6 ms·

Show HN: forthy2 – a higher level Forth-remix in C++

65 points by codr7 7y ago

garmaine 7y ago
Why does pushing drop the stack you are pushing onto? Is there ever any reason to push something onto a stack and then immediately drop the stack? Even if there were, it doesn't seem like the option you'd want to optimize for.
- codr7 7y ago
  I'm sorry, I'm not quite following. I don't get what you mean by dropping the stack in this context. Could you point to where in the README you're having trouble?
  - garmaine 7y ago
    Not from the README. I don't read READMEs. I read code. The issue is this: > (1 2 3) 4 push My expectation is this should yield "(1 2 3 4)". It doesn't. For that you need the following: > (1 2 3) copy 4 push Why is copy necessary here?
    - codr7 7y ago
      Stacks are reference types, the operation doesn't produce any new values. Half the time you want it there to push more items, or to display it in the REPL like in this case, but the other half you're going to drop it anyway. I'm still undecided on the stack ops. There is one case in forthy2 where a reference is left on the stack, and that is splicing values into forms [0], where it's more likely you want to keep working with the same form. More likely is not a very good heuristic for designing API's though, so I might just go with the behavior you expected. [0] https://github.com/codr7/forthy2#quoting https://github.com/codr7/forthy2#quoting
    - codr7 7y ago
      For the record, this should now work as expected without copying.
protomyth 7y ago
The whole concept of a 'body' that isn't on a the stack when the word executes would take a bit of getting used to.
- codr7 7y ago
  The implementation is definitely less Forthy as a consequence of embedding in C++ rather than aiming for rock bottom. Bodies, or scopes, are first class though. You can take their references and/or quote them.
  - protomyth 7y ago
    Postscript does the: num {op} repeat form with the ability to take references of the code blocks using def.
    - codr7 7y ago
      Nice. I get the feeling that PostScript is close in spirit, they were also into pimping Forth using Lisp as inspiration. I think some of the HP-calculators also had similar languages, but never used them myself. For me, generic methods (or words in Forth lingo) is the biggest deal. Followed by scopes and compound literals such as stacks and scopes.
0xdeadbeefbabe 7y ago
CMake Error at CMakeLists.txt:6 (add_link_options): Unknown CMake command "add_link_options".
- codr7 7y ago
  Aight, seems like I need to adjust the required CMake version. It would help if I had any idea which version introduced add_link_options. Anyways, the easiest way forward for you is most probably to upgrade your CMake.
  - slavik81 7y ago
    It was added in v3.13. It would be nice the minimum version required to use a function was directly stated in the CMake documentation. I just keep picking older versions from the dropdown until the page disappears. https://cmake.org/cmake/help/v3.13/command/add_link_options.html https://cmake.org/cmake/help/v3.13/command/add_link_options....
    - codr7 7y ago
      Thank you, I'll update the file accordingly. And I agree: once you have a directive for checking versions, knowing which goes where is pretty critical information. Maybe someone should remind them?
InternalCake 7y ago
There was a burst of interest in concatenative languages a few years back - essentially higher level variations on the features found in forth. https://en.wikipedia.org/wiki/Concatenative_programming_language https://en.wikipedia.org/wiki/Concatenative_programming_lang... They're fascinating and one would hope (well, I did) that the structural simplicity of them would yield something that's amenable to automatic manipulations. In practice the lack of restrictions on what you can do to the stack makes it almost impossible to reason about programs written in such languages.
- codr7 7y ago
  They make great foundations, since their simplicity allows them to be conveniently adapted to whatever purpose. I too find dogmatic Forth a struggle to reason about, which is why we're here I guess. A type system in combination with generic methods help a lot, since most mistakes are caught in dispatch and give reasonable error messages.
- 0xdeadbeefbabe 7y ago
  Forth has a data stack and return stack. Messing with the return stack seems like a practical joke on the programmer sometimes. I guess we don't live with the same constraints that made forth popular in the first place.
  - codr7 7y ago
    I'm curious, what kind of messing with the return stack are you talking about? The only thing forthy2 does that is slightly out of the ordinary is TCO.
    - 0xdeadbeefbabe 7y ago
      In eforth doVAR is R>
      - codr7 7y ago
        Ah yes, hardcore Forthers do seem to take pride in finding the absolute smallest number of fundamental building blocks. There comes a point, though; where pretending everything is the same complicates rather than simplifies.
0xdeadbeefbabe 7y ago
Did you do anything special to make the CPU happy/efficient with forth function calls? I heard a rumor that CPUs find it hard to predict branches with code that doesn't do C calling conventions.
- codr7 7y ago
  One of the main motivations in this case is embedding in C++, rather than trying to replace it. So far the language is implemented as a high level VM/interpreter, I'm planning to eventually emit C++ to get standalone executables and a 20-30% speed up. So no, I didn't do anything :) It's all C++, all the way. I suspect most standard Forths don't give a crap about the CPU, they like to implement everything themselves.

