Computer geeks are difficult people to drag away from their comfort zone. I
was in a PC photography SIG where the discussion got rather deep into jpeg
and pixels. They were making some invalid assumptions for jpeg. I piped up
saying that jpeg worked in the frequency domain and not really pixels. I
was quickly cut off. No opportunity to follow up. Back to the invalid
assumptions. Don't confuse people with facts.

I think that there is no five minute presentation that would work for
everyone. Managers will want to see some pretty finished applications, like
a spread sheet (grid) beside a plot of the data beside it showing how easy
it is to juggle the data to get the graphs to look like they want. Analysts
will want to see pretty reports from tools to process the data so they
don't have to do it themselves. I can't think of any quick presentation for
programmers. They already know what's best. For scientists and engineers
maybe one on more involved math. Generally, finished products with short,
easy for anyone to follow scripts would work best. Like the shorter scripts
in the plot demos. Throw in some easy calculations using one and two
dimension arrays instead of loops. But avoid tacit for such short
presentations.


On Sun, Feb 16, 2014 at 8:05 AM, Raul Miller <rauldmil...@gmail.com> wrote:

> And even in J we sometimes need to use serial thinking.
>
> There are several issues here - computers have limits and so do we. So
> depending on the problem you are working on you might have to deal with
> serial representations of solutions.
>
> And that can be OK.
>
> Meanwhile, I think we could to a lot better about supporting debugging of
> tacit code. One way to think about this involves injecting display
> mechanisms into the code itself, so that when it runs you can see what the
> data looks like or what the shape of the data looks like. But also when you
> get a lot of this going, it can be a problem to match the observable data
> back to the place where it happened.
>
> On the third hand, I often also wish for a style of array coding which is
> perhaps best matched by the phrase "functional reactive programming". Here,
> the data exists across time (some part of the array will be in the future).
> But implementing something like this would require something like a total
> rewrite of the interpreter, and a lot of care to issues that currently do
> not matter (like: if you took the shape of an array whose end exists a year
> from now you might not want to wait that long and would need to be able to
> tell that part of the code to stop working unless you wanted to lose access
> to the data that already exists in the session). This would require a lot
> of reasoning about efficiency and design of the system.
>
> And on the fourth hand (this is one context where friends might be really
> useful), K has a lot of nice structure that I really wish we had in J. We
> might never get the K tree, but I would like K's dependencies and triggers
> and the rapid and simple UI implementation that they make possible. And
> maybe that's more important than FRP (or maybe it's possible to incorporate
> ideas from the one into the other?)
>
> Anyways, it's worth keeping in mind that we have a lot of room for
> improvement. We have some great tools and some incredible capabilities, but
> we should not become too arrogant about them, because we have a long road
> ahead of us. And we are not going to get there if we are afraid to
> experiment nor if we are afraid to embrace the efforts of other people and
> other groups. (Though of course we will at times have to shed some of the
> accumulated clutter.)
>
> Thanks,
>
> --
> Raul
>
>
>
> On Sun, Feb 16, 2014 at 9:42 AM, Don Guinn <dongu...@gmail.com> wrote:
>
> > Things are difficult and useless until they are obvious. J has two
> concepts
> > that are obvious and easy to understand. Array thinking and associated
> > rank. Well, they are obvious once one understands them. Until then they
> are
> > extremely difficult and serial thinking is so much easier because that is
> > what we have been taught from the first grade on. Then throw in a bunch
> of
> > primitives built around mathematical concepts, many of which have been
> > around for generations, but only studied by few people. For these people
> J
> > is great because it does much of the grunt work allowing them to get on
> > with the problem at hand. Just learn the notation.
