Subject: Re: Thoughts on Franz Inc., ACL pricing, etc.
From: Erik Naggum <erik@naggum.no>
Date: 1997/03/28
Newsgroups: comp.lang.lisp
Message-ID: <3068550645247405@naggum.no>


* Cyber Surfer
| You have to convince people who've already bought C++, Java, VB, and all
| the memes that go with these languages.
| 
| If you get paid to use Lisp, then I wish you luck.  Most of us have to
| struggle with lesser tools, but which can strangely deliver the apps that
| we're asked to write, while Lisp cannot.

well, it seems that this is your particular meme and that it isn't going to
die no matter how much evidence you get to the contrary.  frankly, I find
this all very tiring.  it also seems as if you argue mainly to support this
meme and that you would find any new argument to support it should your
previous arguments fail to do that.

| In the case of ACL/PC, the price alone is enough to ensure that it can't
| be used by most programmers.

another meme that won't die with you seems to be that the price of Allegro
for Windows is that of the Professional version, only.  how many times must
you be told that the price of the standard version is less than USD 600?
frankly, I'm getting the impression that you're grasping for straws to
stave off something that looks suspiciously like fear of actually getting a
Lisp environment and having to _prove_ your alleged love of Lisp.  this has
all been going on for too long, now.  it's time to realize that you can't
program in Lisp or that you actually can and take full responsibility of
the consequences of either choice.  you shouldn't just keep going about how
you can't make a decision.  if you do, it certainly isn't Lisp's fault.

| The long list of features supported by C++ but not by ACL/PC should be
| another big problem.

hm.  I can still program in assembly and the amount of extra work to get
the object-oriented features of C++ is not prohibitive.  the way C++ is
designed, it is in fact quite easy to get it.  the main feature of C++ is
that the compiler catches a number of minor problems in the code.  the way
it does a lot of mumbo-jumbo (like calling constructors and destructors as
if the compiler had a free will) is in fact a serious _barrier_ to quality
code.  as for as the syntax goes, the language is mainly automating naming.
the language just isn't smart enough to really warrant the cost of the
complex syntax and its complex semantics; the underlying features are
actually _very_ simple.  in contrast, I would have to work a _lot_ to get
even a semblance of Lispness, and when I need that, I need it _bad_.

C++ really doesn't give much over assembly languages (especially not if you
have a good macro system).  Lisp provides easy access to a lot of highly
complex features that are fundamentally _hard_ to do in assembly.  my
experience is that it's a _lot_ easier to get C++ functionality in Lisp
than to get Lisp functionality in C++.  it seems that some would like
everything to be equally easily accessible in Lisp, despite the obvious
differences of approach, and that doing things natural to a Lisp programmer
would be as prohibitive to a C++ programmer as doing it in assembly.

exempli gratia: implement `apply' in C++.

| You may be very productive with a tool, but if that tool can't use the
| technology (like ActiveX), then some of us won't be able to use it.

this is getting silly.  if you can do it in C++, you can do it in Lisp.
there are many levels of effort needed to get things done.  if it's easier
to talk to ActiveX in C, do it in C, then build a sufficiently advanced
interface to it that you don't need to mess with the minutiae.  Franz'
Foreign Function Interface in Allegro for Windows is very good.  if you
need to write a hundred lines of glue code to get it to work, what's that
compared to the pain of writing all of it in C++?

| Franz may get away with asking $3000 for a product that offers a
| programmer less support for Windows than VC++, but I don't call that a
| justification.  It's just Franz telling us that they don't consider most
| programmers to be potential customers.

Franz is giving away a Lisp environment for free, then charge $600 for the
standard version, and you can buy the individual pieces that go into the
Professional version piecemeal, or buy the whole thing for USD 2500 (new
prices after 1997-04-01).  if you can't absorb a cost of USD 2500 if you
are a professional programmer, the programming language isn't your biggest
problem.

| You don't have to love C++ to use it.  You may even hate it, but you may
| still be forced to use it, or look for another job.  That sounds to me
| like a real "for the love of Lisp" situation.

I wrote a lot of C++ code with Emacs.  the repetitiveness of the C++ code,
with stupid declarations all over the place that the system should have
figured out on its own, the irritating requirement that all methods must be
declared in the class itself, etc, can all be solved mechanically, and
writing a small language that takes care of all this is really simple.  I
later graduated to Common Lisp when I discovered that I was fast removing
myself from C++ semantics, not just syntax.  it was then that I discovered
how assembly-language-like C++ is and how assembly-language-unlike Lisp is,
and consequently why I thought in Lisp in Lisp and in assembly in C++.

#\Erik
-- 
I'm no longer young enough to know everything.