At some point in the next two to six months, Apple will unleash their latest big cat: Mac OS X Snow Leopard. It's no secret - Apple has been talking about Snow Leopard for ages, with particular emphasis on how Snow Leopard will focus on performance, efficiency, and "core innovation" rather than user-focused features.
Specific improvements promised for Snow Leopard include support for up to 16 TB of RAM; improved multi-core support for applications; a next-generation version of QuickTime; out-of-the-box support for Microsoft Exchange in Mail, iCal, and Address Book; and support for OpenCL, which is designed to expose the computing power of modern graphics processing units. Other Snow Leopard promises have included a smaller memory and disk footprint, a faster installation time, and a tweak to Stacks to allow subfolders.
Sound exciting? From a developer standpoint, absolutely. From a user standpoint, not so much. Based on everything Apple has said so far, Snow Leopard won't, on its own, bring any of the marquee features that could change the way you use your Mac, much as past releases of Mac OS X brought us Time Machine, Screen Sharing, Spotlight, Dashboard, Expose, Automator, Front Row, Spaces, Stacks, and more. So how much are you willing to pay for an operating system upgrade that does exactly what your current one does, but uses a little less RAM in the process?
Don't get me wrong. I applaud Apple for taking a break from the feature-based rat race to concentrate on the underpinnings of Mac OS X - along with all those slick features has come bloat. Mac OS X has grown portly, a change largely swept under the rug by increases in CPU performance and decreases in hard drive and RAM costs.
I suspect that some of the changes Apple promises in terms of reduced memory and hard disk footprint are related to the work done at the core of OS X for the iPhone. The world is moving to ever-more-mobile devices, and as a result, toward RAM-based storage that won't compete with rotating disk storage on a price-per-gigabyte basis for some time. If Apple is to be able to innovate in the hardware world - perhaps with much-rumored devices that fit between the iPhone and the Mac in size and capability - a leaner, more efficient operating system can only help.
All this leads to my main point: Assuming that Snow Leopard will indeed feature only the under-the-hood improvements promised so far, Apple should release it for free, instead of the $129 price of most releases. Although I say "free," I could easily be talked into the $29.95 charged for the Mac OS X Public Beta (which could be deducted from the cost of Mac OS X 10.0); Apple's standard $9.95 media cost for those who want to receive it in the mail on DVD also doesn't bother me at all. But it should become a no-brainer to upgrade to Snow Leopard, whether you're running Leopard or Tiger now.
I have no inside information here, and I am not arguing from an "information wants to be free" point of view. But based on what we currently know about Snow Leopard, I think Apple - and the Macintosh industry as a whole - stands to benefit more from making Snow Leopard free for anyone whose Mac meets the hardware requirements than from charging for it. The reasons break down into two basic categories: the benefit of a coherent Macintosh platform and the difficulty of marketing purely under-the-hood changes.
One OS to Rule Them All -- This is the crux of the matter. From a business standpoint, older versions of Mac OS X do nothing but create costs for Apple and for developers, but it's difficult to encourage users to upgrade without an incentive. With the bold move of making Snow Leopard free or very cheap, Apple would attract not just all Leopard users, but every user of Tiger (with compatible hardware) who had put off upgrading to Leopard because the new features weren't worth $129.
Apple wouldn't earn any money from getting laggard users to upgrade, of course, but with Snow Leopard as the sole target platform, users with Macs that were too old for Snow Leopard would have even more reason to buy a new Mac. Let's not forget that new Mac sales are still the core of Apple's business.
Why would this be worthwhile? Developers must continually decide how far back in the evolution of Mac OS X to aim their code. If Snow Leopard became nearly ubiquitous, developers could concentrate their efforts on it, rather than spending resources on Tiger and Leopard as well. That might result in faster development times, better applications, and more total applications, all of which benefit Mac users and Apple too. As more applications begin to require Snow Leopard, the pressure to upgrade would increase on those who had stuck with much older Macs.
The single coherent platform could have other benefits for Apple too. I'm going out on a speculative limb here, but if I were in charge of Snow Leopard, I'd put a lot of effort into improving Mac OS X's security architecture. Were that to happen, Apple might want Snow Leopard to be as widespread as possible to reduce the chance of a high-profile security exploit hurting Mac OS X's reputation for being relatively free of malware.
The final reason I think it makes sense for Apple to move the Macintosh to a single coherent operating system platform is that it has already worked once. Just look at the iPhone and iPod touch, which have sold a combined 37 million units so far. With them, Apple has made major operating system upgrades either free or inexpensive (iPod touch users have had to pay small fees for upgrades). As a result, there's a single target for developers, and a better experience for users. As far as I'm aware, almost no one has passed on the iPhone software updates.
