First off foundation looks pretty. And it probably has some
interesting uses for education. I'm going to send the link to my
sister (an elementary school teacher) and see her opinion of it. (My
guess is that she would defer to her school's administration on the
subject of teaching materials, but she may offer additional insights.)

Anyways, it seems to me that, in the context of J (and APL) that we
would be well served with a variety of treatments of practical and
academic subjects. Some people (Cliff Reiter comes to mind) are
setting really good examples here.

We could also use a wider variety of applications that people find
useful. That would do good things for the community, as would works of
art and media built using J's capabilities.

A related subject, and an important one, is how to deal with the
frustrations which inevitably arise when breaking new ground. That
takes considerable persistence and vision.

And, also, I think the open source J engine needs some cleanup work.
I've spent some time in there myself, but I've some motivation
problems (I know I'm going to mess things up before I understand them
well enough to make them better, and I have to give myself permission
to make all those mistakes).

... anyways... some of that was actually unrelated to Don Watson's
words (though very related to the gist of his message).

That said, I should also mention that there are some real advantages
to text (over visual display). Visual display can be a great starting
point, and a good tool, but it's not the only useful perspective. I am
also quite hesitant about accepting the claim that "Scientists and
engineers don't program because it is too much trouble." While I have
no problem accepting that this is true of some, I am also aware of
numerous counter-examples. I also occasionally stumble across someone
with an obvious background in science (or engineering) who has been
working with J for years.

I think a truer statement might be "Scientists and Engineers typically
do not polish their programming efforts for education and broad
general use". And that is true. But it's equally true that "Computer
programmers typically do not polish their programming efforts for
education and broad general use". Some do, of course, but many are
employed in a variety of tasks where their contribution is a small
part of a whole. Some are bound by contract or state restrictions.
Meanwhile, I think it's also fair to say that the current large body
of open source code is the result of contributions by various
professionals who have needed tools and programs which their company
was not prepared to provide.

J and APL are relative newcomers, here. We've almost no academic
texts, though we have some. We are starting to build open source
software, but we need more work there. And I guess we sort of also
have to understand that all computing environments and languages have
severe problems and shortcomings and to some degree are overhyped.
That probably includes J (though personally I try very hard to avoid
overhyping both J and everything else I'm involved in). And improving
things involves work and time.

Anyways, I guess I am in a sort of half-agreement here.

Thanks,

-- 
Raul


On Tue, May 13, 2014 at 9:51 PM, chris burke <[email protected]> wrote:
> Don Watson asked me to post this on his behalf, and it should be of
> interest to the chat forum:
>
> --------------
>
> There were two characteristics of APL in the 1960s that attracted people.
> It was powerful and it was easy to teach - in fact I wrote a self-teaching
> package that enabled APL to teach people the beginnings of how to use
> itself - it got them started. My interest was in using it to teach
> Mathematics and I was attracted by a 2 day conference Ken Iverson gave in
> 1968 on “The role of the computer in teaching”. When he moved on to J, a
> powerful nature was his priority and Mathematics teaching was left behind.
>
> The use of both Mathematics and Computing by professionals other than
> mathematicians and computer scientists is decreasing in favour of "black
> boxes". Today, few professionals other than Computer Scientists model their
> work by programming any more – as they did when APL was first available.
> There was an interesting blog recently in the ACM that found that they
> would model if it was in their own language. I interpret that as meaning
> that they want something in the manner that spreadsheets provided when they
> spoke the language of managers. Scientists and engineers don't program
> because it is too much trouble.
>
> The language of scientists and engineers is Mathematics. There is a huge
> potential market for a computer language that uses mathematical notation
> where you use tacit expressions. APL was born in the 1960s, when input and
> output was based on one-dimensional typewriters. Mathematics is two
> dimensional. This meant that Iverson was forced to compress a tree
> structure into an artificial train structure. However, the Macintosh was
> mass produced with a two-dimensional screen 30 years ago. I wish Iverson
> had then begun allowing APL notation to become a mathematical tree
> structure and removed the artificiality. Unfortunately, he stated in 1996
> that Mathematics couldn't be processed by a computer - which isn't true.
>
> A new software teaching tool bases Mathematics on functions, removing the
> need for abstract operators in elementary school. It is based upon the use
> of Mathematics expressions as a computer language - which was achieved with
> a single change to Mathematics notation - by adding a different kind of
> variable. Basing Mathematics on functions, where the action is, could unite
> Mathematics and Computing in a common front that could bring about the
> return of both as modelling methods.
>
> I am almost 80 years of age and unable to contribute anything to the needs
> of professionals. I have enough work to do expanding my “Foundation” system
> to meet the needs of elementary and high schools to grade 12. J could
> certainly be adapted for this purpose – all that is needed is to change the
> front end to a tree structure. If J doesn't do it somebody else will. There
> are other issues, like relating documentation to non-mathematicians and
> returning to an elegant control structure as APL originally provided, but
> that is simply achievable.
>
> A paper with embedded videos at the website
> www.FoundationNotation.comdemonstrates how Foundation works and
> explains why it works. So far
> Foundation has been implemented as a Macintosh app and an Ipad app.
> ----------------------------------------------------------------------
> 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