[Jprogramming] How m"n shoulda been defined

[Raul Miller]

Conciseness?

I guess it's debatable whether constants with non-infinite rank are
more or less common than gerunds which need to conform to the shape of
their (eventual) verb argument.

Perhaps it would be worthwhile collecting useful examples of each? I
know I've used low-rank constants within the last few years, but at
the moment I do not remember where...

Thanks,

-- 
Raul


On Sun, Aug 9, 2015 at 2:03 PM, Jose Mario Quintana
<jose.mario.quintana at gmail.com> wrote:
> A constant verb (with rank _) can be produced easily via &[.  For example,
>
>    (1 2 3 &[) _
> 1 2 3
>
> and it could be followed by a suitable rank (") form if necessary.  Am I
> missing something?
>
>
> On Sun, Aug 9, 2015 at 9:20 AM, Henry Rich <HenryHRich at nc.rr.com> wrote:
>
>> All the partitioning modifiers (\ /. \. ;. etc) allow u to be a gerund
>> that is applied cyclically to the partitions.  Thus, +`-/. applies + and -
>> alternately.
>>
>> All the modifiers, that is, except one.  The simplest one.  One that
>> perhaps wasn't thought of as partitioning, though it clearly does.  It
>> partitions the y argument into cells.
>>
>> If m"n had been defined consistently with the other partitioning
>> modifiers, we would be able to write +`-"_1 to have different verbs applied
>> to items.  This would have saved dozens of emails over the years searching
>> for a good way to perform this often-needed operation.
>>
>> The constant-verb would have to move somewhere else: m&n, m at n, and many
>> other places are free.
>>
>> For the next language.
>>
>> Henry Rich
>> ----------------------------------------------------------------------
>> For information about J forums see http://www.jsoftware.com/forums.htm
>>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm