Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #42381 > unrolled thread

Re: Could Forth be getting a bit of traction?

Started by"WJ" <w_a_x_man@yahoo.com>
First post2015-10-06 01:14 +0000
Last post2015-10-08 07:29 +0000
Articles 7 on this page of 27 — 15 participants

Back to article view | Back to comp.lang.forth

Contents

  Re: Could Forth be getting a bit of traction? "WJ" <w_a_x_man@yahoo.com> - 2015-10-06 01:14 +0000
    Re: Could Forth be getting a bit of traction? Jason Damisch <jasondamisch@yahoo.com> - 2015-10-06 10:47 -0700
      Re: Could Forth be getting a bit of traction? Chris Curl <ccurl609@gmail.com> - 2015-10-06 11:01 -0700
        Re: Could Forth be getting a bit of traction? Jason Damisch <jasondamisch@yahoo.com> - 2015-10-06 11:27 -0700
        Re: Could Forth be getting a bit of traction? lehs <skydda.barnen@gmail.com> - 2015-10-06 12:08 -0700
    Re: Could Forth be getting a bit of traction? Howerd <howerdo@yahoo.co.uk> - 2015-10-06 14:06 -0700
      Re: Could Forth be getting a bit of traction? Jason Damisch <jasondamisch@yahoo.com> - 2015-10-06 16:15 -0700
        Re: Could Forth be getting a bit of traction? Chris Curl <ccurl609@gmail.com> - 2015-10-06 21:48 -0700
      Re: Could Forth be getting a bit of traction? stephen@mpeforth.com (Stephen Pelc) - 2015-10-07 12:19 +0000
        Re: Could Forth be getting a bit of traction? Mark Wills <markwills1970@gmail.com> - 2015-10-07 07:40 -0700
          Re: Could Forth be getting a bit of traction? Michael Barry <barrym95838@yahoo.com> - 2015-10-07 08:27 -0700
            Re: Could Forth be getting a bit of traction? Coos Haak <chforth@hccnet.nl> - 2015-10-07 17:40 +0200
              Re: Could Forth be getting a bit of traction? franck.bensusan@gmail.com - 2015-10-07 10:50 -0700
                Re: Could Forth be getting a bit of traction? Coos Haak <chforth@hccnet.nl> - 2015-10-07 22:37 +0200
                  Re: Could Forth be getting a bit of traction? franck.bensusan@gmail.com - 2015-10-07 23:22 -0700
            Re: Could Forth be getting a bit of traction? "Elizabeth D. Rather" <erather@forth.com> - 2015-10-07 09:14 -1000
            Array Style (was Re: Could Forth be getting a bit of traction?) Howerd <howerdo@yahoo.co.uk> - 2015-10-07 13:23 -0700
          Re: Could Forth be getting a bit of traction? Howerd <howerdo@yahoo.co.uk> - 2015-10-07 13:20 -0700
        Re: Could Forth be getting a bit of traction? rickman <gnuarm@gmail.com> - 2015-10-07 12:13 -0400
          Array Style ( was Re: Could Forth be getting a bit of traction? ) Howerd <howerdo@yahoo.co.uk> - 2015-10-07 13:36 -0700
            Re: Array Style ( was Re: Could Forth be getting a bit of traction? ) rickman <gnuarm@gmail.com> - 2015-10-07 17:39 -0400
        Array Style ( was Re: Could Forth be getting a bit of traction? ) Howerd <howerdo@yahoo.co.uk> - 2015-10-07 13:16 -0700
      Re: Could Forth be getting a bit of traction? mhx@iae.nl - 2015-10-07 10:46 -0700
        Array Style ( was Re: Could Forth be getting a bit of traction? ) Howerd <howerdo@yahoo.co.uk> - 2015-10-07 13:39 -0700
          Re: Array Style ( was Re: Could Forth be getting a bit of traction? ) mhx@iae.nl - 2015-10-07 15:01 -0700
            Re: Array Style ( was Re: Could Forth be getting a bit of traction? ) Roelf Toxopeus <rt4all@notthis.hetnet.nl> - 2015-10-09 22:05 +0200
      Re: Could Forth be getting a bit of traction? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2015-10-08 07:29 +0000

Page 2 of 2 — ← Prev page 1 [2]

#42437 — Re: Array Style ( was Re: Could Forth be getting a bit of traction? )

