Open Firmware/OpenBIOS forth on Prop
prof_braino
Posts: 4,313
Has there been any discussion of an Open Firmware/OpenBIOS-type Forth implementation for the propeller? (Don't laugh, bear with me)
I don't see any reason why the current propeller chip could not support a full interactive stand-alone development environment.
A forth kernel for propeller hardware that includes options for TV, VGA, keyboard and mouse would be such.
On such a kernel, options for interpreter, compiler, assembler, disassembler, mass storage support, development utilities, could easily be implemented from existing forth code. A metacompiler would allow for kernel extension / kernel image redefinition.
Target compilation would allow for headerless applications to be generated for dedicated cogs up to the full size of cog memory, as the full development system would not be needed on more than one cog and may be omitted during target compilation at the discretion of the developer.
From this point it could be straight forward to implement a hardware abstraction layer to interface applications written in C or other languages, or go on and develop full applications in forth.
This seems like a (relatively) simple and obvious alternative that offers nearly the execution speed of assembler and the benefits of high level language programming without the learning curve and execution speed penalty for spin, and removes the reliance on an external PC and compiler.
Are there any technical issues that would block this? There already are several forth implementations for the propeller, it looks to me that it is just a question of organizing the community.
Don't get me wrong, I do not mean to cap on Spin or the existing forth implementations. What I describe would be ideal for my own preferences, I imagine others would also find it so.
Anybody interested?
I don't see any reason why the current propeller chip could not support a full interactive stand-alone development environment.
A forth kernel for propeller hardware that includes options for TV, VGA, keyboard and mouse would be such.
On such a kernel, options for interpreter, compiler, assembler, disassembler, mass storage support, development utilities, could easily be implemented from existing forth code. A metacompiler would allow for kernel extension / kernel image redefinition.
Target compilation would allow for headerless applications to be generated for dedicated cogs up to the full size of cog memory, as the full development system would not be needed on more than one cog and may be omitted during target compilation at the discretion of the developer.
From this point it could be straight forward to implement a hardware abstraction layer to interface applications written in C or other languages, or go on and develop full applications in forth.
This seems like a (relatively) simple and obvious alternative that offers nearly the execution speed of assembler and the benefits of high level language programming without the learning curve and execution speed penalty for spin, and removes the reliance on an external PC and compiler.
Are there any technical issues that would block this? There already are several forth implementations for the propeller, it looks to me that it is just a question of organizing the community.
Don't get me wrong, I do not mean to cap on Spin or the existing forth implementations. What I describe would be ideal for my own preferences, I imagine others would also find it so.
Anybody interested?