Asa Dotzler: Firefox and more

December 1, 2008

microsoft can't beat the math?

Just ran across an interesting, though short on supporting details, article by John Martellaro at The Mac Observer titled Apple and Microsoft War Now Dictated by Mathematics.

The premise is that Microsoft, and Internet Explorer specifically, are in a monotonic decline that might be slowed but can't be reversed.

While I don't have math on my side, I do have 10 years in the "browser business" and I'm gonna go way out on a limb and say that nothing's impossible and anything's possible. But. I think he's basically right. There just aren't many opportunities for radical change in the browser marketplace.

The momentum is what it is because of some deeply rooted factors that aren't likely to reverse easily.

Why is that?

I think it's primarily because the investment required to build a capable Web browser that can single-handedly change the market is just too large. It's been done exactly three times.

The first was Netscape, back in the early and mid 90s. The second was Internet Explorer in the mid and late 90's. The third was Mozilla's Firefox which didn't rise to prominence until the mid 00s.

Netscape had the "first to market" advantage and that was huge. That, by definition, won't be repeated.

Microsoft illegally abused its Windows desktop monopoly, something that also probably won't happen again any time soon.

Firefox had a much more difficult time of it and while we're definitely in the top tier, we haven't made the same market share gains as Netscape and Microsoft did, even though Mozilla's investment and effort have been larger and more sustained than both Netscape's and Microsoft's combined.

It's taken 10 years of investment in the development of the code that makes up Firefox and the wide range of tools that support Firefox and the rest of the Mozilla ecology. It's taken thousands of coders, hundreds of thousands of committed testers, and millions of determined advocates just to pull away 20% of the browser market and substantially weaken the I.E. stranglehold.

So, that tells us that being first obviously works. Breaking the law also seems to work if you've got the cajones. And finally, a decade of really hard work by millions of people sharing a common purpose also does pretty well.

Those are the three examples we have of radical browser change in the Web's ~15 year history. No other effort (and there have been literally scores) have been able to effect serious and sustained movement like the three listed.

People have tried with different technologies, different business models, and different motivations but none have really taken off. Growing deep roots just isn't an easy thing to do.

But there's hope for wider competition because we're creating a new era for the Web where there's increased space and opportunity for a broad undergrowth of niche players.

If Mozilla can continue to make the agreed upon web standards a necessity for all browser vendors, if we can make process, performance, and security the major points of differentiation, and if we continue to focus our features for a broad-based appeal -- all the while further eroding Microsoft's dominance on both the Web and the desktop, then I think there's going to be a lot more space for other browsers like Flock, and Safari, and Chrome.

But, yeah, I don't see Microsoft reversing that trend any time soon.

update: I wandered there a little bit and maybe buried the point I that inspired this post.

Let me try to summarize and crystalize my thinking here.

The Web is a large and massive thing and it's very difficult to get it moving in a new direction. Once it gets moving in one direction, it's got a momentum that's difficult to alter. It's not impossible, it's just difficult and either extraordinary circumstances (Netscape and Microsoft) or extraordinary effort (Mozilla) are required.

It's rare that any one entity has caused a major shift in velocity of the web.

Mozilla, having effected a real change in that velocity with Firefox, is acting differently than the other two examples because Mozilla's mission is not to try to consolidate control and become the sole determiner of the direction for the Web, but rather to help more people and projects (and yes, even other browsers) play a much bigger part in determining where this thing goes than they've been able to in the past.

Posted by asa at 4:20 PM

 

reactions, thoughts, comments, etc.

Of course, interestingly, Netscape, um, arguably "borrowed" the code from NCSA Mosaic whereas Microsoft licensed that same code. So arguably the first two "revolutions" were just one revolution repackaged twice.

Mozilla drastically reworked that Netscape codebase and eventually replaced most if not all of it, but Netscape still built from that Mosaic, um, "inspired" base of code and/or knowledge. So maybe it's just one revolution?


And between Safari and Chrome, I'd say that the jury is probably still out on the desktop and laptop. And with the coming proliferation of portable Internet-devices, which will likely outnumber personal computers in relatively few years, I'd say that the jury is still out overall.

Netscape and Microsoft both previously erred by prematurely declaring victory. Let's not have it happen with Firefox as well. No victory laps until the race is over!

reinharden

Posted by: reinharden | December 1, 2008 9:02 PM

Don't forget that Google has an excellent way to push the browser, but they are not using that it the browser is ready.

Posted by: VanillaMozilla | December 2, 2008 2:05 AM

Interesting point VanillaMozilla.

It would be really intersting to see what would happen if Google kept pushing Chrome on google.com over a long period of time. (They did so at the time of release, but only for a short period of time: http://davidnaylor.org/blog/2008/09/nice-google.html )

Posted by: David Naylor | December 2, 2008 4:40 AM

IE Browser share decline is NOT monotonic and the maths presented on their site is a weak outline to a flawed argument at best. I'm not disagreeing with the points raised in it, nor the points in this blog, but it has nothing to do with maths other than the vague statistical observations that IE market share has a overall negative trend.

Posted by: Damian | December 2, 2008 7:39 PM

Actually IE's fairly linear decline in market share completely disproves their mathematical model.

The author also doesn't seem to understand the difference between affine (or linear) and monotonic. A decreasing monotonic function means that any point x f(y), where f here is browser share and the variable is time. But we can clearly see from your own chart Asa:

http://farm4.static.flickr.com/3246/3073834826_f39957e180_o.png

There are many times where IE gains slight market share back before losing it, sure that means the trend is decreasing but it also means it's not monotonic.

Their factorial argument suggests that market share should behave like population growth, although they haven't expanded on the mathematics of this (presumably because they don't know how too). But population growth models have an exponential-like quality to them, so you would expect to see such a quality in market share trends, but in fact you quite clearly don't get this and even at pointed events like the Firefox 3 release the trend doesn't significantly change from linear.

Furthermore I think their comparison is deeply flawed on both with IE and with Windows. Buying database software is not the same as choosing a browser or purchasing a Mac/PC/Single copy of Windows. They also don't seriously analyse any actual market share, they just pull this arbitrary comparison to database software from the 90s.

Posted by: Damian | December 2, 2008 8:01 PM

Mistake above:

The comment page seems to have taken my less than and more than as HTML brackets. I wrote:

A decreasing monotonic function means that any point x < y then f(x) > f(y)

Posted by: Damian | December 2, 2008 8:05 PM

If you are going to use Spanish words, use it correctly.
1. Cajones means big boxes or trunks (plural for cajon or box or trunk in English)

2. Cojones is a vulgar word that means balls or nuts (as in vulgar for testicles). The word is used irrespective of the size of the balls. :)

I believe you meant "cojones". ;)

Posted by: Ramirez | December 6, 2008 10:30 AM










Remember personal info?






asa2008.jpg