On Monday, June 22, 2020 at 1:07:42 AM UTC+2, hughag...@gmail.com wrote:
> Note that the Beez's code doesn't work.
> Normally, ," stores the string at HERE --- embedding the strings
> in the table makes each entry a different length, so a binary-search
> won't work. If we assume that ," stores the string elsewhere and only
> stores the address at HERE with comma, then this still doesn't work.
> The 2/ assumes that the table entry is 2 bytes (on a 16-bit Forth),
> but it is actually 4 bytes because we have both the string address
> and the xt value in each entry.
> This is non-working fantasy code.

You're so VERY wrong. Simply because you don't take 4tH's architecture into consideration.
(1) 4tH doesn't have a dictionary - so compiling code is not linear (like in Forth). Nothing (except initialized data) is allocated and stuff ends up in other places - like a symbol table;
(2) You don't subscribe to the standard "Harvard architecture" definition. That's ok. Everyone is allowed to make their own private definitions that they use in the privacy of their own homes. But that's what 4tH has. It features several distinct segments:
[a] The Integer Segment, that consists of words and houses variables and the stack;
[b] The Character Segment, that consists of bytes and houses writable strings, number generator buffer, TIB and PAD;
[c] The String Segment, that houses character constants and is both non-readable and non-writable;
[d] The Code Segment, that houses compiled Forth words, including integer constants and is non-writable;
[e] The Symbol Table, that contains Forth names and references - usually discarded after compilation;
[f] The Header, which is is both non-readable and non-writable and ties it all together.

The size of each element in each Segment is.. 1. To get from an "address" in the Code Segment to the next one is to add one. Both CELLS and CHARS are dummies, because their reference is always.. multiply by ONE.
Hence you don't read "table entry is 2 bytes", you have to read "table entry is 2 (Code Segment) elements" - which is completely correct.

And yes, only the string reference is stored in that element. The string is a string literal, which is stored in the String Segment. Therefore, you will need @C to retrieve it (CROSS wordset). As a matter of fact, you won't even get a reference to the String Segment, you will get a reference to the PAD in the Character Segment, where this string literal has been copied to, since the String Segment itself is non-readable in 4tH.

There are plenty of 4tH programs in 4tH's Sourceforge repository, which will prove this is NOT "fantasy code".

Of course, if I wanted to make this thing portable, I would have written stuff like "2 CELLS /" (which is perfectly valid 4tH - or Forth - code). But sometimes I don't need to and can do without the hustle. So I don't.

Hans Bezemer

Le mardi 22 mars 2022 à 15 h 15 min 42 s UTC+1, the.bee...@gmail.com a écrit :
> On Monday, June 22, 2020 at 1:07:42 AM UTC+2, hughag...@gmail.com wrote:
> > Note that the Beez's code doesn't work.
> > Normally, ," stores the string at HERE --- embedding the strings
> > in the table makes each entry a different length, so a binary-search
> > won't work. If we assume that ," stores the string elsewhere and only
> > stores the address at HERE with comma, then this still doesn't work.
> > The 2/ assumes that the table entry is 2 bytes (on a 16-bit Forth),
> > but it is actually 4 bytes because we have both the string address
> > and the xt value in each entry.
> > This is non-working fantasy code.
> You're so VERY wrong. Simply because you don't take 4tH's architecture into consideration.
> (1) 4tH doesn't have a dictionary - so compiling code is not linear (like in Forth). Nothing (except initialized data) is allocated and stuff ends up in other places - like a symbol table;
> (2) You don't subscribe to the standard "Harvard architecture" definition.. That's ok. Everyone is allowed to make their own private definitions that they use in the privacy of their own homes. But that's what 4tH has. It features several distinct segments:
> [a] The Integer Segment, that consists of words and houses variables and the stack;
> [b] The Character Segment, that consists of bytes and houses writable strings, number generator buffer, TIB and PAD;
> [c] The String Segment, that houses character constants and is both non-readable and non-writable;
> [d] The Code Segment, that houses compiled Forth words, including integer constants and is non-writable;
> [e] The Symbol Table, that contains Forth names and references - usually discarded after compilation;
> [f] The Header, which is is both non-readable and non-writable and ties it all together.
>
> The size of each element in each Segment is.. 1. To get from an "address" in the Code Segment to the next one is to add one. Both CELLS and CHARS are dummies, because their reference is always.. multiply by ONE.
> Hence you don't read "table entry is 2 bytes", you have to read "table entry is 2 (Code Segment) elements" - which is completely correct.
>
> And yes, only the string reference is stored in that element. The string is a string literal, which is stored in the String Segment. Therefore, you will need @C to retrieve it (CROSS wordset). As a matter of fact, you won't even get a reference to the String Segment, you will get a reference to the PAD in the Character Segment, where this string literal has been copied to, since the String Segment itself is non-readable in 4tH.
>
> There are plenty of 4tH programs in 4tH's Sourceforge repository, which will prove this is NOT "fantasy code".
>
> Of course, if I wanted to make this thing portable, I would have written stuff like "2 CELLS /" (which is perfectly valid 4tH - or Forth - code). But sometimes I don't need to and can do without the hustle. So I don't.
>
> Hans Bezemer
https://medium.com/p/18f0b431e5

