Kelson's blog

Opera Files Antitrust Complaint Against Microsoft

Fri, 14 Dec 2007 11:07:08 -0800

Opera Software has filed an antitrust complaint with the European Commission against Microsoft. In it they allege that Microsoft has used their monopoly on the desktop OS to obtain and a monopoly on the web browser, and that they have maintained this monopoly by not properly supporting open web standards. (Recall that Microsoft was convicted of this nearly a decade ago in the United States.)

The complaint seeks two remedies:

Unbundle Internet Explorer from Windows, or at least provide alternatives on installation.
Fully support open standards in Internet Explorer.

The Q&A with Opera's Haarvard is worth a look.

So how does this affect Firefox?

Well, if they succeed in getting IE separated from Windows in Europe (like Windows Media Player), computers are still going to have to ship with a web browser. That opens the door for Firefox to come pre-installed on systems.

If they succeed in forcing Microsoft to improve their standards support, then it will be much easier for web developers to target four major browsers with one set of code: IE8, Firefox, Opera, and Safari (or rather IE8, Gecko, Opera and WebKit). Once IE6 and IE7 fade away, the biggest barrier to alternative browsers -- IE-only websites -- will become a relic of the past.

Hey, we can dream!

It will be interesting to see how this turns out, and what impact it will have on the web browser market.

If you have time to read a zillion responses, you can check out commentary at CSS3.info, Web Standards Project, Slashdot, Asa Dotzler, and Opera Watch.

(And before anyone starts posting "Opera sux," please go read LIJI's excellent post, 3 Browsers, One Enemy, or my site, the Alternative Browser Alliance, then come back and try to post something a little more insightful.)

Firefox is Back on the Attack

Mon, 05 Mar 2007 15:26:20 -0800

According to Net Applications, Firefox is again gaining on Internet explorer: Firefox, Safari Back on Browser Attack. After a slight dip in January, Firefox has climbed to 14.18% in February -- still ahead of December's 14.00%.

Month	IE	Firefox	Safari	Opera
December, 2006	79.64%	14.00%	4.24%	0.87%
January, 2007	79.75%	13.67%	4.70%	0.73%
February, 2007	79.09%	14.18%	4.85%	0.79%

This shows that January's drop was only temporary. Firefox has caught up to its previous position and passed it, as all three major alternatives to IE gained ground.

Source: Net Applications Statistics.

The value of auto-updates

Mon, 15 Jan 2007 13:13:33 -0800

I just noticed something interesting on my website stats. For the first 15 days of January (well, 14½ days), Firefox accounted for 22.5% of my hits. The top three versions were:

Firefox 2.0.0.1	9.7%
Firefox 1.5.0.9	9.2%
Firefox 1.0.7	0.8%

Notice anything interesting? The first two are the most recent security releases in major versions that have had auto-update. 1.0.7, however, is not. In fact, 1.0.8 (which is the most recent security release in the 1.0 series) comes in at only 68 hits, less than any other non-beta release, and registering 0%.

It looks like a lot of people stopped at 1.0.7 -- the last release before Firefox 1.5 -- and never updated to 1.0.8. By contrast, people on the 1.5 series have mostly updated to 1.5.0.9 (which came out after Firefox 2). What's different? Auto-updates.

Firefox 2 vs. IE7: A Factor of 2

Fri, 05 Jan 2007 16:39:47 -0800

I noticed something interesting about my website's browser stats over the last few days. I'm specifically looking at IE7 and Firefox 2 as compared to the total IE hits and total Firefox hits.

IE7 has been running about 21,000-29,000 hits/day. Firefox 2 has been running 12,000-17,000 hits/day. So by that measure, there are almost twice as many IE7 users as Firefox 2 users.

Compared to other versions, though, IE7 accounts for ~23% of all IE hits, and Firefox 2 accounts for ~44% of all Firefox hits. So nearly twice the percentage of Firefox users have upgraded to the latest version compared to IE users.

I found it very interesting that the ratio reversed itself so exactly when switching from raw numbers to per-browser percentages.

Firefox 2 Is Taking Off!

Tue, 31 Oct 2006 11:24:44 -0800

For most of October, the vast majority of Firefox users hitting my website were using Firefox 1.5, The percentage using Firefox 2 (betas and RCs) hovered around 3% for the first three weeks. The percentage of IE users with IE7 hovered around 2%.

The IE7 percentages didn't change much when IE7 was released. It climbed from ~2% to ~4%. But Firefox 2 jumped from 3% to 8% the day it was released, and has already climbed to 22% in less than a week!

