Forest list archive: msg00064

[Prev][Next][Index][Thread]

Re: APL



Rolfe,

>Is there anyone out there who knows the current status of the APL computer
>language?
>

APL is alive and well.  Perhaps not as well-known as it should be, probably
due in part to IBM keeping it for a long time for their own in-house use
instead of commercializing it.  In fact, at least until recently, to get
IBM's APL2 PC the manual of the demo version instructed to call some obscure
telephone number used for ordering IBM supplies, ask for product number XXX,
and "do not mention APL or software"!

But there are now a number of good implementations.  And clusters of
enthusiastic users, although for some reason mostly outside the US.  Get the
APL FAQ (Frequently Asked Questions) through FTP at watserv1.uwaterloo.ca,
under /languages/apl.  Also software and other info there; with a www
browser point it to ftp://watserv1.uwaterloo.ca/languages/apl/Welcome.http

You can also try the computer.languages.apl News user group, and several www
pages mentioned in the FAQ.

There have been extensions to the language implemented on some systems, for
example APL2 from IBM and APL-III from STSC/Magnugistics.  And Iverson got a
new dialect called "J", using plain ASCII (would have been great years ago,
not important now with graphic cards and configurable keyboards, IMHO).  The
extensions are cute, and often useful, but my experience until now is that
the added complexity/obscurity may not be worth it; can't always predict the
results and are surprised too often.

>What platforms does it run on?  How much does a compiler cost for personal
>computers -- if it is available for PC?
>

There are versions for almost anything.  I am currently using IBM's APL2 PC
(around US$500 when I got it).  Can "package" (compile) applications into
royalty-free stand-alone executables, and includes a 32 bit version if you
need gigantic workspaces (up to the size of your extended memory).  A bit
fragile, at least my version crashes more frequently than I would like.  The
packaging feature made me switch from the STSC (later Magnugistic) APL*PLUS
PC that I used before (versions 6 and 9).  I also have Magnugistic's APL-III
for Windows.  Not cheap, but allows to develop stand-alone Windows
applications, and the interface might be more palatable to those that can
tolerate (even enjoy?) Windows.  I have not really got into it yet.  I heard
that Magnugistics has just sold their PC APL business to somebody else.

FREE versions:  A demo version of IBM's APL2, called TryAPL2.  Crippled on
the file input/output and interfacing department, but may be useful for
teaching.  I have seen a PD statistical package that manages to get around
the file I/O limitations.  Then there is I-APL, available for many micros,
including the PC.  Slow and with only 32K workspaces, but quite usable and
you can't beat the price.  I used it succesfully (I think) in a Forest
Mensuration course.  Last minute news:  a few weeks ago Magnugistic released
a public domain "APL Special Edition" (APLSE), based on their APL*PLUS
Version 10.  I just had a play with it, and it looks great.  You would need
some documentation from their (formerly?) commercial APL*PLUS to use it to
full advantage, though.

All these free interpreters are available from waterloo.  For APLSE look
into .../apl/apl-plus/aplse.html

>Are books still printed about APL?
>

A few, see the FAQ.

>What is the best strategy for converting a program written in APL to a more
>widely used  language (FORTRAN, TrueBasic, Pascal, C)?
>

Why would you want to do that?  "Packaging" in APL2 or APL-III makes the
programs available to the unfaithful.  Of course, APL is great for
prototyping, and if you are a masoquist can later translate by hand into the
gruesom detail required by FORTRAN, C, Assembler, etc. (that is how IBM
developed their System/360, by the way).

>Thanks,
>

De rien, and happy APLing.

Prof. Oscar Garcia   -   ogarcia@nancy.engref.fr
Unite Dynamique des Systemes Forestiers
Ecole Nationale du Genie Rural, des Eaux et des Forets (ENGREF)
14 rue Girardet, 54042 NANCY Cedex, FRANCE  -  Fax: +33 83 30 22 54




[Metla] [Main Index] [Thread Index]

Mail converted by MHonArc 1.1.0