Fromrickman <gnuarm@gmail.com>
Date2015-10-07 17:39 -0400
SubjectRe: Array Style ( was Re: Could Forth be getting a bit of traction? )
Message-ID<mv439t$3f2$1@dont-email.me>
In reply to#42433
On 10/7/2015 4:36 PM, Howerd wrote:
> On Wednesday, October 7, 2015 at 5:13:38 PM UTC+1, rickman wrote:
>> On 10/7/2015 8:19 AM, Stephen Pelc wrote:
>>> On Tue, 6 Oct 2015 14:06:15 -0700 (PDT), Howerd <howe...@yahoo.co.uk>
>>> wrote:
>>>
>>>> Jason, Anton, Bernd, Marcel, Stephen et al : is it possible to use local va=
>>>> riables in a child definition? Does this depend on the implementation, and =
>>>> is it defined/allowed by ANS?
>>>
>>> There are no current mechanisms for Forth locals to be referenced in
>>> words other than the one in which they are defined.
>>>
>>> Some systems, e.g. VFX, do permit you to obtain an address of a local
>>> variable. That address can then be passed into another word. This
>>> feature is useful when interfacing to operating systems but is
>>> non-standard. Other systems, e.g. iForth, use the limitations to
>>> enhance the performance of locals.
>>
>> I believe there was a rather lengthy thread on this not too long ago.
>> People proposed a number of mechanisms to limit the use of the locals so
>> it would not cause problems when referencing them in other definitions
>> than the one they were defined in.
>>
>> --
>>
>> Rick
>
> Hi Rick,
>
> All of this is not convincing me that locals are a good thing.
> They are way to complicated for a simpleton like me ;-)
> Even though it is easy enough to take some syntactic sugar "{" and "}" and make it point to an allocated area on the return stack, there is an increase in complexity if you try to support them fully.
>
> It reminds me of why Forth uses RPN : you could easily add a few "BL WORD"s so that you define :
> : add ( n -- )   BL word count number + ;   \ BL is $20
> then you can type :
> 1 + 1 . 2
> But if you do this you block the simple use of  add  , and have to start processing strings. Maybe you have re-invented Lisp?

I'm not sure what you are saying.  I don't have a horse in this race.  I 
was merely pointing out that others are interested in the sharing of 
locals and others have discussed this before.

I don't currently use locals but am open to learning about this if I 
find the stack to be a problem at some point.  I have written code that 
seemed to have trouble with the stack because of a large number of 
parameters which needed to be passed to a word.  But I eventually worked 
that out with different factoring.  So for now, I'm good.

-- 

Rick

[toc] | [prev] | [next] | [standalone]

#42430 — Array Style ( was Re: Could Forth be getting a bit of traction? )

FromHowerd <howerdo@yahoo.co.uk>
Date2015-10-07 13:16 -0700
SubjectArray Style ( was Re: Could Forth be getting a bit of traction? )
Message-ID<473bbad4-7515-4df2-94f9-0f0aa4a6b71c@googlegroups.com>
In reply to#42420
On Wednesday, October 7, 2015 at 1:19:39 PM UTC+1, Stephen Pelc wrote:
> On Tue, 6 Oct 2015 14:06:15 -0700 (PDT), Howerd <how...@yahoo.co.uk>
> wrote:
> 
> >Jason, Anton, Bernd, Marcel, Stephen et al : is it possible to use local va=
> >riables in a child definition? Does this depend on the implementation, and =
> >is it defined/allowed by ANS?
> 
> There are no current mechanisms for Forth locals to be referenced in
> words other than the one in which they are defined.
> 
> Some systems, e.g. VFX, do permit you to obtain an address of a local
> variable. That address can then be passed into another word. This
> feature is useful when interfacing to operating systems but is
> non-standard. Other systems, e.g. iForth, use the limitations to
> enhance the performance of locals.
> 
> Stephen
> 
> -- 
> Stephen Pelc, stephenXXX@mpeforth.com
> MicroProcessor Engineering Ltd - More Real, Less Time
> 133 Hill Lane, Southampton SO15 5AF, England
> tel: +44 (0)23 8063 1441, fax: +44 (0)23 8033 9691
> web: http://www.mpeforth.com - free VFX Forth downloads

Hi Stephen,

Yes - I thought that this would be difficult, because the compiler would need to "postpone" the action of the local keywords "{" and "}", somehow...

