Recently a developer was going over the browser baselines for a project, they listed them as: IE 6.0, Mozilla 1.5, Netscape 8.0, Opera 9.02 and Safari 1.2. I saw this and thought WAH? Mozilla and Netscape? No IE 7? Safari 1.2? Opera? Let’s take a look and see what should be the baselines for browsers in 2007.
First stop was to check out some global stats:
Looking at these two graphs, I think it’s safe to limit the browser testing to IE 6, IE 7, IE 5 (still has more market-share then Safari) Firefox and Safari.
A couple of points:
The next question I hear from developers is “Why not just develop for IE? Look at the marketshare?”
Well let me tell you…
1) You should be developing for a Web Standards compliant browser (Firefox) then “breaking” it for IE and other browsers. This is the current trend in modern web development, get on the train.
2) Firefox now has a large enough browser share to matter.
3) Mac users (although low in % for browser share) are a segment of the market you want to cater to. They are willing to pay a premium for better experiences, and a large number of them are part of the creative class, making them a very vocal part of the web (writers, producers, designers, bloggers, etc…). A group you don’t want badmouthing your products.
4) Ok, but why Opera? Well Opera really has the most potential out of the lot. This is due to Opera’s strong push to be included in embedded devices. I’ll just take one of the many examples – the Nintendo Wii – it comes with the Opera browser built in, and there are now 4+ million Wiis out in the wild with more and more being sold every day (or how about the 300+ million Nintendo DSs with Opera?). Opera is even being included in some DVD players.
So what should the browser stats be?
Here is my idea for some general guidelines when browser testing:
|Internet Explorer||7.x (100%)||6.x (100%)||5.5 (Functional Only)|
|FireFox||2.x (100%)||1.5.x (85%)|
If you notice, I’ve included %s next to the version numbers. This tells the developer what percentage “correct” it needs to be from the supplied mock-ups. Where as it needs to look 100% in Firefox, it only needs to be functional in IE 5.5 and 95% in Safari.
How do you judge 95% in Safari? This should be up to the designers, if a image is off 10px off or some text is a little larger then in Firefox, then that should be OK – but if the tabs are broken and text is hovering outside a div, that may not be OK – it’s really up to the architect of the experience.
I’ve attached a PDF to this post with some more information, something to pass on to your developers if you see they are testing for AOL 4.2 or something equally as awful.