I'm starting a J Study Group this Friday morning in my department of
analysts, and I'm looking for tips on ways to make programming in J both
interesting and understandable rapidly.  We've agreed to dedicate 30
minutes once a week to "guided exploration"; I hope that at least some of
the group will decide to learn more on their own because they find J
worthwhile in their work.

In the past, I've advised people to start with the Primer.  Now there are
also useful labs, and there's Learning J (there's the great JfC, too, but
I think that's a bit much for this group at the start).  I'm curious if
any of you have any suggestions on what has worked well to get people into
J quickly in ways that make it stick.

I'm thinking I'll say a couple of words and then point them to a lab or
the Primer or LJ and set them loose, answering questions that may come up
and asking them questions to encourage their experimentation.

Do any of you have suggestions?  Do you have experience teaching this
stuff somewhere that might be useful to me in planning what we do?  Are
there NYCJUG experiences worth considering?   Feel free to be as explicit
as "Start with the Primer through Checkpoint A (or whatever) and then do
the "A Taste of J 1 and 2" labs, followed by ...."  I'm hoping to stay
with the Primer, LJ, or labs instead of creating something on my own.  I
know there are many other good sources of information on various people's
Web sites, too.

I conjecture that one of the hangups for newcomers is getting data into
and out of J.  It's easy to teach the avg fork, but people don't want to
type in the data they need to have to make use of various verbs -- they
want to read a spreadsheet or database (or text file or HTML table
perhaps).

To that end, I'm considering doing two things, and I'm curious in your
feedback.  For one, I'm thinking of deploying JOD on a shared drive, so
that we can put J words there that all might find useful.  (I presume that
works; I've never used it because I used Linux in my own business.)

Second, I'm thinking of including a few typical verbs to read a
spreadsheet using tara or taraxml and grab the data.  Yes, one can use
tara directly, but it might make life easier for newcomers to have a verb
that will read a worksheet, extract a rectangular numeric section between
two cells, and convert that to an array (or to extract a partial row or
column of string data and put it into a different J array).

Any thoughts on that?  I want to keep it simple, but JOD seemed like a
useful way to go.

Thanks,

Bill



----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to