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

Groups > comp.lang.forth > #53318 > unrolled thread

Win32Forth EMPTY bug

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

Contents

  Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-14 21:59 +1100
    Re: Win32Forth EMPTY bug november.nihal@gmail.com - 2016-12-14 05:02 -0800
      Re: Win32Forth EMPTY bug Coos Haak <htrofhc@gmail.com> - 2016-12-14 16:06 +0100
      Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-15 16:42 +1100
    Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-14 09:09 -0800
      Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-15 16:43 +1100
        Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-15 03:21 -0800
          Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-16 09:50 +1100
            Re: Win32Forth EMPTY bug Coos Haak <htrofhc@gmail.com> - 2016-12-16 00:34 +0100
              Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-16 03:50 -0800
                Re: Win32Forth EMPTY bug Coos Haak <htrofhc@gmail.com> - 2016-12-16 13:31 +0100
                Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-17 11:04 +1100
                  Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-17 04:38 -0800
                    Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-18 10:57 +1100
    Re: Win32Forth EMPTY bug rickman <gnuarm@gmail.com> - 2016-12-14 12:13 -0500
      Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-14 09:51 -0800
        Re: Win32Forth EMPTY bug rickman <gnuarm@gmail.com> - 2016-12-14 13:10 -0500
          Re: Win32Forth EMPTY bug josv@planet.nl - 2016-12-14 11:11 -0800
      Re: Win32Forth EMPTY bug november.nihal@gmail.com - 2016-12-14 09:58 -0800
        Re: Win32Forth EMPTY bug november.nihal@gmail.com - 2016-12-14 10:00 -0800
          Re: Win32Forth EMPTY bug Coos Haak <htrofhc@gmail.com> - 2016-12-15 02:41 +0100
            Re: Win32Forth EMPTY bug "HAA" <someone@microsoft.com> - 2016-12-15 16:48 +1100
              Re: Win32Forth EMPTY bug albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) - 2016-12-15 12:56 +0100

Page 1 of 2  [1] 2  Next page →

#53318 — Win32Forth EMPTY bug

Win32Forth 6.15.04

: sub ;  ok

: test
  sub ;  ok

empty  ok

: sub ;  ok

: test
  sub ;  ok

empty  ok

: sub ;  ok

: test { a }
  sub ;
  ^^^
Error(-13): SUB is undefined

[toc] | [next] | [standalone]

#53319

On Wednesday, 14 December 2016 10:59:19 UTC, HAA  wrote:
> Win32Forth 6.15.04
> 
> : sub ;  ok
> 
> : test
>   sub ;  ok
> 
> empty  ok
> 
> : sub ;  ok
> 
> : test
>   sub ;  ok
> 
> empty  ok
> 
> : sub ;  ok
> 
> : test { a }
>   sub ;
>   ^^^
> Error(-13): SUB is undefined

I rarely use locals and currently avasts blocks win32forth (so I am unable to look this up or test this.) 

However, I think the problem is in your definition of local there should be a '--' in there. Could you have a look and see if this resolves the problem please ?

NN

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

#53321

Op Wed, 14 Dec 2016 05:02:06 -0800 (PST) schreef november.nihal@gmail.com:

> On Wednesday, 14 December 2016 10:59:19 UTC, HAA  wrote:
>> Win32Forth 6.15.04
>> 
>>: sub ;  ok
>> 
>>: test
>>   sub ;  ok
>> 
>> empty  ok
>> 
>>: sub ;  ok
>> 
>>: test
>>   sub ;  ok
>> 
>> empty  ok
>> 
>>: sub ;  ok
>> 
>>: test { a }
>>   sub ;
>>   ^^^
>> Error(-13): SUB is undefined
> 
> I rarely use locals and currently avasts blocks win32forth (so I am unable to look this up or test this.) 
> 
> However, I think the problem is in your definition of local there should be a '--' in there. Could you have a look and see if this resolves the problem please ?
> 
> NN

The -- is optional. But the problem persists:

"
: sub { a } ; : test sub ; empty  ok
: sub { a } ; : test sub ; empty 
                     ^^^
Error(-13): SUB is undefined 
"

groet Coos

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

#53340

november.nihal@gmail.com wrote:
> ...
> I rarely use locals and currently avasts blocks win32forth (so I am unable to look this
> up or test this.)
>
> However, I think the problem is in your definition of local there should be a '--' in
> there. Could you have a look and see if this resolves the problem please ?
>
> NN

The same thing happens when using LOCALS| .  ANEW (i.e. MARKER) doesn't
exhibit the problem - only words based on MARK it would appear.

