2014-05-12 19:17 GMT+02:00 Eliot Miranda <eliot.miranda at gmail.com>:

>
>
>
> On Mon, May 12, 2014 at 10:05 AM, Nicolas Cellier <
> nicolas.cellier.aka.nice at gmail.com> wrote:
>
>> I think I remember it was either the first or last parameter that was
>> returned in old implementation
>>
>
> the last. ([:a :b|] value: #first value: #last) == #last
>
>
>>
>> It's a known problem: http://bugs.squeak.org/view.php?id=5793
>>
>> but I would have thought that this was fixed with Eliot's closure...
>>
>
> I kept the old semantics, trying to rock the boat as little as possible.
>  I just checked the draft standard.  The end of 3.4.4 "blocks" reads
>
> "If a block has no <block body> or no <statements> in its <block body>
> then the value of the block is undefined."
> I've found the returning the last value bizarre for a while.  If others
> feel strongly we could fix it for 4.6/5.0.  But I'd want agreement with
> Pharo.  Marcus what does Opal do with:
>
>     [:a :b|] value: #first value: #last
>
> ?
>
>
As I reported in the bug tracker, the NewCompiler (former nick name for
Opal) answers nil.
This can be considered as an extension of [] value...


> Squeak & VW answer #last.
>
>
>>
>>
>> 2014-05-12 17:58 GMT+02:00 blake <dsblakewatson at gmail.com>:
>>
>> >> | myBlock |
>>> >> myBlock := [ : arg1 : arg2 | ].
>>> >> myBlock value: 'hello' value: 'there'
>>>
>>> nil
>>>
>>> Because the block isn't returning a value and the last line is calling
>>> the block.
>>>
>>>
>>>
>>>
>>
>>
>>
>>
>
>
> --
> best,
> Eliot
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20140512/dfea3859/attachment.htm