MacEdition's CodeBitch wrote: > Daniel, > > This is all very interesting. See also > http://www.macedition.com/cb/cb_20011231.php for an analysis of MacEdition's > stats, with more detail on Mac browsers, and an Analog file nice article > (http://www.macedition.com/cb/manconf.cfg) that cleans up a lot of search > engine spiders and concentrates on the humans. Note that the default Analog very thorough list of bots and search engines in there, nice! > settings don't adequately capture (non-spoofed) Omniweb or iCab, or many > browsers on non-standard devices, like AvantGo and the Dreamcast browser. I > find that the sub-version browser report setting is quite illuminating, and > I'm actively seeking reports from other sites on the trends in browser usage > in their own stats. heres the reworked browser report: 67.36%: 1062211: MSIE 15.41%: 243054: SearchEngineSpider 10.66%: 168143: Netscape 2.42%: 38161: Opera 0.20%: 3108: Powermarks 0.19%: 2976: libwww-perl 0.21%: 3278: Offline Explorer 0.28%: 4464: PageDown 0.13%: 2094: Come to papa v1.0 1.14%: 18021: Netscape (compatible) 0.07%: 1179: AvantGo 0.08%: 1316: Radio UserLand 0.20%: 3117: Konqueror 0.16%: 2509: WebWasher 0.04%: 623: Lynx 0.07%: 1154: iCab 0.09%: 1497: OmniWeb 0.17%: 2667: DA 5.0 0.06%: 1011: FlashGet 0.04%: 681: ColdFusion interesting seeing it that way..i'm suprised so many people use Konq.. looks like i have a couple more user-agents to block too. btw, since we're talking about it.. do you see Inktomi's slurp bot accessing directories that are disallowed in the robots.txt file? I've been seeing it in the last month off and on.. would hate to block them out like the rest of the spambots, but they're archiving things they shouldn't be.. just wondering :) > Are you sure that the jump up in Netscape isn't really the recent change in > Slurp's browser tag? Also, I have a humble request to use percentages in > the reports, as this makes your stats easier to compare to other sites. I'm not sure that it *isnt* slurp, but the number of hits from slurp bots didn't seem high enough to explain the jump in NS browsers.. And percentages are now on for future reports, good suggestion(i'm easy to convince! :) .djc.