The alternative - charging the full $129 price for Snow Leopard - could have deleterious effects. Were Apple to charge a significant amount for Snow Leopard, a high proportion of users wouldn't upgrade, further fragmenting the installed base, and making it harder for developers to justify new Mac products that take advantage of Apple's latest technologies. This could also hurt the overall reputation of the Macintosh platform, much as the security problems plaguing Windows XP still count as a strike against Microsoft's reputation for security, even though Windows Vista offers much better security.
It's hard to know exactly how the Macintosh user base breaks down right now. The Omni Group tracks the version of Mac OS X reported by their Omni Software Update technology, and their stats show that only in February 2009 did Leopard's installed base overtake Tiger's. (These stats are specific to The Omni Group's customers, of course, but other numbers, such as the 87.5 to 12.5 ratio of Intel to PowerPC processors, seem reasonable. Plus, since The Omni Group's applications are likely to be used by early adopters and power users, the stats would seem especially relevant to this discussion.) If Leopard, with all its user-focused features, managed to capture only half the installed base in 18 months, a full-price Snow Leopard would have even more trouble.
Selling Ice to Eskimos -- On a more practical matter, I think marketing a Mac OS X release that doesn't offer significant user-focused features would be tricky at best. It's not that Apple couldn't describe the advantages of Snow Leopard - larger RAM ceilings, better multiprocessing support, QuickTime X, faster installation time, and so on - but that those improvements largely address problems most users don't have. When was the last time your average Mac user thought, "If only Mac OS X installed faster!" or "I sure wish I could put a terabyte of RAM in this Mac"?
The entire point of Snow Leopard is to focus on improvements that will make future innovation possible, but it's hard, especially in this economy, to sell something based entirely on deferred benefits.
Worse, if done poorly, pushing the under-the-hood features of Snow Leopard could conceivably undermine any benefit-based marketing Apple might want to use to promote the next version of Mac OS X. In particular, Apple could risk being seen as nickel-and-diming users, which could in turn hurt Snow Leopard's adoption rate.
Finally, although a strong case could be made for making Snow Leopard free for Leopard users and charging Tiger users the full $129 price that they would have paid for Leopard, I'd argue that if someone running Tiger hasn't upgraded to Leopard yet, they're not going to, unless Apple makes the upgrade compellingly cheap. Plus, many Tiger users are probably running on PowerPC-based Macs, and the scuttlebutt is that Snow Leopard will run only on Intel-based Macs (so the only possible way to get Snow Leopard would be to purchase a new Mac anyway). While splitting the upgrade path would avoid sour grapes on the part of those who purchased the Leopard upgrade, anyone who purchased Leopard receives the benefit of using it until Snow Leopard ships, so it's not like it was wasted money.
Arguments Against Free -- There are a number of reasons why Apple might still choose to charge for Snow Leopard despite the arguments I've laid out above. And, to be clear, I wouldn't be upset if Apple charged just enough - somewhere between $10 and $30 - to cover the materials and distribution costs of a boxed product to the retail channel.
Some sort of a charge might be necessary if Snow Leopard proves too large to download. Also, for non-Apple retail stores, a high-enough price would be necessary for them to carry the box at all. A fee may even be necessary to meet accounting rules surrounding products like the Mac that are not accounted for on a subscription basis, like the iPhone and Apple TV.
We've also become accustomed to paying for major updates, and Apple may not want to break that habit, even if the price is somewhat lower than normal. Though of course, selling something for which people don't see the value could also break that habit and hurt Snow Leopard's adoption rate.
Lastly, although Apple has never released retail sales numbers for Mac OS X that I'm aware of, the company undoubtedly makes tens or even hundreds of millions of dollars on upgrade fees. Despite posting record profits in recent quarters, Apple may be unwilling to leave that money on the table, even if there's a chance such a strategy might not be in the long-term interests of the platform. It's also possible that Apple's internal accounting requires upgrade revenue to pay off Snow Leopard's development costs.
Free the Snow Leopard -- In the end, I believe that making Snow Leopard available for as little as is feasible - perhaps a free automatic-update download and a low-cost mailed media or retail box option - would help create a single coherent Macintosh platform that Apple and independent developers could build upon without worrying about supporting the past. Some short term profit would be missed, of course, but it would offer numerous long-term advantages and put the Mac on a firmer competitive footing with the upcoming Windows 7, especially given Microsoft's recent cost-based advertising and recent announcement of an optional virtualized Windows XP for Windows 7 users (see "Windows 7 Adds Optional Virtualized XP," 2009-04-24).
Besides the simple benefit of a Mac that works better (in theory, of course), users would also gain from software that would take advantage of Snow Leopard's features and would be easier and faster to develop without support for legacy versions of Mac OS X. And anything that makes users and developers happy benefits Apple in the end, through the sales of ever more Macs.