Article 3773 (25 left - 1 in thread) of comp.lang.apl (48 lines):

Newsgroups: comp.lang.apl      From: Raul Miller 

Subject: APL,  0 K?            Date: 20 Sep 96 19:01:31 GMT

Reply-To: rdm@TAD.MICRO.UMN.EDU



Steve Apter:

   I hate this tool metaphor.  In the first place, programming

   languages aren't tools, unless you mean to broaden the meaning of

   that term to the point where it means very little.  Programming

   languages are, if you please, *languages*.  If a language is a

   tool, then it is a tool of a very special sort; viz., a tool of

   thought.  If that is so, then if we need a multiplicity of such

   tools, it must be because our thoughts about computing do not

   possess unity, are not all of a piece.  For example, thoughts about

   string searching must differ radically from thoughts about sorting.

   I don't believe this; but, even if it were true, the conclusion

   that we should use different languages to express those thoughts

   would not follow, since e.g. it would thereby become more difficult

   both to discover and to express what searching and sorting had in

   common.



There is quite a bit of justice to what you say.  But the devil is in

the implementation.



For example, what APL product would you recommend to produce something

that addresses the economic niche of Adobe's recent PrintGear

(mass-produced vendor-independent high-speed high-quality printing)?



Alternatively, what APL product would you recommend for

**implementing** K?  Compare this to what is actually being used?



Fundamentally, the machines we're working with are rather quirkey

state machines.  A language which reflects those quirks, and exposes

this state-machine nature has a high learning curve, and can be

difficult to design with, but allows for very economical use of the

machine.  [Languages I'm familiar with which expose different aspects

of the underlying machine are assembly (different on different

machines), forth, and C.]



APL is good for problem sets where machine economies aren't the issue.

[Though I'm sure Bob Armstrong has some good insights on low-level APL

implementations.]

                              * * * * *

By the way, is APL a language, or maybe a class of jargon?  How would

rate it compared to, say, English?  Musical notation?  Topological

maps?

In my opinion, APL is excellent for solving computational problems.

--     Raul



Article 3780 (24 left - 0 in thread) of comp.lang.apl (30 lines):

Newsgroups: comp.lang.apl      From: Steve Apter 

Subject: APL,  0 K?            Date: 20 Sep 1996 21:50:39 GMT



Raul Miller  wrote:

>By the way, is APL a language, or maybe a class of jargon?  How would

>rate it compared to, say, English?  Musical notation?  Topological

>maps?



APL is a programming language.  Programming languages are something

(relatively) new under the sun, the way musical notation, topographical

maps, and for that matter natural language were at one time.  (See

Julian Jaynes, _The Evolution of Consciousness and the Breakdown of

the Bicameral Mind_, Princeton UP, for compelling -- in my view, over-

whelming -- arguments that consciousness arose just over the horizon

of historical time).



A programming language serves three purposes (two, on some counts):

        1.  a set of instruction codes for a machine

        2.  a language in which to formulate and understand algorithms

        3.  a language in which to convey ideas about algorithms to others



It's by virtue of properties 2 and 3 that we are justified in calling

coding systems like APL languages.  We think in them, through them,

with them.  (Telephone numbers satisfy property 1, but no one would

think of calling 914 555 1212 a statement (no jokes about it being a

query).)

>-- >Raul

 My reply to above posts  /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\ /\

  ++++++++++++++++++++++++++++: SUN.SEP,960929 :++++++++++++++++++++++++++++



 Steve ( & Raul ) :                       | DDT   |>| SUN.SEP,960929,15:16



 In the beginning was the word .



  Natural languages are Universal . Thus English , chosen by historical

  [ 19:32| evolutionary ] accident , can become the WorldWide language .

  Different subject matters are just different argots within the entire word

  web .

  A core goal of the  CoSy/Notecomputer  is to provide a open universal

  linguistic environment which is all one must learn ( because it is all

  there is ) to make their machine do anything it can . When Someone extends

  the vocabulary in their own direction , they can post it on CoSy.com .



  Windows95 ( & NT ? ) has separated the ' User ' from the language of the

  machine so completely that you HAVE to chose a language : C , VisualBasic ,

  Delphi , .. APL , .. ; to access its vocabulary .

   Rad thing abt Web is that if you want be player , it say

     <B> Here is Language </B>  .



 Programming challenge :

 As time goes on , I see the essence of computing more and more implementing

 a Turing Tape processor .

 Since HyperTextMarkupLanguage is now the Lingua Franca , I want to embed

 CoSy/APL expressions in my page descriptions to do things like update the

  ' Modified : ' time stamp I try to maintain in each page .

 Here`s a problem I need to do in some manner for myself which obviously

 applies to any language .

  Execute a string  " <HTML> . . . "

   ` PARSE occurances of HTML comments delimited by " <!-- " and " -->' "

   ` DO  string between returning the , perhaps modified , HTML string .

  In particular , Construct a ` DO which replaces the text after the comment

  up to some third delimiter after the end of the comment with the result of

  the execution of the string .



 I`m of the camp which to quote JeffMultach :

   never accepted J. Too hard on my eyes. Too confusing. They gave in on the

   visual aspects of APL .



 Whatever symbol set one`s language uses , it is very useful to have either a

 running phonetic transliteration of the tokens sequences , or at least , a

 key to their pronunciation and meaning . For instance , in the specification

 above , while just suggestive , not implemented , the symbols used " and `

 follow the FORTH convention that they are delimited by whitespace and mean :

       "       Quote   Return ( stack ) the string between - not including

                       first and last SPaces

       `       Tick    Return ( stack ) the dictionary address of the

                       following token ; used here allogorically .



 PreHistoric consciousness :

 I suggest that anyone who believes language is a precursor of consciousness

 watch a dog in REM sleep .



 To me the only explanations which count are mathematical in character .

 I think FrancisCrick :



        \The Astonishing Hypothesis : The Scientific Search for the Soul

        ; by Francis Crick ; Paperback List: $14.00 -- www.Amazon.com Price:

       $12.60 -- You Save: $1.40(10%) / Published by Touchstone Books

       Publication date: July 1, 1995 / ISBN: 0684801582



 Is much closer to approaching the problem in a meaningful manner .

 See  http://cosy.com/views/psykofiz.txt  for a summary of my

 understandings .

                                  -- BobA --

          BobArmstrong                                     CoherentSystems

   212-285-1864 ø X:-732-0244                              http://CoSy.com

  42 PeckSlip 4B | NYC NY 10038                              bob@cosy.com

-- Posted on end of thread : 18:56 EDT 29-Sep-96 /\ /\ /\ /\ /\ /\ /\ /\ /\