- Author
- Will Crichton
- Published
- 9 September 2018
- Kind
- Personal essay with a short historical survey
- Original
- willcrichton.net/notes/systems-programming/
- Subject
- Systems and hardware
What it is
Will Crichton, then at Stanford, wrote a personal essay on the history of the phrase. It asks why low-level programming and large-system design share one name. It aims to argue for splitting the two.
What it says
- The post says "systems programming" merges low-level programming with systems design.
- It cites a 1972 paper listing system program traits, none of them required.
- Ousterhout's 1998 article casts systems languages as component builders and scripting as glue.
- Panelists in 2014 gave varied answers, including resource limits, hardware control and cloud infrastructure.
- The post says the 1972 traits read like software engineering issues, not low-level performance issues.
- The post calls OCaml and Haskell more systems-oriented than C or C++.
- The post concludes that what is called systems programming should be called low-level programming.
What evidence it gives
The essay quotes three 1972 papers, a 1998 article and a 2014 panel. It offers no experiment or data. The case rests on the readings and on the author's judgment.
Interests and limits
The author says they do not know who coined the phrase. The survey rests on a handful of sources. The "hot take" about functional languages is stated as opinion. No one is shown to have adopted the proposed names.