Bartosz about Chapel

[Timon Gehr]

On 11/13/2011 03:13 PM, Nick Sabalausky wrote:
> "Timon Gehr"<timon.gehr at gmx.ch>  wrote in message
> news:j9mtth$d26$1 at digitalmars.com...
>> On 11/12/2011 09:05 PM, Somedude wrote:
>>> Le 12/11/2011 17:27, Timon Gehr a écrit :
>>>>> I tend to believe D is conceptually more complex than C++. Which means
>>>>> it allows for more expressivity.
>>>>
>>>> What makes you think it is conceptually more complex? Most features that
>>>> make D more expressive than C++ are conceptually very simple. And what
>>>> is conceptually simple in D is usually also implemented in a simple way
>>>> (as opposed to C++).
>>>>
>>>>
>>> I think D's type system is more advanced and more detailed than C++ or
>>> as any C derived language. This takes a while to grok. Of course, I
>>> guess at the beginning one can program in D with only a subset of the
>>> language.
>>
>> Ok, I see. Probably we are missing a nice tutorial about the type
>> qualifiers (const/immutable/shared/inout) ?
>>
>> Or is there another type system feature that takes a while to grok?
>>
>
> I'd imagine D's type system might sometimes seem more complex than it is
> simply because highly complex types involving fancy templates are used so
> much more. And, of course, they're used more because D makes it so much
> easier to use them.

Maybe, I don't know.

>
> And then, there's also complexities that are for the better. For example,
> C++ AFAIK doesn't have IFTI. IFTI *is* an additional set of
> rules/complexities *in the language*, but they're complexities that make the
> language easier to use. SImilar thing with the implicit conversion rules for
> primitive types: The rules for that are somewhat complicated, but they
> usually make using the language easier and safer.
>
>

C++ allows instantiating function templates without explicitly giving 
the type parameters. Or what exactly do you think C++ does not have?