eberard@ajpo.sei.cmu.edu (Edward Berard) (09/16/88)
A while ago, I posted a query to the net looking for interest in an
Ada compiler which ran under the original Macintosh OS (i.e., _not_
A/UX). I received a good number of responses on the net, as well as a
number of phone calls. All respondents seemed genuinely interested in
such a compiler.
Some quick observations:
- Interest was not confined to the U.S. Europeans also seem
interested. [There is at least one large European Ada project
which is making heavy use of Macintoshes.]
- Academia was also interested in such a compiler.
- An interesting observation is that those who would buy the
fewest would pay the least, while for those who were interested
in buying large quantities (dozens to thousands), price was not
a consideration. (Quantity discounts were expected.)
- There seems to be room for a number of different approaches.
Some were only interested in using such a compiler for teaching
purposes. Others were interested in using the Macintosh as a
serious software development platform. Still others were
interested in the hardware capabilities of the Macintosh. A few
were interested in "a cheap compiler to fool around with."
- One theme I heard over and over again was "we have been beating
on Ada compiler vendors for a Macintosh OS version for some time
with little response."
I did receive more information than I can print here. However, I, like
most of you, have my own job to do. Read and enjoy.
----------------------------------
Geoff George (frobozz@pyr.gatech.edu) writes:
I have been programming in Ada professionally for three years now; I have been
involved with the language for about six. My MS thesis concerned a
generalization of the Ada rendezvous mechanism.
I would certainly buy a copy of an Ada compiler for my Mac II, if a decent one
were available for under $200. My personal group here at GTRI would probably
buy 2-5 copies of such a beast; I cannot speak for the rest of GTRI or
Georgia Tech in general.
...
(PS - My group would probably not impose the <$200 restriction; that's just
all the market will bear for me personally. I'll stick with C if it's any
more than that.)
----------------------------------
From CSD., University of Erlangen, W - Germany (decwrl!fauern!msn.rmi.de!
m_urlichs@ucbvax.Berkeley.EDU) comes the following:
Biggest requirement probably would be the ability rto run the compiler
under MPW, to and to link with other languages. (i.e. use the MPW
linker and really document parameter passing, and/or use Pascal
standard).
The company I am now working for probably would by a few copies if ...
----------------------------------
Michael Peirce (claris!peirce@ames.arc.nasa.gov) of Claris Corporation,
Mountain View CA says:
I would be very interested in getting an Ada compiler for native Mac O/S!
I spent about two years working with Ada inside Digital Equipment Corp.
(doing Factory Automation work). Ada proved to be a great benefit to
us there.
Now that I'm working full time on the Macintosh, I truely miss many of
the abilities that Ada provided. I'm sure I could convince people here
at Claris to buy a good, native macintosh, Ada compiler.
I would hope that this product could work with MPW, since this is Apple's
common platform for compilers and provides a decent environment for
development (though I won't complain about a LightSpeed-Ada - type
environment either).
Pascal and C seem so primitive after working in Ada...
----------------------------------
Frank Kuiper (mcvax!cwi.nl!frankk@uunet.UU.NET), from CWI, Amsterdam, observed:
> 1) Who is interested in Ada compilers for the Mac OS?
Some of my colleques probably are.
>
> 2) Of those who are interested, how many copies do you think you,
> your company, or organization might purchase?
2 or 3 (5 max)
>
> 3) What are your major requirements and considerations regarding
> such compilers?
Should be competely compatible with other ADA compilers. In our case people
would use it in their Mac at home, and we have sufficient other machines at
our office available with ADA compilers. For my collegues to use MacADA. it
sure has to be compatible.
----------------------------------
A. Joseph Rockmore (rockmore@ads.com) said:
1. Who is interested in Ada compilers which run on Macintoshes
under the original Macintosh OS?
advanced decision systems is developing and delivering knowledge-based
systems in ada for dod. we are currently using alsys on suns (pretty
bad), and a beta copy of the karlsruhe compiler on a vax. a good
environment on either suns or, better yet, macs, is of interest to us.