Best regards,
Howerd

[toc] | [prev] | [next] | [standalone]

#42426

Frommhx@iae.nl
Date2015-10-07 10:46 -0700
Message-ID<b11dba72-6a56-4be0-968e-6819a2507cff@googlegroups.com>
In reply to#42404
On Tuesday, October 6, 2015 at 11:06:20 PM UTC+2, Howerd wrote:
[..]
> Jason, Anton, Bernd, Marcel, Stephen et al
> : is it possible to use local variables in a child definition?
> Does this depend on the implementation, and is it 
> defined/allowed by ANS?
 
Does this answer 33% of your question?

: det 	create , ,  ( a c -- )  
	does> ( b -- det ) 
		@+ swap @ locals| a c b |  
		b b *   a c * 4 * -  sqrt  
		b -  a 2* / ; 
FORTH> 1 1 det b
FORTH>  2 b . -1  ok
FORTH> -2 b .  1  ok

-marcel

[toc] | [prev] | [next] | [standalone]

#42435 — Array Style ( was Re: Could Forth be getting a bit of traction? )

FromHowerd <howerdo@yahoo.co.uk>
Date2015-10-07 13:39 -0700
SubjectArray Style ( was Re: Could Forth be getting a bit of traction? )
Message-ID<9d5dd9e4-6bed-48b9-99ec-77758c5b0e38@googlegroups.com>
In reply to#42426
On Wednesday, October 7, 2015 at 6:46:32 PM UTC+1, m...@iae.nl wrote:
> On Tuesday, October 6, 2015 at 11:06:20 PM UTC+2, Howerd wrote:
> [..]
> > Jason, Anton, Bernd, Marcel, Stephen et al
> > : is it possible to use local variables in a child definition?
> > Does this depend on the implementation, and is it 
> > defined/allowed by ANS?
>  
> Does this answer 33% of your question?
> 
> : det 	create , ,  ( a c -- )  
> 	does> ( b -- det ) 
> 		@+ swap @ locals| a c b |  
> 		b b *   a c * 4 * -  sqrt  
> 		b -  a 2* / ; 
> FORTH> 1 1 det b
> FORTH>  2 b . -1  ok
> FORTH> -2 b .  1  ok
> 
> -marcel

Hi Marcel,

I think it answers 100%, but only 33% in the positive.
So it can be done, its not ANS and its behaviour in a standard system is ambiguous?

Best regards,
Howerd

[toc] | [prev] | [next] | [standalone]

#42438 — Re: Array Style ( was Re: Could Forth be getting a bit of traction? )

Frommhx@iae.nl
Date2015-10-07 15:01 -0700
SubjectRe: Array Style ( was Re: Could Forth be getting a bit of traction? )
Message-ID<164063ad-3ca4-4397-994a-8d4e8270fe98@googlegroups.com>
In reply to#42435
On Wednesday, October 7, 2015 at 10:39:24 PM UTC+2, Howerd wrote:
> On Wednesday, October 7, 2015 at 6:46:32 PM UTC+1, m...@iae.nl wrote:
> > On Tuesday, October 6, 2015 at 11:06:20 PM UTC+2, Howerd wrote:
> > [..]
> > > Jason, Anton, Bernd, Marcel, Stephen et al
> > > : is it possible to use local variables in a child definition?
> > > Does this depend on the implementation, and is it
> > > defined/allowed by ANS?
> >
> > Does this answer 33% of your question?
> >
> > : det 	create , ,  ( a c -- )
> > 	does> ( b -- det )
> > 		@+ swap @ locals| a c b |
> > 		b b *   a c * 4 * -  sqrt
> > 		b -  a 2* / ;
> > FORTH> 1 1 det b
> > FORTH>  2 b . -1  ok
> > FORTH> -2 b .  1  ok
> >
> > -marcel
>
> Hi Marcel,
>
> I think it answers 100%, but only 33% in the positive.
> So it can be done, its not ANS and its behaviour in a
> standard system is ambiguous?

