http://www.jsoftware.com/help/dictionary/dx003.htm

Above describes "boxed" as an internal type a noun can possess.

Another definition of "boxed", from 
http://www.jsoftware.com/help/dictionary/dicta.htm ,
is

produced by the verb box denoted by < .

There may be some wiggle room in the word "produced".  Is array A below 
produced 
by <?

    ]A =: <"1 i. 2 3
+-----+-----+
|0 1 2|3 4 5|
+-----+-----+
    3!:0 A
32

3!:0 says the type of A is 32, which the documentation for 3!:0 says is 
"boxed", 
an attribute of the noun A.  But I believe Raul would say only the atoms of A, 
for example

    < i. 3
+-----+
|0 1 2|
+-----+

are "boxed".  That atom is clearly a result of < .  Array A is a result of <"1 
not < .

Now consider

    ]B =: 'a';A
+-+-----+-----+
|a|0 1 2|3 4 5|
+-+-----+-----+

and

    (<'a'),A
+-+-----+-----+
|a|0 1 2|3 4 5|
+-+-----+-----+

and the definition of Link ; from 
http://www.jsoftware.com/help/dictionary/d330.htm :

x;y is (<x),y if y is boxed, and (<x),<y if y is open.

It appears that Link ; considers array A to be boxed.

As my original post was about Link ; I used (32 = 3!:0) as my definition of 
"boxed".


Raul Miller wrote:
> On Wed, Jul 29, 2009 at 9:26 AM, Kip Murray<[email protected]> wrote:
>> I think we have agreed that (32 = 3!:0) tests
>> whether y is boxed.
> 
> I have agreed to no such thing.
> 
>    (32 = 3!:0) y
> 
> tests whether elements of y are boxed.
> 
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to