Are Web Standards Important? - January 5th 2007

WHO NEEDS THEM?

The W3C (World Wide Web Committee) issues and maintain a number of standards for the coding of web pages as well as a number of validation tools to see if your web page actually meets those standards.

I submitted the Activeminds home page to the validator which generated 27 errors as well as various warnings.  Most of these errors were due to the inappropriate use of <div> tags, extraneous end tags (all in the code supplied by Google for search use!) and a couple of uses of & instead of &amp; in the code.  None of the “errors” prevent the page being displayed correctly on all major browsers.  So, should I be worried by this failure to comply?

The W3C states the following reasons why you should ensure your site complies with the standards:

• It ensures that web sites display properly on all browsers on all computers, not just on the popular combinations.
• You are protected from a law suit from someone who says your site isn’t accessible to their particular disability.

The first of these arguments is of course much stronger than the second since the use of standard code could reduce the amount of testing you need to do on different platforms and browsers.  From that point of view (and that alone) it is worth knowing how and when you are breaking the standard and what the implications of that might be.

The not being sued argument is pretty trite and is largely irrelevant to the vast majority of web sites.  Perhaps if you are a Microsoft (five errors on their home page) or a Shell (three errors) or McDonalds (validation crashes because of invalid characters on the page) you should be a little more concerned than the rest of us.

So, are standards important? Standards are useful as guidance.  I like the quote most often attributed to Douglas Bader. “Rules are for the obedience of fools and the guidance of wise men”.  However, in order to be guided by the rules, you have to know what they are.  You can’t simply pretend that standards don’t exist.

Comment by Martin Fleming on January 27th 2007

Web standards has always been a hot topic and many people have questioned their relevance. I was quite lucky last year to get an hour of Andy Bud's time to discuss this issue and it turned out to be quite an eye opener.

In all reality there is no such thing as web standards. Infact the W3C only publish recommendations for browser vendors and web developers to follow, they are not actually standards.

We also have other recommendations by the What Working Group and Microformats. The W3C is listening to the What Working Group and has led the continuation of HTML to bring us HTML5 in the future, even, if controversially, the chair has been given to Internet Explorer's architect.

These recommendations have been very important as it has stopped the old "you need IE or Netscape to view this page" back in the old browser war days. Web pages have returned back to their original ideal of letting any user on any platform read and enjoy the content.

Web Recommendations move on since then and are now focused on correct semantic meaning and the separation of content from style (something that will mean more when XHTML 2 and CSS 3 become mainstream). Do not forget an HTML document is mark-up and not just meant to be read by web users but by other web accessing systems as well. A website looking correctly does not necessarily mean it may be coded correctly. How will the upcoming XHTML+Voice be able to correctly know what its reading to the blind user if a list has been created using *br /* instead of correct *ul*, also how will the screen reader be able to convey whether this list is ordered in importance of points or unordered?

The semantic web is the future and recommendations by the W3C is the way to achieve this goal. In this respect web recommendations matter hugely.

Sometimes a web page cannot validate correctly, due to software such as live help, that have code to embed that cannot be changed due to proprietry reasons, or maybe using a CMS system that is not as tidy as it could be, and yes a couple of errors is not the end of the world, but to blatently ignore all errors is wrong.

In fact its the Frontpage and Dreamweaver cowboys that see no need for standards as they rely on a complete "what you see is what you get" environment but instead that should be "what you see is what you care about".

Also customers are becoming increasingly aware of web recommendations and lately some of the companies that do not bother have been disappearing or being forced by the market trends to care.

The web is changing from just a simple document medium to a place for applications, without recommendations this new breed of functionality may not be available to Microsoft, Apple or Linux users using their preffered browser of choice, or maybe even browsing by command line. web authors have a duty to allow their content to be read by how the user chooses. This is something that comes with the flexibility of the web and without this attitude towards it, these developers may be better off just building VB applications that can only run on Windows for example.

This is purely my opinion and there are also opposite arguments to not bother with web recommendations, but for myself, my clients, my staff and my company - this is an area we will not ignore.

Comment by David Sawers on January 27th 2007

Thank you for your comment. A well argued piece that I have to say I'm coming round to agreeing with.

I wouldn't want to give the impression that I ignore standards, I'm certainly very aware of them and their importance in improving the web for everyone. My attitude has softened somewhat since I wrote this piece.