#expressionproblem — Public Fediverse posts

🎩🔬 Behold, the groundbreaking revelation: software engineers have thought about data types and operations since... always! 🤯 Eli Bendersky bravely tackles the "Expression Problem" in a never-before-seen saga of #C, #Haskell, and #Clojure, despite the entire internet solving it ages ago. Truly, an epic quest for solutions no one asked for. 🏆
https://eli.thegreenplace.net/2016/the-expression-problem-and-its-solutions #softwareengineering #ExpressionProblem #techhumor #HackerNews #ngated

The Expression Problem and its solution

https://eli.thegreenplace.net/2016/the-expression-problem-and-its-solutions

#HackerNews #ExpressionProblem #Solution #Programming #Languages #SoftwareDevelopment #TechInsights

Eli Bendersky has bravely tackled yet another "age-old" programming conundrum—because clearly, no one thought of solving the Expression Problem before 🤔. Somehow, the solution involves a mystical blend of languages like #C, #Clojure, and #Haskell, as if the more languages you throw in, the more impressive it sounds 😏.
https://eli.thegreenplace.net/2016/the-expression-problem-and-its-solutions/ #EliBendersky #ExpressionProblem #ProgrammingLanguages #HackerNews #ngated

The Expression Problem and its solutions

https://eli.thegreenplace.net/2016/the-expression-problem-and-its-solutions/

#HackerNews #ExpressionProblem #Solutions #Programming #Concepts #SoftwareEngineering

@evan oh no! I only just realised that when you say "Noun" you mean that types are sort of like nouns and terms of the types are sort of like names, whereas functions over types are sort of like verbs and interfaces are sort of like "verbers"!

I thought that we're taking about language processing related domain!

If we aren't, then I'm heavily in "it depends" camp. There is #ExpressionProblem, so while picking tools to encode your problem domain, you have to pick the closeness of extending "nouns" with more "verbs" or allowing a "verb" to work on more "nouns" based on the problem domain. There's no right answer and no reasonable way to prefer one over the other.

But from the standpoint of ease of understanding, facilitated the most by #DeclarativeProgramming, I would pick option 1 because it doesn't rely on traits, is explicit and separates "algorithms" (verb) and "data structures" (Noun) so that people can understand them independently.

@evan oh no! I only just realised that when you say "Noun" you mean that types are sort of like nouns and terms of the types are sort of like names, whereas functions over types are sort of like verbs and interfaces are sort of like "verbers"!

I thought that we're taking about language processing related domain!

If we aren't, then I'm heavily in "it depends" camp. There is #ExpressionProblem, so while picking tools to encode your problem domain, you have to pick the closeness of extending "nouns" with more "verbs" or allowing a "verb" to work on more "nouns" based on the problem domain. There's no right answer and no reasonable way to prefer one over the other.

But from the standpoint of ease of understanding, facilitated the most by #DeclarativeProgramming, I would pick option 1 because it doesn't rely on traits, is explicit and separates "algorithms" (verb) and "data structures" (Noun) so that people can understand them independently.

@evan oh no! I only just realised that when you say "Noun" you mean that types are sort of like nouns and terms of the types are sort of like names, whereas functions over types are sort of like verbs and interfaces are sort of like "verbers"!

I thought that we're taking about language processing related domain!

If we aren't, then I'm heavily in "it depends" camp. There is #ExpressionProblem, so while picking tools to encode your problem domain, you have to pick the closeness of extending "nouns" with more "verbs" or allowing a "verb" to work on more "nouns" based on the problem domain. There's no right answer and no reasonable way to prefer one over the other.

But from the standpoint of ease of understanding, facilitated the most by #DeclarativeProgramming, I would pick option 1 because it doesn't rely on traits, is explicit and separates "algorithms" (verb) and "data structures" (Noun) so that people can understand them independently.

@evan oh no! I only just realised that when you say "Noun" you mean that types are sort of like nouns and terms of the types are sort of like names, whereas functions over types are sort of like verbs and interfaces are sort of like "verbers"!

I thought that we're taking about language processing related domain!

If we aren't, then I'm heavily in "it depends" camp. There is #ExpressionProblem, so while picking tools to encode your problem domain, you have to pick the closeness of extending "nouns" with more "verbs" or allowing a "verb" to work on more "nouns" based on the problem domain. There's no right answer and no reasonable way to prefer one over the other.

But from the standpoint of ease of understanding, facilitated the most by #DeclarativeProgramming, I would pick option 1 because it doesn't rely on traits, is explicit and separates "algorithms" (verb) and "data structures" (Noun) so that people can understand them independently.