The Present

Merger

By early 1996, Apple had realised that they had a problem, and had hired Gil Amelio from National Semiconductor. He in turn had brought in Ellen Hancock, also from National Semiconductor, and previously from IBM. By July, Hancock as Chief Technology officer had decided to cancel the Copland project, and was shopping for a replacement. BeOS was considered, and was very much in the public eye; not because it was a serious contender, but because the head of Be spent all his time leaking information about the talks to the press.

Be
Be was an attractive operating system, but still in very early releases, without even the ability to print. It was far too unstable and immature; the native development tools were also late arriving. Being good looking did not necessarily mean that it was well designed, either.

Jean-Louis Gassee, the ex-Apple head of Be, was reputedly asking $400 million and a seat on the board of Apple for himself. Apple had gone far enough to offer $120 million, probably an indication that they saw Be as a partial solution, not the complete answer.

Others under consideration were Solaris from Sun, and the PowerPC port of Windows NT, developed by IBM. Before NeXT entered the picture, Hancock was apparently favouring Solaris. Although Hancock was respected at Apple, it was obvious that many people at Apple saw her as the personification of their long time competitor, IBM, and the clash between corporate cultures was clearly felt. Even Amelio appears more as an outsider than as a part of Apple.

The story goes that NeXT was introduced into the picture by an engineer at NeXT (John Landwehr, actually in Product Marketing) calling some colleagues at Apple, and hearing that Hancock would be receptive to taking a call from NeXT to discuss possibilities. NeXT had been quietly offered for sale many times (this isn't unusual; most private companies will be continually appraising their value on the market), and it was clear that their private ambition to reach an IPO was ludicrously unrealistic. Steve Jobs had spent less and less time at NeXT in the preceding year and more, giving several very gloomy and uncharacteristic interviews with Red Herring and Wired. Several people who had met him in the preceding year had said that he was very happy to talk about Pixar, but would change the subject away from NeXT and refuse to discuss it.

Landwehr, on hearing the news, passed the news up to Mitch Mandich (VP Sales at NeXT), who refused to make the call himself, instead asking Garrett Rice, also in Marketing, to make the prearranged call to Hancock. It can be assumed from this atypical detail that this possibility wasn't being taken very seriously. The call went well, and it an arrangement was made for a team of Apple people to come over to NeXT to take a look.

NeXT steps in
The story continues that when the deal makers were in the boardroom at NeXT, Steve Jobs called to speak to Mitch, and heard who he was meeting with. This was apparently the first he heard of the news. The discussions took about five days from the first meeting to completion. At all points the Apple people focused on their three questions: "does it do: multitasking, multiprocessing and protected memory?". Steve demoed NeXTSTEP using the basic Interface Builder demo that has been part of his on-stage presentations for eight years, and the NEXTTIME demo of showing several movies running on screen at once (usually a Star Wars compilation). As this was a close copy of a Be demo that had apparently already impressed Apple, it was a clincher.

The deal was to buy NeXT in its entirety for around $400 million (the total ended up at $430 million), and to offer Steve Jobs a position as a part time consultant for Apple. This reflects both the fact that Apple saw this as a complete solution, and that the importance, for both good and bad, of Steve Jobs role at Apple was fully understood.

Rhapsody
NeXT was bought to supply the new, next generation operating system for Apple; the role that Copland should have been able to fulfill. Although details were vague, the delivery schedule was precise. An initial version for developers would be released in the middle of 1997 (with dates guessed to be between June and August); an initial user release at the end of 1997 (between November 1997 and February 1998), and a final release in the middle of 1998. It would run on all "currently shipping" (January 1997) Macintoshes.

The initial developer release would be mainly NeXTSTEP running on Power PC Macs with a NeXTSTEP user interface, with Macintosh user interface details to be moved over as required.

Pages designed by:

Paul Lynch
paul@plsys.co.uk

Last updated April 24, 1997.
[P & L|Paul's Home]