apl/windows 
Author Message
 apl/windows


Quote:
> I continue to disagree vehemently with Gosi and others who seem to
> be implictly arguing that J has rendered the language of the 1979
> standard obsolete and useless to the point where it no longer needs
>  a distinct name and identity.

I think you are jumping to conclusions.

I have never said and never intended to imply that the old APL standard
was obsolete and useless.

What my intentions have been is to say that I am happy that the
APL operations are working in J without the need for special APL characters.

The fact that J has also fixed some other problems with old APL is
a bonus.

I do not think that the people agreeing on the standard for old APL
ever intended the standard to mean that things should be frozen in time
and no evolution should take place.

As revolutionary as J may seem then it is basically an evolution from
the old APL standard. I am pretty sure that now that J has proven to
be such an important evolutionary step and it has revived APL as to
becoming much more useful and practical that the standards community
will sit down again and revise those old APL standards.

You are welcome to stand in front of the train and see if you can stop it
but I do not think the evolutionary chain reaction of changes will stop
and remain on the 1979 level forever.

/Gosi



Mon, 21 Jun 1999 03:00:00 GMT  
 apl/windows


Quote:


> > I continue to disagree vehemently with Gosi and others who seem to
> > be implictly arguing that J has rendered the language of the 1979
> > standard obsolete and useless to the point where it no longer needs
> >  a distinct name and identity.

> I think you are jumping to conclusions.

> I have never said and never intended to imply that the old APL standard
> was obsolete and useless.

We're not talking here about the old APL standard, we're talking about a
programming language in its present form, the language that I call APL,
and for which Gosi has no name.

Quote:
> What my intentions have been is to say that I am happy that the
> APL operations are working in J without the need for special APL characters.

As am I.  I am also happy that many of the APL operations are now
available in fortran, without the need for special APL characters.  I
simply do not, however, believe that this makes it appropriate to
misrepresent Fortran as "an improved version of APL".

Quote:
> The fact that J has also fixed some other problems with old APL is
> a bonus.

It is a very good thing that J has fixed numerous problems with old APL.
It is also a very good thing that new APL has fixed numerous problems
with old APL.  But that just doesn't make J and new APL the same thing.

Quote:
> I do not think that the people agreeing on the standard for old APL
> ever intended the standard to mean that things should be frozen in time
> and no evolution should take place.

Nor has anyone suggested that they did, or should have.

Quote:
> As revolutionary as J may seem then it is basically an evolution from
> the old APL standard. I am pretty sure that now that J has proven to
> be such an important evolutionary step and it has revived APL as to
> becoming much more useful and practical that the standards community
> will sit down again and revise those old APL standards.

But if they do, they will undoubtedly produce a new standard for APL.  
They will not replace it with a standard for J.  Perhaps they will, in
time, produce a standard for J, but, if so, they will call it a new
standard for J, not a revised standard for APL.

Quote:
> You are welcome to stand in front of the train and see if you can stop it
> but I do not think the evolutionary chain reaction of changes will stop
> and remain on the 1979 level forever.

I have no interest in stopping the evolution of APL.  The only way to
stop the evolution of APL would be to convince those who would bring
that evolution about to work to improve some other language instead.  To
do that, one would need to convince them to work on J, or some other
language other than APL.

If the "J is APL" crowd were to succeed in convincing the world at large
that J is in fact the current state-of-the-art APL, only then would the
evolution of APL cease.

And if I were to stand in front of that train, it would undoubtedly
crush me to a pulp in 10 seconds.  That would not, however, justify the
claim that that (or any) train was an improved horse-cart.

Eric Landau, APL Solutions, Inc.
"Sacred cows make the tastiest hamburger" - Abbie Hoffman



Mon, 21 Jun 1999 03:00:00 GMT  
 apl/windows

Keywords:

<snip>
<As revolutionary as J may seem then it is basically an evolution from
<the old APL standard. I am pretty sure that now that J has proven to
<be such an important evolutionary step and it has revived APL as to
<becoming much more useful and practical that the standards community
<will sit down again and revise those old APL standards.
<
<You are welcome to stand in front of the train and see if you can stop it
<but I do not think the evolutionary chain reaction of changes will stop
<and remain on the 1979 level forever.

I have no ojection to (an even applaud) the evolution of APL into J.  
What _I_ object to is the notion that the term APL should be applied to
anything but fairly direct implementations of the original language.  
We have all evolved from lizards, and I would object just as strongly if
someone insisted on labeling me a lizard.  I think the lizards would
agree.

