Tale from the dinosaur era
Once upon a time, Lotus 1-2-3 was the world’s premiere spreadsheet application. It was the first killer application for IBM PC, and its ground-breaking functionality helped convince the corporate world that personal computing was the way to go.
Trying to maintain its lead, Lotus wanted to upgrade its product for the next 3.0 version with plenty of new features, including 3-D spreadsheet. They also wanted to conquer the market more by creating Symphony, an integrated software that would include word processing, spreadsheet, charts, similar to what the MS Office suite is today.
They worked really hard for 18 months trying to build both product while keeping up with the DOS memory limitation of 640 kilobytes and the up-to 16 MHz Intel 80286 processor (yes, this is a story from the ancient times). Their strategy, basically, was to build for the most used hardware at that time. Intel already started shipping 80386 processor, but it was deemed too expensive with too little market to pursue.
Try as they might, their strategy just didn’t work. The hardware limitation was too strong, so they wasted a lot of time and had to cut down features. To add insult to the injury, by the time their softwares were released, hardware prices were down thanks for Moore’s Law and everybody had 80386s with 2 or 4 MB of RAM! Lotus was optimizing for nothing, and their softwares were out-of-date.
From that time Excel started to take over the market, because Microsoft optimized it for the better hardware and were patient to wait until the prices catched up with their software requirement. 1-2-3 and Symphony never recovered from that blow and eventually got discontinued.
Moral of the story
Now, do you want to be 1-2-3, or do you want to be Excel?
You can spend hours working on old browser compatibility, or you can spend that same amount of time making amazing themes using HTML5 and CSS3. By the time you’re done, there’s a pretty good bet that modern browsers will be better, not worse. Mozilla now adopts a rapid release policy for Firefox so they can update it much faster with better HTML5 features. Google already did that first with Chrome.
Also, let’s just admit that HTML5 and CSS3 make web development fun while dealing with old browser bug sucks big time. They’re the standard of the near-future and there’s no other alternative to worry about. The choice is obvious as far as I’m concerned.
Joel Spolsky’s Strategy Letter VI, where I first read the Lotus 1-2-3 story.