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

#53322

Hi,

Strange. No problem here.
If you use an editor does it use Windows line endings (CR/LF) ?

Jos

On 14 december 2016 11:59:19 UTC+1 HAA wrote:
> Win32Forth 6.15.04
> 
> : sub ;  ok
> 
> : test
>   sub ;  ok
> 
> empty  ok
> 
> : sub ;  ok
> 
> : test
>   sub ;  ok
> 
> empty  ok
> 
> : sub ;  ok
> 
> : test { a }
>   sub ;
>   ^^^
> Error(-13): SUB is undefined

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

#53341

Well that's interesting.  You are using the same version?

I tested with text copy & paste (ctl-c ctl-v) into Win32f (hence the 'ok's
in the output) also via an 'include' file.  The file does have cr/lf line endings

'see' EMPTY reveals it's doing something with locals but I didn't want to dig
any further.  I notice Coos is getting the same problem so I'm relieved it's
not just me :)

josv@planet.nl wrote:
> Hi,
>
> Strange. No problem here.
> If you use an editor does it use Windows line endings (CR/LF) ?
>
> Jos

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

#53351

Hi, 
I use version: 6.15.04 on Windows 10 Pro
When I copy and paste the following from notepad into Win32Forth:

: sub ;  

: test 
  sub ;  

: sub ;  

: test 
  sub ;  

: sub ;   

: test { a } 
  sub ; 

\ I see:

  ok
: sub ;    ok
  ok
: test  
  sub ;    ok
  ok
: sub ;   
  ^^^
Warning(-4100): SUB is redefined  ok
  ok
: test  
   ^^^^
Warning(-4100): TEST is redefined 
  sub ;    ok
  ok
: sub ;  ok  
  ^^^
Warning(-4100): SUB is redefined  ok
  ok
: test { a }  
  ^^^^
Warning(-4100): TEST is redefined 
  sub ;   ok
  ok
\s
Note: I also use the IDE to edit files on a headless RPI.
That forces me to change the line endings in the IDE.
When I use it again for Win32Forth and the line ending are not right then 
the results are unpredictable.

I take problem serious, I am just trying to isolate the problem.
Jos

On 15 december 2016 06:47:38 UTC+1 HAA wrote:
> Well that's interesting.  You are using the same version?
> 
> I tested with text copy & paste (ctl-c ctl-v) into Win32f (hence the 'ok's
> in the output) also via an 'include' file.  The file does have cr/lf line endings
> 
> 'see' EMPTY reveals it's doing something with locals but I didn't want to dig
> any further.  I notice Coos is getting the same problem so I'm relieved it's
> not just me :)
> 
> jos...nl wrote:
> > Hi,
> >
> > Strange. No problem here.
> > If you use an editor does it use Windows line endings (CR/LF) ?
> >
> > Jos

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

#53364

josv@planet.nl wrote:
> Hi,
> I use version: 6.15.04 on Windows 10 Pro
> When I copy and paste the following from notepad into Win32Forth:
>
> : sub ;
>
> : test
>   sub ;
>
> : sub ;
>
> : test
>   sub ;

Insert EMPTY here and try again

> : sub ;
>
> : test { a }
>   sub ;
>
> \ I see:

If you still see no problem perhaps it only manifests with certain Windows
versions or memory configurations.

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

#53367

Op Fri, 16 Dec 2016 09:50:21 +1100 schreef HAA:

> josv@planet.nl wrote:
>> Hi,
>> I use version: 6.15.04 on Windows 10 Pro
>> When I copy and paste the following from notepad into Win32Forth:
>>
>>: sub ;
>>
>>: test
>>   sub ;
>>
>>: sub ;
>>
>>: test
>>   sub ;
> 
> Insert EMPTY here and try again
> 
>>: sub ;
>>
>>: test { a }
>>   sub ;
>>
>> \ I see:
> 
> If you still see no problem perhaps it only manifests with certain Windows
> versions or memory configurations.

