I have been creating Web sites using Web Standards, Valid XHTML/CSS since Spring of 2003. Over 3 years now. It was a wonderful change in my career.
I remember some early CSS reading where someone posited the following:
Imagine a world where CSS had been implemented properly by browsers when it came out as a recommendation. Most Web designers would look at a table-based layout, scratch their head and wonder: “why in the WORLD did they do it THAT way??”
I don’t remember who said it, and I remember chuckling at it, but this last week I came face to face with the reality of it.
I was doing some subcontracting work for a client inside of a CMS that had a text-editor that:
1. Is A java applet
2. Produces crap code
3. CHANGES THE CODE THAT I GIVE IT
4. Is nested inside of a DOCTYPE-less template that had numerous validation errors, assuming an HTML 4.01 trans DOCTYPE.
I was attempting to massage my lean, CSS-based design into this system and having absolutely no luck. After hours and hours of cursing, tweaking, bloating my markup and CSS out of control, it dawned on me that a simple table would do the trick just fine. But here’s the deal:
THE CODE MAKES ABSOLUTELY NO SENSE.
On this side of the fence: understanding the LANGUAGE aspect of htmL, understanding what tags mean, and how to use them properly, the “simple” solution is to use code that is utterly incoherent. It produces something that is relatively stable across browsers, and uses a bit less markup than my CSS based version was turning out to use to get this wonky rich-text editor to understand it, but looking at it from a semantic markup perspective, it is incomprehensible. If I were to read it as a screen reader sees it, it would be difficult to follow at best, misleading at worst. It’s going to do them no favors, if not be detrimental to SEO and Accessibility. Attempting to maintain this in the future will be as fun as a root canal, but by golly, it appears like it’s supposed to in the browser.
I’m done venting. Thanks for listening.
Leave a Reply