Hundreds of new computer languages are invented each year. By 1966 there were already 700 languages.
Ought we to have standardised on a language, instead of inventing new ones?
Cobol was invented around 1960. It ran on NCR 315 with 32K of memory at NCR in Sydney. (I worked there as a computer operator from late 1967).
Cobol could certainly be used as a “general purpose programming language”, in the sense that any computer programming language is “Turing complete” (or can be extended to be so) and hence is able to be used for general computations.
One could treat Cobol as a base, and write processors in Cobol for other languages – even interpreting a command line may mean implementation of a processor, perhaps for a simple computer language of our design.
Fortran IV has been used as a language processor. William Waite wrote general code in Fortran IV, called STAGE2.
STAGE2 was used to build processors for many languages, e.g. it is said to have been used on the first non-Xerox version of Smalltalk. STAGE2 could easily be implemented in Cobol (although I do not know if this has been done) so this would be a practical way of developing language processors based on Cobol.
Maginnis wrote a compiler generator in Fortran IV. Due to scarcity of tools for the Apple II at Dept of Health in the 1980′s, I rewrote this in Basic and used it. (In hindsight, I think it is possible that the area I was working in was given non-IBM equipment to work with, e.g. Apple II with no software, on the tacit understanding that nothing useful would be developed. Perhaps a more civilised version of a Siberian labour camp).Duration : 0:0:15
| 
