And if we're running into constant snags with CSS that is killing us in IE6, should we ignore it in the spirit of efficiency so that we can get a version up, or always work to get it right?
37Signals will begin phasing out support for IE 6 today as it happens (<a href="http://37signals.blogs.com/products/2008/07/basecamp-phasin.html" rel="nofollow">http://37signals.blogs.com/products/2008/07/basecamp-phasin....</a>). I note that even they had originally planned to phase out support for IE 6 in August this year, but extended the deadline.<p>I think this is a decision that you really need to make after analysing your particular market. If you have a lot of government/corporate clients who may be slow to move to new technology, then possibly the extra effort to get things sorted for IE 6 may be worth it.<p>But if like 37Signals, your market is more made up of small/medium business and tech savvy users who move relatively quickly to new technology, then possibly the time to say goodbye to IE 6 is now.<p>Only you know for sure.
It's usually better to get something out sooner. You can always go back and "tweak" IE if you have time.<p>Companies can thrive on less than 100% of the market; look at Apple as one example. You don't have to win <i>everyone</i> over to succeed -- and in this case, IE6 share will continue to decrease with time, so your market segment will increase likewise.<p>Don't kick IE6 users out, but don't go out of your way to make the experience flawless. If the issues are fairly noticeable, you can put a note saying "You're using an outdated browser; for the best experience please upgrade."
As I'm far too lazy to find my last comment re: IE6 let me dazzle you with stats for this month.<p>For a classifieds site with broad appeal:<p>UBs: 4,828,839<p>IE6%: 24.82%<p>As a point of interest -<p>FF%: 17.17%<p>Safari: 3.97%<p>Chrome: 0.70%<p>Unless you are catering to a predominantly tech audience, you've got some pretty hard decisions to make in not explicitly supporting IE6.<p>If you're working towards your first release I'd be inclined to trade IE6 compatibility for velocity.
At the company I work at, we are gradually inching down towards around 25% IE6 usage out of all IE users. Overall, about 75% of total traffic is Internet Explorer. IE8 is about 1%, IE7 is about 74%. Firefox is around 20% overall. The other 5% contains Safari and other minor browsers, which includes Opera, old IE versions, Chrome, Netscape, etc.
Our site gets about 3.5 million hits a month. My coworker and I are the lead front-end engineers for the site and we are actively pushing on all fronts to abandon IE6 support entirely. We are making gradual headway, but its a long slog. Unfortunately for us IE6 users still comprise too large a percent of our audience to abandon completely.