Tim Bray has an entry that is a continuation from Gilad Bracha, and others have continued from there. Dynamic languages, continuations, and the JVM... the thing about continuations is they are not immediately obvious because they are not available in most languages.
Those that have them have found they are not useful for the everyday programmer's applications. But they do provide a fundamental building block that simplify the several of the building blocks that *do* end up in the everyday programmer's applications.
By the way, continuations *are* implemented in the Rhino Javascript system and the SISC Scheme system for the JVM. I believe they are also in the JVM and dotnet backends for the Bigloo Scheme system. They would be more efficient in the JVM per se, but I'm just saying...
You're not going to see an overwhelming number of programmers saying, "we want continuations!" That will not be an effective gauge for determining their value at the VM level.
No comments:
Post a Comment