[comp.lang.apl] APL Novice Looking fer Wisdom

keith@curry.uchicago.edu (Keith Waclena) (09/23/89)

I have just acquired I-APL/PC (from SIMTEL20) for my clone, and am
happily trying to learn some APL (partly to use it a calculator but
mainly to compare APL idioms to functional programming idioms).  Now I
have some questions about APL I was hoping the net could answer.

1. Versions / Standards of APL: briefly, what are the significant
difference between ANSI/BSI/ISO Standard APL, APL2, and ``Dictionary
APL''?

2. Is there any standard for the stored format of APL workspaces? 

3. Is there any standard for the APL character set?  

4. Is there any standard for the layout of APL characters on a
keyboard?

5. Is there any standard for an ASCII transliteration of APL
characters?  Does Dictionary APL have anything to do with this?

6. Can anyone give me an exact citation for a recent article by
Iverson on Dictionary APL?  I think it was in the *Quote Quad*, but I
need the exact volume/number or date to get it out of our library's
bindery...

7. I'm interested in math textbooks that use APL notation; does
anybody have any good ones to recommend?  Has anybody reading this
ever taken a math class that used APL as the main notation?

(When I say ``standard'' here in lowercase, I mean ``de facto standard
or otherwise''.)

Answers to these questions should help me to understand some of the
remarks I've read in the *Quote Quad* and elsewhere; if anybody could
answer with specific reference to I-APL, that would be great!

Thanks in advance,

							Keith

--
Keith WACLENA                             keith@curry.uchicago.edu
GLS / TIRA / U of Chicago                 keith%curry@uchimvs1.bitnet
1100 E.57th.St Chi IL 60637 USA           ...!uunet!curry.uchicago.edu!keith

"An assignation is the commonest form of confrontation."
	  -- Lindsey and v.d. Meulen, _Informal Introduction to Algol 68_

ljdickey@water.waterloo.edu (Lee Dickey) (09/25/89)

In article <5509@tank.uchicago.edu> keith@curry.uchicago.edu (Keith Waclena) writes:


>1. Versions / Standards of APL: briefly, what are the significant difference
> between ANSI/BSI/ISO Standard APL, APL2, and ``Dictionary APL''?

This is a quick look at differences, which are deep, in one sense, but
not so great in another.  Sadly existing implementations now use some
same symbols for different meanings, and this contributes to confusion.

Dictionary APL is a conforming extension to Std APL.  It has BOX to make
generalized arrays, it has the rank operator, and some other new operators.  

APL2 has ENCLOSE to make generalized arrays.  It has some non-conforming
extensions to Std APL, due to some syntax changes.  Functions pervade
the structure of enclosed arrays.

>2. Is there any standard for the stored format of APL workspaces? 

No.  But there are standards for workspace interchange.  WSIS0 and WSIS1.

>3. Is there any standard for the APL character set?  

Yes, but there are several answers to your question.
First, there is the minimal set, are those needed to define the
standard.  Second there are those in the APL-ASCII overlay, and
certain defined composites.  Finally, there is a pending APL
plane in a forthcoming Character Set Standard, which consists of
all known chararacters ever used in any APL character set, and
maybe some more.  See "APL Quot Quad", 14/2, Dececember 1983
for the first two.  The other exists only in the working papers
of character set standards committees.

>4. Is there any standard for the layout of APL characters on a keyboard?

A defacto standard for the layout of the IBM PC keyboard, subjected
to the APL-ASCII overlay, with composites done as overstrikes.
Every mainframe APL I know has "front end" tables that accept this
convention.

Two other popular layouts are the Union and the Unionized keyboards,
that use SHIFT and ALT keys to get the APL and composites.

>5. Is there any standard for an ASCII transliteration of APL
>characters?  Does Dictionary APL have anything to do with this?

Yes, two.  One is the APL ASCII overlay, an appendix to the Standard,
given in APL Quote Quad 14/2, Dec 83, the other is Iverson's
codes, APL Quote Quad, 18/1, Sept 87.
There are also those done by Xerox, DEC, Honeywell, but to the
best of my knowledge, no two manufacturers use the same.
There are more.

No, not directly, but Iverson's codes are in the back of this
edition of the dictionary.  (APL Quote Quad, 18/1, Sept 87).

>6. Can anyone give me an exact citation for a recent article by
>Iverson on Dictionary APL?  I think it was in the *Quote Quad*, but I
>need the exact volume/number or date to get it out of our library's
>bindery...

Volume 18, Number 1
September, 1987

>7. I'm interested in math textbooks that use APL notation; does
>anybody have any good ones to recommend?  Has anybody reading this
>ever taken a math class that used APL as the main notation?

	Sims: Algebra (?)
	Orth: (?)
	(?): Calculus in a new key
	Iverson: Tangible Math      (1989)

-- 
    L. J. Dickey, Faculty of Mathematics, University of Waterloo.
	ljdickey@water.UWaterloo.ca	ljdickey@water.BITNET
	ljdickey@water.UUCP		..!uunet!watmath!water!ljdickey
	ljdickey@water.waterloo.edu	

jaxon@uicsrd.csrd.uiuc.edu (09/26/89)

> 1. Versions / Standards of APL: briefly, what are the significant
>    difference between ANSI/BSI/ISO Standard APL, APL2, and ``Dictionary
>    APL''?

