> From: Don Watson
> J is, quite rightly, directed towards the use of tacit programming.
> It does not conceive of a different audience who can get by for
> most of their work with explicit programming.
I don't agree with this - the point has been made several times that there are
long-time J users that use explicit code almost exclusively.
> The one thing J lacks for my purposes is a simple, one
> line way of creating an explicit function in
> exactly the same way that a tacit function is created.
I agree that this seems to be the thing you don't want to let go of.
> This would make a considerable difference to what I am trying to
> initiate. To demonstrate, I have written the documentation as if
> such a facility already existed. Its implementation would enable
> me to continue with the documentation.
I disagree. So far there is no evidence that not being able to assign an
explicit phrase without "trappings of computer science" presents a significant
barrier to using J as a tool for exploring numbers and mathematics. Actually I
think that your treatment of verb assignment using a hypothetical form of
explicit J is more contrived and awkward to explain than if you just introduced
one of the methods for assigning an explicit sentence to a name.
On the document itself:
I think that it is better to use full words not abbreviations for verb names.
If you are going to make them learn cryptic abbreviations, you may as well just
use the standard symbols.
I think you should choose a simpler formula than Standard Deviation to start
showing how to assign verbs to names.
I think the best thing to convince teachers that this is something useful is to
show them how J can be used to help illustrate/illuminate mathematical concepts
to students. In my opinion, the document currently seems is show off how to use
the functions of a fancy calculator, it doesn't actually show how using J will
help the student/teacher learn to think about Mathematics in a better/different
way.
In other words I think that your document would be best aimed at students, and
should be primarily about illustrating mathematical concepts, not how to do
stuff in J. That should "just happen" along the way.
In my opinion, the first step towards your goal should be working with some
students to see what works with them, and then use that evidence to help you
convince teachers that this is something worthwhile. You may have to get
alongside a supportive teacher to make this happen, but it sounds like you have
a number of suitable contacts there.
> In addition, students are not
> writing J themselves, but they need to be convinced that the J language
> they see is equivalent to the mathematical formula it models.
Hmmm... it sounds to me that the learning paradigm you are thinking of is
missing the real potential of introducing a tool like J. It should be more
about "letting students explore" than "showing students how to do", so students
should certainly be writing J phrases and they should be convincing themselves
of relationships between numbers and operations.
> I accept that I have no chance of persuading anyone to allow the
> replacement of two character symbols with one character.
Funny, I thought you had made a better case and had a better reception for this
idea than you had for S.
Some resources that you might find interesting:
* J Tutorial and Statistical Package, 2003, 67 pp.
At http://www.cs.ualberta.ca/~smillie/Jpage/Jpage.html
* The Arithmetic Lab (Studio|Labs...|Arithmetic)
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
