We are writing to offer the CRTC an international perspective on Web accessibility for people with disabilities.
Who we are
We are all experienced Web developers who have worked on small-, medium-, and large-scale projects. Some of us have written books on Web standards and Web accessibility. Many of us have written articles and given presentations and seminars. We have direct knowledge of what it takes to make a Web site accessible, and it usually isn’t as complex or expensive as interveners make it sound.
Statements about cost
Clarification would have helped
First, the CRTC should have clarified what it meant by “W3C compliance.” The World Wide Web Consortium (W3C, not “WC3”) publishes dozens of standards (“recommendations”). But for Web accessibility, the relevant standard is the Web Content Accessibility Guidelines (WCAG).
There’s also an independently-developed set of errata for WCAG 1.0, the WCAG Samurai Errata, that bring WCAG 1.0 reasonably up to date with contemporary development methods.
In broad terms, then, telecom companies and broadcasters have a choice of two accessibility standards – WCAG 1.0 (with or without the WCAG Samurai Errata) and WCAG 2.0. It isn’t true that one standard implicates another standard which implicates yet another standard, turning the simple task of making an accessible Web site into an endless trail of interlocking standards documents.
The CRTC could have simplified its own hearing if it had asked interested parties to comment on complying with WCAG 1.0 or 2.0 at a certain conformance level.
On the subject of cost, it’s true that retrofitting a site for accessibility is always more expensive than creating an accessible site from scratch. But since Web standards and accessibility have been widely understood since at least 2000, any Web sites that telecom companies or broadcasters have put in place since then must represent companies’ own choices. Companies had the option all along of creating standards-compliant and accessible Web sites. If they opted not to do so, we don’t understand why they are objecting to the cost of catching up. This is, after all, a question of service to their own customers.
We know that some of the large-scale software applications used on large sites – like content-management systems (CMSs) – are costly, hard to replace, and usually incapable of producing compliant or accessible Web sites. But not all CMSs have those problems, and again, companies have to live with the ramifications of having chosen a CMS or other software that interferes with accessibility. One of those ramifications is the cost of cleanup, retrofitting, and remediation.
Even without inspecting any of the sites in question, we find it unlikely that a stem-to-stern rewrite of a very large site would cost $20 million. A $3 million cost could be possible if, in effect, every vestige of an existing site were thrown out and replaced with new content, markup, and scripting – but that is not strictly necessary. Recall that WCAG has different conformance levels. Even the lowest conformance level of WCAG 1.0 (Priority 1) is much better than nothing and provides adequate accessibility for many groups. WCAG 1.0, Priority 1 compliance is easy to achieve even on many sites that use generally noncompliant CMSs or other software.
We believe the fundamental problem is one of human resources and know-how. Hence we urge the CRTC and telecom and broadcasting operations to hire only the most qualified Canadian developers – those who are knowledgeable about Web standards and accessibility. Often these will be the youngest developers with the newest skills. Having qualified staff on the job is the most reliable way to provide high-quality accessible Web sites at reasonable cost. It could have avoided most, if not all, of the problems being discussed at the hearing.
No conflict of interest
As developers working outside of Canada, we feel we have no financial conflicts of interest in providing this advice.
Response to Canadian Association of Broadcasters
The Canadian Association of Broadcasters, in its reply comments of January 12, 2009, writes:
As observed by the group of international Web developers… [WCAG documents] do not establish actual standards but constitute recommendations under the title of “Web Content Accessibility Guidelines”…. Both versions of the guidelines are operational simultaneously and, between them, offer six different possible conformance levels (three each).
This is a misreading of the facts and of our statements.
W3C guidelines are standards. Once a set of guidelines becomes a W3C “recommendation,” it becomes a standard. WCAG 1.0 and 2.0 are both W3C recommendations, hence they are, in fact, standards. W3C guidelines are not lists of hints, tips, or suggestions.
WCAG 1.0 and 2.0 are not “operational simultaneously” on the same document or site. In WCAG compliance, developers choose to conform to:
- WCAG 1.0
- WCAG 1.0 with WCAG Samurai Errata
- WCAG 2.0
There are no other options. CAB implies that there are six conformance levels that Web developers may mix and match. You cannot select a little from column A and a little from column B.
There are only two sets of three conformance levels, and developers have to comply with everything in a conformance level and all levels below it. Compliance with Level 2 means you comply with everything in Levels 1 and 2, for example.
CAB also writes that we “acknowledge that retrofitting existing sites for accessibility would be costly and, for larger sites, could be impossible.” True – but we made that statement in the context of broadcasters’ and telcos’ deliberate choice to code inaccessible sites in the first place. Those choices have costs, and now may be the time to pay them.
We want the CRTC and other readers to understand that we do not endorse the interpretation advanced by CAB. If broadcasters and telcos follow our advice, the result is accessible Web sites. If they follow CAB’s advice, the result is no change whatsoever.
Director of User Experience, Nomensa
Director, Web Accessibility Tools Consortium
Posted December 15, 2008. Updated January 18, 2009.
Location of this document
Indefinitely archived online at