Briefly: ISO describes 'flat' arrays of uniform type elements and operators
that take only the built-in dyadic scalar functions.  The others are competing
definitions with 'nested' arrays of mixed element types where operators can
take any function and (in APL2) users can define operators.  

> 2. Is there any standard for the stored format of APL workspaces?

There is a recognized 'convention' for workspace interchange.  Many APLs have
provided this as the )IN and )OUT commands, or as an application workspace.
It is not efficient enough for everyday use as )LOAD and )SAVE, the format
does not permit random access such as is needed for )COPY.  

With this and all your standards questions, definitive answers can be found
in the ISO APL standard document itself.  ANS required the X3J10 committee to
list there all other standards with which their's interacts.  I don't know
the document number on the last public version, perhaps Lee Dickey will
respond here also (ljdickey@water.waterloo.edu knows these things).

> 3. Is there any standard for the APL character set?

Their graphic appearences are registered with ISO.  The idea of forming new
characters by overstrikes is mentioned in the standard, and a minimum set of
useable characters is prescribed in ISO APL.
  
   Their order in #AV >>> IS NOT <<<  specified by ISO! ... and is different
in every APL I ever used or wrote.  APL2 allows more than 256 characters -
good for them, someday we'll have user-defined characters, so that everybody
(not just the lucky langauge designers) can create wierd squiggles.

> 4. Is there any standard for the layout of APL characters on a keyboard?

  Indirectly yes.  There is a standard for replacing ASCII keycaps by APL
  keycaps, and there are some loosely followed standards for arranging
  ASCII keycaps on a keyboard.  There are a few recommendations for arranging
  a large set of the overstruck APL chracters on IBM PC keyboards, they
  come from IPSA and STSC, but have not so far as I know been voted on by
  X3J10.

> 5. Is there any standard for an ASCII transliteration of APL
> characters?  Does Dictionary APL have anything to do with this?

There is no standard, there are many systems.  No scheme satisfies all of 
a user's needs.  The ones of the form <escape char> <mnemonic> are the most
common, but the <escape char> is annoying and with a good choice for the
<mnemonic>s, would usually be unnecessary (choose rarely used identifiers
and then require the <escape char> when typing those identifiers, but not
when typing the <mnemonic>).

The <mnemonic>s are best expressed in some natural language, which 
makes ISO standardization very difficult.

> ... a math class that used APL as the main notation?

Many APLers take their notes down in APL, then snooze awhile until the
prof finishes writing out the next monster SIGMA i=1 to n a sub i,j ...
The exercise of translating checks your understanding, but if you misread
your prof, you don't have a fall-back.

billh@hplsla.HP.COM (Bill Harris) (10/11/89)

In response to the query about standard keyboard layouts, I see in the
April 1989 issue of "The Education Vector" (from the British APL
Association) a letter to the editor mentioning the keyboard layout adopted
by I-APL.  The writer claimed to find the layout easy to remember without
stickers, using some memory aids that Anthony Camacho ("Old World Chairman"
of I-APL) apparently gave in an earlier issue.

Does I-APL have one of the "standard" layouts mentioned earlier?  Does
anyone know what those memory aids are?

Thanks,

Bill Harris
billh%hplsla@hplabs.hp.com

jaxon@uicsrd.csrd.uiuc.edu (10/21/89)

> [an APL keyboard] layout easy to remember without stickers.
> Does anyone know what those memory aids are?

The IBM 2741 APL keyboard set the standard by rationally laying out 88 keys.
The 94 key ASCII standard complicated things a little bit.  I don't have an
APL keyboard in front of me, and I haven't seen one for two years... so lets
test the memory aids I learned!

Question Mark    Q	Modulus          M
Omega            W	Circle		 O      
Epsilon          E	Tilde		 T
Rho              R	Ceiling		 S ??
Iota             I	Floor		 D ??
Power            P	Del/Delta        G,H (which order?)
Drop             U	Take		 Y (think T, then miss the key)
Alpha            A	Semicolon	 ,
Quad             L	Colon		 .
Quote            K	\	         /
Jot              J	+/-  on same key, x/divide on next key.
Base             B	Overbar          2
Encode (Rep.)    N      Relationals in order < <= = >= > ne  above 3-8

ljdickey@water.waterloo.edu (L.J.Dickey) (10/21/89)

In article <49700016@uicsrd.csrd.uiuc.edu> jaxon@uicsrd.csrd.uiuc.edu writes:
>
>> Does anyone know what those memory aids are?

>Rho              R	Ceiling		 S ??

					 "S"eiling  ;-)

>Iota             I	Floor		 D ??

                   	     		 Down

>Drop             U	Take		 Y (think T, then miss the key)

					 the Y looks like an up arrow caught,
					 like an umbrella, in a gust of wind 

-- 
    L. J. Dickey, Faculty of Mathematics, University of Waterloo.
	ljdickey@water.UWaterloo.ca	ljdickey@water.BITNET
	ljdickey@water.UUCP		..!uunet!watmath!water!ljdickey
	ljdickey@water.waterloo.edu