Home
Ask HN: Could C find a future in being used instead of pseudocode in papers?
I recently came across a comment of someone asking the author of a blog post to consider writing his code examples (could have been also pseudocode) in Rust instead of C.<p>The answer of the blog author was: “I'm using C in lieu of imperative pseudocode. The benefit it has over pure pseudocode is that it compiles and doesn't rely on some syntax I pulled out of nowhere. The benefit over higher-level languages is that it shows the mechanics of the calculation in a way that's trivial to transpose to any other dialect. This can't be said about code that uses cute but language-specific array manipulation and iteration idioms, e.g.:<p>WalshMatrix = Nest[ArrayFlatten@{{#, #}, {#, -#}} &, 1, #] &;”<p>Do you think C or some future version/adaptation of it could be used to formally substitute all pseudocode examples in the future, especially in proofs/papers?
1 comment
eesmithabout 1 year ago
No.<p>We have historical precedence in ALGOL. From its Wikipedia page, it "was the standard method for algorithm description used by the Association for Computing Machinery (ACM) in textbooks and academic sources for more than thirty years."<p>When was the last time you saw ALGOL?<p>In parallel, we can also look to APL, which arose out of a mathematical notation for arrays. From <a href="https://codegolf.stackexchange.com/questions/162254/generate-a-walsh-matrix" rel="nofollow">https://codegolf.stackexchange.com/questions/162254/generate...</a> the Walsh matrix in APL (Dyalog Unicode) is (⍪⍨,⊢⍪-)⍣⎕⍪1. For those who think in APL, I suppose that makes obvious sense.<p>ALGOL users and APL users would never agree on a single language.<p>You will not get C and Mathematica users to agree in a single language.<p>You will not get Python and Rust users to agree on a single language.<p>People doing computer assisted proofs use tools like the Coq Theorem, which has its own language.