Re: A new notion: stronglly-typed-by-user language

["J-P. Rosen" <rosen@adalog.fr>]

Maciej Sobczak a �crit :
> On 15 Kwi, 10:59, "J-P. Rosen" <ro...@adalog.fr> wrote:
> 
>> "C++ is a strongly typed language, if conversions between logically
>> unrelated types are avoided".
>>
>> Ain't that cute?
> 
> http://en.wikipedia.org/wiki/Strong_typing
> 
> "these terms have been given such a wide variety of meanings over the
> short history of computing that it is often difficult to know, out of
> context, what an individual author means when using them."
Of course, there is obviously no clear-cut

[...]
> The statement that you cited is therefore not very convincing.
> 
> In particular, what are "conversions between logically unrelated
> types"?
> Is Integer'Image such a conversion? What about arbitrary casts between
> numeric types? What about unchecked casts?
I was expecting someone to come up with Unchecked_Conversion, so you win :-)

Do not confuse "weakly type" and "strongly type with a mean to disable
it in a controlled way when absolutely necessary". Having no safety
belts in a car is not the same thing as having belts and not putting
them (as far as the car design is concerned).

What I find amusing (or characteristic of the C/C++ spirit) is the idea
that "if you are careful, it is good enough".

I generally conclude my presentations of Ada with two quotes. The first
one is from K&R, in one of the first books about C:

"C has been designed under the hypothesis that the programmes is
reasonable and knows what he's doing"

The other one from ARM's introduction:
"Concern for the human programmer was also stressed during the design"

C is not for humans...
-- 
---------------------------------------------------------
           J-P. Rosen (rosen@adalog.fr)
Visit Adalog's web site at http://www.adalog.fr