Date	Fx2/Fx	IE7/IE
Oct 10	5%	2%
Oct 11	3%	3%
Oct 12	5%	2%
Oct 13	5%	2%
Oct 14	1%	3%
Oct 15	2%	2%
Oct 16	3%	3%
Oct 17	3%	2%
Oct 18	2%	3%	Internet Explorer 7 Release
Oct 19	3%	3%
Oct 20	4%	3%
Oct 21	4%	4%
Oct 22	3%	4%
Oct 23	5%	4%
Oct 24	8%	3%	Firefox 2 Release
Oct 25	12%	4%
Oct 26	12%	3%
Oct 27	15%	3%
Oct 28	14%	5%
Oct 29	17%	4%
Oct 30	22%	3%

IE7 will probably take off more next month, when automatic updates start offering it. But it's interesting to see: when Firefox offers a new version, nearly 1/4 of its users have upgraded within a week. When IE offers a new version, everyone wants to wait and see.

(I've got even higher FX2 numbers on the Alternative Browser Alliance, but traffic's so much lower that the stats are very erratic. On that site, the second-highest percentage of IE7 hits (14%) is actually a week before the release!)

IE7 Requirements: Let's Target Win2K and Win9x Users!

Thu, 19 Oct 2006 10:31:30 -0700

Well, IE7 is out. As we've known for over a year, it's only available to people using Windows XP or later, and only those on the latest service pack. Windows XP is the most commonly used operating system out there right now, but there are still quite a few users of Windows 2000 (especially businesses that have standardized on it) and Windows 98 (mostly home users, I'd expect). You can still find people using Windows Me and even Windows 95!

As an example: My website gets 83.2% of its hits from Windows XP, 4.5% from Windows 2000, 2.1% from Windows 98 and 1% from Windows Me.

All these people with older versions of Windows can't upgrade to Internet Explorer 7. Here's where Firefox comes in. Firefox will run on all of these operating systems. One component of the Firefox 2 promotion should be a campaign aimed at these Win9x/Me/2k users who can't upgrade Internet Explorer, but can still get a modern web browser by installing Firefox. (I wrote a post on my own blog on this subject a few days ago.)

The only problem with this scenario is that Firefox 3 won't run on Win9x/Me. We might risk alienating people if we encourage them to switch to Firefox since they can't upgrade IE, then 8 months from now they can't upgrade Firefox either.

Voyage of the FyreFawkes

Tue, 19 Sep 2006 16:43:42 -0700

In observance of Talk Like a Pirate Day, I've written The Voyage of the FyreFawkes -- a tale of the browser wars on the high seas.

Nothing to see here

Thu, 29 Jun 2006 10:41:24 -0700

Just staking a claim for my Technorati Profile.

Move along, nothing to see here. Though if you really want to see some content for your trouble, please check out the Alternative Browser Alliance.

Know Your Enemy

Mon, 15 May 2006 15:21:47 -0700

There's a lot of misinformation out there about various browsers. Firefox can/can't do this. Opera can/can't do that. There's only so much you can do to promote one product when you only know rumors or outdated facts about another.

If someone told you Opera was better than Firefox because Firefox won't let you reorder tabs, you wouldn't take them seriously. You'd know that Firefox 1.5 did just that, and you'd wonder what else they have wrong. Similarly, you won't convince an Opera user that Firefox is better because it doesn't have ads (Opera doesn't anymore). And you won't convince an IE fan that Firefox is better because of tabs and a built-in search box because they'll tell you that IE7 has both.

When you're trying to convince someone that X is better than Y, and they know Y very well, you'd better know Y well enough not to make statements that the other person knows are false. When you do, you'll lose credibility, and the rest of your argument -- the part you do know well -- will suffer for it. (I suspect a lot of software flame wars get started this way!)

So here's my suggestion: Everyone here, go and download the Opera 9 beta and (if your OS will run it) the IE7 beta. If you're on a Mac, fire up Safari. Mess around with them enough that you're familiar with how they work, what you can do with them, and how they handle your favorite web pages. That way the next time you face an IE fan (to the extent that IE has fans), or an Opera fan, or a Safari fan, you'll be armed with accurate information.

As for the post title -- I don't think it's necessary for the major browsers to be enemies. I think there's plenty of room for cordial competition rather than a cutthroat struggle. But "Know Your Enemy" is a better attention-getter than "Familiarize yourself with the competition." ;-)

Firefox 1.5 Approaching 60%

Thu, 02 Mar 2006 11:33:30 -0800

I've got a script that checks the website logs for Firefox users, then sorts them into 0.x, 1.0.x, and 1.5.x bins. In mid-January, Fx 1.5 accounted for 45% of Firefox visits to my website Hyperborea, up from about 15-20% at the beginning of December.

