Stanley Allen [SMTP:[log in to unmask]] wrote:
> First question: do we really believe that programmers will be attracted to
Ada on the basis
> of reliability? I know that some will, but what evidence is there to
persuade us that these
> are more than a small minority of programmers? It appears to me that, in
> especially, programming is an ego-enhancing activity, the vast majority of
> being males interested in the creative and fun aspects of programming,
with a focus on
> new or exotic technology features. The satisfactions that the average
> to find in developing software are not very closely related to the goals
we often list as
> "software engineering principles": reliability, maintenance, etc.
I'm guessing you are right about most ... parenthetically, is your use of
the word "programmer" intentional here, over "software engineer"? Are there
really two different classes of jobs here? I know that perception was very
much the case 19 years ago when I was a College Senior. My fellow CS grads
weren't interested in stooping to accepting a job with the title
"Programmer". I basically picked up that perception and assumed there was
some accuracy to it.
So... might it be accurate to say that "programmers" are attracted to
languages as you describe, while "software engineers" are attracted for
other reasons? Maybe that's a trivial point, I don't know. Either way,
those who are looking for a language that's fun to program in are probably
in the majority.
> Furthermore, it seems to me that our attempts to sell Ada on the basis on
> safety, "critical" systems certification, etc. have a touch of bad faith
in them. By this I
> mean that I distrust you people whenever you claim that "reliability" or
"safety" are the
> main reasons that you like to program in Ada. It's a bit like hearing
people say they
> subscribe to Playboy for its fine journalism. You just want to say, "Come
on, get real!"
> My thesis is: programmers (like you) program for pleasure first; other
> second. What aspect of Ada first appealed to you, struck your fancy,
> interest? For me it was tasking. For you perhaps it was private types,
generics, or other
> various features that lend themselves to the creative impulse in
programming. I doubt it
> was range checks.
Speaking as a "software engineer" ;-)
When Ada came in to my life, I was at what a coworker of mine likes to call
stage one on the path to Ada Nirvana: "It's just another programming
language". Well, sort of. What I mean is that I did not have the hubris to
go around saying, "I will only program in language <such and such>". I had
just joined a new project with some significant degree of company
visibility, and frankly it was an exciting position that made me say,
"program in xyz? Sure, why not!" Actually the project was directly tied to
the new DoD policy requiring Ada in all new projects. The project I was
joining was a core technology group charged with developing a software
engineering environment for Ada, and all software we wrote had to be in Ada.
I learned Ada because it was required for the job, not because I was
attracted to it. I was attracted to the idea of improving the quality of
software development in the company, as well as the idea of belonging to a
group that other projects depend on (we all like to feel needed ;-) ).
As for what feature of Ada was the most sexy to me, well, let's see. I
liked the syntax better than both Pascal and C, but we won't go there ;-) I
really don't know. If anything, the fact that implicit operators were not
automatically directly visible was actually a pain in the butt. Ada was the
first language I had run into that had that "feature". I quickly developed
a hatred for the "use" clause, and to use the "goto" was literally
unthinkable. I can't really say there was anything sexy about Ada. I
mostly found myself poo-pooing the complaints of others. My favorite has
always been "I have to type too many characters" in defense of "use"
clauses. Maybe as a touch typist, I'm a bit prejudiced, but this complaint
has always struck me as the epitome of laziness.
To me, Ada is a means to an end, not an end in and of itself. Furthermore,
the most disturbing implication to me over the last couple of years with the
termination of the Ada Policy is that somehow all the work that was done to
design the best language is worth doodly-squat. Call me a trusting soul,
but when I heard that Ada was a language designed over a number of years
with the interests of the DoD in mind, I just assumed that Ada had SOMETHING
going for it. Who am I to question that? So I didn't. And nowhere in my
11+ years with Ada has the language failed to live up to that trust. It
seems like there's this mentality out here in Aerospace and Defense Land
that automatically dismisses Ada and all the work that was put into it. I
think I get it; I just don't buy it.
> Second question: what's wrong with selling Ada as the real-time language
> instead of the safety-critical language of choice? Jeff Burns says:
>> Ada may be better at real time, but C is the major player
>> there and well established. It's an uphill battle to
>> establish Ada as better than C for real time.
> Unlike Tucker, I tend to believe that focusing on "critical" systems has
> psychological results. The underlying message of such an approach is "Ada
is good for
> systems that absolutely must work -- so don't consider it for less
Doesn't selling Ada as the "real-time" language of choice also have the same
type of underlying message? Indeed, no matter what focus you pick, by your
argument, aren't you limiting Ada's outreach?
James Squire Send my Spam to mailto:[log in to unmask]
MDA^H^H^HBoeing St. Louis http://www.boeing.com
Opinions expressed here are my own and NOT my company's
"Were you like this when you were married?"
"The woman was a saint."
-- Ivanova and Sheridan, "A Race Through Dark Places"