Yesterday I had a meeting with some graphic and interaction designers, and we were discussing how best to code something up. There was a suggestion to place text onto the background image itself. I hedged. “Yeah, we can do that..where I’m from I don’t really like images, but it looks like we’ll be using them anyway..”

The designer’s jaw dropped. “You don’t like IMAGES??” Perhaps images to him are akin to puppies or babies. How can someone not like such a thing?

I back-pedaled.. “well, from an accessibility standpoint, it’s not indexable or selectable…but in this case, yeah, we can do it. It’s just not something I’m used to”.

I still think I frightened him. And I felt bad for saying it. After all, this is an agency. The whole look and feel is what MAKES our products. It is just a departure from what I’ve been doing for years.

I thought about it all last night — and it was actually a useful exercise to see where those thoughts come from.

Background: I started coding using the geocities home page editor: all code, no images. Then I got photoshop and every headline was custom made. Which meant with every redesign, every headline was REmade. I was hired at Maxim, and text was all displayed as text. Contract work was a mix.

I spent four years at LexisNexis. We have huge web applications in multiple languages. We serve government agencies and accessibility is a true concern. There is a huge amount of rapidly changing content. It needs to be served up in different languages and formats.

We used resource bundles to serve up our content in different languages. When I left we were still using images for buttons, but there had been much research done into CSS-styled buttons (I was actually involved in several drafts of the UX Recommendation in this regard) What was interesting was that the benefits to using CSS were the same as the drawbacks. Users could resize the font on the buttons — a benefit for some, but there were possible implications on the overall layout of the page. (Granted, this was related to international content in general, not solely buttons.)

Ultimately it came down to an issue of control: using sliced up images and absolute positioning provides the designer control, not the user. It also necessitates the designer having strict control over the content.

Now, you know I’ll say it… web2.0 is about user-created and manipulated content. If we are overly rigid in our designs, can we support a user with a login name of 50 characters? Does it fit within our vision? The site should be for the user, so if they need to disable images, or maginify their screen 4 or 5 times, are we still meeting their needs?

Perhaps some of the sites we build can afford to be more rigid: we are defining the users brand and presenting it. Therefore, we should ensure that it is properly presented to site visitors. That being said, I still see the web as an inherently fluid medium and I would love to see some development using ems and fluid positioning, to provide a more flexible and robust solution.

Ultimately, I think I need to move out of my “web application” mindset. The sites we build are informative and engaging, but they are not terribly content-heavy. We can constrain the data to fit within the framework we design, and modify the design at will. To some extent, it reminds me of my contract work, when a site would be launched with a certain design which would then have to be modified to accommodate new features.

However, when we are talking about brand and identity, perhaps we need to re-evaluate the design when we add new features, to ensure integrity. There may be a conscious desire or even need to revisit the overall design, rather than simply push content out..