Upgrades continue, and now nearly 60% of Firefox visitors are on 1.5 or later. In fact, it only dipped below 50% twice in the entire month of February, and those were both 49%. (I can get higher numbers by running the script on the Alternative Browser Alliance, but there are two problems: the target audience is highly tech-savvy, so they're more likely to upgrade quickly, and the traffic is much lower, so the results are very erratic. Though on busy days, when there's enough traffic to factor out noise, it seems to approach ~80% regularly.)

Three months for 60% of the user base to upgrade? That's not bad at all! I don't remember how long it took Internet Explorer 6 to outstrip IE 5.5, but I seem to recall it was more than 3 months... and as much as I'd wish otherwise, there's no way IE7 will replace IE6 as quickly, either.

Based on the stats I've collected, here's a graph of the upgrade rate:

Upgrades do seem to be slowing down a bit, which is to be expected.

Finally, in case anyone's interested, here are the statistics for the past month.

Day	%Fx visitors on 1.5
Feb 1	49%
Feb 2	52%
Feb 3	52%
Feb 4	51%
Feb 5	53%
Feb 6	51%
Feb 7	56%
Feb 8	49%
Feb 9	51%
Feb 10	53%
Feb 11	51%
Feb 12	54%
Feb 13	58%
Feb 14	55%
Feb 15	55%
Feb 16	56%
Feb 17	53%
Feb 18	57%
Feb 19	56%
Feb 20	56%
Feb 21	57%
Feb 22	54%
Feb 23	52%
Feb 24	55%
Feb 25	56%
Feb 26	53%
Feb 27	57%
Feb 28	61%

Firefox 1.5 uptake reaches 45%!

Tue, 17 Jan 2006 11:24:15 -0800

I've been using a script to check the percentage of Firefox-using visitors who have upgraded to 1.5. Things have vastly improved since the last time I posted upgrade stats. Of course, I am trying to help things along by targeting Fx/Upgrade/SFx banners based on the user's browser.

Back in early December, only about 15-20% of Firefox users hitting my website, Hyperborea, were using the latest version. Six weeks later, it's been holding at 44%-45% for the last four days. (Percentages are even higher for the Alternative Browser Alliance, but the sample size is too small to draw any conclusions.) If this rate holds, in another three months nearly all Firefox users will have upgraded.

It won't be 100%, of course, and I'm sure it'll taper off. I'm still seeing a few hits from 0.x versions. There are people who don't want to be bothered, people who don't have the system access, people who are using the version that came with their Linux distro, etc. I'm sure there will still be a few people on 0.x and 1.0 at the end of the year, but the upgrade rate is definitely promising!

Day	%Fx visitors on 1.5
Dec 1	20%
Dec 2	15%
Dec 3	18%
Dec 4	17%
Dec 5	23%
Dec 6	21%
Dec 7	21%
Dec 8	25%
Dec 9	20%
Dec 10	22%
Dec 11	20%
Dec 12	19%
Dec 13	25%
Dec 14	30%
Dec 15	26%
Dec 16	27%
Dec 17	31%
Dec 18	31%
Dec 19	33%
Dec 20	34%
Dec 21	32%
Dec 22	34%
Dec 23	29%
Dec 24	36%
Dec 25	29%
Dec 26	29%
Dec 27	37%
Dec 28	31%
Dec 29	35%
Dec 30	40%
Dec 31	37%
Jan 1	35%
Jan 2	35%
Jan 3	37%
Jan 4	40%
Jan 5	41%
Jan 6	37%
Jan 7	38%
Jan 8	39%
Jan 9	43%
Jan 10	43%
Jan 11	39%
Jan 12	41%
Jan 13	45%
Jan 14	44%
Jan 15	45%
Jan 16	45%

Conditional Firefox/SFX/Upgrade Banners using JavaScript (DOM-compliant)

Wed, 04 Jan 2006 00:30:21 -0800

This is a follow-up to my earlier post that used old-school JavaScript.

It's kind of redundant to post a "Get Firefox!" banner to someone already using Firefox, but it's useful to show them an upgrade banner if they're on an old version. It's also useful to show Firefox users a banner for Spread Firefox.

This can be done easily with PHP or other server-side scripting languages, but sometimes you have to use static HTML files. That's where client-side scripting becomes necessary. Last month I posted some sample code that used document.write, which doesn't work with XHTML. (On top of that, the <noscript> blocks I used ended up causing validation errors because of their location!)

