Re: Anybody in US using ADA ?

[kcline17@hotmail.com (Kevin Cline)]

"Marin David Condic" <mcondic.auntie.spam@acm.org> wrote in message news:<av1c7j$ctl$1@slb3.atl.mindspring.net>...

> Maybe that's part of why there was/is a lot of resentment toward Ada. It
> walked in the door with a mandate and everyone had big investments in other
> things and the feeling was "This is going to cost me truckloads in
> discarding my existing investment and for what? An uncertain 'technical
> advantage'? Why are they trying to make my life harder?"

It walked in the door with an overly broad mandate and almost nothing
else.
The compilers were buggy.  The debuggers were a joke.  Tasking was
useless
except on an Ada RTOS.  Bindings to system facilities were
compiler-specific.
Without function pointers, binding to a GUI was nearly impossible.
Work that could be done in hours in C suddenly expanded to take
several days.
Most programmers with any breadth of experience quickly labelled Ada
a non-starter.

> 
> This is why Ada's best hope would be to attach itself to some new, emerging
> technology where it doesn't have to displace existing technology and can
> itself become the entrenched status quo. Barring that, it has to find some
> groundswell of usage out in the hacker world so that there is some large
> infrastructure of practical things in use that keep it afloat. 

LOL.  Look at the languages that have been adopted by the 'hacker
world'
over the past fifteen years: C++. Tcl/Tk. Perl. Java. Python. Ruby.
PHP.
The drive is to do more and more while saying less and less.  Safety
is good, but expressiveness and compatibility with existing technology
(i.e. the Web, relational databases, and GUI) are the real drivers. OO
is still going strong, but functional programming is gaining
mindshare.
Compile-link-run is on the wane.

Also, all these languages provide built-in associative containers.  
Arrays are fine for small embedded systems, but not very useful
for database-driven text processing applications.