Programming language Readability is of, at least, two layers:

  1. Human Readability
  2. Machine Readability.

To date, programming language designers have been emphasizing human readable syntaxes.

To date, we think of programming to consist of only 2 layers:

  • machine code (assembler, etc.)
  • human-readable PLs.

Machine readable languages enable automatic manipulation of source code.

Machine readable languages enable layering, e.g. allowing more than 2 layers for software designs.

Human-readable languages can be automated using parsers. (REGEXP is a subset of parsing).

The belief that only 2 layers are possible, leads to:

  • the belief that PLs can be designed and implemented by only a select few (interpreter and compiler people)
  • the belief that PL technology is split into only 2 parts - compile-time and run-time
  • mixing implementation details with DI (Architectural) details
  • an uneven blend of DI vs. Implementation details in a language (depends on the language designer). For example, data abstraction is a DI concern, whereas low-level operations, like “+”, are implementation details.1

Human Readable

Basically, any language that “has syntax” is meant for human consumption.

Syntax is a way to error-check programs written by humans.

Examples:

  • Pascal
  • Python
  • HTML
  • spreadsheets

Machine Readable

Examples:

  • assembler
  • bash, *sh
  • CL (Common Lisp)
  • JS (JS is schizophrenic - it contains syntax, but also contains machine-level concepts)

See Also

Blog
References

  1. Note that user-defined data structures are actually Implementation details. ↩