65 points by codr7 7y ago

Why does pushing drop the stack you are pushing onto? Is there ever any reason to push something onto a stack and then immediately drop the stack? Even if there were, it doesn't seem like the option you'd want to optimize for.

codr7 7y ago
I'm sorry, I'm not quite following. I don't get what you mean by dropping the stack in this context. Could you point to where in the README you're having trouble?
- garmaine 7y ago
  Not from the README. I don't read READMEs. I read code. The issue is this: > (1 2 3) 4 push My expectation is this should yield "(1 2 3 4)". It doesn't. For that you need the following: > (1 2 3) copy 4 push Why is copy necessary here?
  - codr7 7y ago
    Stacks are reference types, the operation doesn't produce any new values. Half the time you want it there to push more items, or to display it in the REPL like in this case, but the other half you're going to drop it anyway. I'm still undecided on the stack ops. There is one case in forthy2 where a reference is left on the stack, and that is splicing values into forms [0], where it's more likely you want to keep working with the same form. More likely is not a very good heuristic for designing API's though, so I might just go with the behavior you expected. [0] https://github.com/codr7/forthy2#quoting https://github.com/codr7/forthy2#quoting
  - codr7 7y ago
    For the record, this should now work as expected without copying.

protomyth 7y ago

The whole concept of a 'body' that isn't on a the stack when the word executes would take a bit of getting used to.

codr7 7y ago
The implementation is definitely less Forthy as a consequence of embedding in C++ rather than aiming for rock bottom. Bodies, or scopes, are first class though. You can take their references and/or quote them.
- protomyth 7y ago
  Postscript does the: num {op} repeat form with the ability to take references of the code blocks using def.
  - codr7 7y ago
    Nice. I get the feeling that PostScript is close in spirit, they were also into pimping Forth using Lisp as inspiration. I think some of the HP-calculators also had similar languages, but never used them myself. For me, generic methods (or words in Forth lingo) is the biggest deal. Followed by scopes and compound literals such as stacks and scopes.

0xdeadbeefbabe 7y ago

CMake Error at CMakeLists.txt:6 (add_link_options): Unknown CMake command "add_link_options".

codr7 7y ago
Aight, seems like I need to adjust the required CMake version. It would help if I had any idea which version introduced add_link_options. Anyways, the easiest way forward for you is most probably to upgrade your CMake.
- slavik81 7y ago
  It was added in v3.13. It would be nice the minimum version required to use a function was directly stated in the CMake documentation. I just keep picking older versions from the dropdown until the page disappears. https://cmake.org/cmake/help/v3.13/command/add_link_options.html https://cmake.org/cmake/help/v3.13/command/add_link_options....
  - codr7 7y ago
    Thank you, I'll update the file accordingly. And I agree: once you have a directive for checking versions, knowing which goes where is pretty critical information. Maybe someone should remind them?

InternalCake 7y ago

There was a burst of interest in concatenative languages a few years back - essentially higher level variations on the features found in forth. https://en.wikipedia.org/wiki/Concatenative_programming_language https://en.wikipedia.org/wiki/Concatenative_programming_lang... They're fascinating and one would hope (well, I did) that the structural simplicity of them would yield something that's amenable to automatic manipulations. In practice the lack of restrictions on what you can do to the stack makes it almost impossible to reason about programs written in such languages.

codr7 7y ago
They make great foundations, since their simplicity allows them to be conveniently adapted to whatever purpose. I too find dogmatic Forth a struggle to reason about, which is why we're here I guess. A type system in combination with generic methods help a lot, since most mistakes are caught in dispatch and give reasonable error messages.
0xdeadbeefbabe 7y ago
Forth has a data stack and return stack. Messing with the return stack seems like a practical joke on the programmer sometimes. I guess we don't live with the same constraints that made forth popular in the first place.
- codr7 7y ago
  I'm curious, what kind of messing with the return stack are you talking about? The only thing forthy2 does that is slightly out of the ordinary is TCO.
  - 0xdeadbeefbabe 7y ago
    In eforth doVAR is R>
    - codr7 7y ago
      Ah yes, hardcore Forthers do seem to take pride in finding the absolute smallest number of fundamental building blocks. There comes a point, though; where pretending everything is the same complicates rather than simplifies.

0xdeadbeefbabe 7y ago

Did you do anything special to make the CPU happy/efficient with forth function calls? I heard a rumor that CPUs find it hard to predict branches with code that doesn't do C calling conventions.

codr7 7y ago
One of the main motivations in this case is embedding in C++, rather than trying to replace it. So far the language is implemented as a high level VM/interpreter, I'm planning to eventually emit C++ to get standalone executables and a 20-30% speed up. So no, I didn't do anything :) It's all C++, all the way. I suspect most standard Forths don't give a crap about the CPU, they like to implement everything themselves.