From: Yukihiro Matsumoto Date: 2006-10-09T21:07:57+09:00 Subject: Re: [ruby-cvs:18323] ruby: * eval.c (splat_value): use "to_splat" instead of "to_ary" to Hi, In message "Re: [ruby-cvs:18323] ruby: * eval.c (splat_value): use "to_splat" instead of "to_ary" to" on Mon, 9 Oct 2006 11:28:48 +0900, Mathieu Bouchard writes: |> Every language grows in more recent version, as Ruby does, and PL/I did, |> until it grows too much to collapse into a black hole by its own |> gravity. It's kind a like destiny. Otherwise the language will die. | |And when it finally collapses into a black hole, it dies as well, unless |you think of the black hole as some kind of afterlife. (but I don't think |that trapping matter and emitting blackbody X-rays is your idea of an |afterlife) Yes, languages are mortal. I'd rather see Ruby being a black hole, rather than having thermal death. |> Implicit is better than explicit (under some condition), since |> succinctness is power. | |Almost all method calls in Ruby are not that implicit. I don't know why |you'd make a special case out of to_a if you don't want to_a to be the |default while not wanting to_ary to apply all of the time. But then I |don't understand why to_a is not good enough for you because I don't |understand what's too general about it and I don't know what's the |difference between an array, an array, an array, and an array. nil, hash, IO, Socket have to_a. I don't think they are not the target of splat expansion, whereas range and enumerator are. matz.