Speaking of Browsers... Hooray!

We had a meeting at work this afternoon and decided to drop product support for release 4 browsers! I'm so happy. If you do browser-based software, you should give this serious consideration too; it buys you a lot and doesn't cost very much.

The essay earlier today comes out of the research that went into this decision. If you make the decision to drop release four browsers, the following good things happen:

• You can use ECMAScript/JavaScript to animate your interface and interact with the user with only a very moderate amount of crazy-making pandering to incompatibility.
• You can cram a lot of the styling information into CSS, which gives you better, more professional-looking pages, and makes them smaller and faster.
• You can cut out a lot of design compromises that are probably lurking in your HTML-generation code.
• You can ship valid, well-formed XHTML, which may surprise you by coming out usable on a variety of non-PC wireless devices right now today.

Of course, there's no free lunch. You pay the price that people still running Netscape 4.* can't use you. Four years ago, that would have been intolerable. Two years ago, it would have been somewhat painful. Today, it's not an issue.

Heh-heh, every software company has a few modules that cause nose-wrinkling and head-shaking. Let me speak out of school a bit and say that at Antarctica, one of them is called asi2D.c, and it has the browser-sniffing and control flow that lives inside Apache and marshals the map-generation code and legend-writer and javascript-decorator for our maps in a way that comes out looking not-too bad on all those browsers. I originally wrote it in a hurry and it's been re-re-re-rewritten with only modest refactoring, and well... here be dragons. But as of today, we're gonna slay a few of 'em.