> >
> > But the rest of us have to struggle not only with the J notation, but the
> > underlying concepts. Concepts that took a lot of time to fully
> understand.
> > Maybe years. Then throw in the serial approach in programming computers.
> > Necessary because today's computers are serial. Tell an experienced
> > programmer to ignore what he knows and spent years learning because APL,
> J
> > and other array languages do things differently and it is a turn-off.
> >
> > Math for most people is add, subtract, multiply and divide. Talk about
> > anything beyond that and most people stop listening. Spread sheets are
> OK.
> > Just don't call it an array. Matrix is cool and geeky.
> >
> > J is more than a notation. It is a "Tool for Thought". Most people want a
> > tool that is "easy" so they can get on with their problem without
> thinking.
> > Just "turn the crank".
> >
> > But then, I am rambling.
> >
> > Perhaps showing J to do spread sheet type calculations is a way to show J
> > in a way that would appeal in five minutes.
> >
> >
> > On Sun, Feb 16, 2014 at 5:56 AM, Bo Jacoby <bojac...@yahoo.dk> wrote:
> >
> > > Learning a little J is fast, and is worth while. I like this example
> > > exxplaining addition.
> > >    7#1
> > > 1 1 1 1 1 1 1
> > >    14#1
> > > 1 1 1 1 1 1 1 1 1 1 1 1 1 1
> > >    (7#1),(14#1)
> > > 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
> > >    #(7#1),(14#1)
> > > 21
> > >    7+14
> > > 21
> > >
> > >
> > > Learning more J is still worth while.
> > >
> > >
> > >
> > >
> > > Den 13:37 søndag den 16. februar 2014 skrev R.E. Boss <
> > r.e.b...@planet.nl
> > > >:
> > >
> > > From
> >
> http://www.grammarphobia.com/blog/2009/07/steep-learning-curves.html'Thus,
> > a “steep learning curve” in common parlance refers to the difficulty
> > > of learning something.'
> > > >
> > > >
> > > >R.E. Boss
> > > >
> > > >(Add your info to
> http://www.jsoftware.com/jwiki/Community/Demographics)
> > > >
> > > >> -----Original Message-----
> > > >> From: programming-boun...@forums.jsoftware.com [mailto:programming-
> > > >> boun...@forums.jsoftware.com] On Behalf Of Robert Knight
> > > >> Sent: zondag 16 februari 2014 12:59
> > > >> To: programm...@jsoftware.com
> > > >> Subject: Re: [Jprogramming] J in 5 minutes
> > > >>
> > > >> If J's learning curve were indeed "steep," then learning it in "5
> > > minutes"
> > > >> would be  feasible.  Unfortunately, it is quite shallow.
> > > >>
> > > >> *​*​
> > > >> -Robert Knight
> > > >>
> > > >> On Sun, Feb 16, 2014 at 6:21 AM, R.E. Boss <r.e.b...@planet.nl>
> > wrote:
> > > >>
> > > >> Let me be the odd one out and disagree with those who want J in 5
> > > minutes
> > > >> or
> > > >> > presenting different types of nouns in different colors or
> formats.
> > > >> >
> > > >> > IMHO the steep learning curve of J is very useful (and, who knows,
> > > >> > intended), it selects the right kind of people. That's how
> evolution
> > > works.
> > > >> > To paraphrase Mark Twain: "If we try to convince stupid people,
> they
> > > will
> > > >> > drag us down to their level and beat us by experience".
> > > >> > So let's not try to make learning J easier, we will end up
> > > simplifying the
> > > >> > language to a level we don't want.
> > > >> >
> > > >> > And for people who want to teach J to kids, read
> > > >> >
> > > >> >
> http://www.scientificamerican.com/article/why-gaming-could-be-the-
> > > >> future-of-
> > > >> > education/<http://www.scientificamerican.com/article/why-gaming-
> > > >> could-be-the-future-of-education/>
> > > >> >
> > > >> >
> > > >> > R.E. Boss
> > > >> >
> > > >> > (Add your info to
> > > >> http://www.jsoftware.com/jwiki/Community/Demographics )
> > > >> >
> > > >>
> ----------------------------------------------------------------------
> > > >> For information about J forums see
> > http://www.jsoftware.com/forums.htm
> > > >
> > > >
> > > >----------------------------------------------------------------------
> > > >For information about J forums see
> http://www.jsoftware.com/forums.htm
> > > >
> > > >
> > > ----------------------------------------------------------------------
> > > For information about J forums see http://www.jsoftware.com/forums.htm
> > >
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> ----------------------------------------------------------------------
> 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