Marketing Ada: Is the Sky Falling?

Marketing Ada: Is the Sky Falling?

[Richard Pattis]

The new issue of ACM LISP Pointers just arrived, with a nice article by
Richard E. Waters titled, "The Survival of Lisp: either we share, or it
dies". It explains the "C/C++ versus the World" battle from the view of
[one person from] the LISP community. There are interesting parallels to
what is going on in the Ada community. 

Is the dominance of C/C++ making all other languages marginal, leading
them to an accelerating spiral of decline. Will these language survive? Will
they survive in niche markets but not prosper (too few companies devoting too
little effort to improving products)?  Is this the equivalent of "survival of
the fitest" or "language genocide"?  We already know the PC implications [pun
intended] of such a course of events.

Maybe it is time for a serious meeting to discuss how to contain C/C++, not
because it is a bad language [insert your own version of a smiley here], but
because we don't want it to grow so large that it becomes the only
well-supported language.

Rich


-- 
------------------------------------------------------------------------------
  Richard E. Pattis                     "Programming languages are like
  Looking for a Job                      pizzas - they come in only "too"
					 size: too big and too small.

Re: Marketing Ada: Is the Sky Falling?

[Scott McLoughlin]

pattis@cs.washington.edu (Richard Pattis) writes:

> Is the dominance of C/C++ making all other languages marginal, leading
> them to an accelerating spiral of decline. Will these language survive? Will
> they survive in niche markets but not prosper (too few companies devoting too
> little effort to improving products)?  Is this the equivalent of "survival of
> the fitest" or "language genocide"?  We already know the PC implications [pun
> intended] of such a course of events.

Howdy,
        I think that they will survive (not all of them, but some
of them).  There are alot of hurdles, though. C/C++ has really
high _visibility_. It's on store shelves and advertised in popular
magazines.  It compiles well and makes programs that run well on
"low end" hardware. Implementations are often dirt cheap or free
(gcc).  You can drop down to assembly to do weird stuff easily in
many implementations.
        I think you have to aim right for the sweet spot of the
market: a brain dead CPU running a brain dead OS.  Other 
"alternative language" vendors seem to be doing ok: Digitalk's
Smalltalk/V, Borland's Turbo Pascal, MS's Visual Basic. I think
it helps to offer not just another language, but some other 
advantage as well. So Digitalk has various MIS related add ons
(PARTS, I think it's called). Borland is making a really cool
"Visual Pascal" that acts like VB in many ways but makes fast
compiled EXE's. Etc.
        I don't know how fun it would be to "succeed". Think 
of all the OS/2 designers/implementors at IBM sitting around
for god knows how long just _sweating the fat_ out of "Warp"
so it will run ok on a 4 meg machine. Blech. Probably lot's
more fun trying to get your favorite weirdo language 
compiler to generate code that will beat fortran on a 200
meg Silicon Graphics box, right?
        Anyway, most of the world doesn't _want_ a language
like C/C++. Too hard.  That's just what's "there" in the
market place.  So get yourself a few million dollars and
a sweat shop and a bunch of dumb windows boxes and go to
town ;-)

=============================================
Scott McLoughlin
Conscious Computing
=============================================

Re: Marketing Ada: Is the Sky Falling?

[Robert Dewar]

In response to Richard Pattis' comment on C++, I certainly feel that
it makes sense for Ada implementations to have very good interfaces
to C++ so that C++ bindings can be used with little or no effort.

Re: Marketing Ada: Is the Sky Falling?

[Marco Antoniotti]

In article <3cmr8l$q3q@gnat.cs.nyu.edu> dewar@cs.nyu.edu (Robert Dewar) writes:

   Path: thecourier.cims.nyu.edu!thecourier.cims.nyu.edu!nobody
   From: dewar@cs.nyu.edu (Robert Dewar)
   Newsgroups: comp.lang.ada,comp.lang.lisp
   Date: 14 Dec 1994 08:17:09 -0500
   Organization: Courant Institute of Mathematical Sciences
   Lines: 4
   References: <D0rLos.2pB@beaver.cs.washington.edu>
   NNTP-Posting-Host: gnat.cs.nyu.edu
   Xref: thecourier.cims.nyu.edu comp.lang.ada:12362 comp.lang.lisp:5641

   In response to Richard Pattis' comment on C++, I certainly feel that
   it makes sense for Ada implementations to have very good interfaces
   to C++ so that C++ bindings can be used with little or no effort.

The main problem with C++ is that each different compiler imposes its
own demangling scheme on the environment. This is an issue not only
for Common Lisp (for which a "standard" FFI would be more important at
this time) or Ada, but for C++ compiler developers themselves.

I would tend to look at a "language independent" approach as ILU as a
better solution toward integration of C++ stuff in other languages
(and viceversa).

--
Marco Antoniotti - Resistente Umano
-------------------------------------------------------------------------------
Robotics Lab		| room: 1220 - tel. #: (212) 998 3370
Courant Institute NYU	| e-mail: marcoxa@cs.nyu.edu

...e` la semplicita` che e` difficile a farsi.
...it is simplicity that is difficult to make.
				Bertholdt Brecht

Re: Marketing Ada: Is the Sky Falling?

[Robert Dewar]

I don't see the different mangling schemes as a significant problem for
Ada binding to C++, yes it is a problem, but relatively minor.

Re: Marketing Ada: Is the Sky Falling?

[Marco Antoniotti]

In article <3cnksb$6vn@gnat.cs.nyu.edu> dewar@cs.nyu.edu (Robert Dewar) writes:

   Path: thecourier.cims.nyu.edu!thecourier.cims.nyu.edu!nobody
   From: dewar@cs.nyu.edu (Robert Dewar)
   Newsgroups: comp.lang.ada,comp.lang.lisp
   Date: 14 Dec 1994 15:34:19 -0500
   Organization: Courant Institute of Mathematical Sciences
   Lines: 3


   I don't see the different mangling schemes as a significant problem for
   Ada binding to C++, yes it is a problem, but relatively minor.

I might be paranoid, but I think it is a minor problem as long as we
remain in relatively friendly environments as G++ (considered that the
coordinator is here at NYU :) ).

Since the major vendors (i.e. Borland and Microsoft) tend to ship
their system complete with new libraries etc. etc. They could in
principle change the mangling scheme in each release.

This may turn out to be a headache.

On top of that, some major vendors may even be tempted not to reveal
their mangling scheme (in order to prevent linkage by other vendors)
or to even patent their specific name mangling scheme (this is even
more paranoid, but conceivable).

--
Marco Antoniotti - Resistente Umano
-------------------------------------------------------------------------------
Robotics Lab		| room: 1220 - tel. #: (212) 998 3370
Courant Institute NYU	| e-mail: marcoxa@cs.nyu.edu

...e` la semplicita` che e` difficile a farsi.
...it is simplicity that is difficult to make.
				Bertholdt Brecht