Apple knows transitions

Macworld UK | Happy 10th Birthday, Power Mac!

Sunday was the 10th anniversary of the introduction of Apple's Power Macintosh line. Based on IBM's POWER line of processors, the original PowerPCs gave Apple a big performance leg up on Intel, one that Apple only regained (however fleetingly) with the introduction of the G5.

One of the great things about working at Coke (and later, at CNN.com) was that companies would provide software and hardware before it was released to the public.

In late 1993 or so, we got a couple of Top Secret boxes from Apple. They were prototypes of a new generation of machines we read about in the Mac press, and which would eventually be the first Power Macs.

Apple sent us the top two models, and the cases weren't even finished yet. The small one, the Carl Sagan, was in a sheet-metal box about the size of my Quadra 650, and the big one, Cold Fusion, was in a Quadra 800/880 case. We knew that the new boxes needed optimized code to perform at their very best, but we had also heard about "fat binaries," which allowed them to run older code with good performance.

Bigger fish than I got first dibs on both of them, but eventually, the time came for us to test KO/Office, Coke's internal office automation software, and since I was the support specialist for the Mac version, I finally got some seat time with the newest God Box.

We were worried about KO/Office, because it had one control panel in particular that worked on some fairly low-level routines. It was a security piece that managed logins and a secure screensaver, and we were trying to figure out what would be involved in rewriting that piece to work with the new hardware.

So I sat down with Cold Fusion, the box that eventually was introduced as the 8100/80, fired it up, and ... was immediately at home. The first thing I noticed was how little I noticed. Everything looked and ran exactly like it did on the Quadras. I had kind of hoped the PowerPC Macs (which is all we knew them as) would have some new bells and whistles, but Apple was focused on easing the transition as much as possible.

I installed KO/Office, rebooted so the control panel would work, and gave an exploratory sniff for fried hardware. No? Can I login to a network server and authenticate? Yep. Fire up the (in-house, custom) e-mail program? No errors. In short, everything worked. We were agog.

I was reminded of how seamlessly Apple made that transition recently, when I helped a friend upgrade to OS X. The friend in question has worked through 4 Macs: an SE, a IIsi, a Power Mac 6100, and a Blue and White G3/350, the target box for OS X.

My friend (I'll call her "N"), has some old, old code on her machine. She uses an address book desk accessory that was last updated in 1995, and barely appears on the web. She's happy with Freehand 7, from 1996, and doesn't want to buy an upgrade. As a result, she became the first OS X install I've done where Classic is likely to be up all the time.

Once again, I expected some fireworks. Maybe it would come from the third-party SCSI card, or the print-shop level Epson inkjet, or the comparatively low-powered CPU, or some of the old, old code. No way is this one going to be a smooth upgrade.

Once again, Apple proved me wrong. The old code happily runs under Classic, which itself runs quickly. The unusual inkjet is one of the stock printer drivers, and the SCSI card just works.

I don't know if anybody else in technology has so adeptly handled two such radical technology shifts, but it's enough to make you wonder ... If Apple decided it needed to, would it be any harder than these transitions to bring OS X to x86?