W3C logo Web Accessibility Initiative (WAI) logo | EOWG Home

EOWG Minutes, August 10, 2001


Outreach Updates

WL: Eric Miller, chair of Semantic Web working group, is a Yuri Rubinsky fan. Reminds us that accessibility is an imperative, not a add-on.

LL: Attended a meeting of the Consumer Disability Telecommunication Access Committee of the SBC. Very little content on access. Issue, how distribute the cost for universal access between users and providers. No longer on committee.

HB: Excellent webcast of keynote presentation by Mike Paciello for a MacroMedia workshop on Section 508. Mike introducing accessibility, ICADD, WAI, WCAG, to show precedents for Section 508. Audience was government attendees. Mike has consented to my posting this. I played it on Real Player running on Windows. Even the real-time captioning is readable at 2x magnification. Link: Macromedia Accessibility Webcast

Evaluating Web Sites for Accessibility

JB: (continuing from where we left off last week) Link:
dated 3 August 2001. starting with Section "3. Conformance Evaluation to WCAG 1.0".

WL: Preliminary review doesn't cover any validation. There is no expectation that documents are valid there.

JB: Emphasis there was to allow novices who are not markup-aware to do it.

LL: 3.2.2 expects user to use at least two accessibility evaluation tools. Would like to be able to depend on valid document.

WL: Should be able to depend on valid documents, so state.

JB: Clarify accessibility evaluation tools including WAVE, BOBBY, A-Prompt.

LL: Bad html can cause many accessibility problems.

JB: Should we add a validity check in the preliminary review?

HB: I believe Tidy will likely find validity problems in most sites.

JB: Validity should be optional in the preliminary review, keep it simple.

LL: If you discover accessibility review problems, then you should also evaluate your markup for validity. Some validity problems may not affect accessibility.

MRK: Concern for effects of invalidity.

JB: 2.5 Add in first summary, "if validation is included..."

3.3 Manual Evaluation

LL: There are many combinations of versions and platforms.

WL: Operating Systems (OS) are Mac, Linux, Windows.

HB: [Note added: We are not now considering the simple OS for hand-held devices.]

JB: Examine page selection sample development.

JB: Test with at least three browsers [MSIE, NSN, Opera, Amaya, Home Page Reader, Lynx, WidowEyes, ...].

WL: Designers focus on keeping up with the different releases of MSIE and NSN.

LL: Cross-platform and browser testing would be a future ideal; many businesses aren't doing this testing today.

MRK: "Browser tricks" [HB: comment missed, believe it was about browser sniffing to vary content to overcome differences.]

JB: Could be 3 OS x 3 browsers x 2 levels = 18 variants for test. Simplify to suggest testing many (5) combinations.

LL: Many in business ignore Mac and linux. Putting a testing lab together may not be needed for WCAG 1.0.

JB: Good for usability, but not needed in our focus.

DS: Agree to test with variety. People want to use their available resources. Small business are unlikely to have variety.

LL: Agree. How to help them is important.

JB: Suggest "choose among."

LL: MSIE and NSN are large part of market.

JB: Minimum: Choose three among choices of browsers, their versions, and the operating system.

HB: Bobby has choice for selecting and reporting results that would occur among several browsers and their versions.

DS: Three choices seems skimpy. Consider available resources.

JB: Select three different configurations among different graphical user interfaces, and browser versions running on different platforms. For example, MSIE, NSN, Opera, in recent versions.

JB: Section 3 is trying to give whole lot of choices.

LL: Conformance with WCAG is first step.

WL: If do that, wouldn't need to do anything else.

LL: Makes first step possible.

WL: Make this testing seem like what they already do. Normally do "performance" check. Make it "performance/conformance" check.

DS: Great point. Practical.

JB: Call it something that looks broader. We haven't been concerned with "performance" as goal, build this into the site's normal review process.

JB: Concern for reordering checklist. For some people, walking through the list is hard.

