> The reality is that Ada does not meet the needs of these
> programmers, largely because
> 1) There are no Ada compilers,
When true, I will agree Ada does not meet their needs. But this article
did not say there was no Ada compiler. Tt simply dismissed without name
anything but C, assembler, and Java.
> 2) They see anything other than assembler
> or C as too far from the hardware, and
> therefore, too inefficient.
Better described as Ada does not meet their misinformed _wants_
> For many of these embedded systems programmers, even C is too high
> level ...... I see Ada as ill-suited to applications such as those
> that typify the kind of project targeted to the I-8051. Assembler
> and Forth are still the better choices for such applications. Notice,
> I did not include C in that set of choices.
Forth is a good choice when you need assembler-like efficiency AND the
productivity of not having to do bit-twiddling. And Forth, although a
bit cryptic, does not encourage disastrous errors as much as C does.
(Disclaimer: The Forth I know and the Forth of today are two different
languages. I last used Forth twenty years ago.)
But I maintain my criticism of that article: Asked to choose a language,
the author rejected without thought everything but C and assembler, then
set out, not to decide between them, but to support the decision he had
already made. There was apparently a JDK available, since he gave Java
a sentence or two. If there's Java, it's hard to believe there isn't
C++, Ada, and at least one niche language.