On Browser Share
A lot of people working on web applications these days seem to be oblivious to the rest of the world beyond their own team. Naturally, most of these people are using Firefox to do their browsing, and subsequently to do their development and testing as well. This makes sense, given, well, just about everything about Internet Explorer (and the fact that only about 5% of us own macs). Unfortunately, the browsing public at large is nothing like these people.
Depending on who you ask, Firefox’s share of the browser market hovers somewhere between 15 and 20%, while Internet Explorer dominates with 70% or more (and Safari is slowly building ground at around 6%). Given this, I have to ask, why on earth do developers think its okay to write Firefox only software? If asked, most of them will tell you, “Oh, I’m just a startup, I don’t have time to invest in IE,” or “the market I’m going after all uses Firefox.” I’m not buying.
Except for the most demanding of projects, making something work in IE is hardly the burden people make it out to be. Sure you have to spend extra time to workaround the missing pieces, or the flawed implementations, but if you wait until after you’ve launched a product, you’re going to have to spend weeks or months in “compatibility” development, where you write zero new features. Meanwhile, you’re turning away 80% of the potential users of your product, and many of them will never come back.
At the other end of the spectrum, it’s highly doubtful you’re really targeting people who only use Firefox. Even if you had done market research, which you almost certainly haven’t at this stage in the game, at the very least you’re likely talking about a group of people who use Firefox and Safari, in which case you really should have put in the extra effort to at least make Safari work. The differences between Firefox and Safari, at this point, are miniscule — in most cases it is trivial to make something that works in Firefox work in Safari. And if you buy the argument that Safari users are very similar on average to Firefox users, you are still excluding as much as a quarter of your potential market not working in Safari. But, like I said, this probably doesn’t even apply to you, because it’s highly unlikely you can know for sure what percentage of potential users are using Internet Explorer.
Of course, there may be exceptions. Legitimately complex applications, or incredibly focused markets, for example, might validate focusing on Firefox alone (or preferably with Safari). A good example of this is recently launched Heroku. This team has tackled a hard problem on the client side (an in browser text editor) that could be difficult to reproduce faithfully in Internet Explorer. Furthermore, by targeting Ruby on Rails developers specifically, they probably are talking about a smaller market that is likely to use Firefox more than Internet Explorer. But even Heroku, who’s goal is to bring Rails to the kind of people who don’t want to go through the hassles of deployment on their own, is probably shutting out at least some potential users who would honestly try it out, if it worked in the browser they were using. It is difficult to say how many, but analyzing their own traffic, they may be able to get a good idea. (Aside: since the rails community seems particularly Mac friendly, this might make an even stronger case for having Safari support, though, in some of my own research, it has seemed that some of the Firefox features they rely on for text selection may legitimately not have duplicates in Safari).
Ironically enough, many of the same people who would advocate Firefox only support, would outright refuse to use a service that was Internet Explorer only. Yet somehow, they cannot manage to imagine a user who would not be immediately willing to go download a new browser, and fundamentally change their online behavior, just at the opportunity to try out the “next big thing” on the web.