Web Access Centre Blog

Comments (9)

1. Ben | 02/11/2006 at 15:15 | Permalink

   Henry, may I pick a few nits? You say:

   “Read our JK Rowling case study to find out how the site uses captioning for the hearing impaired, a sound glossary for the visually impaired, and tab ordering techniques to ensure everybody can enjoy the site.”

   Please correct me if I’m wrong, but isn’t it only LightMaker’s inclusion of a text version that ensures interoperability? The Flash version is accessible only to users of Internet Explorer for Windows and JAWS or Window-Eyes. That’s obviously better than nothing, but there are other screen readers, browsers, and operating systems that arguably have better accessibility features for non-Flash web content. Don’t we need to pressure Adobe to support more a broader range of assistive technologies, especially now that they’ve got a new Flash Player 9 Beta out for Linux?

   It’s also worth noting that JK Rowling’s home page doesn’t validate, which means it flunks WCAG Priority 2 Checkpoint 3.2.

   You also say:

   “It is a legal requirement to make your website accessible in the UK just as it is in the States where a blind student, supported by National Federation of the Blind, is taking the US website Target to court (read more about the Target case here).”

   I realize you’re primarily targeting businesses with this post, but just to be clear about this, doesn’t this legal pressure apply to the websites of government, businesses, and organizations, but not private individuals?

   Postscript: It would be useful if the “Post a comment” section for this blog included some help about whether any markup and special characters are allowed in these comments to make them more accessible. Also, I can’t understand why the “Name” field has been limited to 20 characters - 1 character less than my full name, which is what I usually use when commenting on blogs.

2. Henny | 03/11/2006 at 18:19 | Permalink

   Hi Ben,

   Thanks for your comments and views, much appreciated.

   For me I think the exciting thing about the JK Rowling site is that is “pushes the boundaries” of accessible Flash and has made the most of the accessibility features offered in the Flash authoring tool. It’s a great example of what can currently be done with Flash and how far you can take it. The inclusion of a text version is something that we still advise because there are users who may have issues (not least those who do not have Flash supported) in still accessing Flash. One day I hope that the necessity for an alternative version will not be the case - and we are heading in the right direction for this. Indeed the Web Content Accessibility Guidelines (WCAG) version 2, has been written in such a way that it can apply to non-W3C technologies such as Flash.

   As for validation you are right, it doesn’t and this is something that could be looked into by the designers. But in this case the validation errors are not causing specific accessibility issues; all users who go to the page can access it and link through to whatever version of the site they want and this is what is important here.

   “Don’t we need to pressure Adobe to support more a broader range of assistive technologies, especially now that they’ve got a new Flash Player 9 Beta out for Linux?”

   Absolutely Ben and this is what RNIB and other organisations have been doing for a while. We have worked very closely with the teams in Adobe who deal with both Flash and PDF assessing and influencing the development of their technologies to accommodate for the needs of people with disabilities. This pressure also extends to the software vendors for access technologies. The team who work on Flash have also worked closely with such vendors to progress how these assistive technologies can handle Flash.

   “I realize you’re primarily targeting businesses with this post, but just to be clear about this, doesn’t this legal pressure apply to the websites of government, businesses, and organizations, but not private individuals?”

   Yes, businesses, government and other organisations offering goods and or services online are obliged to make their websites accessible by law. That is very clear. If you are an individual however who has a personal website, let’s say blogging about your day to day life, you do not fall under that legal requirement. If, however, as an individual you decided to start providing goods or a service through your site then you would be legally obliged to make it accessible.

   Thanks for the Postscript too, useful feedback. We are slightly restricted by Wordpress but are always tweaking what we can so we’ll have a look at the points raised and see what we can do.

   Henny

3. Ben | 04/11/2006 at 19:08 | Permalink

   Henny,

   “We have worked very closely with the teams in Adobe who deal with both Flash and PDF assessing and influencing the development of their technologies to accommodate for the needs of people with disabilities. This pressure also extends to the software vendors for access technologies. The team who work on Flash have also worked closely with such vendors to progress how these assistive technologies can handle Flash.”

