Thanks, you're right and my memory was fuzzy. But the strong claims like that in the paper are about trying to completely erase the local/remote difference, whi

Thanks, you're right and my memory was fuzzy. But the strong claims like that in the paper are about trying to completely erase the local/remote difference, which neither E nor Erlang did. Instead they reduced the pain in a way that this paper leaves an impression is not worth pursuing: > Rather than using those resources in attempts to paper over the differences between the two kinds of computing, resources can be directed at improving the performance and reliability of each. > ... it is a mistake to attempt to construct a system that is “objects all the way down” if one understands the goal as a distributed system constructed of the same kind of objects all the way down. (E.g. in E the objects are the same kind; it's the references to objects that have two kinds.)