Gforth 0.7.0, Copyright (C) 1995-2008 Free Software Foundation, Inc.
Gforth comes with ABSOLUTELY NO WARRANTY; for details type `license'
Type `bye' to exit
: sqrt s>f fsqrt f>s ;  ok
: det   create , ,  ( a c -- )    compiled
        does> ( b -- det )   compiled
                dup @ swap cell+ @ locals| a c b |    compiled
                b b *   a c * 4 * -  sqrt    compiled
                b -  a 2* / ;  ok
1 1 det b  ok
  2 b .  -1  ok
 -2 b . 1  ok


Looks pretty standard to me :-)

-marcel

[toc] | [prev] | [next] | [standalone]

#42527 — Re: Array Style ( was Re: Could Forth be getting a bit of traction? )

FromRoelf Toxopeus <rt4all@notthis.hetnet.nl>
Date2015-10-09 22:05 +0200
SubjectRe: Array Style ( was Re: Could Forth be getting a bit of traction? )
Message-ID<56181e00$0$9583$e4fe514c@textnews.kpn.nl>
In reply to#42438
On 08/10/15 00:01, mhx@iae.nl wrote:
> On Wednesday, October 7, 2015 at 10:39:24 PM UTC+2, Howerd wrote:
>> On Wednesday, October 7, 2015 at 6:46:32 PM UTC+1, m...@iae.nl wrote:
>>> On Tuesday, October 6, 2015 at 11:06:20 PM UTC+2, Howerd wrote:
>>> [..]
>>>> Jason, Anton, Bernd, Marcel, Stephen et al
>>>> : is it possible to use local variables in a child definition?
>>>> Does this depend on the implementation, and is it
>>>> defined/allowed by ANS?
>>>
>>> Does this answer 33% of your question?
>>>
>>> : det 	create , ,  ( a c -- )
>>> 	does> ( b -- det )
>>> 		@+ swap @ locals| a c b |
>>> 		b b *   a c * 4 * -  sqrt
>>> 		b -  a 2* / ;
>>> FORTH> 1 1 det b
>>> FORTH>  2 b . -1  ok
>>> FORTH> -2 b .  1  ok
>>>
>>> -marcel
>>
>> Hi Marcel,
>>
>> I think it answers 100%, but only 33% in the positive.
>> So it can be done, its not ANS and its behaviour in a
>> standard system is ambiguous?
>
> Gforth 0.7.0, Copyright (C) 1995-2008 Free Software Foundation, Inc.
> Gforth comes with ABSOLUTELY NO WARRANTY; for details type `license'
> Type `bye' to exit
> : sqrt s>f fsqrt f>s ;  ok
> : det   create , ,  ( a c -- )    compiled
>          does> ( b -- det )   compiled
>                  dup @ swap cell+ @ locals| a c b |    compiled
>                  b b *   a c * 4 * -  sqrt    compiled
>                  b -  a 2* / ;  ok
> 1 1 det b  ok
>    2 b .  -1  ok
>   -2 b . 1  ok
>
>
> Looks pretty standard to me :-)

SwiftForth i386-Mac OS X 3.5.8 24 Sep-2015
: sqrt s>f fsqrt f>s ;  ok
: det   create , ,  ( a c -- )
         does> ( b -- det )
                dup @ swap cell+ @ locals| a c b |
                b b *   a c * 4 * -  sqrt
                b -  a 2* / ;  ok
1 1 det b  ok
2 b . -1  ok
-2 b . 1  ok


VFX Forth for OS X IA32
  © MicroProcessor Engineering Ltd, 1998-2014

  Version: 4.71 [build 0255]
  Build date: 23 October 2014

  Free dictionary = 7767740 bytes [7585kb]

: sqrt s>f fsqrt f>s ;  ok
: det   create , ,  ( a c -- )
         does> ( b -- det )
                dup @ swap cell+ @ locals| a c b |
                b b *   a c * 4 * -  sqrt
                b -  a 2* / ;  ok
1 1 det b  ok
2 b . -1  ok
-2 b . 1  ok

[toc] | [prev] | [next] | [standalone]

#42450

Fromanton@mips.complang.tuwien.ac.at (Anton Ertl)
Date2015-10-08 07:29 +0000
Message-ID<2015Oct8.092943@mips.complang.tuwien.ac.at>
In reply to#42404
Howerd <howerdo@yahoo.co.uk> writes:
>Jason, Anton, Bernd, Marcel, Stephen et al : is it possible to use local va=
>riables in a child definition? Does this depend on the implementation, and =
>is it defined/allowed by ANS?

Forth-2012 (and -94) says:

|13.3.3.1 Compilation semantics
|
|d) Separate sets of locals may be declared in defining words before
|DOES> for use by the defining word, and after DOES> for use by the
|word defined.

I.e., the following is standard:

: foo 
  create {: A :} \ A visible here
does>
  {: B :} \ B (and only B) visible here
;

Whether systems other than Gforth implement this correctly I have not
tested.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2015: http://www.rigwit.co.uk/EuroForth2015/

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | comp.lang.forth

csiph-web