   As far as I can tell, the assistive technology vendors are if anything ahead of Adobe in their support for user choice. For example, both Window-Eyes and JAWS (the two screen readers supposed to be compatible with the Rowling site) support Firefox on Windows. But the Flash Player plugin only exposes accessible information in Internet Explorer. That appears to be an Adobe decision, and one with ramifications that go beyond the Windows platform. There’s currently a thread on the mozilla.dev.accessibility mailing list about this very issue:

   http://tinyurl.com/y68flt

   “If, however, as an individual you decided to start providing goods or a service through your site then you would be legally obliged to make it accessible.”

   That raises an interesting question. Many private citizens have websites that include Google Ads to help recoup costs. Would that count as “providing goods or a service through your site”?

4. Ben | 04/11/2006 at 19:27 | Permalink

   Henny,

   Talking of valid and invalid HTML, it would be good if this blog page validated. Judging by the errors found by W3C’s validator, it looks as though the template designer is trying to force WordPress to serve HTML instead of XHTML. That’s essentially a good idea since HTML is far more interoperable, but it’s not a task WordPress makes particularly easy. At any rate, one needs to really understand the difference between HTML and XHTML in order to get it right. If getting WordPress to output valid HTML proves too tricky, my draft guide to serving XHTML 1.0 with a text/html MIME type may offer a second-best alternative:

   http://tinyurl.com/esods

   I think it’s crucial for accessibility advocates like the RNIB blog to be promoting best practice by example. Valid markup may be especially important for sophisticated user agents like Emacspeak, which sometimes run transformations on content. Obviously, invalid markup increases the risk of garbling such transformations.

5. Ben | 08/11/2006 at 9:35 | Permalink

   Hi Henny,

   Just to let you know, I left a couple further comments here a few days ago, but they never appeared. I see from an earlier thread you’ve had some trouble with spam; I wonder if your spam filter ate my comments because they contained links. One comment pointed to a thread on the Mozilla dev-accessibility mailing list discussing the inaccessibility of Flash content in Mozilla browsers; and asked whether including Google Ads on a personal site counted as “providing goods or a service”. The other comment suggested that it would be good if this blog page set an example by validating as (X)HTML.

6. Henny | 21/11/2006 at 10:46 | Permalink

   Hi Ben,

   Your comments did get eaten by our over zealous spam filter I\’m afraid, apologies. We\’ll keep an eye out for any other future comments that go the same way.

   I read with interest the thread about the inaccessibility of Flash with Mozilla, thanks for that. JK Rowling was built before the release of Flash 9 so there may well be glitches. I think the comments in the thread illustrate an important point in that developments in Flash and it\’s ability to produce accessible content together with screen reader and browser support are constantly changing. Assistive technology vendors, browser developers and Flash content producers have to keep working together to ensure progress in the right direction.

   This reinforces how a truly accessible web is dependant an several components of web development working together: content (in this instance a Flash website), authoring tools (such as the Flash authoring tool), and user agents (screen readers, browsers etc). The Web Accessibility Initiative (WAI) goes into some detail about this in Essential components of web accessibility, and essentially backs up your point Ben.

   The question of providing Ads by Google on an \”individual\” website being classed as \”providing goods and services\” has never been tested in court and no stringent guidelines exist.

   It is a question that has popped up before so my colleague Donna Smillie asked the DRC (Disability Rights Commission) last year for an answer. She was refereed back to the part of the DDA that talks about \”goods and services\” and told that the DRC were not allowed to provide any interpretation on the DDA.

   It really is a grey area and one that can\’t be concretely commented on without any case law (not even by the DRC!).

   As for validation and Wordpress, yes there are some issues but we are confident that users can access the blog - this is the main thing. We\’re aware of a number of tweaks that we\’d like to make that are under development.

   Henny

7. Tom | 21/11/2006 at 12:13 | Permalink

   Ben

   To reiterate what Henny has previously said, the accessible Flash version of JK Rowling’s website is an important milestone. It helped dilute the myth that Flash is inaccessible.

   Through elements such as a real-time caption box, sound glossary and the use of Progressive Disclosure, Flash has shown that in many areas it offers greater levels of accessible features than regular HTML.

