I was taking a quick look at the browser stats on my personal site for February:
MS Internet Explorer 31.4 %
Mozilla 19.5 %
Firefox 18.9 %
Google Chrome 10.9 %
Unknown 8.7 %
Safari 6.3 %
Opera 0.9 %
IPhone (PDA/Phone browser) 0.7 %
Android browser (PDA/Phone browser) 0.5 %
NetNewsWire (RSS Reader) 0.4 %
Others 1.4 %
The detailed breakdown tells me that the most popular MSIE is version 8, with 13.7%. Then MSIE6 with 11.3%, and MSIE7 with 6%. Virtually nobody’s using MSIE9, with 0.2%. Worryingly there are a handful of hits from other older MSIE versions, although they’re all at 0%: versions 5.5, 5.01, 5.0, 4.01, 3.02 and would you believe it 2.0 all get a mention.
In Firefox land the biggest is version 3.6.13, with 13.2%, followed by a few on version 4.0, and other variations of 3.
Most of the Chrome users are on the current version 9.
There were a small number of hits (all less than 0.1%) from such rare beasts as SeaMonkey, Blackberry, Nokia browser, and the various versions of Netscape — everything from version 0.91 (!) up to 8.1. A bunch of various RSS readers are also in there.
But the real mystery is the figure of 19.5% for Mozilla. What does it mean in this context? Is it a munged reading for more Firefox browsers, a generic compatibility claim from various unidentified browsers, or something else? The detailed breakdown doesn’t tell me anything.
You might be forgetting that there used to be a Mozilla browser that was quite popular – Furefox was born out of a desire to make an unbolted version of it… Which makes it all the more ironic that it is now such a bloated browser!
I knew that, but I find it unlikely that it would be attracting 19.5% of hits; more than all the other versions of Firefox put together!
Yes but depends how up to date the user agent strings are in your stats software, and how it makes decisions upon them.
An older package I used to use years back used to make a wrong assumption about the newer (but now obviously obsolete) “Netscape” browser, thinking it was the (even older again) former Navigator browser. It just didn’t realise that the newer and similar strings were actually a different browser.
Likewise I’d put money on the table that your software probably doesn’t recognise certain/latest versions of Firefox – and as such all it would be left with in the transmitted user agent string to identify with would be Firefox’s admission that it is based upon Mozilla and using the old Gecko renderer… that would therefore round off as “Mozilla”.