I tried Windows 10 and Vista and XP, both 32 bits. All say they
can't find SUB

groet Coos

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

#53377

Hi,
I looked further into the problem.
Thanks to HAA I could see the problem.
The following might give a solution for now:

1) Remove from: ~\win32forth\src\Extend.f the line with:
mark empty

2) Insert in ~\win32forth\Win32ForUser.cfg after 
the LAST >Win32ForthIDE the line:
mark empty

3) Start ~\win32forth\Setup.exe
Run option: B (Hit Build without changing anything)

I wonder if this indeed solves the problem.
Jos

On 16 december 2016 00:34:39 UTC+1 Coos Haak wrote:
> Op Fri, 16 Dec 2016 09:50:21 +1100 schreef HAA:
> 
> > josv...nl wrote:
> >> Hi,
> >> I use version: 6.15.04 on Windows 10 Pro
> >> When I copy and paste the following from notepad into Win32Forth:
> >>
> >>: sub ;
> >>
> >>: test
> >>   sub ;
> >>
> >>: sub ;
> >>
> >>: test
> >>   sub ;
> > 
> > Insert EMPTY here and try again
> > 
> >>: sub ;
> >>
> >>: test { a }
> >>   sub ;
> >>
> >> \ I see:
> > 
> > If you still see no problem perhaps it only manifests with certain Windows
> > versions or memory configurations.
> 
> I tried Windows 10 and Vista and XP, both 32 bits. All say they
> can't find SUB
> 
> groet Coos

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

#53378

Op Fri, 16 Dec 2016 03:50:16 -0800 (PST) schreef josv@planet.nl:

> Hi,
> I looked further into the problem.
> Thanks to HAA I could see the problem.
> The following might give a solution for now:
> 
> 1) Remove from: ~\win32forth\src\Extend.f the line with:
> mark empty
> 
> 2) Insert in ~\win32forth\Win32ForUser.cfg after 
> the LAST >Win32ForthIDE the line:
> mark empty
> 
> 3) Start ~\win32forth\Setup.exe
> Run option: B (Hit Build without changing anything)
> 
> I wonder if this indeed solves the problem.
> Jos

Thanks, that works!

groet Coos

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

#53383

josv@planet.nl wrote:
> Hi,
> I looked further into the problem.
> Thanks to HAA I could see the problem.
> The following might give a solution for now:
>
> 1) Remove from: ~\win32forth\src\Extend.f the line with:
> mark empty
>
> 2) Insert in ~\win32forth\Win32ForUser.cfg after
> the LAST >Win32ForthIDE the line:
> mark empty
>
> 3) Start ~\win32forth\Setup.exe
> Run option: B (Hit Build without changing anything)
>
> I wonder if this indeed solves the problem.
> Jos

That also works for me. Thanks.

Looking at the change it's not obvious why losing the definitions in
Win32ForUser.cfg should have created the problem.  However being so,
'mark empty' should probably be deferred to the last moment e.g. just
before Start-Interpreter in ForthBoot.  Executing it within the config files
isn't assured as they may be missing (which Win32for.exe allows).

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

#53387

Hi,

Using ForthBoot looks like a good idea.
That deferring empty is a bit hard. 
Where should the execution vector be defined?
It seems the following ForthBoot could be used.

: ForthBoot     ( -- )
                cls
                only forth also definitions
                set-shell
                stack-check-off
                config$ LoadConfigFile
                s" mark empty" evaluate 
                Start-Interpreter ;

A better way is welcome.

Jos

On zaterdag 17 december 2016 01:04:24 UTC+1 HAA wrote:
> jos...nl wrote:
> > Hi,
> > I looked further into the problem.
> > Thanks to HAA I could see the problem.
> > The following might give a solution for now:
> >
> > 1) Remove from: ~\win32forth\src\Extend.f the line with:
> > mark empty
> >
> > 2) Insert in ~\win32forth\Win32ForUser.cfg after
> > the LAST >Win32ForthIDE the line:
> > mark empty
> >
> > 3) Start ~\win32forth\Setup.exe
> > Run option: B (Hit Build without changing anything)
> >
> > I wonder if this indeed solves the problem.
> > Jos
> 
> That also works for me. Thanks.
> 
> Looking at the change it's not obvious why losing the definitions in
> Win32ForUser.cfg should have created the problem.  However being so,
> 'mark empty' should probably be deferred to the last moment e.g. just
> before Start-Interpreter in ForthBoot.  Executing it within the config files
> isn't assured as they may be missing (which Win32for.exe allows).

