On October 17, 1998 Apple released Mac OS 8.5, the first operating system that ran solely on Macintoshes with PowerPC processors. As far as system software upgrades go, this was the end of the line for any Mac built before the Power Macintosh 6100, introduced in March 1994. Earlier Macs ran on some variation of 680x0 processors and were supported mostly via emulation in a PowerPC environment. Emulation works, but it also slows things down. By 1998, Apple decided it just couldn't support 680X0 emulation for a number of reasons, but chiefly among them was speed.
What happened was just what you would expect. In user groups, USENET and the Internet (which was only starting to explode), apoplectic non-PowerPC Mac owners threatened class action lawsuits and the rending of garments. Of course, most Power Mac users loved the newfound speed introduced in Mac OS 8.5, thanks to code optimized for PowerPC processors and jettisoned emulation support.
It took Apple only four years to introduce the PowerPC chip and make any Mac without it obsolete. Technology moved on.
In September Apple will release Snow Leopard, which will only run on Intel based Macs, thus cutting off PowerPC support. This time it took eleven years from inception to extinction (well, three for the Intel transition), but even so I can hear the hue and cry machine cranking up. Once again, the major reason for dropping legacy support is speed. Technology has moved on.
Whenever something like this happens there is a potential for a marketing meltdown, but this time Apple is doing something brilliant. It is going to sell Snow Leopard for $29. When I saw this on the video stream of the WWDC keynote address my jaw dropped, my eyes glazed and only later did it start making sense to me. Apple first introduced Mac OS X in 2001, and excluding the free update to Mac OS X 10.1 from Mac OS X 10.0, a new version of the OS has been released roughly every 18 months, always at a price of $129. The sales pitch is always the same: with each new version, OS X gets new features and an "enhanced computing experience" which largely depended upon how much you like the new features.
Mac OS X 10.6 will be the fifth major release in eight years, and some users are complaining about feature overload. There will always be users who want four ways to do the same thing, but for others, feature-laden releases are overwhelming and the glimmer and excitement of a new OS X release has faded. What a perfect time to work under the hood, set up the core of the operating system for the future and stabilize what's already there! But of course you can't make everyone happy. I would expect a large group of users to not be mollified by a nicer QuickTime and an improvement to Stacks. In effect, where's the beef? The beef is under the hood this time.