Sue Denem

During a discussion on web standards, I cited an article on Mike Davidson's blog.  Someone disagreeing with the points I raised retorted with:

You can't really refer to a 4 year old article on this.

Ha!  What does the age of an article have to do with things?  Are books written by Zeldman and Meyer irrelevant after a couple of years?  Does all good practice change after a few years?  Of course not - marquee and blink tags were a bad idea even back when popular, and anyone with a modicum of understanding about the web knew that.  Table layouts and font tags were a good idea in the abscence of the more structured CSS model, and are only now bad practice because there is something far superior to replace them.

So the specific details of web design can evolve with time, as can design philosophies.  However, it does not mean that every thought a designer (or anyone else, for that matter) had pre-2004 is invalid.  To suggest otherwise is, quite frankly, idiotic and ignores the knowledge and wisdom gained through sticking with and evolving a craft over time.

Mike's post still raises some very valid points.  By all means disagree with him (and me!) and put up a reasoned counter argument, but simply dismissing the views out of hand because of the age of a blog post is closed minded and will limit your understanding of the subject.

I'm sold on the idea of web standards. The continual tirade of rhetoric coming from standards evangelists doesn't help with the business case, though.

Do web sites which don't validate make less money, or inherently make them less usable or accessible? Not necessarily. A few examples:

• Google
• Amazon
• One and One
• Apple
• Microsoft
• BBC News

These are some of the biggest web sites in the world, and are all worth millions, if not billions (with the exception of .net, of course). At some point they've taken a pragmatic look and decided that it was worth sacrificing the holy grail of fully validating code for something else - better support across browsers (CSS is unfortunatly flaky in many browsers, adding a development and testing overhead), ease of maintenance, integration with existing systems and so on.

The assumption that none-validating code is automatically "bad" is as short-sighted as the assumption that standards don't matter at all. As with most things, there's a middle ground and a good web professional will use discretion.

More Reading...