wmb@MITCH.ENG.SUN.COM (02/27/91)
> AT&T and other big companies maintained very strict control over C,
Say what? Evidence?
Evidence to the contrary: Lots of crappy, partial, C implementations,
and not once did AT&T complain.
Why C did not diverge:
For a long time, nearly everybody that used C used it on Unix machines.
In that environment, it was a nearly complete language with most of
what people really needed.
By the time that it became popular, there was a useable de facto
standard (K&R C in the Unix environment). Subsequent implementations
for other environments were judged in the marketplace based on how
well they conformed to the de facto standard.
In other words, C essentially had a single environment in which to grow,
and by the time it "broke out", it was already essentially complete.
Also, the original designer of C had the good sense to make it pretty
much complete before he starting passing it around.
On the other hand, Forth "hit the streets" before it was anywhere near
being complete, and in order to make it useable for real jobs, everybody
was forced to extend it even in areas which are generic and should have
been standardized from the word go.
The same think pretty much killed Pascal.
I think Chuck and Niklaus Wirth are both too proud and too stubborn to
give up their freedom long enough to make hard engineering choices in
the interest of usefulness for other people. Wirth may have learned
his lesson (but perhaps too late; evidence: Modula-2). I not sure the
same is true for Chuck.
C is a boring idea with excellent engineering behind it; Forth is an
awesome, inspired idea, with lousy engineering. One thing you learn
in business is that ideas are a dime a dozen, and its the execution
that makes or breaks you.
Mitch Bradley, wmb@Eng.Sun.COM