LL: If done all WCAG 1.0 level 1 (and possibly level 2) on several browsers, this goes a long way.

JB: Manual evaluation section now 3.3.

LL: Lots of it is usability, not just accessibility.

MJK: Move 3.2.3 target conformance level of WCAG 1.0 section into usability section 3.4.

LL: Split up Usability section 3.4. Rewrite NOTE there.

JB: Problems and best practices.

DS: [from email: learned that Andrew, Carlos, and Jean-Marie are on a separate bridge call now.

JB: Bizarre! [got assistant to clear a bridge, add it, and allow them to join.]

HB: Add to 3.5: Summarize and followup, include validation.

HB: 5.1 Not necessarily "each URL", but each page type and representative page of the type (the database-generated pages needn't all be checked.).

Conformance Assertion

DS: Is there any meta for assertion of conformance?

HB: Not in Dublin Core.

WL: Proposed material. Question how it is recommended. WCAG asserts that a conformance statement should be included in the document.

DS: Like a "flag" often at the end of the content of the document.

WL: The entire EARL assertion methodology should apply to this.

HB: Send pointer to EARL Primer draft to group.

WL: Analogous to PPP, Likewise appropriate for any XML, so should be in XML guidelines.

Business Benefits

[Phone bridges joined...]

CV: Joined at 10:01

AA: Joined at 10:02

MDA: Joined at 10:04

AA: Business case demographics show the summary of what The three of us [who had been on a separate call] agreed to put in some potential demographics: eneralize 15 to 20% of population in developed countries have some disability.

AA: Changed heading 2.

AA: Table "YES" entries link generally to preceding discussion, some are missing..

JB: Re: demographics. Saw pointer to statistics section. It is counted differently in each country. Numbers were low, under 5%. Having been quoting 10% to 20%. Nervous about such assertions.

HB: Many have glasses, without them would be disabled. Cop-out to quoting a high percentage.

CV: the figure 20% with disability is used in Europe.

WL: Add: In a competitive world, can any audience be ignored.

JB: Discuss disability demographics with reference. Why just to developed countries.

AA: We use those because that's where we have statistics.

JB: Take to the list.

AA: Has posted.

AA: Three issues to be clarified, then can go live.

JB: Problem with printing under Opera NT 4.0 Opera 4.02.

AA: No problem with NT 4.0 and Opera 5.01.

WL: The links on the "YES" in the table are great.

AA: Have solutions for the open ones.

JB: Re: "@@" will get resolved/reviewed. Will review this weekend.

Return to Evaluation

JB: Evaluation of "frozen" sites.

HB: Disclose up front if links to available legacy sites may be inaccessible.

WL: "frozen" is ok, don't emphasize legacy, as this is a copout.

JB: Ongoing Monitoring -- "provide assurance" is too strong.

JB: Add validation.

JB: Add "qualities of reviewer."

WL: Rather, "qualifications of reviewer."

JB: Conformance: familiar with multiple markup languages, browsers.

CV: Not sure need to know markup,

JB: Expectation: can do preliminary review without needing to know markup.

HB: Preliminary reviewer need not be the fixer.

JB: Issue raised previously by HB: Use ordered list for ordered steps, bullets if no order.

LL: Trace 508 list has major discussion, often referencing WAI. Will post to list.

Future meetings

JB: Haven't found a local host for September Berlin meeting. Instead focus on US meeting. Try again in October-December for Europe, possibly Brussels. Possibly October in DC Oct 15 to 19.

CV: Plan to do in Berlin, to minimize local travel. Bonn 5 hours drive away.

JB: We wouldn't get a critical mass of EOWG members there.

Last updated 17 August 2001 by Judy Brewer (jbrewer@w3.org)

Copyright  ©  1998 - 2001 W3C (MIT, INRIA, Keio ), All Rights Reserved. W3C liability, trademark, document use and software licensing rules apply. Your interactions with this site are in accordance with our public and Member privacy statements.