Re: Let's talk strategy (Netscape's 75% figure)

Bob Cunningham (bob@mano.soest.hawaii.edu)
Wed, 2 Aug 1995 12:27:34 -1000


After taking a closer look at detailed statistics from some local
sites here over the past 4-6 weeks, I've come to the conclusion
that the aggregate numbers are misleading in the sense they
don't truly reflect how very fragmented the browser population
really is. Aggregate raw hits (cutting off arbitrarily at 5%)
show like this:

74.0% Mozilla (Netscape)
13.6% NCSA_Mosaic

...followed by 63 (!) other browsers (robots, cachers, etc.).
[Lynx was at .8% FYI.]

But...looking at the detailed statistics carefully, I see 771 different
combinations of browser make/model/computer OS. With 406 of those
being Mozilla (i.e., ALL versions of Mozilla--including those which I'd
thought would have expired by now, multiplied by virtually all the
combinations of computer OS types). Here's the top of the detail
summary (arbitrarily breaking off at 5%):

13.0% Mozilla/1.1N (Macintosh; I; 68K)
11.8% Mozilla/1.1N (Windows; I; 16bit)
6.5% Mozilla/1.0N (Windows)
5.8% Mozilla/1.1N (Macintosh; I; PPC)

...followed by 767 other combinations...

I was amazed to see so many hits by older versions of Mozilla,
especially for Windows (I can only guess that Macs showed up so high in
my statistics is that Mac users for some reason simply upgrade their
software faster:-). For example, use of Mozilla1.2b2 (Windows; I;
16bit) was nearly the same as Mozilla/0.9 Beta (Windows), both being
around 1%. Indeed, as the numbers above hint at, almost 1/3rd of the
hits from Mozilla come from old versions.

Bottom line: Simply because you see 74% hits from Mozilla (Netscape)
doesn't mean those all use features in the current version of Mozilla.