Sorry for the long piece, but this seemed interesting. It's an interview
by Andrew Gore of MacUser of Avie Tevanian.

Neil Ticktin
MacTech Magazine
publisher@mactech.com

http://www.mactech.com

-------------------------------------------------------

MacUser Online February 24, 1997

A tough year awaits Avie Tevanian, former vice president of engineering at
NeXT Software. Recently anointed as Apple's senior vice president, system
software engineering, Tevanian has what is quite possibly the most
challenging job in the computer industry: creating the next Mac OS.
MacUser Editor Andrew Gore sat down with him to get his take on Rhapsody
(Apple's code name for its upcoming operating system) and on the rumors
that future PCs may come equipped with a "Mac Inside" sticker.

MacUser: Avie, I wanted to start off with a basic question: In the new
organization in Apple, what part of the OS organization are you
responsible for?

Tevanian: I'll be responsible for the Rhapsody product and all the
deliverables that fall out of that. So that's essentially the new modern
OS that Apple's building based on the NeXT technology that they've just
acquired, and based on integrating in some of the key pieces of their
existing technologies. And I'll also be responsible for the ongoing
projects we've had at NeXT, including things like our WebObjects
technology.

Could you take a few more minutes and just explain what Rhapsody is? We've
heard a lot of stories about Rhapsody, and we've written a couple
ourselves. From your standpoint, what is Rhapsody?

Rhapsody is, at a high level, basically the future operating system for
Apple. It's the next generation in the evolution of operating systems at
Apple, and it's the piece of technology that's going to take Apple into
the 21st century with computing.

More specifically, Rhapsody is a modern operating system. It has all the
features that you expect of a modern operating system, like symmetric
multi-processing, multi-tasking, memory protection, all those types of
things, and it does all those things in not only a modern way, but a very
good way compared to the way other systems on the market do today. So it's
definitely best-of-breed in all those categories.

But it's much more than that. It's a set of APIs that exist above the core
operating system that make it very easy for developers to write new
applications. It's a compatibility system -- which we call the Blue Box --
which allows customers to run old Macintosh applications compatibly on
PowerPC platforms. And it's an advanced look and feel that people expect
when they use a Macintosh.

There's been some talk about combining the Mac OS look and feel with
NeXT's look and feel. In your mind, what currently exists in the NeXT look
and feel that you'd like to see carried over into Rhapsody? When somebody
boots up this machine, what should it feel like? Should it feel like a
Macintosh or should it feel like a NeXT machine?

The issue here stems from the fact that both the Macintosh and Nextstep
have very good user experiences.

But they're a little different, too.

They are a little bit different, but conceptually they're very similar and
they're both very good. Our goal is that when we ship the final product,
people feel as though they're getting the Macintosh experience -- it's
really to use, things work the way they expect them to, and they have no
complaints about the user interface.

To get there, what we need to do is we need to acknowledge that both the
Macintosh of today and Nextstep of today have major strengths. I think
every one of your readers understands the strengths of the Macintosh user
interface. They probably don't understand NextStep's as well, but to give
you a flavor for the power, the Nextstep user interface has been more
designed to match a multitasking environment. [In that scheme] it works
work better than, say, a Macintosh.

So our goal here is to take the best paradigms from both and put them
together in a way that makes sense -- not just arbitrarily mix them. We'll
put them together in a way that makes sense, and have the best of both
worlds.

Let's take a specific example. The most obvious aspect of the NeXT
interface that's different than the Mac is the Dock. Do you see the Dock
being something that you'd want to have in the Rhapsody interface?

I think that's a good example. We haven't completely decided if the Dock
will or will not be there, but that's a great example where both Apple and
NeXT have looked at the various products and agreed: gee, the Dock is one
of the things that doesn't exist on the Macintosh today, but that we
should probably really consider making sure is part of the final product.
That's the kind of thing we're doing.

Could you take a minute and talk about the Rhapsody schedule? There's been
a public schedule, and it's been pretty specific. As the guy who's
ultimately going to be expected to deliver this, what is your confidence
is in that schedule? What should users' expectations should be for what
they'll have, what they'll be able to do with it, and when?

We have three major milestones in our planned rollout. The first one,
which will occur around the middle of the year, will be our Developer
Release. The goal of this release is to seed developers with the APIs, so
they can start creating applications that will run when the final product
ships.

So that's the goal of that. It's not in general release to lots of
different customers. It's for developers and for really, really early
adopters who want to get a look at the technology and just understand it.

