MCL/PowerPC reality check

[jmeehan@mendoza.adobe.com (Jim Meehan)]

Let me play the devil's advocate for a minute. (Since I've had this
same discussion 1-on-1 with several members of the newsgroup, I
thought I might as well make my argument a little more public.)

Everyone believes that MCL is A Good Thing. Oh sure, it could stand
some improvement in this or that area, but the testimonials flowing
through this newsgroup confirm that, all things considered, it's right
up there with Sliced Bread.

Everyone also believes that MCL sells dozens and dozens of Macs, all
over the world. Maybe hundreds. But thousands? I don't have any data
from Apple -- perhaps someone from Apple will supply some -- but I
don't think it's thousands, and the problem is that unless MCL sells
thousands and thousands of Macs, it's a money-loser for Apple,
considering what it has cost Apple, over the years, to support the
product and the team at Apple ATG East that built it and maintain
it. MCL might be a money-maker for a small, lean, and hungry
one-product software house (remember Coral?), but apparently not for
Apple.

Did MCL have poor marketing? Sure. Poor enough to matter? Naaah. There
just aren't that many places that are enlightened enough to use MCL
for development or that build mass-market products where thousands of
end-users are affected, albeit indirectly, by the quality of the MCL
environment. Folks, we are a drop in the bucket of programmers. You
know what they're all programming in. It ain't Lisp.  It never was.

On the other hand, I think the reports of MCL's demise have been
greatly exaggerated.  Here's what Fleischman actually said:

  As long as there is a market for MCL 2.0.1, we will continue to
  maintain it, support it, and make it available through APDA and our
  other resellers.  Our maintenance plans for MCL 2.0.1 include bug
  fixes and ensuring that it runs on future Macintosh computers,
  including the Power Macintosh in emulation mode.  However, the current
  sales volume of MCL does not justify the resources that would be
  required to continue new feature development on MCL.

I don't believe that the appearance of the PowerPC is going to cause
all previous Macs to disappear in 1994. Or 1995. Or 1996. Do you know
when Apple finally dropped the Apple 2E?  Last November.

Of course we all want to move to the latest, greatest, fastest
machine, AND we want to believe that the machine we have today will be
here forever, a double standard of sorts. But what's the expected
lifetime for any machine used by professional developers? 5 years? 3?
It's part of the reality of our world that we change fairly often.
It's nice when the upgrades are similar, but hardware is volatile.
It's a remarkable time; the rate of advancement in hardware speed and
capacity, matched by a corresponding drop in the price per cycle or
MB, is astounding, and the hardware wizards at DEC, at least, predict
no dramatic change in this curve in the next decade or two. "50%
better per year" is their rule of thumb. That's why they're so happy
about the Alpha; it was specifically designed to be able to follow the
curve for the next 20 years.

In this respect, software isn't much different. Nobody forced us to
use Lisp. Nobody signed a contract saying that they would support MCL
"forever, no matter what, at no more than $495." When we chose MCL, we
took a gamble. We may have considered it a safe bet; it may still BE a
reasonably safe bet. (It's a safer bet than Dylan, today at least.)
But Apple is a for-profit business. Apple is not DARPA. Apple is not a
university. Apple is not a shareware collective. There is no moral
imperative for Apple to support a product on which they believe
they're losing money, and 1993 was a very difficult year at Appa product on which they believe
they're losing money, and 1993 was a very difficult year at Apple.

We all wish there were an easy way for a company to make money
supporting a rapid-prototyping environment for the kinds of things
that WE like to build, or that there was some company or agency that
decided they could afford NOT to make money on it, at least not right
away, but vould support it anyway, as a long-term investment in the
future, in Good Things, as places like DARPA used to do on a greater
scale than they can do now.

So, keep those cards and letters coming. I'm sure the folks in
Cupertino are reading them, and I'm sure they like hearing messages
that range from "I like MCL" and "I bought my 2 Macs because of MCL"
through "We use MCL in all of our classes" and "We're solving
important problems with relative ease and/or grace because of MCL."

But let's tone down the high dudgeon: "How could they DO this to us?"
"Apple abandoned its best product!" and "I'll never trust them, or
their machines, ever again!" The sky has not fallen, Jesse Helms has
not introduced federal legislation banning Lisp from the information
highway, System 8 is not being designed to prevent MCL programs from
running, and MCL may yet be "rescued," by Apple or by someone else,
perhaps by you.


Jim Meehan
jmeehan@mv.us.adobe.com