Jackson: Unrepentant

[libertango]

Thomas Penfield Jackson apparently still doesn't let evidence get in the way of his conclusions. From that interview, a relevant quote:

"Microsoft has won the browser war in the United States. Netscape Navigator, if it is still available at all, has only a small fraction of the browser market."

Guess he's never heard of Firefox. Or how well it's doing in the market. For example, this report -- where IE and Mozilla are neck-and-neck -- is typical of a whole bunch of Google hits on the search terms, "browser market share".

Comments

[darkmane.livejournal.com]

But Ummm... he's right.

NN 4.7x is gone, while mozilla, firefox etc, were built with Netscape resources. they are completely different.

One Browser War I is over. Browser War II (or the War on Open Source) is just beginning and it will last forever.

And while Firefox is growing in market share, it's still got to grow about 5 times to reach majority status. The only feature they can keep as thier own is that they aren't integrated with OLE/WinDNA/COM on Windows, keeping them from being the prefered virus delivery platform. All the others can and will be co-opted by IE.

Market share

[hal-obrien.livejournal.com]

One of the things MSFT has been superb at, over the years, has been what John Boyd would call the OODA Loop -- Observation, Orientation, Decision, Action. It's his base model for how military units make decisions on the battlefield, and one side usually wins when they are able to sustainably execute their loops faster than their opponents. The bloody miracle of MSFT has been that even as they got bigger, they were still able to execute their loops quickly. And, let's face it -- MSFT's two largest opportunities came when the other guy blinked: First when Phillipe Kahn decided to re-write Paradox for Windows after he acquired Ashton-Tate (allowing enough time for MSFT to acquire FoxPro, develop Access, and generally create the market for office suites), and when Netscape decided to re-write their browser from the ground up. (See Joel Spolsky on this.)

The problem for MSFT now is: a) I think the Firefox folks already realize this, and won't let up in their loops, and b) MSFT is nowhere near as spry as it used to be. I may well be wrong, but I think MSFT has finally gotten to the point where its size is now a hindrance to getting things done.

What that means is, sure, perhaps Firefox has features that "can and will be co-opted by IE." The question is, in what kind of time frame? And will the customers even care? Technical chops alone don't define this business, or we'd be talking about the great Holy War between Amiga and Be.

Re: Market share

[darkmane.livejournal.com]

I have no information about the Paradox situation, but I can talk about Netscape.

The final blow to Netscape actually came much earlier than thier decision to re-write Mozilla from the ground up. I put it about the time HTML 4.0 was finalized and their proposal for "Layers" was not adopted. That decision moved them out of the market leader position and into a follower position, not to mention making the effort they had put into NN 4.7 largly wasted. In my opinion that was when NN really lost the Browser War. And that was in 1996-7.

The decision to re-write Mozilla came later in 1998, and it was kind of the opening salvo in the "Browser War II" or a new phase in the "War on OpenSource". MSFT has been sitting on it's laurels for the past 3-4 years, we'll see what happens with new versions of IE.

Re: Market share

[blufive.livejournal.com]

(blundering in three days later upon return from holiday)

I think Netscape's mistakes were even earlier. Going from Netscape 3 to Netscape 4, they bolted on new HTML features like there was no tomorrow. There were several consequences:

- The "features" were often hurried and poorly designed. As a result, they were passed over by the standards bodies. Microsoft suffered from this too, but I think they did better than Netscape.

- The features were poorly implemented. Several ex-Netscape developers are on the record commenting that the old NN4.x rendering engine code was a complete mess, bug-ridden and crufty to the point where it was almost unsalvageable.

- While they were rushing to implement all their own cool stuff, they lost track of what everybody else was up to - and dropped the ball completely on CSS support. At a stroke, that cost them support of many web authors.

Re: Market share

[blufive.livejournal.com]

Addendum: as a result of all the poor design/implementation, NN 4.x was a lousy user experience. I bailed after about 4.7, because the thing crashed so damn oftem it was almost unusable. Faced with that mess or IE4 (which had problems, granted, but...) it was a no-brainer for most users.

