Flipping the Bird

Friday, 27 June 2003

One of the problems with major media news organizations is their tendency to emphasize conflict above all else. This vs. that. Conflict is interesting, and ostensibly, reports of conflict increase ratings and readership.

Sometimes conflict is real. But many times it is not, or at the very least, it is greatly exaggerated in news reports. Apple Computer has long suffered under this — how many articles have you read over the years, for example, claiming that Apple is going out of business? Not just from columnists and pundits, but from news articles, ostensibly written by objective reporters. One technique for this is to get quotes from “industry analysts”, i.e. wankers on Wall Street who will never be convinced that Apple is not in the same business as Dell.

The tendency to cast every story in terms of conflict is also fueled by laziness. It is easier and less work than doing the research to gain a deep and genuine understanding of the issues at hand — not to mention the difficulty in conveying this genuine understanding to your readers.

Which brings us to a despicable story by Paul Boutin published Monday in Slate, “Flipping the Switch: Linux’s new popularity may hurt Apple more than Microsoft.”

The gist of Boutin’s article is that certain industry analysts (them again) are projecting that Linux will pass the Mac OS in desktop operating system market share sometime next year, and that this means Apple is in serious trouble. What is maddening about this article is not just that it is misleading tripe — so utterly wrong-headed that it might be aptly described as completely wrong — but also that Boutin, who writes regularly for Wired, Slate, and other publications, should know better. He’s a good writer and generally treats Apple fairly, but this article is way beneath his usual standards. (And Slate’s as well — yes, they’re owned by Microsoft, but I’ve never detected even a whiff of pro-Redmond slant in their coverage of their parent company.) In fact, I can’t help but suspect that Boutin knows the premise is false, but that he wrote the article anyway because it sounds like a juicy story.

Dissection

From paragraph 1:

But while listing the new machine’s impressive specs, Jobs left out a related, eye-popping statistic: Business Week columnist Alex Salkever dropped the bomb last week that next year, “Linux should pass Apple in market share for desktop operating systems on computers.”

In what way is Linux’s desktop market share “related” to the debut of the PowerMac G5? I realize Boutin is being facetious, but it doesn’t even make sense as a snarky aside.

From paragraph 2:

A few calls to industry analysts confirmed that they’ve come to the same conclusion as Salkever: Steve’s new babies have been born into third place behind both Windows and Linux, which had been dubbed a desktop flop just two years ago.

At this point it is also worth pointing out that while Boutin links to Salkever’s Business Week column, neither he nor Salkever link to any actual sources for the statistics behind this claim. Boutin doesn’t even name the “industry analysts” he called who back this up, let alone quote them. What kind of journalism bases entire articles on statistics without sources?

But for the sake of argument, let’s assume the premise is completely true. If Linux is going to move into second place next year, how will “Steve’s babies”, which are scheduled to ship in August, be born into third place? At least according to iCal, August is in this year.

From paragraph 3:

The days when a new Mac on your desk was considered the stylish geek’s protest against Microsoft’s ubiquitous software (unless you could afford a $10,000 Sun workstation) have ended. There’s a new way to Think Different in town.

Ignoring the fact that I have never in my life met anyone whose proclaimed reason for buying a Mac was to protest against Microsoft, this is laughable. Boutin is trying to indicate that geeks are moving to desktop Linux and away from Macs? If Boutin seriously believes this, what planet is he on? Here on Earth, Unix geeks are switching to Mac OS X en masse. It’s almost startling how quickly Mac OS X has ascended to weapon-of-choice status among Unix geeks. Tim O’Reilly, in a keynote address at least year’s WWDC, said, “[Alpha geeks] are choosing Mac OS X in overwhelming numbers.” See also James Duncan Davidson on the rise of Macs at the Wisconsin Software Symposium.

This is not to say Linux is not a popular choice for Unix geeks. It is. But Mac OS X is the OS with momentum in that crowd.

Still in paragraph 3:

IBM, DEC, SCO, and finally Sun have lost the non-Windows portion of the server market to Linux, and no wonder: Linux is basically a better version of their Unix products, for free.

How can IBM have lost server market share to Linux, when they are selling Linux servers? Linux’s non-proprietary nature means it isn’t really against anyone. Because of Linux, IBM is surely selling fewer servers running AIX (their own proprietary Unix flavor), but that doesn’t mean they’re selling fewer servers overall. And DEC? How is DEC relevant when it was consumed by Compaq over five years ago?

And did I miss the announcement that Sun stopped selling Solaris servers?

From paragraph 4:

By blending gorgeous design with user-friendly software, Apple lets you buy your way out of the Microsoft world — aided by a hand-holding deal with Microsoft to help the two brands work well together.

