26 Nov
2007
26 Nov
'07
7:32 p.m.
On 26 Nov 2007 at 18:59, Fred Cisin wrote:
I started a tangent that isn't likely to stay on
topic.
I'm sorry.
I have a crackpot theory about the effect on software development of
having similar or different syntax between spoken and programming
languages, and therefore why Forth would be better than FORTRAN for some
cultures, etc. Unless we want to talk about what effect that had in the
1970s and 1980s, this isn't the place for it.
I think I can put this back on topic, being one of the guilty parties
(sorry, Jay).
While some have suggested BASIC and FOrth, no one has yet suggested
COBOL. Yet, COBOL is the closest of common programming languages to
English, unless you'd like to count some of the contrived "natural
programming languages" such as Metafor. In any case, COBOL might be
closer to the language already understood by a young person.
However, natural language is rarely rigorous, as many have pointed
out and attempting to add rigor to conversational natural languages
is probably folly.
I submit that any high-level programming language--formal or natural--
while a shortcut way to get a machine to "do something", obscures the
inner workings of a computer to such an extent as to give little clue
as to precisely how the thing operates. In particular, "structured
programming" elements can really hide inner workings.
So, I propose that programming be taught first in machine language,
then assembly. That's how I learned to do it.
Cheers,
Chuck