2. What criteria are the most important in the possible selection
of such compilers?
the usual: speed, bug-free-ness, completeness, etc. also, a good
development environment, not only tools but also a good user interface
(my programmers are used to lisp environments, and would like
equivalents).
3. If such compilers were available, how many would you, or your
organization, be likely to purchase?
immediately, several. long term, many more (depends on how our
business grows).
----------------------------------
Richard Pattis (pattis@june.cs.washington.edu) provided an academic slant:
Phil Miller, at CMU, is interested in building a programming
environment on Macs (similar to the Karel and Pascal environment).
They are looking for a Mac compiler on which to build their
environment.
Macs are available at lots of schools now, and I think educational users would
be happy to see Ada software on them. We are switching our introductory
programming classes to Ada (on a VAX) this quarter; ultimately we will go the
PC/workstation route.
----------------------------------
Warren J. Madden (rabbit@eddie.mit.edu), of MIT, EE/CS Computer
Facilities, Cambridge, MA., writes:
My organization is just starting to procure Mac II's for various
projects. We have decided not to get A/UX, as we want to wait for it to
mature before trying to develop systems on it. Currently we are planning
on using Smalltalk and C as our main languages, but if a good Ada compiler
were available I could easily see us buying ten or so copies for our Macs.
I also have an SE at home for my personal enjoyment. I happen to
think Ada is a fun language to program in, and would certainly consider
buying an Ada compiler for myself. However, the prime consideration here
is cost. If the compiler is priced over, say, $500, then it effectively
prices itself out of the casual programmer market. (Actually, that price
is more likely around $300, but given Ada's complexity I am being more
generous.)
Now many companies would say that losing the casual programmer
market would not be a sufficient reason for keeping the cost of their
compiler within the above limits. However, I feel this would be a
short-sighted and grave error. One of the reasons that C is such a popular
and widely used language is that it is available to so many new
programmers. Before Ada can hope to become as popular as C, it needs to
establish a large base of programmers interested in programming in it. In
my opinion, the best way to do this is to make it as widely available as
possible at a reasonable price. The money lost in keeping the cost down
would be made up for in the long run by the increase in compilers sold. Or
perhaps sell the compiler cheaply, but make the money back on separately
marketed CASE tools.
Of course, perhaps the optimum solution is to convince Richard
Stallman to write an Ada compiler for his GNU project. He could call it
Ida (Ida does Ada) :-).
I don't have much hope of any company looking past the short-term
profit motive to see the benefits of this proposal, but I thought it should
be aired.
----------------------------------
Alan Kaminsky P. O. Box 9887
School of Computer Science Rochester, NY 14623
Rochester Institute of Technology 716-475-5255
ark@cs.rit.edu
provided the following:
I teach in the Software Engineering master's degree program at the
Rochester Institute of Technology. This program started one year ago. We
are committed to teaching and using the Ada programming language in our
program. We are now planning a software engineering laboratory. My
current thinking is that the laboratory will have Apple Macintoshes that
run the Macintosh OS (_not_ A/UX), possibly networked with a DEC VAX
mainframe or some VAXstations. These Macintoshes will be equipped with
the Ada programming language as well as CASE tools.
2. What criteria are the most important in the possible selection
of such compilers?
My "must-have" criteria:
- Complete implementation of all Ada capabilities, especially generics,
tasking, and separate compilation.
- Able to run under the Macintosh Programmer's Workbench.
- Able to generate object code directly. (At present, Meridian's Ada compiler
generates C code which must be compiled by the MPW C compiler. This is
unacceptable.)
- Able to handle large programs.
- Object module library manager with capabilities like those of DEC's ACS
command for VAX/VMS Ada.
My desirable criteria, but not absolute musts:
- Source-level debugger.
- Transcendental math function libraries.
- Interface libraries to the basic Macintosh Toolbox routines.
- Interface libraries to AppleTalk.