Isn’t the opposite just as true? That by blending bland design with ubiquitous software, Microsoft and Intel let you buy your way out of the Apple world? And what does Boutin mean by “hand-holding”? That the five-year collaboration agreement (which expired last year) between Microsoft and Apple was, what?, Microsoft just being nice and keeping Apple afloat? It was a two-way deal, from which Microsoft and Apple both benefitted. Microsoft, for example, got Apple to agree to make IE the default browser for the entire five-year term of the deal.

From paragraphs 4-5:

But it comes at a hefty premium: Apple’s new desktop models start at $1,999.

Linux takes the low road, price-wise. It’s not pretty, but it’s free, plus it’s lean and fast enough to run on a yellowed old PC from the storage room if you’re willing to spend a few hours getting the software installed and running. Or, for $248, you can buy a brand new, ready-to-use Linux desktop computer from Wal-mart.com. The bargain-basement price is possible because Torvalds and other Linux programmers don’t demand license fees.

What is the implication here? That the difference between Wal-Mart’s $248 Linux computer and Apple’s new $1999 PowerMac G5 is $1751 in operating system licensing fees?

Let’s assume this $248 computer from Wal-Mart can be considered “low-end”. A fair assumption, I think. The PowerMac G5 is clearly Apple’s high-end offering. If you’re going to make a comparison, the only fair thing to do would be to compare against Apple’s low-end offering, the eMac, which starts at $799. But it also includes a built-in 17” display and a built-in modem; Wal-Mart’s computer has neither. Throw in $125 for Wal-Mart’s cheapest 17” CRT and $25 for a modem, and you’re up to $400.

$800 is still twice as much as $400, but the difference isn’t nearly as dramatic as Boutin’s straw-man $1999-vs.$248 comparison. And the eMac is a vastly superior computer. The truth is that Apple doesn’t sell “low-end” machines; they stoop no lower than mid-range, and their mid-range computers are very competitively priced. Nor is this corporate elitism; it’s just good business. The Apple brand stands for quality, and Apple’s business model depends on reasonable profit margins. The low-end PC business is a commodities game with razor-thin margins.

More from paragraph 5:

As a brand, Linux is anti-corporate and anti-consumerist, but skip the neo-Marxist gift economy theories that have sprung up around Torvalds. He’s more like the Crazy Eddie of software: His prices are insaaane!

The Linux “brand”, such that it is, is not anti-corporate. It’s not anti-anything. It’s simply pro-freedom. This is why it is so confounding, both to the press and to Microsoft. Microsoft’s favorite tactic to use against competitors is to “cut off their air supply”.

Linux is like an opponent that doesn’t need to breathe. There is nothing to cut off.

Fact: Apple is in the business of selling computers. They do not sell operating systems. Their operating system is, without question, the main appeal of their computers, but that does not make them an operating systems company. Thus, Apple does not compete against “Linux”, which is neither a computer manufacturer, or a company, or even a product. Apple does compete against companies that sell computers running Linux, but somehow I doubt they’re overly concerned about $250 boxes from Wal-Mart. Analogies comparing the computer and automobile industries are overdone, but here goes anyway: Is the emergence of very low-cost Korean imports (Kia, Hyundai, etc.) a threat to BMW?

Still paragraph 5:

Linux is fast, cheap, and reliable, in defiance of the old engineer’s adage that you can only have two out of three.

That’s a misinterpretation of the adage. Fast isn’t in reference to performance, but rather to scheduling. The adage is better stated as “Soon, cheap, good — pick two.” Which if you think about it makes intuitive sense. Once restated, Linux no longer defies the adage — it is cheap and good, but it does not gain new features quickly. Salon’s Scott Rosenberg has more on this point.

From paragraph 6:

Sure, the new Mac operating system (code named Panther) is pretty slick — it’s also based on Unix, and partially open-sourced. And Jobs has made it clear he doesn’t compete on price, but on the more complex curve of price/performance, which includes factors such as ease of use, customer support, and interoperability with Microsoft — areas where Apple is way ahead of Linux. But with technology budgets frozen or slashed in most offices and homes, it’s getting harder to compete with free — unless you’re Microsoft.

How? How does Microsoft have any advantage when competing against free?

Isn’t this completely backward? $400 computers may be cheap, but they aren’t free. It is the Linux software that is free, not the computers it runs on. Microsoft is a software company. Yes, they sell mice and keyboards and the Xbox (but they sell Xbox consoles at a loss to remain competitive with Sony and Nintendo), but their empire is built on software — the Windows operating system and Windows software.

