Feed Sign in with OpenID OpenID

Simon Willison’s Weblog

Gecko beats IE!

I haven’t looked at the statistics for this site in a few months. It turns out I was in for a pleasant surprise:

Top 10 browsers visiting this site
BrowserPercentage
Netscape 5.x44.02%
MSIE 6.027.29%
Feedreader6.76%
NetNewsWire/1.0.33.74%
Internet Explorer 5.5x3.69%
Internet Explorer 5.x3.63%
SharpReader/0.9.0.23.34%
NewsGator/1.22.60%
SharpReader/0.9.0.32.56%
NetNewsWire/1.0.22.37%

Yup, Gecko engine based browsers appear to be more common than Internet Explorer, at least for this small corner of the web.

This is Gecko beats IE! by Simon Willison, posted on 17th June 2003.

View blog reactions

Next: HTML Definition Lists

Previous: Eldred Act Reasoning

11 comments

  1. I too have seen this trend, with Mozilla beating out MSIE on a regular basis.

    I consider it a very good thing, but wonder how much of it has to do with the audience my site attracts, as the stats for other sites I'm involved with don't show the Gecko numbers I'd like to see...

    pete - 17th June 2003 20:57 - #

  2. Gecko-based browsers heavily outnumber IE on my site, too. But, in my case, no doubt it has something to do with the MathML content.

    The practical upshot is that I don't need to worry much about working around IE's numerous rendering bugs. Reason enough to rejoice ...

    Jacques Distler - 17th June 2003 21:10 - #

  3. I'm sure the trend is purely down to my site audience being a particularly well informed subset of the internet population at large: you can bet IE still rules the roost for the "commercial" web. There are sites like thecounter.com which publish reports on browser coverage for all sites that use their free web counter (March 2003 stats here), but even these can't be thought of as an accurate sample as they only reflect the audience of sites that use free counters. At the end of the day the only stats that should be used to inform design decisions are the stats of the site you are currently working on - for example, if you have a high percentage of NS4 users (unlikely but possible) you may need to keep with table layouts for a bit longer.

    Simon Willison - 17th June 2003 21:11 - #

  4. Hey, don't forget that KHTML-based browsers (well, at least Konqueror) identify themselves as Netscape 5.x too. Oh, and thecounter.com stats are a bit suspect, if you look at some of their other figures. I'd much rather believe Google Zeitgeist, but they've shown some very strange behaviour with the IEs recently too.

    Jim - 17th June 2003 21:55 - #

  5. Whenever I look at my logs, I see about the same trend, and I chalk it up to Simon's point about the audience being slightly better-informed; it's especially obvious when looking at the dichotomy between regular visitors (Gecko, Opera, and Safari) and people who get referred from search engines (IE, IE, some Netscape 4).

    James - 17th June 2003 22:19 - #

  6. Oh yeah, and your "rememberMe" script doesn't work (at least, not when served as application/xhtml+xml).

    Took me a while to get my Javascript version working.

    Jacques Distler - 17th June 2003 22:27 - #

  7. "Remember me" works when you post immediately, but not when you preview first, I think.

    Jim - 17th June 2003 23:01 - #

  8. Where's Opera? I use Opera7.11but have my preferences set to ID as MSIE6.0 for my online banking site and rarely bother to change it back for regular browsing. The only benefit I can see to IDing as my real browser is to up opera's stats in webmaster's logs, but then again I'd rather see a webmaster code their pages to the W3C standards instead of the browser that visits them most often. So while the stats might be interesting in a generalized way, I wouldn't put too much stock in them. Three types of lies...three types of lies...

    greg - 18th June 2003 04:44 - #

  9. Well, if I were at home I’d be posting this from Safari on Mac or Firebird on Windows. But I’m not - so I don’t get to choose the browser.

    Michael - 18th June 2003 08:56 - #

  10. One could easily make a JavaScript counter that sees Opera for what it is. Simply check if window.opera is true and voila! UA spoofing has no effect on the result. The only drawback is of course the fact that it all depends on JavaScript to work, which could be turned of. Nevertheless it's a nice indication of the actual number of visitors using Opera.

    Bas - 18th June 2003 14:19 - #

  11. I haven't quite got to where Mozilla is beating IE yet, but I'm gratifyingly close. They're practically even-steven.

    Dorothea - 18th June 2003 17:37 - #

Comments are closed.

Previously hosted at http://simon.incutio.com/archive/2003/06/17/tabularData

A django site