Re: Hoare's gripes about Ada (was Re: Ada and C++ ...)

[agate!howland.reston.ans.net!darwin.sura.net!haven.umd.edu!cs.umd.edu!aft]

In article <1414@fedfil.UUCP>, news@fedfil.UUCP (news) writes:
|> In article <1993Aug19.120801.18134@jarvis.csri.toronto.edu>, blaak@csri.toro
nto.edu (Raymond Blaak) writes:
|> *Someone, somewhere writes:
|> 
|> *>>One of the supposed beauties of Ada is that it cannot ever change.  The A
da 
|> *>>which Hoare claimed was unfit for any use with serious consequences is th
e
|> *>>same Ada which is out there now, to the best of my understanding.
|> 
|> *Regardless of what Hoare thinks about Ada now, what were his original
|> *complaints about it?

Ted must have seen a somewhat different speech than I've got in my Turing
Lectures book.  In his 1980 Turing Lecture, Hoare talks about his involvement
in the ALGOL language design.  That's the main thrust of his speech and
13.5 of its 16 pages (in my copy).  At the end he talks about the problems
that ALGOL 68 had due to its size and complexity.  He goes on to indicate
that he sees Ada taking the same course as ALGOL 68.  His main concern is
with the safety of a large, complex language... Subsequently he has
indicated that it didn't turn out so bad.

|> That the language itself was far too large for starters...  Hoare correctly
|> noted that the big success stories of recent years, C and Pascal, included
|> in the base language only those things which figured to be used in every
|> program which ever got written using them, or very nearly only those, and
|> that everything else and the kitchen sink, rather than being part of the
|> language, got put into libraries to be linked in by the user who needed
|> them, only when he needed them.  C doesn't even include (in the language)
|> any notion of IO and, while this may have appeared extreme at one time,
|> it appears far-sighted now.  There is simply no reason why a program which
|> uses MS-Windows should also contain all of the code for stdio.h, X Windows,
|> MS-DOS screen IO (conio.h), etc. etc., when it doesn't use them.

Hoare doesn't mention C at all...  only Pascal and ALGOL.

|> Hoare noted that the language was far too complex, and that users were
|> going to spend more time working around Ada than solving their own 
|> problems;  in theory, most programmers are being paid to do the later and
|> not the former.

NOPE.  Doesn't say this at all...

|> And there were a number of other things.  It's not as if he gave a long
|> speech on another topic and then, at the end, said "Gee!  this Ada thing
|> looks like a bad idea..."

He did in fact give a long speech on another topic (ALGOL) and then related
the bad things that happened in ALGOL 68 to what might happen in Ada "today"
(where "today" was 1980).

I think the best summary of his feelings is the statement (remember in 1980),
"For none of the evidence we have so far can inspire confidence that this
language has avoided any of the problems that have afflicted other complex
language projects of the past."  The language did change a bit between
1980 and 1983 :-)

|> Were that the case, I could easily imagine his being convinced somehow or ot
her
|> to change his mind.  In fact, however, the entire speech was on the topic
|> of Ada, it was a hell of a long speech, and there was a buildup which took
|> into account a number of predecessor languages which failed for the same
|> reasons which Ada has failed for, and this included most notably PL/1.

Pure, unadulterated BS.  The speech was on ALGOL.  PL/I was never mentioned.
His concern with Ada was genuine, but it totalled only 5 written paragraphs.

|> In the speech, Hoare very clearly laid out what should be guiding principals
|> in the design of programming languages, and Ada was a kind of an ultimate
|> opposite example to everything which Hoare figured was right.

Not true.  Hoare does give the principals which he based the design of
ALGOL on.  He does not relate this to Ada anywhere.  He does comment that
a language without subsets should be made small.

|> So, assuming the people aren't simply lying in claiming that he has since
|> come around to being an Ada admirer, your guess is as good as mine, but
|> I am simply not able to believe that he simply was convinced that he had
|> been wrong.  That would imply that everything he had ever learned or
|> believed about computer science prior to 1980 had been 100% in error,
|> and that he'd have been better off selling used cars for a living prior
|> to that time, and only starting work in computer science AT that time.

Well, Ted, since he expressed a concern and hadn't passed final judgment --
the language wasn't complete yet and wasn't implemented for many years to
come.  I find it ironic that this message which seems to have many
"factual short-comings" ends accusing others of lying...

|> I figure they probably tied him to a tree and forced him to listen to
|> rapp music until he succumbed, but that's ONLY a conjecture.

|> Ted Holden      Ada is to computer science
|> HTE             As rapp is to music

-- 
Christopher A. Warack, Capt, USAF
Computer Science Department, US Air Force Academy

cwarack@kirk.usafa.af.mil                (719) 472-2401