   There are two points you’ve raised during your string of posts above which I would like to respond to in turn:

   1. “The Flash version is accessible only to users of Internet Explorer for Windows and JAWS or Window-Eyes. That’s obviously better than nothing, but there are other screen readers, browsers, and operating systems that arguably have better accessibility features for non-Flash web content.”

   Henny has already responded to this, but it is an important point. The decision was taken when developing the accessible JK Rowling site to work with screen readers that were known to work well with Flash at the time, and have good penetration levels in the market. At this time it was Window Eyes and JAWS. Since then IBM’s Home Page Reader has launched, and the latest version of HAL by Dolphin now has very good support for Flash.

   2. “It’s also worth noting that JK Rowling’s home page doesn’t validate, which means it flunks WCAG Priority 2 Checkpoint 3.2.”

   It is important that you understand there is a significant difference between valid compliance and accessibility. A site that gets the green light from the W3C validator doesn’t mean it is accessible. This is a very common misconception. Accessibility is tied to usability, and can only be truly tested manually. The JK Rowling site doesn’t validate for this reason, it was a conscious decision to highlight this fact and promote the separation between a valid site and an accessible one.

8. Gary Hides | 21/11/2006 at 17:48 | Permalink

   On the failure of validation point. It’s not exactly a big deal, but the points that it fails on are mainly to do with the way the flash file is embedded in the page, and the remaining failure points are easily avoided, minor things.

   Well done to them for going to such lengths to get such a variety of different user interfaces for the site, ensuring good accessibility. But, of course, not every business has huge budgets to commission sites like the JK Rowling one with many different versions.

9. Ben | 24/11/2006 at 14:52 | Permalink

   Tom,

   Thanks for a robust defence, but I’m afraid I’m not at all persuaded.

   You claim: “Through elements such as a real-time caption box, sound glossary and the use of Progressive Disclosure, Flash has shown that in many areas it offers greater levels of accessible features than regular HTML.”

   Flash’s accessibility features depend on users:

   1) Buying, running, and upgrading Microsoft Windows.
   2) Buying, running, and upgrading expensive screen readers like Jaws, Window-Eyes, and HAL.
   3) Running Internet Explorer instead of more competent and accessible alternatives.

   By contrast, HTML’s accessibility features don’t depend on any particular operating system, screen reader, or browser.

   Also, in comparing feature sets, you’re comparing apples to oranges. HTML is a hypertext publishing format. Note the “text” in hypertext. To state the obvious, textual content may embed audio-visual content like Flash, but it has in itself no need for captioning or sound glossaries. And how would you substantiate your claim that Flash is more capable of “Progressive Disclosure” than HTML? It seems to me to depend entirely on how you use either technology. But this isn’t really about HTML versus Flash. It’s about the inaccessibility of Flash itself.

   You explain: “The decision was taken when developing the accessible JK Rowling site to work with screen readers that were known to work well with Flash at the time, and have good penetration levels in the market. At this time it was Window Eyes and JAWS. Since then IBM’s Home Page Reader has launched, and the latest version of HAL by Dolphin now has very good support for Flash.”

   How well screen readers work is obviously a concern, but I can’t make head or tail of your chronology. When was “this time”? The final version of Home Page Reader (3.04), which claimed to support the accessibility features of Flash Player 7, was released in January 2005:

   http://www-306.ibm.com/able/news/hpr304.html

   Your “accessible” relaunch took place in July 2005:

   http://weblogs.macromedia.com/accessibility/archives/2005/07/jkrowlingcom.cfm

   On Page 4 of his whitepaper, “Best Practices for Accessible Flash Design”, published in August 2005, Regan claimed that Flash 6 and higher content was also accessible to Home Page Reader 3.04, HAL 6.50, and KDS PC Talker.

   http://www.adobe.com/resources/accessibility/best_practices/best_practices_acc_flash.pdf

   Are you really saying that Flash content was even less accessible than Macromedia literature claimed?

   Your reference to “good penetration levels in the market” makes no sense to me, since I believe web content should ideally be open to all comers, not just those using the dominant user agents. Do you think websites with notices stating “Requires Internet Explorer 7 or Firefox 2.0″ are the height of good design? Far from promoting accessible Flash, by employing such pseudo-commercial reasoning you discourage Adobe from bothering to expose Flash content in other browsers and on other platforms.

