Well said Paul. I learned APL in 1968 using a 1050
terminal WITHOUT an APL type element and two key
strokes to enter a line of input... One of the strange
results of things like the 1050 is carried forward
into MS Windows text files having <CR><LF> at the
end of each line, even though they are entered by
a single keystroke - although the two characters
on the 1050 were a return followed by EOT. One fun
aspect was that you could send a multiline "instant
message" to the operator )opr or another user. Some
of the kids doing IM on their cell phones might be
surprised to know that it started more than 40 years
ago...

As regards APL characters. Seeing uppercase M overstruck
by uppercase O doesn't really generate a very pretty
icon for rotate, and grade up (H over M) was better than
grade down (G over M)... When I did get a real APL type
element, I too quickly fell in love with the graphics.

I have spent many hours rendering the APL glyphs
and presented a paper on the subject, "APL Pi" at
APL81 in San Francisco. But just looking at the APL
page in wikipedia shows that there isn't much agreement
on what the symbols are. Even more annoying to me is
that even in this enlightened age of high tech, messages
with APL in them typically render like a ransom note
where each character seems to have been clipped from
a source to obscure its origins -- and that is if you
are lucky enough to get output that is even readable!

Someone said that when they see j's $ they think rho,
but for the dyadic case of $ that isn't very helpful.
It took me some time to be comfortable with +/ which
is equally beautiful in APL and j, except with quite
different results... I believe that much of the
distress expressed by APL users when they encounter
j is that they have to change the way they think,
just as a BASIC or COBOL has to change they way they
think to use APL or j (or LISP for that matter).

People resist changing the way they think - or even
admitting that it might be a good thing to do...

- joey


At 11:27  -0500 2007/12/18, Paul Gauthier wrote:
I had several discussions with different people mostly old APL programmers
about the APL symbols being so missed...

With today's technology one can easily (with motivation) build a unicode
front end to J and customize it so that graphic characters of his/her choice
would be used to write underlining J code. This graphic layer would probably
be different for each person so somewhere along the lines one needs a dog's
breakfast to hookup to his/her flavor of the day of special inspiring
graphics.

These criticisms are really showing differences in taste rather then
substance. For instance, way back in time before J even existed, I was
saying to a friend that I would not care to use series of colored spots to
program in APL because what was important was what I could do with the
underlying concepts concealed behind the APL symbols. Well, with J it is so
much more so that I challenge anyone to come up with a nice set of graphics
that would satisfies everybody. The dog's breakfast could very well be the
basic ingredients to all the worlds meals recipes rather than a meal... ;O)

Maybe someday, someone will come up with a graphic layer that would generate
J code and it will catch up as a new fad, but for now the basic ingredients
are sufficient to see the wonders one can cook.

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of BobGraf
Sent: Tuesday, December 18, 2007 11:00 AM
To: [email protected]
Subject: Re: [Jchat] J readability


Please see my comment in the Chat thread "Readible J" by James C. Field.
Thank you,
Bob from Boynton Beach, FL


--
View this message in context:
http://www.nabble.com/J-readability-tp14345905s24193p14399835.html
Sent from the J Chat mailing list archive at Nabble.com.

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

Reply via email to