Web Design in Halifax Nova Scotia, Canada

Show Navigation 

What it truly means to build with web standards

Posted On March 18th, 2010 Author Kyle Racki Filed Under Business Development, Coding & Application Development,

Many of the web projects that we bid for specify in their RFP that the vendor must build with web standards as specified by the W3C. However, in the past, some of our clients have wondered why sites we produce do not always validate, and some clients have even been challenged by losing bidders that they apparently made the wrong choice in a web partner. We were once even told that because a website didn’t validate, it would not rank highly in search engines!

The attempt of this post is to set the record straight and dispel the myth that web standards and validation are the same thing (though they are closely related).

Without getting into the intricacies of web standards, it is important to note that web standards have been devised by the W3C to separate content, presentation and behavior. HTML tags are “marked-up” around textual content in a semantic manner (using tags to describe the meaning and format of the content), and in a separate file, CSS is used to instruct the HTML how to look (ie: fonts, colours, layout etc.). For a more detailed, but completely simple and comprehensible explanation of web standards, see the post on Boagworld.

The W3C offers a validation service for free that will automatically scan a web page for inappropriate HTML tags. This makes sure that critical things like missing closing tags are spotted and corrected by the web developer. Validation is very useful and an important check to make before launching a website, but is it the same as building a site with web standards?

Remember that the ‘wrong’ way of building websites was often with HTML tables for layout, image tags for headings, and inline style-sheets. The ‘right’ way is with semantic HTML separated from presentation code. However consider this; You can build the old way, and your page will still validate, as long as your tables are correctly written, your image tags have alt attributes specified and your inline style-sheets are correctly formatted. So for anyone who says validation is some sort of certification or that it is official “proof” that a website is built to standards—the validator would in effect be approving a badly constructed website, and incorrectly stating it is built to standard.

Conversely, a properly coded, web standards-compliant website may be missing an ALT attribute on an image (which basically describes the image to users who have images disabled in their browser). Or the parameter that opens a link in a new window (target=“_blank”) may be in use. Because of these minor offenses, the web page generates validation errors. Should the errors be corrected? Probably, but sometimes it is the lesser of two evils to use degraded tags that break validation, while keeping the intended functionality of your website. (For all you geeks out there; I know you can use javascript to open links in a new window, but it’s not my point. Once a website is in a CMS, clients will often add images with ALT, or use the target parameter, in which case it’s out of your hands).

The point is, a non-standards based website may validate, while a beautifully coded, standards-based website may not validate. Validation is a best-case-scenario checklist, and not the be-all-end-all decider or whether or not a website was built with web standards.

View Kyle Racki's profile on LinkedIn

What do you think?

comments powered by Disqus