   You assert: “A site that gets the green light from the W3C validator doesn’t mean it is accessible.”

   Of course not. It doesn’t even mean that your markup is conformant. It’s just a baseline consisting of extremely simple parsing tests. On the other hand, it does pick up on a few things that are clearly important to accessibility and have been built into the document type for that reason, such as missing ALT attributes.

   You say: “This is a very common misconception.”

   Do you have any evidence to back that up? I meet folks who have no idea about validation, folks who have no idea about accessibility, and folks who trust too much in markup validators to prove conformance and automated checkers like Bobby to prove accessibility. But I can’t recall ever meeting anyone in any danger of confusing HTML validation with a certification of accessibility. I have however met plenty of folks who seem to think that producing broken markup is somehow a necessary or clever thing to do.

   You affirm that “Accessibility is tied to usability”, as though that alone helped differentiate it from conformance, which is also “tied” to usability. You also say that accessibility “can only be truly tested manually”, which I agree with. Of course, if you only bother to support a couple user agents, manual testing is a lot easier.

   You attempt to justify your invalid markup as “a conscious decision to highlight this fact and promote the separation between a valid site and an accessible one.” So why wasn’t this explained in the RNIB case study? It’s certainly not obvious from the site itself.

   To be honest, I think it is irresponsible to deliberately insert errors in the markup of a client’s production site. You cannot guarantee that no user agent will fail to display your site properly because of your caprice, and if it happens you would have no one to blame but yourself.

   Even if Rowling site’s invalidity created no problems in any present or future user agent, it would contribute absolutely nothing new to the debate about whether validity contributes to accessibility. Like Tolstoy wrote of family happiness, valid pages are all alike; every invalid page is invalid in its own way. Just because the Rowling site happened to work would be no guarantee that other invalid sites would work too. No user or user agent can be reasonably expected to make all variations of broken markup work: it is your job to fix it.

   And yet, even though there’s very little markup on the homepage (most of the content is Flash), your decision to write non-validating tag soup still manages to create new accessibility problems. If you add an IMG element to a page without an ALT attribute, the validator will register that as an error. If you add an EMBED element to a page without an ALT attribute, as you do not once but twice, the validator cannot warn you because EMBED is undefined in HTML; all it can do is tell you there’s no such element in the first place. As a result, users of Lynx get to read “[EMBED]” twice. If you read the Disability Rights Commission study cited by Henny, “ALT tags on images non-existent or unhelpful” was one of the key problems experienced by blind users (see Page 29).

   But even if you were to add the ALT text, users of ELinks would still have to read “IFrame” twice and users of W3M to read a page cluttered by “embed(f7b3qash/accessibility_button.swf)” and “embed(f7b3qash/language_menu.swf)”. And then users of screen readers like Orca, Emacspeak, LSR, and SpeakUp get to hear this mess. They would experience no such problems with the standard OBJECT element.

   I simply can’t reconcile your cavalier attitude to WCAG checkpoints with Henny’s complaint that too many UK sites fail them. Indeed while I picked on validity as easy to test and a good sign of basic competence, the Rowling homepage clearly fails other good checkpoints in Priority 2. By using deprecated presentational attributes it flunks 11.2, and by including TARGET=”_blank” on the anchor link to Lightmaker it arguably flunks 10.1. More embarrassingly, it also fails checkpoints in Priority 1. It flunks 1.1 by not including a text alternative for EMBED and seemingly not including any text alternative for Japanese content whatsoever. And it flunks 4.1 by failing to markup link text in other languages as such (using the LANG attribute). This prevents sophisticated screen readers like JAWS adopting the correct pronunciation for those languages. Against this background, doesn’t the site’s claim to double-A status seems somewhat deceptive?

   I think your fight for broken, non-standard markup that depends on error correction to work at all is an odd battle to choose. Occasionally, people argue that achieving certain WCAG checkpoints sometimes come at the cost of accessibility, not its gain. Anything is possible, but I’ve yet to see that demonstrated of validity. It’s a regrettable Lightmaker don’t use their influence to encourage Adobe to make Flash more accessible, rather than taking obscure potshots at the very web standards that help them market their services.