Quote:Java/1.5* - 63,231
Java/1.4* - 88,259
Java1.3* - 5,324
Java1.2* - 0
Java1.1* - 473
hm, those numbers are pretty different from what i've been getting.
for the last few weeks at http://acg.media.mit.edu/people/fry/zipdecode i've got:
11698 - not installed
7742 - 1.1
329 - 1.3
5154 - 1.4
7668 - 1.5
which i think are a reasonably accurate sampling of generic web users, since it's been posted to some fairly random places and is a ways up in google for "zip code", so a lot of traffic comes from that.
compare that to the processing home page, which for the last few days or so is giving me:
2019 - not installed
2091 - 1.1
417 - 1.3
2760 - 1.4
2365 - 1.5
3 - 1.6
which makes sense because there are a lot less people on 1.1 (because it's an audience more likely to have a later version of java) as well as more people with java actually installed. the higher number for 1.4 vs 1.5 i think it's because of all the mac users. and the 1.3 number stays pretty consistent.
these numbers match what you've got, but i think your 1.1 (and maybe 1.3) numbers are off.. if the piece is built in p5, unless it's done in 91 it's likely to have issues with some 1.1 and 1.3 java vms, and may not be reporting numbers properly (meaning that the applet fails to load, or locks up before it gets a chance to phone home with the result).
in the end, i think 1.1 is still very much around (sadly.. ugh) the numbers for 1.4 and 1.5 are a good sign (though 1.5 is an annoyance since it's so friggin bulky and has so many quirks). however the "not installed" number seems to be on a measured increase over the last year or two because of msft no longer including even java 1.1 in winxp as of sp1a.