Re: Generic-Package Elaboration Question / Possible GNAT Bug.

["Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>]

On Wed, 23 Nov 2011 09:32:20 +0000, Simon Wright wrote:

> "Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de> writes:
> 
>> It is lousy language design placing implementation (executable code
>> ensuring the desired behavior, e.g. raising exception) into
>> declarations.
> 
> Are you just talking about [post]conditions? because it's not uncommon
> to use complex declarative regions for run-time array bounds etc, which
> will include the possibility of exceptions.

Such constructs are certainly not welcome. However declaration of a local
variable requiring some elaboration that may fail is one thing, placing
executable semantically loaded code into the interfaces is extreme.
Implementation belongs to the bodies.

If you mean dynamic constraints, yes it is very dangerous stuff, which must
be handled by the language with a great care. I think Ada 83-95 found a
balance here.

P.S. I hope everybody agrees that dynamic pre-/post-conditions are a part
of the implementation? The rest is simple. [A comparable case is barriers
of entries.]

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de