Apple might be competing against cheap computers, but Microsoft is competing against free software. The person buying a $250 Linux computer probably wasn’t even considering a Mac. But if it weren’t for Linux, they probably would have bought a computer with a Windows license instead. But Boutin would have you believe these PCs are skin off Apple’s back, not Microsoft’s.

There are always going to be cheap PCs that sell for way below the cost of even the cheapest Mac. There were, there are, there will be. The difference is that Microsoft used to get a nice cut on every one of these, and now they’re not.

From paragraphs 6-7:

Every field in software seems to thin out to Microsoft and Someone Else. Usually, it’s Microsoft and Second Place, but this year’s game console wars illustrate the point, too: The entry of Microsoft’s Xbox hasn’t hurt first-place PlayStation 2. Instead, it bumped second-place Nintendo to third.

As the Unix wars proved, the software biz doesn’t have time for No. 3.

Boutin presents this premise — that all software fields narrow to Microsoft and Someone Else — as fact, but with no examples other than gaming consoles, where all three big guns are doing at least OK. Where else is this true?

Microsoft does tend to eliminate its competition, but I’ve never seen any evidence that two is a magic number in the competing-with-Microsoft ecosystem. But the concept sure does suit the “Apple is in trouble” premise of Boutin’s article.

Paragraph 7 continues:

Apple still has software applications not available on Linux — such as Quark for publishing, or Photoshop for graphics — but if Salkever’s analyst buddies step forth and pronounce Linux the No. 2 platform, software companies will re-evaluate their commitments.

Wow, that’s a lot of crap for one sentence.

First, QuarkXPress and Photoshop sell for $900 and $600, respectively. Don’t hold your breath waiting for them to be ported to run on $250 Linux computers. High-end software is sold to people who buy high-end computers. The Mac is always going to have commercial software applications not available on Linux.

Second, “Salkever’s analyst buddies” may hold sway over sheep-like institutional investors, but they hold little influence over Macintosh software executives. Geeks use Linux on the desktop because it’s a good system for nerdery; but the masses who are using Linux on the desktop are doing so because it is free, not because it’s good. Linux users don’t buy software; the whole appeal is that the software is all free. Trying to sell QuarkXPress or Photoshop to Linux users would be like trying to sell Cadillacs to farmers in China.

Third, “Linux” is not a single platform. Technically, Linux is just a kernel, but complaints from Richard Stallman notwithstanding, it’s also become a catch-all description for any operating system built around the Linux kernel. But it is not, in any way, a desktop GUI environment such as Mac OS X or Windows. There are several such environments for Linux, and they’re not compatible. E.g. even if Adobe went institutionally insane and ported Photoshop to Linux’s KDE desktop environment, it wouldn’t run on a Linux machine running the GNOME desktop. So it doesn’t even make any sense to claim that “Linux” will become the second most-popular desktop operating system — KDE and GNOME are entirely different desktop platforms.

But not only is Boutin wrong about major commercial developers like Quark and Adobe, he completely overlooks the Unix-layer technologies where Mac OS X and Linux are working together, hand-in-hand. Hugely popular open source Unix software — such as Apache, Perl, Python, and PHP — runs great on both Mac OS X and Linux (along with the various BSDs). Mac and Linux desktop software are very different; but their server software is very much the same. A web application written using PHP and MySQL should work the same on a Linux server as it does on Mac OS X.

Back to paragraph 7:

No doubt the graphic designers and multimedia artists who have remained loyal to Macs will continue to buy them, but to grow Apple needs more Switchers to abandon Windows — and not for Linux.

But why can’t there be both? Isn’t it plausible that low-end users — those on tight budgets and the genuine “all I want to do is email and the web” crowd — might switch from Windows to Linux, while others — hobbyists, web developers, appreciators of good design, iPod users, owners of digital video cameras, etc. — might switch to Macs?

Did Boutin find a single actual person who considered a Mac, but went to Linux instead? If so, shouldn’t the article mention them? If not, then isn’t this entire article utter bunk?

Finally, paragraph 8:

Is the new Mac the fastest personal computer ever? Maybe, but that was Sun’s line, too. I’d love to do my work on a shiny new G5, but $248 is a lot closer to a free-lancer’s purchasing power these days. Unless Jobs unveils a better, faster economy at his next keynote, my next desktop computer will come from Wal-Mart.

Ah-hah, so the author of the article is just what we’re looking for. That’s terrific reporting.

I anxiously await reading about Boutin’s experiences with his new PC. Surely he didn’t fudge this just to include evidence of someone, anyone, eschewing a Mac for a $248 Microtel box from Wal-Mart.