This is derived from part of a D.learn post, maybe it's of your interest.

This program shows two similar versions of the same loop:


import std.stdio, std.algorithm, std.typetuple, std.range;

void main() {
    auto items = [[10, 20], [30]];

    foreach (_; 0 .. 2) {
        foreach (sub; items) {
            iota(sub.length)
            .map!((i){ writeln(sub); return 0; })
            .array();
        }
    }

    writeln();

    foreach (_; TypeTuple!(0, 1)) {
        foreach (sub; items) {
            iota(sub.length)
            .map!((i){ writeln(sub); return 0; })
            .array();
        }
    }
}


Its output:

[10, 20]
[10, 20]
[30]
[10, 20]
[10, 20]
[30]

[10, 20]
[10, 20]
[30]
[30]
[30]
[30]


Do you see why the output of the two foreach(_) is different?

Bye,
bearophile

On 03/01/2013 11:07 PM, bearophile wrote:
> ...
>
> Do you see why the output of the two foreach(_) is different?
>

They are not different.

Timon Gehr:

> They are not different.

Really? Sorry, I don't understand.

Bye,
bearophile

On 03/02/2013 12:23 AM, bearophile wrote:
> Timon Gehr:
>
>> They are not different.
>
> Really? Sorry, I don't understand.
>
> Bye,
> bearophile

According to the spec the two outputs are the same.

On Friday, 1 March 2013 at 23:25:31 UTC, Timon Gehr wrote:
> On 03/02/2013 12:23 AM, bearophile wrote:
>> Timon Gehr:
>>
>>> They are not different.
>>
>> Really? Sorry, I don't understand.
>>
>> Bye,
>> bearophile
>
> According to the spec the two outputs are the same.

Are you saying they should output the same or that they are the same?

You and I have a different definition of "same" if the latter.

Timon Gehr:

> According to the spec the two outputs are the same.

I have filed this:
http://d.puremagic.com/issues/show_bug.cgi?id=9628

Bye,
bearophile

On 03/02/2013 01:17 AM, Brad Anderson wrote:
> On Friday, 1 March 2013 at 23:25:31 UTC, Timon Gehr wrote:
>> On 03/02/2013 12:23 AM, bearophile wrote:
>>> Timon Gehr:
>>>
>>>> They are not different.
>>>
>>> Really? Sorry, I don't understand.
>>>
>>> Bye,
>>> bearophile
>>
>> According to the spec the two outputs are the same.
>
> Are you saying they should output the same or that they are the same?
>
> You and I have a different definition of "same" if the latter.

With a correct implementation of the language, the outputs are the same. Bugs should go to the bug tracker.