Here’s an idea that needs smothering in the crib immediately: Andy Bubb gone into apologist mode for table-based layout. Bubb’s rationale is that there is a learning curve for CSS.
Sites are not developed by designers alone. There are engineers behind them. Designing with tables does not reduce your cost. All it does is shift it from design to engineering.
Suppose that instead of using a CSS layout, you lay out the site with tables. You hand the tables to the engineering staff who will write code in some language to generate the layout. Any change to the design now requires engineering resources.
The client changes their mind, or you discover that something’s broken when you put the UI in front of real people and you:
- Spend money on designers to rework the design.
- Spend engineering resources implementing reworked design.
Suppose, for the sake of argument, that designers and engineers cost the same. You have doubled the cost of the change, and you may not be able to pass it through to the client.
Meanwhile, the engineers have not added features or new functionality. Instead, you have wasted scarce resources, and have treated them as if they were a copy of Pagemaker. Congratulations, you now have overworked, resentful engineers. I speak from experience.
I don’t get Bubb’s argument for going back to tables. It seems to come down to “but we know how to use tables”. So what? Sit down and learn CSS. Engineers had to go through a learning curve when we switched from CGI to servlets, mod_perl, and php. Should we had continued to use a broken, inefficient way to run applications on the server because we had to learn a new methodology?
Finaly, when you try to do something and the medium doesn’t let you do it easily, it’s a good indication that you shouldn’t be trying to do it in the first place.
