On Tuesday, July 25, 2023 at 4:48:28 AM UTC-4, none albert wrote:
> In article <eefa62f9-eab0-4c2b...@googlegroups.com>,
> gobli...@gmail.com <gobli...@gmail.com> wrote:
> >From Veltas repo of Christopher Leonard I use this scriptin function to
> >add words to external file
> >
> >\ github.com/Veltas/demo/blob/main/script.4th
> >\ Copyright 2023 Christopher Leonard - MIT Licence
> >
> >: CREATE-FILE! CREATE-FILE ABORT" failed to open file" ;
> Normally these kind of stuff doesn't go wrong.
> Routinely I do
> CREATE-FILE THROW
> Only in the debugging face it fails. Other wise you are advised
> to generate more specific messages if
>
> 'complicate-stuff-that-creates-init-file CATCH DUP 31 =
> ABORT" You probably tried to create an init file where you have not \
> write access on the directory"
>
> This also show why I prefer numeric identification for error situation
> and I abhor the reuse of same error numbers for vastly different errors.
>
> It is a myth that a one line abort" message does cut it.
>
> This is the documentation of error 121 for my bsl package.
>
> "
> 121 : line is not stable before character
> A character is expected, so we wait for a start bit.
> If the line is already in a start bit / break state, that
> means that it is stuck in that state, or that we were too slow
> and the sbc has already started sending.
>
> If you get this error consistently then we are stuck in that
> state.
> This is indicative of
> 1) A wrong cable connection.
> Use the -t option to diagnose your cable. Check the connections.
> 2) your parallel port doesn't use the standard address.
> Use ` -a #### ' to specify the proper address.
> (Mostly you get 134 with that problem.)
>
> If the fault is intermittent, at least the msp430 can control
> the line, but it does so erratically.
> 3) Your pc is too slow. Terminate time consuming processes, like
> a gui.
> 4) A defective MSP430 or faulty bsl in MSP430. This has been
> seen in combination with a wrong patch of the bsl.
> "
>
> If bsl fails you see "error 121", but you know where to look.
> Groetjes Albert
> --
> Don't praise the day before the evening. One swallow doesn't make spring.
> You must not say "hey" before you have crossed the bridge. Don't sell the
> hide of the bear until you shot it. Better one bird in the hand than ten in
> the air. First gain is a cat spinning. - the Wise from Antrim -
Thank you for helping give credit where credit is due. i loved that video and wished to find a whole series by this person ..... amazing skill both programming and teaching.

On Tuesday, July 25, 2023 at 10:39:22 AM UTC-4, David Gipson wrote:
> On Tuesday, July 25, 2023 at 4:48:28 AM UTC-4, none albert wrote:
> > In article <eefa62f9-eab0-4c2b...@googlegroups.com>,
> > gobli...@gmail.com <gobli...@gmail.com> wrote:
> > >From Veltas repo of Christopher Leonard I use this scriptin function to
> > >add words to external file
> > >
> > >\ github.com/Veltas/demo/blob/main/script.4th
> > >\ Copyright 2023 Christopher Leonard - MIT Licence
> > >
> > >: CREATE-FILE! CREATE-FILE ABORT" failed to open file" ;
> > Normally these kind of stuff doesn't go wrong.
> > Routinely I do
> > CREATE-FILE THROW
> > Only in the debugging face it fails. Other wise you are advised
> > to generate more specific messages if
> >
> > 'complicate-stuff-that-creates-init-file CATCH DUP 31 =
> > ABORT" You probably tried to create an init file where you have not \
> > write access on the directory"
> >
> > This also show why I prefer numeric identification for error situation
> > and I abhor the reuse of same error numbers for vastly different errors..
> >
> > It is a myth that a one line abort" message does cut it.
> >
> > This is the documentation of error 121 for my bsl package.
> >
> > "
> > 121 : line is not stable before character
> > A character is expected, so we wait for a start bit.
> > If the line is already in a start bit / break state, that
> > means that it is stuck in that state, or that we were too slow
> > and the sbc has already started sending.
> >
> > If you get this error consistently then we are stuck in that
> > state.
> > This is indicative of
> > 1) A wrong cable connection.
> > Use the -t option to diagnose your cable. Check the connections.
> > 2) your parallel port doesn't use the standard address.
> > Use ` -a #### ' to specify the proper address.
> > (Mostly you get 134 with that problem.)
> >
> > If the fault is intermittent, at least the msp430 can control
> > the line, but it does so erratically.
> > 3) Your pc is too slow. Terminate time consuming processes, like
> > a gui.
> > 4) A defective MSP430 or faulty bsl in MSP430. This has been
> > seen in combination with a wrong patch of the bsl.
> > "
> >
> > If bsl fails you see "error 121", but you know where to look.
> > Groetjes Albert
> > --
> > Don't praise the day before the evening. One swallow doesn't make spring.
> > You must not say "hey" before you have crossed the bridge. Don't sell the
> > hide of the bear until you shot it. Better one bird in the hand than ten in
> > the air. First gain is a cat spinning. - the Wise from Antrim -
> Thank you for helping give credit where credit is due. i loved that video and wished to find a whole series by this person ..... amazing skill both programming and teaching.
As to the original question about saving progress in swift forth i looked in the pdf documents and there is an excellent section talking about the words forget remember marker and the word G along with an example ~/.SwiftForth-editor script to run vim after using locate or when using G.

On Tuesday, March 14, 2023 at 9:25:50 AM UTC-7, Yaroslav Veremenko wrote:
> For example yesterday I installed SwiftForth, I read the documentation
> and it has few pages on source management. The only word that seem to matter
> to me in this case is INCLUDE. But it doesn't seem to go deep into how to actually use it.

In my novice package I have this:

only forth definitions \ start from a known configuration

marker novice.4th \ don't use TRY on this
\ this is provided primarily for Gforth that lacks EMPTY

: try ( -- ) \ stream: marker-name
bl word
dup find if execute else drop then \ execute the marker to get rid of the file's code
count included ; \ include the file (marker name must be file name)

\ The marker-name is required to be the same as the file-name.
\ TRY will then remove the file's code if it has already been compiled, then compile the file.
\ TRY can't be used on NOVICE.4TH because removing the file's code removes TRY
\ and TRY will then go haywire.

: require ( -- ) \ stream: marker-name
bl word
find if drop else count included then ;
\ include the file if it is not already there (marker name must be file name)

\ REQUIRE is used to make sure a needed file is compiled.

I've been using the novice package since 2010 and this works fine.
Occasionally I will crash SwiftForth and have to reload it,
and reload the floating-point support that for reasons unknown is
not automatically loaded, then reload NOVICE.4TH etc..

> SwiftForth doesn't seem to have ...
> a way to reset the state of a interpreter.

This is a bug in the ANS-Forth design. If your compilation or your program
aborts, the system doesn't restart in a known state. If you have word-lists
changed, BASE changed, etc., this doesn't get reset. The solution is to have
code at the beginning of your file that sets everything to a known state.
TRY is in the FORTH word-list so it should be available, then your initialization
code fixes any other lingering problems.