Re: Type Conversion in an Assignment Statement

[Carsten Freining <freining@informatik.uni-jena.de>]

Ted Dennison schrieb:

> In article <3B2DB8B9.1A26F90F@informatik.uni-jena.de>, Carsten Freining says...
> >This is a correct program. I thought it would raise a constraint_error,
>
> You shouldn't ever *count* on a constraint check for proper operation of your
> code. They can be unexpectedly removed or moved around on the whim of the
> compiler or user (via pragma suppress or a command-line parameter).

I usually wouldn't. I am the Tutor of an Ada-course. Just simple things and I
needed some Exercises for accesstypes. This Example is really interesting and
tricky, because it is not that obvious. But then, as long as there is no proof for
this in the Ada LRM it could possibly be the Compiler with a bug.
So you can't *count* on anything, without a proof that makes the facts clear. But
we don't have to teach this stuff to the next Generation, well, not from the start.

Carsten Freining.
--
----------------------------------------------------------------------------
Carsten Freining

e-Mail:         freining@informatik.uni-jena.de
Tel.            03641 9 46344
Adresse:        Friedrich-Schiller-Universit�t Jena
                Fakult�t f�r Mathematik & Informatik
                Institut f�r Informatik
                Lehrstuhl f�r Programmiersprachen und Compiler
                D-07740 Jena, Germany

----------------------------------------------------------------------------