BCS wrote:
> Reply to Bruno,
> 
>> BCS wrote:
>>
>>> That is a good point, but how else you you get a real closure that
>>> has mutable state?
>>>
>>> int delegate() Seq()
>>> {
>>> int at = 0;
>>> int Next() { return at++; }
>>> return &Next;
>>> }
>> Err.... just have a closure like D has now?... one that allows access
>> to any visible variable?
>>
> 
> Franks local delegates won't work because they aren't valid after the other function returns, his heap delegates won't work because the variables they can access can't be altered.
> 
> 

I feel there is a misunderstanding here. I wasn't talking about Frank's closures, but the one D currently has. What exactly did you mean with "how else [do] you get a real closure that has mutable state?" ?

-- 
Bruno Medeiros - Software Developer, MSc. in CS/E graduate
http://www.prowiki.org/wiki4d/wiki.cgi?BrunoMedeiros#D

Reply to Bruno,

> BCS wrote:
> 
>> Reply to Bruno,
>> 
>>> BCS wrote:
>>> 
>>>> That is a good point, but how else you you get a real closure that
>>>> has mutable state?
>>>> 
>>>> int delegate() Seq()
>>>> {
>>>> int at = 0;
>>>> int Next() { return at++; }
>>>> return &Next;
>>>> }
>>> Err.... just have a closure like D has now?... one that allows
>>> access to any visible variable?
>>> 
>> Franks local delegates won't work because they aren't valid after the
>> other function returns, his heap delegates won't work because the
>> variables they can access can't be altered.
>> 
> I feel there is a misunderstanding here. I wasn't talking about
> Frank's closures, but the one D currently has. What exactly did you
> mean with "how else [do] you get a real closure that has mutable
> state?" ?
> 

The problem with the current ones is that they aren't valid in some cases if the outer function leave the scope the delegate is defined in.


void Foo()
{
  int delegate() Bar;
  for(int m = 3; m; m--)
  {
      const int i = m;
      Ber = {return i;}
  }
  {
      int j = 6;
      int k = Bar();
      //k could be 6
  }
}