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

#53398

josv@planet.nl wrote:
> Hi,
>
> Using ForthBoot looks like a good idea.
> That deferring empty is a bit hard.
> Where should the execution vector be defined?
> It seems the following ForthBoot could be used.
>
> : ForthBoot     ( -- )
>                 cls
>                 only forth also definitions
>                 set-shell
>                 stack-check-off
>                 config$ LoadConfigFile
>                 s" mark empty" evaluate
>                 Start-Interpreter ;
>
> A better way is welcome.
>
> Jos

No, that's looks fine.  Perhaps the Win32F developers will have
more to say.

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

#53323

On 12/14/2016 5:59 AM, HAA wrote:
> Win32Forth 6.15.04
>
> : sub ;  ok
>
> : test
>   sub ;  ok
>
> empty  ok
>
> : sub ;  ok
>
> : test
>   sub ;  ok
>
> empty  ok
>
> : sub ;  ok
>
> : test { a }
>   sub ;
>   ^^^
> Error(-13): SUB is undefined

I have to say, Forth is not for the casual programmer.  I only use Forth 
occasionally, mostly win32forth.  So when I use it I typically have to 
relearn many things.  One of those things is figuring out how to view 
the source for a word I am not familiar with.

I tried "see", but that returns just the raw definition which is full of 
words I am not familiar with (opening a rabbit hole of even more words I 
don't know if I try to figure them out).  I tried "view" and get the 
following odd result...

view empty  loaded from: (console) ok

How do I find the source code that defines "empty"?

-- 

Rick C

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

#53325

Hi,
The definition of empty is not meaning full since it is not defined.
But if you would like to see the source of words you could enter in Win32Forth:

view words

or

v words

Jos

On 14 december 2016 18:13:00 UTC+1 rickman wrote:
> On 12/14/2016 5:59 AM, HAA wrote:
> > Win32Forth 6.15.04
> >
> > : sub ;  ok
> >
> > : test
> >   sub ;  ok
> >
> > empty  ok
> >
> > : sub ;  ok
> >
> > : test
> >   sub ;  ok
> >
> > empty  ok
> >
> > : sub ;  ok
> >
> > : test { a }
> >   sub ;
> >   ^^^
> > Error(-13): SUB is undefined
> 
> I have to say, Forth is not for the casual programmer.  I only use Forth 
> occasionally, mostly win32forth.  So when I use it I typically have to 
> relearn many things.  One of those things is figuring out how to view 
> the source for a word I am not familiar with.
> 
> I tried "see", but that returns just the raw definition which is full of 
> words I am not familiar with (opening a rabbit hole of even more words I 
> don't know if I try to figure them out).  I tried "view" and get the 
> following odd result...
> 
> view empty  loaded from: (console) ok
> 
> How do I find the source code that defines "empty"?
> 
> -- 
> 
> Rick C

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

#53329

On 12/14/2016 12:51 PM, josv@planet.nl wrote:
>
> On 14 december 2016 18:13:00 UTC+1 rickman wrote:
>> On 12/14/2016 5:59 AM, HAA wrote:
>>> Win32Forth 6.15.04
>>>
>>> : sub ;  ok
>>>
>>> : test
>>>   sub ;  ok
>>>
>>> empty  ok
>>>
>>> : sub ;  ok
>>>
>>> : test
>>>   sub ;  ok
>>>
>>> empty  ok
>>>
>>> : sub ;  ok
>>>
>>> : test { a }
>>>   sub ;
>>>   ^^^
>>> Error(-13): SUB is undefined
>>
>> I have to say, Forth is not for the casual programmer.  I only use Forth
>> occasionally, mostly win32forth.  So when I use it I typically have to
>> relearn many things.  One of those things is figuring out how to view
>> the source for a word I am not familiar with.
>>
>> I tried "see", but that returns just the raw definition which is full of
>> words I am not familiar with (opening a rabbit hole of even more words I
>> don't know if I try to figure them out).  I tried "view" and get the
>> following odd result...
>>
>> view empty  loaded from: (console) ok
>>
>> How do I find the source code that defines "empty"?
>
> Hi,
> The definition of empty is not meaning full since it is not defined.
> But if you would like to see the source of words you could enter in Win32Forth:
>
> view words
>
> or
>
> v words
>
> Jos

Ok, then I am very confused.  If I type

see empty

I get a long complicated definition.  Why do you say
"empty" is not defined?

