I examined two Web sites, Vancouver2010.com
and CTVOlympics.ca
, for the first five days of competition (2010.02.13 through 2010.02.17).
In the 2000 Sydney human-rights complaint, a major bone of contention was the presentation of results. The tables used to list who won which competitions were coded so badly they couldn’t be used with a screen reader. Tables at the two Vancouver Olympic sites use good, if verbose, HTML and should be completely accessible to screen-reader users (if they can actually reach them on the page – see below).
There are some questionable uses of foreground and background colours for text and table cells, but only testing with human subjects could determine if those amount to barriers.
The following problems are so serious they make the sites inaccessible and unusable by people in different disability groups.
Dozens of images don’t have an alternate text. If you’re blind, or if the device you’re using cannot show images, or if the image just fails to load because of a technical error, you can use a text replacement (an alternate text or alt
text) instead. It’s just a sentence or two (or just a phrase or a couple of words) that replaces the image if you can’t see it for whatever reason. You write out in words what the image means or does.
Alternate text is the most basic technique in Web accessibility. There is no excuse for missing alt
texts on images. If you’re blind and you find an image with no written version, you can’t understand the image. (You might be able to make a guess from a filename. There are other rare exceptions.)
The Vancouver2010.com
homepage can have as many as 44 images without alternate text. Most are unimportant spacer images that are used purely for visual layout (an outdated and unprofessional practice in itself). Some are actually quite important, like the Olympic slogan. Even though it would have had to be handled specially because it’s coded as a background image, text of the slogan WITH GLOWING HEARTS is completely unavailable to a blind person.
In another serious example, photos on news stories (most provided by the AFP news agency) also don’t have alt
texts. News photos are not decoration and have actual information value, which blind people are completely denied.
Videos don’t have audio description. Talking just about prerecorded videos for a moment (both sites have dozens of those), there is never a version with audio description available. Audio description is an added narration track that talks you through the video. It explains whatever might be happening onscreen that you can’t understand just from the main soundtrack.
Live video (e.g., streaming sports events) could be audio-described, but it’s barely ever been attempted. Arguably the Olympics would have been a good place to attempt it. But if you look at the near-live videos on both sites, like videoclips of events from the previous day and earlier, those don’t have description either – and they could have. (Considering the size of these sites and their budgets, I’d go so far as to say they easily could have had description.)
Olympic Web sites are not special offenders here. They are doing what nearly every other site in the world that offers video does. They aren’t offering video accessible to blind people.
Related issue: The news spinner on the front page (which scrolls you through a set of news highlights) is a Flash application that probably isn’t accessible.
Result: Nearly all content that isn’t actual text is inaccessible to blind people.
You can’t use either site just by keyboard. This seems to be true across the board and on all pages I surveyed. If you can’t use the mouse, you can step through the items on a Web page (usually by pressing the Tab key to move forward and Shift-Tab to go backward, but there are other methods). I never found any page on either site, including the homepage, where every item on the page could be selected or used by keyboard. In fact, I couldn’t navigate anywhere.
Navigation bars or navbars on the top of most pages are particularly troublesome.
You can’t use those via keyboard, but there is really no other way to go to other parts of a site. You may be trapped on the homepage, unless you go to heroic lengths like using Google to search for inside pages and bookmarking them. Nondisabled people don’t have to go to that much trouble.
Some advanced users may be able to use their software (usually built into the computer operating system) to emulate a mouse – using the keyboard to move the mouse cursor, click, and so on. Use of these so-called mouse keys is a last-ditch effort for most keyboard users, and is pretty much impossible for a blind person who can’t see where the mouse cursor is.
A particularly serious example of keyboard inaccessibility is submenus shown as their own scrollable boxes:
You can’t actually get the cursor into these submenus, let alone scroll them, in my testing.
Useful information is provided solely by JavaScript. CTVOlympics.ca
uses scripting (small bits of computer programming) to display useful information. The master results table, though it has good table code, relies on JavaScript to display informative pop-up messages when you hover the mouse over certain objects. If you can’t use a mouse or are just using a keyboard or if you’re using a device that cannot run JavaScript, you don’t see that content and have no alternate method to gain access to it.
Menus have a mind of their own. In a basic Web-development error, most pull-down menus on both sites can’t be used by keyboard. If you use, say, arrow keys to move down a menu, the first item you move onto may be selected. Expert users and some software can overcome the problem. Well-coded sites avoid it entirely (by adding a simple Go or OK or Submit button near the menu).
