> My impression is that Rational's products are aimed at making tools
> to help programmers more easily make more solid programs. They may
> have decided that the demand is for "chewing gum and baling wire" to
> hold together poorly designed products, but it seems unlikely that
> even the best maker of chewing gum and baling wire will have a large
> market, or a good reputation, as it becomes clear that poor design
> plus CG&BW doesn't cut it.
Rational's products are not all "chewing gum and baling wire." True,
they do have a very good debugger, but some of them do a lot to help
programs be less poorly designed.
There are people who think that Ada is "doomed." Let's just try
to look through their eyes for a moment.
IF (the big IF) Ada is a sinking ship, will swimming around it trying to
hold it out of the water do anyone any good? The C++ ship has bigger
quality leaks, but they also have much bigger bilge pumps (and they're
pumping a LOT of bilge). Personally, I don't think the Ada leaks are sososo
bad, but if they were, maybe humanity would be better served by a
change of tactics. Some possibilities:
1. Promote Java. Hey, we know Ada is a lot better, but if Ada can't win,
Java is certainly the lesser of two evils, right? And while you're
teaching Java, you can subtly raise the awareness of Better Ways...
"Now no language is perfect, and the need for a 'break' here is one
of the old lessons Java missed. If you can't be careful with this,
you might have to switch to Ada."
"Now Java doesn't have safe enumerated types like Ada, but here's a
way to emulate them that's not TOO bad..."
(Maybe you could title the course "How to Write Java as safely as Ada")
2. Promote tools that will reduce the probability of the disasters that
C & C++ are so prone to (like Rational Apex). Again, sneak in a few
subtle hints along the way....
"Everybody knows* this C construct is very error-prone, but the XXX
feature of our tool, if used properly, can make your code almost as
reliable as Ada in this area."
* a little hyperbole--half the problem is that NOT everybody knows.
(The other half is those who know act like they don't care.)
I've not given up on Ada myself (yet), but if supporting Ada is a war,
let's try not to shoot parties who seem to be neutral or less dedicated
than we are.