Re: Market share

[darkmane.livejournal.com]

Those decisions lead to the point I was talking about.

The code was unsalvageable, they got about a 5 fold increase in teh number of developers and after 6 months of trying to salvage it, they gave up and started what we now know as Mozilla.

They had CSS support, but they tried to do what they did in the HTML 3.0 time frame and define standards, of course when the decision was made to use a different definition they were screwed, it could have gone the other way and we'd be talking about IE like it was Bob. What killed them was largely arrogance and political machinations of MSFT.

Re: Market share

[blufive.livejournal.com]

Those decisions lead to the point I was talking about.

Indeed.

They had CSS support, but they tried to do what they did in the HTML 3.0 time frame and define standards, of course when the decision was made to use a different definition they were screwed, it could have gone the other way and we'd be talking about IE like it was Bob.

What CSS support they had was a nasty JS-based hack on top of the rendering engine; it was stretched practically to breaking point by NN4.7x.

So, in the long term (12+ months) the rewrite was an inevitable consequence of those decisions, not a decision in itself. If the CSS standards decisions had gone their way, it might have delayed things for a while, but they still would have been in deep trouble.

What killed them was largely arrogance and political machinations of MSFT.

On that, I agree totally.

Re: Market share

[blufive.livejournal.com]

Like darkmane, I don't think NS lost the war when they decided to re-write; they were already in deep doo-doo for other reasons, and that was their last, desperate hope of escape from the oncoming juggernaut. For Netscape as a company, it didn't work. For the underlying software, however, it worked.

To be really nit-picky, they didn't re-write the browser, they re-wrote their layout engine - the stuff that turns HTML into what you see on screen - a critical component of a browser, but not a browser. I believe significant lumps of old Netscape code (e.g. plugin handling, javascript engine, image rendering, network code, and more) lived on in the mozilla codebase, getting further bug fixes and enhancements. Several of those areas have since had their own (less visible) ground-up re-writes in turn.

Netscape (the company) didn't die because of the Layout engine re-write; they died because they were a (relatively) small company that relied (mostly) on a single revenue stream. Microsoft comprensively annihilated that revenue stream, they died. Just another small company getting squished by a much larger competitor moving aggressively into their market. (I'm going to stay out of the arguments about the legality of MS's behaviour right now)

All that said, I think the immediate future will be interesting. Recently, it's MS that's been complacent, and the mozilla team have grabbed the initiative. How long they can keep the initiative, now that MS is waking up, is another question.

[replyhazy.livejournal.com]

Well, that guy's report is only for HIS website. Isn't it?

Firefox is doing very well indeed, but I don't think, globally, that it's neck and neck with IE.

Reports

[hal-obrien.livejournal.com]

No, that original report is only for the author's own web site. But he notes some interesting things:

"I get hit by lots of referer-spambots (idiotically, since I don’t post referers) and they universally claim to be IE."

Then, later:


"Here’s a hypothesis: on weeks when I wrote mostly short pieces that fit entirely into the RSS feed, Mozilla spikes down and IE up. People who use RSS are probably in aggregate a bit more geeky, and such people are more likely to use a non-IE browser; so short articles preferentially filter them out.

Note that Mozilla spikes upward, and the other lines sag, every time ongoing gets Slashdotted. Similarly, Safari spikes, and others sag, when I write something that gets noticed in the Apple community."

This all goes along with this blog post, that takes note of the market segmentation that seems to be going on: People who use the Net a lot use Firefox, those who use it more casually use IE (assuming they're not automated -- see above).

But even in the broader marklet, Firefox is growing like crazy: See this report, or this one.

Re: Reports

[blufive.livejournal.com]

The second of those stats links is for another single-site report, which actually shows gecko declining slowly over the last few months.

That's the first report I've seen from the other source (janco), and I'm still trying to figure out how much I trust them. There's some odd figures in there (the decline in the figures for Netscape and AOL from Jan-April looks funny, for starters)

Finally, may I take this opportunity to plug my posts over on stats_weenie where I wibble at length about stats in general.