� wrote: > And the idea of the MOP is that the meta object should have the same > interface for all dynamic languages.
That is the intent of Attila's project. It provides a set of common interfaces/classes for method invocation and property lookup (both with either indexed or named parameters), type coercions (like GString to String and back), and composability (to allow a runtime to e.g. failover to Java dispatch if it optz out of handling a call itself). The primary things missing from dynalang that I was were: * First-class representation of a closure or passed function. This would help make closures a bit more translatable across languages * Ability to query for a "callable" that could then be cached for future calls. This would fit better into inline caches, method handles, indy, and so on. Along with this would probably need to come a way for "callables" to be invalidated if things change. - Charlie _______________________________________________ mlvm-dev mailing list mlvm-dev@openjdk.java.net http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev
