Revision as of 10:58, 17 February 2011 edit Ruud Koot (talk \| contribs) 31,416 edits →Does Haskell allow functions as keys in a map?: re ← Previous edit		Revision as of 11:01, 17 February 2011 edit undo Ruud Koot (talk \| contribs) 31,416 edits →Does Haskell allow functions as keys in a map?: add link to paper Next edit →
Line 294: : The most common definition you'll find in the literature is that functions can be passed as arguments (higher-order functions, downwards funarg problem) and returned and stored as values (upwards funarg problem). : Now I've thought about it bit more, related to the your question is the concept of function equality ([[intensional equality]] vs. [[extensional equality]] vs. ~~ad-hoc~~reference equality, see e.g. http://www.cs.princeton.edu/~appel/papers/conteq.pdf). Its not directly related to first-classness, but as you raised it in this context it might be useful to discuss in the article. —''[[User:Ruud Koot\|Ruud]]'' 10:01, 17 February 2011 (UTC) : I'd prefer to have a simple table comparing language features (support for higher-order function, function types as the return type, anonymous functions, anonymous function that can refer to various outside their body, etc.) over a plain statement that language ''X'' has "first-class" functions or not. —''[[User:Ruud Koot\|Ruud]]'' 10:54, 17 February 2011 (UTC)

Talk:First-class function: Difference between revisions