Re: Access type to member procedure of instance (Object Oriented programming in Ada)

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

On Mon, 19 Nov 2012 14:19:46 -0800 (PST), Adam Beneschan wrote:

> On Monday, November 19, 2012 1:26:58 PM UTC-8, Dmitry A. Kazakov wrote:
>> On Mon, 19 Nov 2012 10:23:13 -0800 (PST), Adam Beneschan wrote:
>> 
>>> This doesn't look at all like what I'm talking about, since &Example::f
>>> still refers to a method defined for the class Example but without
>>> reference to a particular object.  Something equivalent would be if you
>>> could say &object.f (or maybe &object.*p) and store the result of that in
>>> a pointer that could be called later.
>> 
>> What would you do for:
>> 
>>    type T is tagged ...
>> 
>>    procedure Foo (X, Y, Z : T) ;
>> 
>> Would the closure carry three objects with it?
> 
> What I'm proposing is a way that one could say something like
> 
>    Object.Operation'access
> 
> In Ada, with regard to your procedure Foo, you can say
> 
>    X.Foo (Y, Z)
> 
> That's just the way the language is.

Two wrongs does not make one right.

>> If closures to be supported then they better be a bit more universal than 
>> just one tagged object + one operation.
> 
> Why?  Is there a rule that says that if a proposed change doesn't fix the
> entire world, then we shouldn't propose something that improves part of
> the world?

No idea. It is a great mystery to me what makes into the language and what
does not. Considering this one:

1. Why closures at all?

2. Why some arguments are more important than others, unless the programmer
specifies so?

3. Why closures must support only tagged types?

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