Doug White



Mon, 21 Jun 1999 03:00:00 GMT  
 apl/windows

As an occasional visitor to this usenet group, I chanced upon
this line in one of the responses.

Quote:
>> As am I.  I am also happy that many of the APL operations are now
>> available in Fortran, without the need for special APL characters.  I
>> simply do not, however, believe that this makes it appropriate to
>> misrepresent Fortran as "an improved version of APL".

Which fortran compilers offer APL operators? Where can I get info
on these features? Will they do matrix multiplications and inverses?
Without the need to define work matrices? etc. etc.
Eagerly awaiting any replies.
Regards.                RAF


Wed, 23 Jun 1999 03:00:00 GMT  
 apl/windows


Quote:

>>> As am I.  I am also happy that many of the APL operations are now
>>> available in Fortran, without the need for special APL characters.  I
>>> simply do not, however, believe that this makes it appropriate to
>>> misrepresent Fortran as "an improved version of APL".

True. In fact, they did, IMHO, a crummy job of picking up the
APL primitives. Check my SIGPLAN Notices article (Feb 1991)
for a critique of the FORTRAN 90 work on array primitives.

Quote:
>Which fortran compilers offer APL operators? Where can I get info

I can't help you with specific compilers. However, look for
"Fortran 90" and "HPF" (aka High-Performance Fortran).
Note that they have discovered lower case, as well.

Quote:
>on these features? Will they do matrix multiplications and inverses?

These capabilities are available (and have been for years)
via library subroutines (BLAS, LINPACK, etc). The lib routines are,
for many users, superior to those available in {*filter*} APL because they
support things like:
  - sparse arrays (Try to do domino on a 300000x300000 array in APL).
  - symmetric and upper-diagonal, tridiagonal, etc for better performance
  - special case code to support arrays that don't fit in nicely to
    domino (they get bad answers...)

Quote:
>Without the need to define work matrices? etc. etc.

Sort of -- check out the paper cited above.
Bob


Fri, 25 Jun 1999 03:00:00 GMT  
 apl/windows


Quote:
> I could not
> have done this in ANY other language (including J). This is only because
> ths char set is GOOD.

You are confusing the char set with the language operations.
The APL operations are the important issue.

The char set happens to be BAD but the operations are
really GOOD.

I take ANY language (including J) anytime above ANY other
language because of the operations.

The problem is that the problems with the char set very
often get in the way and are a hindrance. That is the main
reason I started to look at J but now it is only one of many.

Actually I have become an avid fan of C and mostly because
I can so easily mix the two (J and C). I can use the same editor and I
can pass arguments and code between the two without problems.

And incidentally there are more differences between many
C compilers than between ANY APL interpreters/compilers
(including J) that I know of.

Nowadays I sometimes see J as an extension to C and
other days I look at C as an extension to J.

/Gosi



Fri, 25 Jun 1999 03:00:00 GMT  
 apl/windows


Quote:
>Nowadays I sometimes see J as an extension to C and
>other days I look at C as an extension to J.

Then in that case the J discussion belongs in the C newsgroup and not here
in the APL one. APL and J are rather incompatible.

Don.



Mon, 05 Jul 1999 03:00:00 GMT  
 apl/windows


Quote:

>>Nowadays I sometimes see J as an extension to C and
>>other days I look at C as an extension to J.

>Then in that case the J discussion belongs in the C newsgroup and not here
>in the APL one. APL and J are rather incompatible.

>Don.

While J and APL are somewhat incompatible, J is much closer to APL than
it is to C and most J postings would be inappropriate for the C newsgroups.
C really functions like a portable assembler for many developers.
This might be a good argument for starting a J specific newsgroup, though
given the low volume I like having J and APL together.

Regards,
Steve



Tue, 06 Jul 1999 03:00:00 GMT  
 
 [ 10 post ] 

 Relevant Pages 

1. apl/windows

2. Searching for an APL Windows font

3. apl/windows [Longish & Rambling!]

4. Shareware APL, APL FAQ, J for Windows

5. I-APL, Vanguard APL, and APL.68000

6. APL*Plus III for Windows

7. APL and Windows-development

8. APL+Win Version 2.0 and Windows XP

9. Dowload free APL for windows

10. APL Plus for Windows

11. Looking for APL for Windows

12. MONDAY, 10 OCT: DYALOG APL FOR WINDOWS

 

 
Powered by phpBB® Forum Software