FORTH DIMENSIONS
JUNE/JULY 1978
EDITORIAL: WHAT IS THE FORTH INTEREST GROUP?
The Forth Interest Group, which developed in the fertile ground of the computer clubs of the San Francisco Bay Area, grew in a few months from nothing to where we are now getting several letters a day from all over the country. With this increasing public interest we need to let people know what we are doing and why, what we would like to see happen, how others can be involved, and what we can and cannot do.
We are involved because we believe that this language can have a major effect on the usefulness of computers, especially small computers, and we want to see it put to the test. Increasingly software is becoming the critical, limiting factor in the computer industry. Large software projects are especially difficult to develop and modify. Few are happy with prevailing operating systems, which are huge, hard to understand, incompatible with each other, and without unity of design.
The Forth language is its own operating system and text editor. It is interactive, extensible (including user-defined data types), structured, and recursive. Code is so compact that the entire systen (mostly written in Forth) usually fits in 6K bytes, running stand-alone with no other software required, or as a task in a conventional operating system. One person can understand the entire Forth system, change any part of it, or even write a new version from scratch. Run-time efficiencies are as little as 30% slower than straight machine code, and even less if the system’s built-in assembler is used. When the assembler is not used, programs can be almost completely transportable between machines. Any large Forth program is really a special-purpose, application-oriented language, greatly facilitating maintenance and modification. We don't yet have conclusive data, but typical program development times and costs seem to be a fraction of those required by Fortran or assembly. Forth is especially useful for real-time, control-type applications, for large projects, and for small machines.
The problem is availability. Users have shown an ease of learning after they have a system available. The Forth characteristics of postfix notation, structured conditionals, and data stacks are best understood by use. To encourage Forth programmers, we need readily available systems even of modest performance. We hope that three levels will be available:
- Demonstration—free (or under $20.) introductory version without file structure which compiles and executes from keyboard input.