Incidentally, self-activating menus like these are hard to use for a lot of nondisabled people; it’s too easy to pick the wrong item.
Result: Even when blind people can read the available information, they will probably not be able to get to it. Most physically disabled people who can’t use a mouse also can’t use either site without resorting to extreme countermeasures.
Videos aren’t captioned. From what I can tell, not a single video on either site has captioning for deaf and hard-of-hearing people. As with audio description, prerecorded and near-live videos could be presented with captioning.
Live streaming video could also be captioned, at little cost. There are any number of ways to include captioning with streaming video, but CTVOlympics.ca
didn’t have to bother. As all its TV coverage is captioned, the site could simply offer a separate feed that decodes and displays the TV closed captioning in the online stream. (Your browser wouldn’t display the captions; they’d be burned into the picture. You’d just select the captioned stream vs. the uncaptioned one.)
This option produces results at least as good – or as bad – as what TV viewers will see. And it’s cheap: I worked on a similar project about ten years ago whose total hardware cost was $600.
The CTV site would still inevitably show some live video without captioning, because there’s more live video on the Web site than on TV. But there’s no excuse to provide no captioning.
The Vancouver 2010 official Web site merely links to different national broadcasters and doesn’t offer live streaming video (that I could find).
Result: There’s never been more online video available at an Olympic event. If you can’t hear, you can’t understand any of it.
This topic is much less clear-cut than others. People with cognitive impairments have a vast range of requirements, some them conflicting. It may just be impossible to present as much information as an event like the Olympics generates without causing inaccessibility for some people who cannot read that much information all at once.
Nonetheless, one very basic mistake could have been avoided. The news spinner should not run automatically. In fact, nothing should automatically move, animate, or blink on a Web page; that also means videos should not autoplay. (Content that moves attracts your attention. Many people with cognitive disabilities can’t afford to have their attention distracted.) Both sites make the mistake of autoplaying moving content on their pages.
The Vancouver 2010 site works in English or French, but the underlying code never declares the language. So if you’re using a screen reader and pick the wrong language, your screen reader may try to pronounce English text as though it were French or vice-versa. (Proper language coding lets a screen reader use the right language without any extra effort.)
Neither Web site has particularly high-quality code. Both sites have hundreds of validation errors. Though most of them are minor and few have any bearing on accessibility, the fact remains that the two sites’ underlying code is not grammatically correct. Browsers display it well only because browsers have been programmed, at great effort and expense, to display incorrect content well. (Athens 2004 Web pages had nearly perfect code, so it’s quite possible.)
The video plug-in used on CTVOlympics.ca, Microsoft Silverlight, caused no end of bother for users of the site, disabled or not, since it involves adding yet another layer of complexity to one’s Web browser – and from the highly imperfect Microsoft at that. It’s a hassle, and it does nothing but give the finger to Adobe, whose Flash plug-in, for better or worse, is now the lingua franca of Web video.
It may or may not be vaguely possible to make Silverlight video player controls accessible to some keyboard and screen-reader users under Windows. (If I could find definitive documentation, I’d link to it.) Those controls surely won’t work on Macs. This isn’t entirely Silverlight’s fault; developers have had to piece together accessible YouTube variations because standard Flash-video controls are inaccessible. In short, the choice of Silverlight made matters worse for all users who don’t already work at CTV, disabled people included. It was a pointless bit of one-upmanship.
This was not a thorough test or debugging of either site, which would have had to be done well before site launch and would have cost a fortune. (Unless either site had hired really qualified developers, in which case it would just have been an expected part of their jobs.) This was an analysis of code and content in which I looked for very big and obvious errors. They were easy to find, and most of them are cut-and-dried: A news-article photo without an alternate text is not accessible to a blind person, for example. Very few of my findings are matters of opinion.
In specific, this was not a user test with real people under real conditions. That would be expensive no matter who did it or when.
I looked at English sites only, with rare exceptions.
I did not attempt to evaluate the huge amount of content on either site. I went only a level or two deep. This is defensible, though: If a disabled person can’t use the homepage, then the rest of the site isn’t usable, either (in part because you can’t get there).
The related official iPhone app and the mobile versions of either site were not tested. Those too could be accessible (in the iApp case, very easily, almost automatically).
Posted: 2010.02.22
You were here: Home → Accessibility → Web accessibility →
Vancouver Olympics Web sites are inaccessible to disabled people → Findings