[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Hoping for an upgrade to MCL

Open letter to Apple re MCL upgrade

I'm not usually a big letter-writer 
or crusader, but since this is an 
issue that has a big impact on my 
life, I thought it would be good to 
give you my two cents' worth 
regarding the rumors flying around 
that Apple is not planning any 
further major upgrades to Macintosh 
Common Lisp.  

Even without the advent of the 
PowerPC, I had been looking forward 
to such possible improvements as a 
more complete Metaobject Protocol or 
an expanded library of CLOS classes 
for interface design and application 
construction.  (I wouln't have spent 
my own hard-earned money on a dead-
end product -- but I felt safe, 
because I was buying from Apple, not 
someone like Ashton-Tate.)  And now 
that faster machines such as the 
PowerPC are becoming a reality, we 
can finally afford the luxury of 
programming in highly expressive 
albeit slower-executing languages 
such as Lisp.   

I understand when buggy or ill-
defined languages fizzle out and get 
dropped by their vendors, but 
Apple's brand of Lisp offers 
excellent features and performs "to 
spec" -- no small achievement in a 
fifty-year-old industry littered 
with failed projects and broken 
promises.  Within its market, 
Macintosh Common Lisp has a lot of 
staunch supporters, including many 
influential people pretty high-up in 
academia, the military and high-tech 
industries such as computer-aided 
manufacturing, expert systems and 
process control.  Here you have, 
with almost no active marketing or 
"evangelism" on the part of Apple, a 
small sophisticated market of users 
who would "rather fight than switch" 
-- I know a lot of marketing people 
who would kill to have a product 
with this kind of fierce "brand 
loyalty," "early adopters,"  and 
"opinion leaders."  I'm sure you 
don't need to be told how valuable 
an intangible asset this kind of 
product differentiation can be to a 
high-tech company such as Apple -- 
the challenge is, can you get this 
message through to the decision-
makers in marketing or senior 
management?  I don't think anyone is 
asking Apple to become a not-for-
profit charity -- we're just 
reminding you that you're not 
selling pork bellies, you're selling 
highly differentiated "knowledge" 

Forgive me if I'm wrong, but the 
message I'm getting from Apple is 
this:  You seem to be tired of all 
the complexity of developing 
cutting-edge technology, and you've 
decided that the way to increase 
sales volume is to drop prices and 
drop quality and just become another 
"me-too" maker of generic PCs.  A 
very dangerous image to create in 
such a sophisticated industry -- 
unless you've calculated that you 
can afford to dump all your current 
customers and replace them with more 
new ones.  Go ahead and woo 
corporate America with pizza boxes 
for word processing, spreadsheeting 
and desktop publishing, but at the 
same time, throw a little marketing 
money and a few programmers at your 
established Lisp product to keep it 
alive as well.  I should think 
getting MCL up to version 3.0 will 
take a hell of a lot less effort 
than getting Dylan up to version 1.0 
-- and won't there be some kind of 
synergy to having two strong Lisp 

Perhaps the most exciting arena of 
computer programming these days is 
not just "object-oriented dynamic 
languages" but simply the gradual 
convergence between the declarative, 
logical, functional and object-
oriented programming paradigms, as 
exemplified by the emergence of new, 
*executable* specification languages 
supporting code re-use and 
"programming-in-the-large."  This 
progress is being fueled by advances 
in architecture and processing 
speeds, as well as advances in 
language and compiler design.  I am 
using my Quadra 800 at home to try 
to port the executable specification 
language OBJ (developed by Jose 
Meseguer, Joseph Goguen et al. at 
SRI International in Menlo Park) to 
the Mac, and I need to use MCL as 
the "underlying language" since the 
original source implementing OBJ was 
written in Austin-Kyoto Common Lisp.  
(By the way, if you know of any 
resources documenting the 
differences between MCL and AKCL 
that would help me to do port, I 
would really appreciate any 

I bought MCL and the 800 because I 
wanted a true object-oriented 
language on a reliable yet popular 
platform, and I trusted that both 
the hardware and the software would 
be upgradable to keep up with the 

I hope that Apple either commits to 
upgrading MCL and making it go 
native on the PowerPC, or at least 
releases the source to another 
vendor or to the public so that 
those of us who love MCL so much can 
continue to get the most out of it.  

Scott Alexander
The San Juan Star (newspaper)
San Juan, Puerto Rico