Death and Resurrection

Copland is dead. Apple has now decided to kill its plans for the much delayed, massive operating-system overhaul that was last expected to ship next summer. Instead, the company will roll out smaller updates to the Mac OS at six-month intervals.

This is a Good Thing.

I'll wait a moment for you to lift your jaws off the floor (pause). OK. Let me give you my reasons for why Copland's death is a good thing and how the killing of Copland may, in fact, save the Macintosh OS.

Reason #1: Copland has become an albatross around Apple's neck, a symbol of all that's wrong with the Mac and an emblem of Apple's inability to effectively compete with Microsoft's now entrenched Windows 95 and newly surging Windows NT. There's a stigma attached to Copland left over from years of mismanagement that the new regime now leaves behind.

Reason #2: Personal computing has changed substantially since the blueprints were made for Copland (I won't call it Mac OS 8, because there will still be a Mac OS 8, just not the one we think of as Copland). For instance, look at the phenomenal growth of the Internet. Customers now expect more-savvy networking and Internet access from their computers. But Copland was started well before the Internet boom and would have to have been significantly overhauled to really integrate the Net. And, no, Cyberdog doesn't cut it; the Internet must be an integral part of how the OS works, not just expressed in some clever bundleware.

In other words, while Apple has been working with mad-scientist-like focus on a big lumbering beast composed of salvaged body parts, the world outside Castle Coplandstein has passed it by.

Reason #3: The sum of Copland's parts is greater than the whole, but waiting for all the parts to be ready for one OS release has been like waiting for another famous no-show, Godot.

Customers want and need preemptive multitasking and protected memory. But because these considerably more complex OS technologies have been components of Copland, other, less complex but equally important parts have had to wait. For example, two and a half years after the first PowerPC-based Macs shipped, the Mac OS is still largely 680x0 code, slowing everything down. Critical bottlenecks, such as the File Manager, are still not PowerPC-native, because those components have been waiting for the rest of Copland to happen. The fastest personal computers on earth have to stop and digitally tap their feet, waiting for a poky OS to get back to them with data.

A decentralized approach to OS development will help put key technologies into users' Macs faster. The best technologies destined for Copland will find their way into the operating system when they're ready and won't have to wait for everything else to be done before they ship.

Reason #4: Accountability will at last be placed where it belongs -- directly on the engineers' heads. According to Apple Chief Technology Officer Ellen Hancock, releases will be scheduled for twice a year; the first will be in January, followed by another next July. Hancock is deploying engineers into teams to work on mission-critical technologies, making it clear that it is their responsibility to get the work done in time. If their particular component isn't finished by that release's deadline, it will have to wait for the next one, which won't look good on that team's next performance review. The concept of schedule slippage no longer exists in Apple's OS group. This hardheaded competitive approach to OS development may be just the kick in the pants Apple needs.

Are there disadvantages to killing Copland? Sure. Public perception is that Apple's handling of its highly touted next-generation operating system -- a.k.a. The Windows Killer -- has been inept at best, and canceling the project will appear to be the last nail in the Mac's coffin and will certainly be reported as such by the media. Also, scratching the project may lead to delays, as engineers focus on delivering the easier pieces, putting off the tough jobs for later. To pull off a smooth transition to the new strategy, Apple's management team has to act swiftly and communicate what Mac users should expect in the next OS release.

Careful management of these system updates is also critical. Apple must not succumb to incessant updatitis. Look at what's happened to 7.5.3 -- there are now four versions of the same OS. Customers won't stand for 7.8.2.1a, nor are they willing to be nickel-and-dimed for each OS update.

Most of all, Apple has to evangelize this approach to OS development to third-party developers. I didn't hear cries of woe and lamentation from developers when Apple shifted its OS gears -- in fact, rejoicing was heard in some quarters -- but developers are already working on products for one Mac and two Windows operating systems. Adding more flavors to the Mac OS may leave a bad taste in everyone's mouth.

Entire contents © copyright 1996 Ziff-Davis Publishing Company, L.P. All rights reserved; reproduction in whole or in part without permission is prohibited. MacUser and the MacUser logo are trademarks of Ziff-Davis Publishing Company.

Please send questions and comments about MacUser's articles, reviews, and presence on the Web via our feedback form. If you have a subscription problem, see our online subscription problem form.