I learned Forth in ‘80 on a Z80. I had first learned Pascal, then Z80 assembly, some Basic… but I loved Forth. Especially for the time because it was super small and fast (and my computer at the time had 56k of memory so small mattered). I later helped IBM implement OpenFirmware. But other than that I haven’t done a lot of Forth commercially, but I still tinker around with it from time to time.
There are two things that I feel that any alien species will have that are more or less identical to earth, the game of Go and the Forth programming language.
You mentioned dedicated Java processors at 4:25, but go on to say that while there are a few physical implementations of the Java VM, they aren't "huge serious chips". Did you know that ARM introduced Jazelle DBX - Direct Bytecode eXecution - with the ARM9, mostly (but not exclusively) found first in the ARM926EJ-S variant, providing direct hardware execution support for a subset of those bytecodes that map well onto existing CPU functions, and providing a mechanism to trap into software handled bytecodes for those it didn't support in hardware. It led to the quip that ARM can hardly be called RISC these days since it actually supports multiple instruction sets. Jazelle in the original ARM9xx implementation wasn't hugely successful, apparently because the quality of JIT compilers meant that even without implementing specifically for Jazelle, the code would not suffer much in performance at all. Even so, we see Jazelle live on in the form of the ARM Thumb EE instruction set, which is still available in the ARM Cortex A9, but again got deprecated by ARM in 2011. Even so, there are still many new and current devices still being made with these two cores, the ARM926EJ-S and the ARM Cortex A9.
Nice to see more research! In my cheap book of computer languages note on Forth said only that author developed it to control astronimic telescope. Fascinating and given in interesting way story. Also i remeber from 8-bit era available languages In my country was BASIC, LOGO and FORTH.
And lets not be too hard on Fortran, according to Episode 55, they did a pretty job creating the first high level language. A different (yes very complex) tool in the toolbox with limited appeal, but darn good at what it does.
Infix for functions of two arguments does make a lot of sense: you keep the function as close to the operands as possible. It doesn't make much sense for functions of more than two arguments though. There are languages focusing on the infix form, mainly APLs and other array-based languages, and they usually restrict functions to up to two arguments, with some asterisks of course.
I learned Forth in ‘80 on a Z80. I had first learned Pascal, then Z80 assembly, some Basic… but I loved Forth. Especially for the time because it was super small and fast (and my computer at the time had 56k of memory so small mattered).
I later helped IBM implement OpenFirmware. But other than that I haven’t done a lot of Forth commercially, but I still tinker around with it from time to time.
Forth is more than just a computer language. It's an entirely different model of computing and how one interacts with a computer.
There are two things that I feel that any alien species will have that are more or less identical to earth, the game of Go and the Forth programming language.
I am glad you spoke Lisp too. I have had similar thoughts.
You mentioned dedicated Java processors at 4:25, but go on to say that while there are a few physical implementations of the Java VM, they aren't "huge serious chips". Did you know that ARM introduced Jazelle DBX - Direct Bytecode eXecution - with the ARM9, mostly (but not exclusively) found first in the ARM926EJ-S variant, providing direct hardware execution support for a subset of those bytecodes that map well onto existing CPU functions, and providing a mechanism to trap into software handled bytecodes for those it didn't support in hardware. It led to the quip that ARM can hardly be called RISC these days since it actually supports multiple instruction sets. Jazelle in the original ARM9xx implementation wasn't hugely successful, apparently because the quality of JIT compilers meant that even without implementing specifically for Jazelle, the code would not suffer much in performance at all. Even so, we see Jazelle live on in the form of the ARM Thumb EE instruction set, which is still available in the ARM Cortex A9, but again got deprecated by ARM in 2011. Even so, there are still many new and current devices still being made with these two cores, the ARM926EJ-S and the ARM Cortex A9.
Nice to see more research! In my cheap book of computer languages note on Forth said only that author developed it to control astronimic telescope. Fascinating and given in interesting way story. Also i remeber from 8-bit era available languages In my country was BASIC, LOGO and FORTH.
I'm kinda fond of the Harvard Architecture. I program PICs in assembler for fun.
And lets not be too hard on Fortran, according to Episode 55, they did a pretty job creating the first high level language. A different (yes very complex) tool in the toolbox with limited appeal, but darn good at what it does.
And a PDF of the article you were looking for has been sent to your email.
The TI 99/4a had a Forth cartridge, and apparently a non-TI Forth cartridge written in TMS9900 assembly is still available.
Great!
The Lovell Telescope I can see from window uses Forth for its systems.
EDIT: I wrote my comment a long time before you mentioned it. 🙂
"Fast processor chip takes instructions directly from forth" sounds like a porn title to me wtf lol
Is infix really logical though? Here is a value, here is an instruction, and here is another value? 🙃😉
Infix for functions of two arguments does make a lot of sense: you keep the function as close to the operands as possible. It doesn't make much sense for functions of more than two arguments though. There are languages focusing on the infix form, mainly APLs and other array-based languages, and they usually restrict functions to up to two arguments, with some asterisks of course.