So, for example, the new interface won't be there at that point?

Certainly not in its final form. There may be a few simple things that
we've already adjusted for, but there's clearly no goal to make that be
the user interface that's going to be final.

Next is the Premier Release. The Premier Release will get out by early
next year, which means in the January/February time frame. We would like
to get it out by the end of 1997 if we can, but we hesitate to commit to
that, because it's still a little bit far out. The goal of the Premier
Release is to be a fairly solid release, so that early adopters can start
to use it, evaluate it. And it will have the first public version of
support for Blue Box (the part of the upcoming operating system that will
run System 7.x applications).

So in terms of what, if you were going to make a recommendation to a
friend or family member, what would you tell they should be able to do
with the Premier Release or not do? In other words, what will it be good
for? Is it basically just for evaluation, or do you think people will be
able to do productive work with it?

I think people will certainly be able to do productive work unless they
depend on a Mac OS application that won't be supported by the Blue Box in
that timeframe. And it's hard to predict exactly which ones will or won't,
but that would be probably the big thing that would come to mind, that
they wouldn't be able to do at that point in time.

But other than that, everything should be clearly functional. There may be
some pieces of the UI that are still subject to revision, but we'll have
at least some of the changes, maybe most of the changes, done by that
timeframe.

Along those lines, we will complete the Rhapsody rollout in the middle of
next year. This will be the Unified Release, which is the complete Yellow
Box (the new Rhapsody API), the complete Blue Box -- everything fully
functional, fully supported. At that point, anyone ought to be able to use
Rhapsody without any restrictions.

What is Rhapsody going to offer developers to induce them to port their
applications to Yellow Box and not just rely on Blue Box?

A number of things. First of all, from a technical perspective, the Yellow
Box frameworks are very powerful, making it extremely easy to build very
nice applications very quickly. For example, if you're a small development
shop and you want to do something to change the world, you're going to be
able to bring your new application to market for far less cost and much
quicker with Rhapsody technologies. You can focus on what your idea is and
your added value instead of how to make menus work.

Now, combined with that needs to be the business case, if you will. So the
technology will be very compelling. What's the business case going to be?
Well, first of all, we will absolutely be leveraging Apple's volumes,
which is in the millions, and in a lot of markets, for example, the
publishing market, the Macintosh owns that. If you have an idea in the
publishing market, Macintosh is without question your best bet.

In addition, we'll be making the APIs and the tools available
cross-platform. So as you're developing for Rhapsody, you're not limited
just to Apple's customer base. You can now also go after NT-based systems
or even sell the software for PCs running Rhapsody.

So you're saying you'll be able to develop a Rhapsody application using
the tools that will be available, and it will be a simple matter to run
that same application on either NT or Macintosh?

That's absolutely correct.

You touched on publishing for a minute. It has, obviously it's a big
stronghold for the Macintosh. It's also a very entrenched marketplace. You
can still go out in publishing houses and find linotypes in use. What are
you going to do to make Rhapsody attractive to this very entrenched
marketplace, so that they will make the conversion sooner rather than
later?

One of the things that might be most interesting to the customers you
describe is the Blue Box support. And so what will happen is all the
applications they're using today on their Macs will continue to work. But
it's more than that, because since they're running in the context of the
Blue Box inside of Rhapsody, they gain all the benefits of Rhapsody
itself, so they get the advantages of the powerful core OS underneath,
which is going to actually, in many instances, make their applications run
faster. They'll gain more protection.

Certainly if you have a native Rhapsody application that uses the Yellow
Box, and you have a problem with that application, it's not going to crash
the Blue Box. If you have a Blue Box application that crashes, it's not
going to crash the Yellow application.

So it's a very nice transition path for these customers who are slow to
move, by simply focusing on running applications first in the Blue Box,
interacting with new applications in the Yellow Box, and doing the
transition between the two.

What's your expectation of the performance of the Blue Box?

Our indications at this point are that applications running inside the
Blue Box should run at least as fast as they do today, if not faster. And
that has been the experience with people who have used MAE (the Macintosh
Application Environment for UNIX) in the past, and we think that will
carry forward with the Blue Box.

The Blue Box is not MAE technology, right?

That's right. Blue Box is not MAE. Blue Box is an environment, but it's
not an emulation environment. It is an environment in which System 7 runs.

It's a virtual machine.

It's a virtual machine, yes.

So it's actually very similar to the way Windows 95 runs Windows 3.1 apps.

Yeah. I wouldn't say that they're exactly the same, but in spirit they're
very similar.