f2u

718525ae-a057-11eb-a679-ebc1ee2a486b Sun, 18 Apr 2021 12:04:59 -0300

"To master riding bicycles you have do ride bicycles"

This is still just a "brainstorm", really is a documentation of changes.

---

Table of Contents

• [Introduction]
• [Size or Speed]
• [Details]

This is a evolution from what I learning with u2forth, ATMEGA8, gcc assembler and forth implementations as 328eforth, eforth, amforth, avr-forth, gforth, flashforth, punyforth, etc

(I need to learn how to use and format with .md)

https://github.com/matiassingers/awesome-readme

https://silentlad.com/how-to-write-beautiful-and-meaningful-readme.md

Introduction

In the chat session Chuck was asked, "How did you come to the conclusion that Forth was too complex, and that sourceless programming was your next move?" His reply was, "Maybe by reading the Forth Standard." Moore, 2002

Forth is language based in thread code, with a dictionary of words as named routines and two stacks for arguments in a forever read–eval–print-loop.

The dictionary is a linked list of words, each with name and code. It have two types of words, those called primitives, aka leafs (leaves), which are coded in specific CPU or MCU instructions, and those called compounds, aka twigs, which are sequences with references to words.

"A most important aspect of FORTH is its ability to define new words. New definitions and code are devised continously. Likewise, new constants or variables are created.", Rather, Moore, Hollis, 1974

I want a forth with:

• use the minimal indirect thread code model, MITC

• the minimal inner interpreter and primitives words (clock, uart, interrupts, stacks, math, moves) dependent of a MCU family;

• all compound words independent of any specific MCU family, without any assembler specifc code inline, like a imutable list with rellocable references.

The dictionary is a unique linked list composed by: a) a link for previous word's link, b) a byte with size of name plus flags, c) a sequence of characters of name (with a pad byte, if need) and d) a sequence of, or references or assembler code.

The word names, until 15 characters, are full preserved, not hashed, not lowered, for trace back compound words, word by word. As well, 15 letters are enought to almost 86% of common words in many languages

I like use the terms leaf for primitives words and twig for compound words, a tree view.

Size or Speed

"Keep it Simple"

Most of Forth implementations goes "runnig for speed" for timming applications or simply to be "the most faster than", but when memory space is the critical limit most of design decisions must take another path.

A MCU 32 bits, as ARM Cortex-M4 STM32L4+ with 2048k bytes flash and 640k bytes sram, really have 512k flash mcu words and 160k mcu words sram. Sure, that is much more than enough to run Forth.

But my choice for design is a small AVR Atmega8, a complete RISC MCU with harvard memory architeture, 4k words (16-bits) program flash memory, 1k bytes (8-bits) static ram memory, 512 bytes of EEPROM, memory-mapped I/O, one UART, one SPI, one I2C, with 32 (R0 to R31) 8 bits registers, with some that could be used as eight (R16 to R31) 16 bits registers.

I intend to use the AVR Atmega8 as if it were a PDP-11, but with some adaptations due to the Harvard architecture and only 1k bytes of SRAM.

Using inner (address) interpreter as dispatcher for code to reference (twigs), and for code to execute (leafs).

All code are made with absolute jumps, imediate or indirected, the common instructions of call and ret are used only when is absolutely need.

UPDATE: 16/11/2022 changed to ATMEGA328, for more Flash (32 kb), SRAM (2 kb) and EEPROM (1kb) memory.

A challenge.

There are many low cost MCU with far more resources and pleny of SRAM and flash. Why use an old MCU for hosting Forth ?

Most to refine language paradigms and understood manage memory, RAM and FLASH, and how forth works inside, looking from behind the stacks.

Many challenges to resolve, how make a minimal bios, what basic word must set, how update flash memory, how access internal resources, etc. Learn from previous many implementations of Forth and adapt to survive.

There are also Forth implementations for many AVR RISCs, mostly for the larger ones like Amforth, Avrforth, CamelForth, Eforth, Flashforth. Studying these codes, as references for models and styles, has been an excellent learning experience for comparing solutions and techniques.

The notes resumes some decisions and details.

the PDP-11

the DEC PDP-11 was the most important tool for creation of Internet, UNIX, language C and language Forth.

For comparation, in 1979, the Von Newman architeture PDP-11, Programmed Data Processor, was eight 16-bit registers, (including one stack pointer R6, and one program counter R7), memory mapped devices, but with a Von Newman unifed memory addressing. .

The PDP-11 had successful implementations of Forth, by Charles H. Moore reported in 1974, and a Fig-Forth by John S. James in 1980.

the 1130

Charles H. Moore also creates Forth in a IBM-1130, back 1968, and "The IBM 1130 was a fairly simple 16-bit computer, with a load-store architecture, a single accumulator and three index registers X1, X2, X3, stored at locations 01, 02, 03 in the core memory."

More Docs ?

please follow to the docs diretory

ForthS

Some special insights from

jonasforth, Richard WM Jones

sectorforth, Cesar Blum

lbforth, Lars Brinhhoff

polyForth, GreenArrays

muForth, David Frech

amForth, Matthias Trute

flashforth, Mikael Nordman

camelforth, Brad Rodriguez

avrforth, Daniel Kruszyna

fig-forth, John S. James

Licensing:

GNU Public License v2 (GPL)