This is a follow-up to my entry on Robert X. Cringely's column on OS X for Intel, which led to a Slashdot story on the topic, which led to (as of this writing) over 700 comments there and 100 comments here, and a temporary increase in blog traffic to 200 times normal volume.
First, I should point out that I had a pleasant e-mail exchange with Cringely -- thanks, Robert, for taking this in good spirits. He politely declined my offer of a bet, on what I thought were reasonable grounds. I'm still interested in making a public bet on this. I'm thinking of a mechanism somewhat like Long Bets, in that bets are made in the open, the judgement process is transparent, and the amounts wagered are paid in advance. However, Long Bets doesn't allow time periods of less than two years (I'm up for a year-long bet), and winnings must be paid to pre-designated charities (which is okay with me, but might not be so with some people). If anyone has any alternatives, please feel free to let me know about them.
Second, at least one Slashdot commenter asked, in so many words, who am I and why should anyone pay attention to me? I can't answer the latter, but as for the former, I'm currently COO and co-founder of a firm building simulation-based e-learning software for the US military. (I don't often talk about my current job in this blog -- I like preserving my editorial independence as much as possible.) In reverse chronological order, I've served as:
- VP Product Marketing at QDesign (digital audio software technologies)
- VP Developer Relations and VP Business Development at Be (alternative operating systems)
- VP Product Management at Red Storm Entertainment (games)
- VP and GM at Virtus (3D modeling and visualization)
- Senior Product Marketing Manager at Adobe Systems (document production and distribution)
At various companies, I was:
- A co-designer of Tom Clancy's Rainbow Six (the first realistic tactical first-person shooter)
- The original designer of Tom Clancy SSN (the first 3D submarine combat simulation game)
- The original product manager of Adobe Acrobat (uh, it's Acrobat)
You can find my complete resume
here on LinkedIn. Does any of this make me especially qualified to comment on the question at hand? That's up for my readers to decide.
Now to the question itself: will Apple ship a version of OS X for Intel as aftermarket software, installable by customers on their existing Windows PCs? Cringely predicted this, and many Slashdot and pseudorandom readers agreed with him. I wrote that it was unthinkable to me, and many readers agreed with that analysis.
Some commenters have said that while I might be right in the main, the issue I cited of limited driver availability wasn't a real issue. These arguments have run along the lines, "If Linux has broad driver support, OS X could, too," or, "Apple could easily induce third parties to write drivers for them." I don't buy either argument, for the following reasons:
- Linux driver support is still far less broad than that of Windows, and what gains it has made have come after years and years of effort by a dedicated open source community. Apple can't imitate Linux: it doesn't have the same open source type of community, and it wouldn't have years to wait for the driver situation to improve.
- The same is true of inducing third parties to write drivers -- again, Apple would be in a situation of shipping a product that wouldn't work perfectly on a large range of machines for many years. This would contradict their reputation as the easier-to-install, easier-to-use operating system, and so would undermine a key piece of their corporate positioning at a fundamental level.
Other commenters have said that Apple should emulate Be and offer an operating system to people to try for free. If it doesn't work on someone's system, no harm, no foul. If it does work, force them to upgrade after a certain amount of time, or to gain access to certain features. There are obvious problems with this example.
- Our strategy at Be didn't work. Now, to be fair, we shifted our efforts to the "zero billion dollar" Internet appliance market, which in the end turned out to truly be worth zero billion dollars, but I don't think we accumulated powerful evidence that our strategy would have worked given time. Yes, we were signing up software developers, but although BeOS ran rings around Windows, Microsoft was able to devote the effort to Windows to make it better -- not in BeOS' range, but better enough to keep developers and users in the fold.
- At Be, we carefully targeted our marketing efforts at customers who had a mission-critical need for what we did. Our greatest successes were in the audio processing market, where we offered low-level performance that was dramatically better than that of Windows (or even of the then-current version of MacOS). These customers tended to be highly technical and willing, in theory, to do some hard lifting to install a second OS. The only reason for Apple to go after Microsoft with aftermarket software would be to gain large market share, which would mean they wouldn't be able to focus their efforts precisely as we could at Be.
- At Be, we were offering customers certain things they simply couldn't get with Windows (high performance video processing in software, predictably low latency system I/O). Apple wouldn't be able to offer customers anything they couldn't get (in one form or another) under Windows. They'd just be offering them a somewhat better experience -- and that would only be in theory. The driver issues discussed above would probably make the experience less satisfactory than Windows.
A few commenters have opined that Apple will switch from Mach to Windows as its kernel, or abandon OS X entirely for Windows. On the one hand, for those in the latter camp, they have John Dvorak
on their side. On the other hand, as one blogger
put it, "John Dvorak is smoking crack." Reading the column in question is like reading a Kennedy assassination conspiracy theorist:
The theory explains several odd occurrences, including Apple's freak-out and lawsuits over Macintosh gossip sites that ran stories about a musicians' breakout box that has yet to be shipped...
This may also explain the odd comment at the Macworld Expo by a Microsoft spokesperson that Microsoft Office will continue to be developed for the Mac for "five years." What happens after that?
This switch to Windows may have originally been planned for this year and may partly explain why Adobe and other high-end apps were not ported to the Apple x86 platform when it was announced in January.
If Dvorak is serious, he has gone off the deep end -- truly.
I haven't read anything in the comments that has changed my mind at all about the likelihood of Apple offering OS X as aftermarket software for PC owners. It was unthinkable to me yesterday morning, and it's unthinkable to me now. So what could I imagine Apple doing?
- Offer Macs with Boot Camp and Windows XP pre-loaded. This isn't a business model change, and to paraphrase, now that Apple has released Boot Camp, we know what they are -- it's just a matter of setting the price. I don't think this is probable, but I don't think it's particularly improbable, either.
- Offer Macs with some sort of Windows compatibility not reliant on Windows, like WINE. Again, this isn't a business model change. Because it would probably go out with every Mac, it would be a threat to Apple's growing OS X developer community. However, I could see Apple justifying the risk if they felt their Windows compatibility solution was solid -- they could sell OS X + Windows PCs without paying Microsoft a dime, and take $50-100 (that's a guess) out of Microsoft's pocket for each OEM sale lost to a customer who buys an Intel Mac. Again, not improbable, but not necessarily probable.
- License OS X to PC vendors like Dell. This isn't unthinkable per se, but it's highly unlikely. Jobs shut down a clone market once -- the Mac clone makers -- because they were stealing sales from him without growing the market. In other words, for every Mac clone sold, Apple was missing out on hundreds of dollars in revenue, with nothing to offset it. For Apple to work with a vendor like Dell, they would have to be absolutely convinced that either:
- Dell would be additive to the existing Mac market -- that every Dell sale would be to someone who would have otherwise bought a Windows PC, or,
- Dell would grow the Mac market to such an extent as to offset the need for Apple to sell its own hardware -- in other words, by a factor of 5-10.
It's very difficult for me to imagine Apple leaders convincing themselves of either of these suppositions. What would the evidence be for them?
This has been a fun exercise. Thanks to all the commenters who have participated -- and please feel free to keep the discussion going. And stay tuned for another Apple thought piece soon.