Browser stats, day #3

Can you tell me what is strange with this picture, which shows the browsers that have been used to view the pages on this site during the three days since I moved?

browser stats

If you say: “That Firefox has almost 50%, and more than IE”, you’re part right. On the other hand: it is not surprising, is it? After all my plugging for it here, one would really have to be a n00b to use IE, right?

It is not either that “Lynx” is represented, with a whopping 0.6%. Lynx is a text-based browser, quintessentially retro, which is fine for a site like this one and great for quick lookups, because it starts immediately — no need to wait for a modern browser to load all its bells and whistles if all you want to do is check for new updates at Things Twice, now, is there? (the only surprising thing is that it’s Lynx and not one of its more capable cousins, like Links or Elinks).

No, what’s really surprising here is the number at the bottom: that Opera is used by 0.3% only of my visitors — even less than Lynx! — that’s a surprise. I know these numbers are not statistically significant or anything, but nonetheless: Opera is the best browser of them all, really: 50% faster than Firefox and 100% faster than IE, according to recent studies; comes with an email program and a newsreader as well; is even more standards compliant than Firefox; works on all platforms, Linux, Mac, and Windows; and has some features that others don’t. The only reason I’m not using Opera is the Vimperator extension to Firefox, which I can’t live without, but if it wasn’t for that, I’d be on an Opera any day.

So, you might ask, if Opera is a such a damn good browser, how come nobody uses it?

Opera is one of the casualties of the Free/Open Software war where Firefox is the true winner. Opera actually used to be a program you’d have to pay for — fairly unthinkable today, and that didn’t last so long either: you could soon get it in a free version with a banner ad at the top which you couldn’t remove. Fair enough, it was still the fastest browser around, and the most configurable one, and banner ads are all over the place anyway, so one more or less . . .

Then even that disappeared. Today’s Opera is totally free of such commercial bindings, and we’re back to the mystery again: why is the fastest and most able browser out there stuck at 0.3%? Well, in the meantime, Firefox had had its tremendous success and had more or less exhausted the field of “Alternatives to IE”. Besides, even though Opera is Free as in beer, it is still not Free as in speech — the source code is not open. For most people, that is completely irrelevant — who has ever looked at the source code of Firefox, other than geeks and software developers? — but ideologically, it apparently makes a difference.

But 0.3% . . . Let’s put it this way: if you’re looking for a blazingly fast, small browser with excellent functionality, I suggest you try it out.

11 thoughts on “Browser stats, day #3”

  1. One thing about Opera is that you can change what browser your browser says it is. Some websites block or otherwise inconvenience visitors using “non compliant” browsers even if that browser can work or mostly work with the website. Opera allows you to say you’re using IE or Firefox to avoid that. Maybe that skews Opera’s statistics down a bit?

  2. Here’s my theory, very ad-hoc, made up after I was Reddit’ed:

    There is a large group of web users who browse with what they’ve got, or, when fed up with popups and spyware, install Firefox because they’ve heard (and then experience) that it’s much better in many ways.
    Then there’s a — much smaller — group of web users who have things like “ALL: ALL: DENY” in their /etc/hosts, who use ssh and know why it’s a good idea with passphrase authentication. People in this group are likely to run Linux, to have Opera installed, and to find out how to change how to make the browser identify as something else.
    Documentation: I’ve watched the site statistics over the past few days, and Opera stayed stable at around 0.5% — even after my fervent advertisement. Then my LaTeX article (from May) ended up on Reddit, and the number of hits skyrocketed. Since Reddit is a tech oriented site, the number of Linux users grew dramatically as well — as did the number of Opera users. Here’s a screenshot of the same statistics right now, after 4,500 hits from Reddit and Slashdot:

    Opera is up where it belongs at 5%, and Firefox is unchallenged at 65%. I included the OS stats as well: Linux has risen from the <2% it had before to c. 20% (all the “others” are Linuxes and other *nixes). All thanks to a pleasant invasion of the Dylan territory by the tech people.
    All this to say that, yes, you’re right, but I don’t think that’s the explanation. :-)
    (Just out of curiosity: I see that your browser identifies as Firefox — or is that really an Opera in disguise?)

  3. thanks for the additional info, the stats are pretty interesting. It’s encouraging to see so many linux users, although like you said they were tech oriented people (even further than the normal definition of tech oriented people,, they’re LaTeX using tech people!). Also interesting that Ubuntu was pretty close to having as many as all the other linux/*nix variants combined (~8 vs ~13), and that linux outnumbered Mac. Is that very skewed from your normal statistics?

    And no, it really is Firefox :) Last time I tried Opera they had ads embedded which really turned me off. I’ve been thinking about giving it another go recently, as Firefox’s stability has plummeted, at least for me.

  4. The Mac – Linux ratio, yes, it’s very skewed. In fact, the numbers for Mac have been fairly stable around 15%, which has been a surprise for me — the numbers I’ve usually seen have been around 3–5%. Perhaps this means that Dylan people (which are my main reader base — at least so far) are more artistic than they are usually given credit for :-)
    However, the site hasn’t been up that long on the new location, so the stats are not really trustworthy — although the numbers after Reddit are significant enough. Of that particular user base, that is.

    About Opera: I’d say: unless you have some can’t-live-without’em Firefox addons, give it a try. It’s faster by far, and has some great features. One thing I didn’t mention in the original post is its approach to zooming: the whole page zooms, not just the text, and the top of the page stays where it is (instead of jumping to the relative position in the document, as in Firefox) (I’m not making myself clear here, but anyone who has been in the middle of a long page and then zoomed in or out will know what I mean).

  5. started using opera after having skimmed through yr comments. worth a try i thought after having been annoyed by firefox being notoriously slow.
    ..never looked back! absolutely happy with it.
    thanks,
    e

  6. Some more causes for Opera to score so low in web statistics:
    – When you press “back” in Opera, then page is reloaded from a local cache, but in other browsers it’s loaded from the server again, thus incrementing hit point counter for that browser. I’m not sure which browsers act that way, but you can test that by loading some site, that contiains dynamic ads. In opera if you’ll navigate back to an ads containing page, then you’ll see the same ad, as before, but in browsers, that reloads pages from the server you’ll see a new one.
    – When you search something in google, Firefox prefetches top 10 pages from the search result, to fasten page load if user opens one of them. Each of those prefetches also counts by web statistics.
    – And of course already mentioned there Opera’s ability to mask itself as another browser.
    ———–
    happy Opera user :)

Leave a Reply to Andrew Cancel reply

Your email address will not be published. Required fields are marked *