I've redone everything with DOM scripting, which will work with either HTML 4 or XHTML 1.0.

1. Put the following in a file called sfxlinks.js:

2. Use your regular Spread Firefox affiliate link and add a unique ID -- let's use id="FxLink" as an example -- to the <a> tag.

3. Load the script in your document's <head> section:

4. Call the function in the body onload event using the ID you chose in step 2:

When the page loads, the script will check the visitor's browser to see if it's an old version of Firefox or a current version of Firefox. If it's an old version, it'll replace your standard button with your upgrade button. If it's a current version, it'll replace it with a Spread Firefox button with your referral link. Otherwise, it leaves the button alone.

This has a lot of advantages over the old version, including XHTML compatibility, no need for <noscript> blocks, easier validation, and it still degrades gracefully (if JS is unavailable or old, it leaves your normal button in place).

You can see it in action on my website, Flash: Those Who Ride the Lightning.

Targeted Banners with JavaScript

Fri, 16 Dec 2005 13:03:13 -0800

If someone already uses Firefox, there's not much point in posting a Firefox banner on your site -- but it might be worth posting an ad for Spread Firefox. If they have an old version of Firefox, you'll probably want to show them an upgrade banner, though.

Someone posted PHP code a while back to handle this, and while server-side parsing is definitely the best way to handle it, sometimes you don't have it available. The busiest section of my website is almost all static HTML pages, so client-side scripting was the way to go.

Note: This method does not work with XHTML served as application/xhtml+xml! I plan on rewriting it using DOM functions, but this was a quick-n-dirty fix with document.write, which does not work in XML. It will work in HTML and in XHTML served as text/html.

/* Post upgrade link for Firefox visitors only */
if (0 <= navigator.userAgent.indexOf('Firefox/0') ||
    0 <= navigator.userAgent.indexOf('Firefox/1.0')) {
    document.write('[Code for Upgrade Fx Banner]');
} else if (0 <= navigator.userAgent.indexOf('Firefox')) {
    document.write('[SFx Affiliate Link and Banner]');
} else {
    document.write('[Code for Standard Fx Banner]');
}

I've used indexOf instead of regular expressions for efficiency.

You can place this inside a <script> tag on each HTML page, or you can put it in a .js file and call that file on each page.

Don't forget that some people will be surfing with scripting disabled! If you want them to see a basic Firefox banner, be sure to add another copy of your banner code encased in a <noscript> tag! Together, they should look like this:

UPDATE: I have rewritten this using DOM scripting, which works with XHTML.

Thoughts on Browser Upgrade Patterns

Mon, 05 Dec 2005 19:04:04 -0800

I thought I’d check my sites’ stats to see how quickly people are upgrading to Firefox 1.5. I’ve got a script I wrote a while back that totals hits by Firefox 0.x, 1.0.x, and now 1.5.x and shows the percentage of the latest version out of all Firefox hits.

I tried it on this month’s logs from the Alternative Browser Alliance first, since it’s a much smaller log file, and saw that Firefox 1.5 accounts for anywhere from 43% to 69% of Firefox users visiting the site on a given day. Then I checked it against Hyperborea, which gets a more mainstream audience, and found that only 15-24% of its Firefox hits were from users who have upgraded.

Sticking with the more mainstream site, I looked at some other statistics. While Firefox as a whole is doing quite well at 18.9% (plus another 1.4% for Netscape and 1.2% for Mozilla), there’s a shockingly large number of people still using Internet Explorer 5 for Windows. MSIE 5.0 and MSIE 5.5 are each pulling 1%. That doesn’t sound like much, but there are more people on outdated versions of IE than any version of Netscape, and each IE version is pulling in more than all versions of Opera combined (0.9%). This, frankly, sucks. MSIE 6 is a free upgrade that will run on any system that can run ether 5.0 or 5.5, and is a significant improvement over either. IE6, Firefox, or Opera will all run on even obsolete versions of Windows—and there are more of those than you might think as well! (I’m seeing 4.8% of traffic coming from Windows 98—more even than Windows Me.)

100m: Still going strong

Wed, 19 Oct 2005 11:29:57 -0700

Ah, the 100 million mark! Time to check the calendar:

November 9, 2004: Firefox 1.0 released
February 16, 2005: 25 million downloads
April 29, 2005: 50 million downloads
July 26, 2005: 75 million downloads
October 19, 2005: 100 million downloads

Each 25-million-download milestone has taken 2½-3 months. For all the doom and gloom we're hearing about adoption rates slowing, downloads don't seem to have slowed down much at all!