Le mardi 11 octobre 2022 à 06 h 03 min 52 s UTC+2, Coach Elsa a écrit :
> Le mardi 22 mars 2022 à 15 h 15 min 42 s UTC+1, the.bee...@gmail.com a écrit :
> > On Monday, June 22, 2020 at 1:07:42 AM UTC+2, hughag...@gmail.com wrote:
> > > Note that the Beez's code doesn't work.
> > > Normally, ," stores the string at HERE --- embedding the strings
> > > in the table makes each entry a different length, so a binary-search
> > > won't work. If we assume that ," stores the string elsewhere and only
> > > stores the address at HERE with comma, then this still doesn't work.
> > > The 2/ assumes that the table entry is 2 bytes (on a 16-bit Forth),
> > > but it is actually 4 bytes because we have both the string address
> > > and the xt value in each entry.
> > > This is non-working fantasy code.
> > You're so VERY wrong. Simply because you don't take 4tH's architecture into consideration.
> > (1) 4tH doesn't have a dictionary - so compiling code is not linear (like in Forth). Nothing (except initialized data) is allocated and stuff ends up in other places - like a symbol table;
> > (2) You don't subscribe to the standard "Harvard architecture" definition. That's ok. Everyone is allowed to make their own private definitions that they use in the privacy of their own homes. But that's what 4tH has. It features several distinct segments:
> > [a] The Integer Segment, that consists of words and houses variables and the stack;
> > [b] The Character Segment, that consists of bytes and houses writable strings, number generator buffer, TIB and PAD;
> > [c] The String Segment, that houses character constants and is both non-readable and non-writable;
> > [d] The Code Segment, that houses compiled Forth words, including integer constants and is non-writable;
> > [e] The Symbol Table, that contains Forth names and references - usually discarded after compilation;
> > [f] The Header, which is is both non-readable and non-writable and ties it all together.
> >
> > The size of each element in each Segment is.. 1. To get from an "address" in the Code Segment to the next one is to add one. Both CELLS and CHARS are dummies, because their reference is always.. multiply by ONE.
> > Hence you don't read "table entry is 2 bytes", you have to read "table entry is 2 (Code Segment) elements" - which is completely correct.
> >
> > And yes, only the string reference is stored in that element. The string is a string literal, which is stored in the String Segment. Therefore, you will need @C to retrieve it (CROSS wordset). As a matter of fact, you won't even get a reference to the String Segment, you will get a reference to the PAD in the Character Segment, where this string literal has been copied to, since the String Segment itself is non-readable in 4tH.
> >
> > There are plenty of 4tH programs in 4tH's Sourceforge repository, which will prove this is NOT "fantasy code".
> >
> > Of course, if I wanted to make this thing portable, I would have written stuff like "2 CELLS /" (which is perfectly valid 4tH - or Forth - code). But sometimes I don't need to and can do without the hustle. So I don't.
> >
> > Hans Bezemer
> https://medium.com/p/18f0b431e5
https://sites.google.com/view/tahmiloghniyatalghazalarayka/
https://sites.google.com/view/terrifier2streaming/
https://sites.google.com/view/ziyadatmotabinhakikiyin/
https://sites.google.com/view/how-to-heal-acne-scars-fast/
https://sites.google.com/view/code-de-triche-sims-4-pc/code-de-triche-sims-4-ps4-argent" rel="nofollow" target="_blank">https://sites.google.com/view/code-de-triche-sims-4-pc/code-de-triche-sims-4-ps4-argent
https://sites.google.com/view/code-de-triche-sims-4-pc/
https://sites.google.com/view/code-de-triche-sims-4-pc/sims-4-high-school-years-trailer-and-review
https://sites.google.com/view/code-de-triche-sims-4-pc/sims-4-ecran-de-chargement-infini

https://sites.google.com/view/tutorialmasters/makeup-by-mario-discount-code
https://sites.google.com/view/tutorialmasters/euro-car-parts-discount-code
https://sites.google.com/view/tutorialmasters/fashion-nova-discount-code-2022
https://sites.google.com/view/tutorialmasters/makeup-by-mario-ethereal-eyes