Oh, I get it... duh!  "view" opens a file name rather than searching for 
a word.

What is the word that lets me see the source code for a given word?  I 
know I have used this in the past.  "See" appears to decompile the word. 
  I recall using a word that would display the source along with a 
couple of lines before it to display comments.  Am I not remembering 
this correctly?

Like I said, I'm an occasional user.

-- 

Rick C

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

#53331

Hi,

Oeps, my mistake.
Normally view should open the file where the definition is defined and point you to that definition. 

Jos

Op woensdag 14 december 2016 19:10:06 UTC+1 rickman wrote:
> On 12/14/2016 12:51 PM, jos...nl wrote:
> >
> > On 14 december 2016 18:13:00 UTC+1 rickman wrote:
> >> On 12/14/2016 5:59 AM, HAA wrote:
> >>> Win32Forth 6.15.04
> >>>
> >>> : sub ;  ok
> >>>
> >>> : test
> >>>   sub ;  ok
> >>>
> >>> empty  ok
> >>>
> >>> : sub ;  ok
> >>>
> >>> : test
> >>>   sub ;  ok
> >>>
> >>> empty  ok
> >>>
> >>> : sub ;  ok
> >>>
> >>> : test { a }
> >>>   sub ;
> >>>   ^^^
> >>> Error(-13): SUB is undefined
> >>
> >> I have to say, Forth is not for the casual programmer.  I only use Forth
> >> occasionally, mostly win32forth.  So when I use it I typically have to
> >> relearn many things.  One of those things is figuring out how to view
> >> the source for a word I am not familiar with.
> >>
> >> I tried "see", but that returns just the raw definition which is full of
> >> words I am not familiar with (opening a rabbit hole of even more words I
> >> don't know if I try to figure them out).  I tried "view" and get the
> >> following odd result...
> >>
> >> view empty  loaded from: (console) ok
> >>
> >> How do I find the source code that defines "empty"?
> >
> > Hi,
> > The definition of empty is not meaning full since it is not defined.
> > But if you would like to see the source of words you could enter in Win32Forth:
> >
> > view words
> >
> > or
> >
> > v words
> >
> > Jos
> 
> Ok, then I am very confused.  If I type
> 
> see empty
> 
> I get a long complicated definition.  Why do you say
> "empty" is not defined?
> 
> Oh, I get it... duh!  "view" opens a file name rather than searching for 
> a word.
> 
> What is the word that lets me see the source code for a given word?  I 
> know I have used this in the past.  "See" appears to decompile the word. 
>   I recall using a word that would display the source along with a 
> couple of lines before it to display comments.  Am I not remembering 
> this correctly?
> 
> Like I said, I'm an occasional user.
> 
> -- 
> 
> Rick C

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

#53326

On Wednesday, 14 December 2016 17:13:00 UTC, rickman  wrote:
> On 12/14/2016 5:59 AM, HAA wrote:
> > Win32Forth 6.15.04
> >
> > : sub ;  ok
> >
> > : test
> >   sub ;  ok
> >
> > empty  ok
> >
> > : sub ;  ok
> >
> > : test
> >   sub ;  ok
> >
> > empty  ok
> >
> > : sub ;  ok
> >
> > : test { a }
> >   sub ;
> >   ^^^
> > Error(-13): SUB is undefined
> 
> I have to say, Forth is not for the casual programmer.  I only use Forth 
> occasionally, mostly win32forth.  So when I use it I typically have to 
> relearn many things.  One of those things is figuring out how to view 
> the source for a word I am not familiar with.
> 
> I tried "see", but that returns just the raw definition which is full of 
> words I am not familiar with (opening a rabbit hole of even more words I 
> don't know if I try to figure them out).  I tried "view" and get the 
> following odd result...
> 
> view empty  loaded from: (console) ok
> 
> How do I find the source code that defines "empty"?
> 
> -- 
> 
> Rick C


* empty is defined as : 

  mark empty in extend.f


* You can use view and see

* you can also use 'find' from WinEd to locate all source files where its been used.

 

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

#53328

actually I think I might just have answered your question haa. Its to do with how empty is defined. 
its not defined as a word but rather using mark !

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

Page 1 of 2  [1] 2  Next page →

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

csiph-web