A MacOS wrapped around a Windows NT core--and capable of launching Windows applications directly, without need of emulation? Apple CEO Gil Amelio has hinted that may be the medium-term answer to Apple's need for compatibility with the Windows world.
At a time when everything Amelio says--and more than a few things he doesn't say--becomes fodder for a Mac market eager to try to divine Apple's OS strategy, Amelio was guarded in a teleconference with New Zealand journalists last week.
But he did appear to lend weight to theories suggesting that Apple may piggyback the PowerPC chip's support for NT. It was suggested in July, shortly after Amelio hired IBM's Ellen Hancock as chief technology officer, that Apple would use a kernel being jointly developed by Microsoft and Motorola.
Cross-compatibility is a major focus under Amelio, and he says in the next year it will be achieved with enhancements and price cuts to the Dos compatibility cards--which have already been credited with keeping Macs in some major sites.
In the long term, he contends that all modern OSes will converge towards an optimal architecture which largely erases compatibility and portability problems, and that Java "will help us move forward in a common environment. I think the key to that is going to be the Internet and the acceptance of Java as the standard of the Internet. What that will do is create this tremendous common playing field on which people can develop applications and I think we're just at the beginning of that era."
But on the way there, he says: "Our objective is to run Windows applications--but not necessarily to run Windows. At the moment, we don't have much choice--we have to be able to run the Windows OS in emulation to be able to run the Windows applications. In the longer term I don't think we'll have to do that. We'll be able to run Windows applications directly, as we move into this new environment.
"What's happening in operating system software is this. Both ourselves and the Windows world have an architecture which was excellent at the time it was introduced but which is getting somewhat long in the tooth. And today I think if we were designing our operating system on a clean sheet of paper, we would design it somewhat differently.
"In fact, I think computer scientists are at a point now where there's a fair degree of common understanding about what represents a truly good and modern OS. The difficulty is neither ourselves nor the Windows world has such an architecture.
"It's my view that for the next five years, both of us are going to have to migrate in that direction. So at that point, when we have a similar architecture--even though the details may be different--the ability to port or even run other applications will be significantly enhanced."
Amelio tiptoes around the other hot topic in the Mac community--exactly how Apple might incorporate Be or the Be OS in its strategy.
"We don't have any formal relationship with Be other than as a developer. Be is a licensed Apple developer and to that extent, it has the same rights and privileges as any other developer. We see Be regularly, we support it where we can. But we have no agreement with it beyond that at this point.
"I think the reason people get excited about that is that the BeOS, which was an OS started with a clean sheet of paper, has many of the attributes that a truly modern operating system must have, so it gets people excited. The difficulty is that we don't have the luxury of starting with a clean sheet of paper, because we can't abandon our installed base.
"Going forward, I would expect that we will develop our own architecture, but if we can get technology that exists out there through licensing or some other arrangement, we'll do that--we don't have to invent everything. But so far we don't have anything other to announce than what I've just said."
And on the option currently attracting the shortest odds among the pundits--Apple taking the BeOS application model and merging it with the MacOS microkernel?
"The Be model is nothing more than what people have been saying for a few years now as to what the applications model ought to look like--so again, I don't want to get too drawn into something that could be misunderstood. Yeah, I think that BeOS, being a cleansheet architecture, has some good attributes regarding its applications model and people find it attractive. I think it's highly likely that we will move in that direction, but so will other people. NT, for example, already has a similar applications model--and I would say that moving into the future you'll see more and more OSes adopting that kind of an architecture."
The most difficult feat for Apple, and its new senior management--which has been hand-picked by Amelio and bears almost no resemblance to the Spindler-era team--may be to bring the Macintosh in from the cold without alienating the Mac user base. After all, it was not only the performance of Microsoft Word 6.0 which drove Mac users mad, but its maddening deviations from the Mac look and feel.
On the other hand, Microsoft has reaffirmed its ability to develop well for the Mac market with Internet Explorer 3.0, which emphatically is not a quick-and-dirty port. But would Microsoft--or anybody else--bother to develop native Mac applications if Windows apps could be simply slotted into a Mac environment? Much does depend on Apple's developer relations--including those with its largest developer, Microsoft. So has the greatly enhanced cordiality between the two been down to Amelio's efforts?
"Yes," chirps the Apple CEO. "Of course it has!"