Sure, if you do not make use of d. is insignificant. However, difference in
behavior because of context could cause a production system to crash and
that would be very significant to me; being extra careful (trust but
verify), in my experience, has prevented grim consequences in similar
circumstances.

I am curious: What you do exactly mean by "Here's a definition for at which
works exactly like @"?

On Fri, Jan 4, 2013 at 11:00 AM, Raul Miller <[email protected]> wrote:
> On Thu, Jan 3, 2013 at 5:53 PM, Jose Mario Quintana
> <[email protected]> wrote:
>>> Here's a definition for at which works exactly like @
>>>
>>>    at=: 2 :'([: u v)"v
>>
>>
>> Rather works almost exactly?
>>
>>    ('*'"_) @ ((+: @ *:) (d.1)) (0 1 2)
>> *
>>    ('*'"_) @ ((+: at *:) (d.1)) (0 1 2)
>> ***
>>
>>    ((+: @ *:) (d.1)) b.0
>> _ _ _
>>    ((+: at *:) (d.1)) b.0
>> 0 0 0
>
> Here, we are no longer comparing the definitions of @ and at
>
>    ((+: @ *:) (d.1))
> 0 4x&p.
>    ((+: at *:) (d.1))
> 0 4x&p."0 0 0
>
> Instead, it's the working of d. that is significant here.
>
> Here's another case where @ and at are different, and I feel that the
> significance of this case is similar to the significance of the d.
> case (though obviously they are not identical cases):
>
>    '@' -: 'at'
> 0
>
> --
> Raul
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to