Re: Ada and Java/C++

[Shark8 <onewingedshark@gmail.com>]

On Monday, August 27, 2012 11:35:28 PM UTC-6, kalvi...@gmail.com wrote:
> 
> 
> My suggestion here is to introduce Ada's good virtues for C and Java programmers in syntax that is familiar to most of the software engineers.

You aren't listening to the replies you're getting; it's a bad idea to try to gloss over (extend) C/Java syntax. Randy's reply is a good one in explaining why syntax itself matters.

> 
> 5. Ada programmers are hard to find
> - For example, DoD has faced the problem of finding Ada programmers, DoD has been force into situation that mission critical software is being implemented in Java

Having been looking for an Ada job, and applying to lots of listings to get lots of non-responses I'm going to call this one bullshit. (Besides, why is it when comparing languages the other way people always say "any programmer can learn C/C#/PHP/Java etc?" shouldn't they also be able to learn Ada if that's the case?) 

> - Java and C programmers see Ada as a complex language

Well, it's complex in some issues because it can't be any simpler. (Ok, that isn't strictly true; but take a good look at the rules and a lot of times the complexity comes from addressing a non-trivial issue. Example: Tasking.)

> 
> - Ada is perceived as military language

So? If you go back a few decades the Jeep was perceived as a military vehicle.

> 
> - Ada is perceived as a complex and heavy language

Again, that's because there are non-trivial problems being addressed. And it could be considered 'heavy' if you're looking at all possible combinations rather than realizing that the structured nature and orthogonality in which they are applied is very consistent (i.e. generics and compilation-units which can be nested in other such compilation-units).

> 
> As you may see from list above, the points 1&2, and points 3-5 are contradicting. Could the situation be improved somehow, so that existing Java and C programmers can adopt Ada-mindset more easily, and can the Java and C syntax be fixed to allow more robust software engineering.

[Warning: From here on sarcasm and hyperbole increase]

Possibly; but I'm going to have to say it's not likely. Why?
Because of an irksome psychological-tendency for programmers to make their life more difficult, especially WRT debugging. (Programming's supposed to be hard, right? And it really shows off how clever you are when you spend hours/days/weeks tracking down an obscure bug... and what's the best way to make a bug obscure? Make it so the compiler accepts the widest range of possible syntax-structures [which should be similar] and outputs a runnable executable: we need to increase the edit-compile-run loop!)

[End Warning]

That is the real reason that Ada (and Delphi and Object Pascal) isn't as popular as C-ish languages; the compiler is so stubborn that it complains with sensible error-messages until you fix things... which leaves the programmer with [almost] only logic-errors. And who wants to admit that they said "not in" instead of "in", "<" instead of "<=" or changed the code so that assumptions were no longer valid?

That's not programming! There's just not enough debugging involved -- the dragons are too small! In fact, if we were really clever we'd use a language consisting only of tabs and meaning could be inferred from context!

But on a more serious note; it might be interesting to build a [basically mufti-lingual] compiler that could, say, take a grammar and a source and compile the source. {Yeah, that'd be kinda like GCC, with all it's front-ends.} Then you could inexpensively try out different grammars until you find your Ada/C hybrid.