Re: Mike Feldman, meet Archie

[priestm@westminster.ac.uk (Mark Priestley)]

In article <1993Mar8.152831.6407@seas.gwu.edu> mfeldman@seas.gwu.edu (Michael Feldman) writes:

>In article <1993Mar8.132419.21952@westminster.ac.uk> priestm@westminster.ac.uk (Mark Priestley) writes:
>>
>>There is another, more fundamental reason in my opinion: Ada is widely regarded
>>as a dead language, both by students and many faculty members, and as a result,
>>although the technical merits of the language might be well-understood,
>>no-one's prepared to make the significant investment that changing to Ada would
>>represent.  By "dead language" I mean an ill-assorted collection of
>>observations, including: "there are very few jobs in Ada"; "why don't Borland
>>have an Ada compiler?"; "why is Ada introducing tagged types, instead ofjoining
>>the OO mainstream?" As a colleague said: "it's like teaching Latin instead of
>>French, on the grounds that Latin's got more grammmar".

>Well, you make some good points. I suggest, though, that you ask your friends
>at the following UK universities why they don't think Ada is dead. In all

> [stuff about takeup of Ada in the UK omitted]

Don't get me wrong - I'm basically on the side of the angels here, and I know
about the spread of Ada in the UK.  We currently teach Ada as a second-year
option for students specialising in Software Engineering: I occasionally moot
the possibility of switching to Ada as a first language (in place of Modula-2)
and the argument I have most difficulty confronting is the "dead language"
argument. It basically has two components:

    * Ada's missed the boat on object orientation.
    * Ada's too big, and hasn't caught on (just like Algol 68, for example)

Now I disagree about 80% with both of these assertions - but I find them very
difficult to counter.  The people who hold these opinions just aren't
sufficiently impressed by Ada's good points, I'm afraid (maybe I'm no good at
arguing :-)

>I believe, as do many other teachers, that one's first language influences
>one's thinking forever. This holds for natural as well as computer languages.
>Ada should certainly not be the only one learned by students, but many of
>us believe it should be the first. Even in the UK, apparently...

I agree wholeheartedly with all this, except that I waver on whether the
benefits of switching from Modula-2 to Ada outweigh the costs ...

>Let's talk a bit about this "investment" to move to Ada. Are you referring
>to universities' perceived investment in compilers? Have you checked the
>prices lately? Are you referring to mental investment? Are you claiming
>that the mental investment to move from, say, Pascal, to ML or C++ is
>lower? I find that difficult to believe.

Compiler prices are certainly an issue: we want to use a Unix network, not PCs,
for most of our programming teaching, and our Ada compiler supplier recently
shocked us with the suggestion that we should pay a lot more than we currently
are!  We're currently looking around: all recommendations gratefully received!

Mental investment in moving to Ada certainly is significant, though less than
moving to ML or C++ (heaven forfend!) would be.

Most significant is the knock-on effect on the resources required for other
courses: a lot of courses use the first language simply as a tool, and in many
cases rely on particular libraries, or interfaces being available.  If we were
to switch to Ada, we'd need to find equivalents for all these.  I'm not saying
it couldn't be done, but it's a *lot* of work.

>As far as Ada's "death" is concerned, we'll have to see where it is in
>five years or so. Obviously nobody has a pure, clear crystal ball. But
>we can all make predictions. Mine is that Ada will, over the long run,
>be seen to have "staying power" and may outlast some of the more faddish
>things as people become disillusioned with the instability of the fads.
>Ada's stability and conservative "waterfall model" design may be thought
>of as stodgy by many of us impatient computer tekkies, but I'll stick 
>out my neck and speculate that the conservativeness and accompanying
>portability will be eventually seen as a distinct advantage.

Well, I hope you're right, but Ada's superiority to competing languages was
IMHO much clearer 10 years ago, and for whatever reasons, it hasn't exactly
swept the floor with them since then.

>Cheers all -

>Mike Feldman

Cheers

Mark Priestley

-- 
Mark Priestley      	    	    Email:  	M.Priestley@uk.ac.westminster
School of Computer Science ...
University of Westminster
115 New Cavendish Street    	    Telephone: 	+44 (0)71-911 5000 ext. 3653