Validation and standards
The confusion increases. Is validation, seen from an accessible perspective, important? Time to set the record straight: The answer is No … and Yes! Confused yet? Keep reading…
There are a lot of free validation tools to be found on the web. From time to time new tools and new websites, designed to show how good or bad websites validate, emerge. That alone is no problem. The problem emerges when you claim that websites with few errors in validation equals high accessibility.
But what is the problem? The problem is that most validation errors don’t create actual problems with accessibility. A webpage with 500 validation errors can work just fine accessibly. A webpage with only 2 validation errors can work terribly.
So what you need is to be able to interpret lists of validation errors to determine if the error constitutes a problem or not. And even if you are able to do so you will soon notice that most errors don’t mean much at all when it comes to working with accessibility. Please note though that many validation errors can create serious problems with you website.
But validation is still important. Perhaps one of thee key factors. How does it all come together? Well, in order to validate towards something you need to have a standard.
Standardisation is what’s important
An absolute must regarding accessibility is that there are standards and guidelines. If there is no standard you cannot, for example, build assistive technology. The developers of such equipment can’t guess how the assistive technology will work. There has to be a standard. The standard, in a way, connects the assistive technology with the apparatus/interface. It all regards to how the actual computer is supposed to work, how the programs in the computer corresponds to websites and browsers alike. The same thing goes for telephones, terminals and other gadgets. Standardization is central to a more accessible society.
So, it is very important that the technology you chose follow standard. Equally important is that the standard you choose to work with supports accessibility. Html for example lack support for accessibility in the version called 3.0. When HTML 4.01 was released however it was more thought through and supported accessibility. With the arrival of xhtml the possibilities increased even more.
That’s why it is important that website developers make the decision to build according to standard. There are other standards apart from html. The standard makes it possible for developers to develop programs and technical aids. Standard also makes it possible write guidelines and instructions for the developers to help them build products that work. What’s written in standard is right. If you don’t believe so you have to alter the standard. When there are a lot of versions of what supposedly is right it always creates serious problem to whomever is trying to develop assistive technology; a problem that has become very evident with mobile phones and other portable devices recently.
Assistive technology tried to consider poorly constructed websites
At the end of the nineties and partly into the new millenium browsers seldom followed standard. Internet Explorer, the most dominant browser of all, was perhaps worst of the lot. As a result developers chose to build websites to look good in Internet Explorer with the negative side effect that the webpages didn’t follow standard. This in turn led developers of assistive technology for the blind to try to compensate these flaws by adapting the aid to one certain browser (IE) rather than standard. Now when most browser do follow standard that proved to be a strategically bad decision. Even if you build a website according to standard, using a browser that can interpret that website according to given standard assistive technology for the vision impaired built today can’t use it to its full capacity.
The distributors of such aids haven’t been able to build in all the different finesses and attributes that the standard permits. Instead they were forced into trying to make understandable interpretations of websites failing to follow standard. What remains to be seen is if they can rectify their products in order for the end users to get the support standard actually has the possibility to grant.
Validation points out technical problems for not-technicians
There is no point in choosing a standard if you’re not going to build accordingly or allow yourself to deviate to far from it. With that in mind validation becomes important. At Funka we claim that validation is a good tool for non-technicians to get a notion of the quality of the code that generates the website. If validation shows that there are 500 deviations from decided standard there is a problem. Even a non-technician can understand that. As to why there are 500 deviations perhaps a technician can explain. The magnitude of the problems that surface may also be up to a technician or an accessibility expert to help sort out.
But if your partner is under contract to deliver code according to certain standards only to permit themselves to deliver code with over 500 deviations you are in serious trouble. You did not get what you ordered. And even more trouble may be lurking up ahead. The probability that there are accessibility issues with a website with 500 deviations than two is obvious. Many validation errors can probably be traced back to negligence, incompetence, faulty systems or a combination of all of the above. Of all the accessibility problems we here at Funka deal with maybe 5-10 % of them are found through automatic validation tools.
Are there way to measure accessibility?
As early as the nineties the first tools that tried to measure accessibility appeared and we who were active during that period remember Bobby. Your website could be “Bobby approved” and Bobby was one of the first true examples of how young people at the forefront of technology tried to create the first tool to automatically measure and control accessibility. Since then similar products have appeared and disappeared. They may be used to find specific certain accessibility issues. The most common are tools that find pictures lacking an alt-tag. What they can’t find though are pictures with bad alt-tags or pictures where the editor has learnt that you can press space in the alt-attribute in order to avoid validation errors. Most tools are also able to point out h-tags for headings are missing or if tables are used mainly to control layout rather than presenting actual tables of data. Tools however can say nothing about the actual quality of the heading. So far only actual people can do that.
When we at Funka analyze accessibility we use automatic tools in the same manner as a doctor would use a thermometer. To find indications to problems. But a serious doctor doesn’t diagnose by a thermometer alone. In the same manner a serious accessibility expert would say nothing about a websites´ accessibility by validation alone.