Re: Why ADA is not prevalent yet?

[aunro!alberta!ubc-cs!news-server.csri.toronto.edu!utgpu!cs.utexas.edu!wup]

pattis@cs.washington.edu (Richard Pattis) writes:
>Borland would write a Turbo-Ada if they thought it were profitable. They may
>be doing so, although I would guess not. Of course, because of Ada's
>standarization, they would have difficulty creating proprietary features
>that would lock users into their system...

That doesn't mean they wouldn't have a product...it just means they'd need
to go about it differently.

The goal is not "locking users into a system" (although admittedly it looks
that way sometimes;-).  The goal is "differentiation in the marketplace"--
which is rendered in English as "creating a reason for people to buy your
product instead of someone elses.  There are various ways to do this.  Pro-
prietary features give one way...but there's also "quality" features (like
fast compilation, good diagnostics, good code) and external features like
editor help, debugger, etc.

In short, there's no end of ways to differentiate a product; I suspect the
reason Borland hasn't produced a "Turbo-Ada" (yet) is tied to Richard's
first reason: they haven't thought it would be profitable.

>...Maybe becaue of Ada's complexity,
>it wouldn't be so "turbo". Maybe when Intel's architecture becomes obsolete,
>Turbo-Pascal, etc. will pass away ;).

But Intel's "architecture" (OK, we'll call it that if we must:-) *IS* obso-
lete.  There are those who would argue it was obsolete when the 4004 was
introduced, but that's a different flamefest...the problem is that it's not
*gone* yet.
-- 
Dick Dunn    rcd@raven.eklektix.com   -or-   raven!rcd
	...Simpler is better.