IMHO, this very discussion is a striking example of problems
using graphic symbols and good reason to get past them.
WADR, if the user is kind enough to neatly line up all their
their NB. (which by the way is defined in most dictionaries -
OK, I concede that dictionaries use N.B. ) and tinted green ...
OH! I think I detect Oleg has tongue in cheek ....
I think the original idea of naming it {lamp} may have resulted
from less than wonderful rendering on a Selectric typewriter,
making the result look something like the filament in an Edison
bulb (and that is a nice mnemonic for illumination) BUT can
anyone render {lamp} on a screen conveying that quaintness.
Still, seems to me there should be more substance in what is
envisioned for "symbolic j" -- Does anyone have comment on
global versus local assignment? (maybe global is bold arrow?)
or the fact that in most APL fonts {circle} is very easily
confused with o ? Surely no one is suggesting restricting
names to upper case only... Any candidate symbols for {:
and }: -- maybe up/down arrow "overstruck" with - ?
That brings up another issue. The extra symbols in APL (e.g.
{lamp} were originally created by overstrikes - does anyone
know if the unicode points for APL include "all possible"
combinations? Or might we have, as I first saw lament of in
1971 on Lou Solheim's wall regarding , overstruck with -
some symbols "Discriminated against because of size."
A final comment (I promise to shut up about all this!) I
personally had a fairly difficult time moving from APL to j
(back in 1990) with the biggest hurdles being the changes in
things like catenation and reshape. Having the same symbol ,
didn't help at all and if rho had been used for reshape, I
think it would have been less than helpful if only by providing
mental stumbling blocks. On the other hand, as Roger points
out, approximately 0.0% of all programmers know of APL.
I think the much stronger reaction of the general population
would be to suggest use of Sigma and Pi for +/ and */ and
in doing that would miss one of the main advantages of j
(and APL). Similarly with inner product etc.
- joey
At 14:24 -0800 2007/02/07, Oleg Kobchenko wrote:
They will know immediately, because the comments
will be nicely lined up at the right of the code,
and in the session manager they will be nicely
syntax highlighted with tilted green.
--- Roger Hui <[EMAIL PROTECTED]> wrote:
How is one supposed to know that the APL {lamp}
symbol is a lamp?
I estimate that 100.0% of all programmers know
nothing about APL. For such people, NB. is at
least as readable/understandable/mnemonic as {lamp},
without the problems associated with entering and
displaying {lamp}.
----- Original Message -----
From: Markus Triska <[EMAIL PROTECTED]>
Date: Wednesday, February 7, 2007 1:59 pm
Subject: Re: [Jgeneral] Displaying J with Unicode Special Symbols
> Roger Hui <[EMAIL PROTECTED]> writes:
>
> > How is the APL {lamp} symbol an improvement over
> > NB. as far as being more understandable/readable?
>
> It's mnemonic: Comments are supposed to be illuminating, no?
>
> All the best,
> > Markus Triska
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
