Re: How to get Ada to ?cross the chasm??

["Randy Brukardt" <randy@rrsoftware.com>]

"Paul Rubin" <no.email@nospam.invalid> wrote in message 
news:87bme2oy91.fsf@nightsong.com...
...
> It's the other way around.  You're constrained from using HOFs and
> immutable data that are simple to reason about, instead of finicky
> control structures and other low-level manipulation that it's far
> preferable to leave to the compiler.

If you're doing that in Ada, you're doing it wrong. Use a container, use 
bignums if you have to (although that is very rarely necessary - integer 
algorithms tend to have small ranges). Futzing with "finicky control 
structures" (what the heck is that anyway, the only control structures Ada 
has are the normal ones like if and case) is unnecessary and "low-level 
manipulation" is something you do only when you need direct hardware acesss.

If you're writing "new" in Ada 2012, you had better be rewriting something 
that proved to be too slow in testing. (Or writing an Ada runtime.:-) 
Otherwise, you are wasting your time and energies. If you're explicitly 
putting "access" into an the interface of an ADT, you probably are copying 
the design of some other language -- it's almost never necessary in Ada. And 
so on.

If Ada has a problem in these areas, it is that you *can* do those 
lower-level things. That causes people (even ones that know better, like me) 
try to do better than the compiler/runtime stuff. You aren't likely to do 
so.

                                   Randy.