On 30.1.2014 17:11, Karl Groves wrote: > >> Apologies for my hasty striking of the send button. Here is the full >> response: >> >> WCAG-EM contains the statement, >> "Methodology Requirement 4.a: Check that each full page (web page and web >> page state) in the selected sample satisfies each of the WCAG 2.0 Success >> Criteria of the target conformance level." >> >> This statement and supporting material make it appear as though an >> *entire* >> web page must be tested in its entirety in order to satisfy the >> requirements. This ignores the fact that websites composed of a series of >> static pages is the exception, rather than the rule. The vast majority >> of >> websites these days are presented through the use of some server-side >> means >> to generate the final interface - from basic SSI to template systems, to >> complex web applications being generated from content management systems >> and everything in between. In some cases the server-side generation >> merely >> wraps content in a global wrapper while in others the server-side code >> merges data and markup as part of a template. This methodology seems to >> ignore that fact and may mislead the reader into believing that the final >> rendered web page(s) must be tested in their entirety. This risks >> requiring >> the tester to perform needless duplicate work. >> >> While the goal is and should remain a full page that is accessible, the >> testing of the entire page is unnecessary in reaching this goal. A >> sensible >> sampling strategy can and should use representative samples of those >> document features which contribute to the page. Multiple instances of UI >> components that are shared across pages do not need to be tested multiple >> times and should be ignored or avoided during the test effort as their >> inclusion will not add any new data but will add needless effort. >> >> >> On Thu, Jan 30, 2014 at 11:08 AM, Karl Groves <karlgroves@gmail.com> >> wrote: >> >> WCAG-EM contains the statement, >>> "Methodology Requirement 4.a: Check that each full page (web page and web >>> page state) in the selected sample satisfies each of the WCAG 2.0 Success >>> Criteria of the target conformance level." >>> >>> This statement and supporting material make it appear as though an >>> *entire* web page must be tested in its entirety in order to satisfy the >>> requirements. This ignores the fact that websites composed of a series >>> of >>> static pages is the exception, rather than the rule. The vast majority >>> of >>> websites these days are presented through the use of some server-side >>> means >>> to generate the final interface - from basic SSI to template systems, to >>> complex web applications being generated from content management systems >>> and everything in between. In some cases the server-side generation >>> merely >>> wraps content in a global wrapper while in others the server-side code >>> merges data and markup as part of a template. This methodology seems to >>> ignore that fact and may mislead the reader into believing that the final >>> rendered web page(s) must be tested in their entirety. This risks >>> requiring >>> the tester to perform needless duplicate work. >>> >>> While the goal is and should remain a full page that is accessible, the >>> testing of the entire page is unnecessary >>> >>> >>> -- >>> Karl Groves >>> www.karlgroves.com >>> @karlgroves >>> http://www.linkedin.com/in/karlgroves >>> Phone: +1 410.541.6829 >>> >>> www.botsmasher.com >>> www.a11ybuzz.com >>> www.mothereffingtoolconfuser.com >>> Day One Wordpress Theme: https://bitbucket.org/karlgroves/day-one-theme >>> >>> >> >> >> > -- > Shadi Abou-Zahra - http://www.w3.org/People/shadi/ > Activity Lead, W3C/WAI International Program Office > Evaluation and Repair Tools Working Group (ERT WG) > Research and Development Working Group (RDWG) > -- Karl Groves www.karlgroves.com @karlgroves http://www.linkedin.com/in/karlgroves Phone: +1 410.541.6829 www.botsmasher.com www.a11ybuzz.com www.mothereffingtoolconfuser.com Day One Wordpress Theme: https://bitbucket.org/karlgroves/day-one-theme
Received on Thursday, 30 January 2014 19:12:36 UTC