- Educational discount pricing and/or site licensing available.
3. If such compilers were available, how many would you, or your
organization, be likely to purchase?
For our laboratory, we would have a dozen or so Macintoshes, and would
require a dozen or so copies of the compiler. We may also encourage our
students to buy their own Macintoshes, Ada compilers, and CASE tools. We
currently have about 60 students, and expect to grow to about 100 within a
couple of years.
All of the above is tentative. Still, I hope this helps convince compiler
vendors to offer Macintosh Ada compilers.
[For those of you who are interested, contact the Ada Information
Clearinghouse, c/o IIT Research Institute, 4600 Forbes Blvd., Second
Floor, Lanham, Maryland 20706-4312, for information about CREASE 5.0
(Catalog of Resources for Education in Ada and Software Engineering).
CREASE 5.0 lists approximately 200 colleges and universities with Ada
course offerings. You may also phone the Virginia office of the Ada IC
at (703) 685-1477.]
----------------------------------
Gary Combs (garyc@tekecs.GWD.TEK), of Tektronix, Inc., in Wilsonvile,
Oregon, said:
> 1. Who is interested in Ada compilers which run on Macintoshes
> under the original Macintosh OS?
I am !!
> 2. What criteria are the most important in the possible selection
> of such compilers?
1) Must at least run under the current Mac+ OS and fit within
the 1 MB of RAM.
2) Toolbox access !!
> 3. If such compilers were available, how many would you, or your
> organization, be likely to purchase?
Currently, I'm doing this on the side so 1 is OK for now.
----------------------------------
David Collier-Brown (geac!daveb@uunet.UU.NET) writes:
> 1. Who is interested in Ada compilers which run on Macintoshes
> under the original Macintosh OS?
I, as an individual, and a former-employer-and-still-friend
as a company.
> 2. What criteria are the most important in the possible selection
> of such compilers?
Ability to run under the Mac version of a APSE/KAPSE, MPW,
without seriously affecting the ability to move to a more
traditional environment at a later date...
> 3. If such compilers were available, how many would you, or your
> organization, be likely to purchase?
One. (Silly question! Mac's interconnect usefully on local
nets, so I reject the "all PCs stand alone" assumption...)
----------------------------------
In addition to the net mail, I received a number of phone calls. One
phone call from GE/RCA in Moorestown, New Jersey came from a caller
who said "if a decent Ada compiler existed for the Macintosh OS, my
organization would be interested in purchasing hundreds to thousands
of copies." When I asked about price, I was told that this was not an
issue. They fully expected to pay "several thousand dollars" for a
good quality compiler. However, they also expect quantity discounts.
According to the caller, his company found the Macintosh II to be a
particularly attractive machine for "board development," e.g., one
reason was the "six slot open architecture."
The caller made it clear that the traditional Macintosh interface was
very attractive. Further, the capabilities provided by Ada, especially
the ability to address low-level hardware architecture, made it a
strong candidate for a development language.
The requirements of the caller's organization for the Ada compiler
were the ones which were most familiar to those charged with using Ada
to develop embedded systems, e.g., full representation specification
implementation. They also were interested in an adequate toolset,
i.e., the usual software tools found in an embedded system development
environment.
----------------------------------
I received another call from someone working with the U.S. Space
Station effort. In case you didn't know already, Ada technology has
been chosen as the implementation language of choice for space station
software. Macintoshes are quite prevalent among those working with
space station software development. (To be fair, Apollos, Suns, and
other machines are in use as well.)
The caller told me that A/UX was not a possibility -- the Ada compiler
would have to run under the traditional Mac OS. The emphasis on
getting to the low-level hardware architecture via Ada was not as
strong as it was with the GE/RCA caller.
When I asked how many copies might be purchased, the answer was
"dozens to hundreds." As with the GE/RCA caller, price was not an
issue.
----------------------------------
If you have any questions or comments (or flames -- I'm used to them),
feel free to contact me.
-- Ed Berard
(301) 695-6960