Re: No Go To's Forever!

[Robert Dewar <robert_dewar@my-deja.com>]

In article <38D9B88B.366CC1AE@acm.org>,
  jrcarter@acm.org wrote:
> Richard D Riehle wrote:
> > When Dijkstra wrote his letter to the Communications of the
ACM
> > in 1968, "Goto Considered Harmful," lots of people read the
> > title but not the letter.

>
> Quite true. For example, IIRC the letter was about program
correctness
> proofs, and argued that the goto made such proofs much harder,
not that
> the goto made human understanding or modification of software
harder.


That's not really a fair formulation of the letter (please
reread it).

But what is important is to realize that in the environment
that EWD was operating (Europe) he was not saying something
new. Everyone who ever wrote in Algol-60 knows that you don't
often need goto statements. Really the letter was an interesting
commentary on the damage that was being done in the US by using
junk languages like Fortran that lacked decent structural
alternatives to the excessive use of gotos.

It is interesting to note that in 1971 (I think that was the
year), there was a survey of programming language use in
unversities in the British Journal of Computing (I think that's
the right citation, I no longer have this at hand). It asked
what language you are teaching now, and what would you like
to teach.

Overwhelming response was

England, now teaching Algol-60, would like to teach Algol-68 (*)

Europe (certainly at that time, England did not consider itself
part of Europe, whether it does now is debatable :-)
now teaching Algol-60, would like to teach Simula-67

USA, now teaching Fortran, would like to teach Fortran

:-)

(*) a reflection of the govt policy of forcing universities to
buy ICL machines, and the excellent Algol-68 compiler available
for the ICL series at that time.


Sent via Deja.com http://www.deja.com/
Before you buy.