I've released a new version of the scripting library. Most of the changes done for this version is transparent to the application, and is in preparation for the new template types, however there are a few improvements that applications can take advantage of. The tokenizer now recognizes the UTF-8 byte order mark as whitespace, which allows the application to properly compile unicode scripts. The upcoming versions will improve the international support even further. The script compiler now permits prefixing the enum constants with the enum name for the situations where multiple enums declare the same name. The value cast behaviour can now be used by the compiler to instanciate objects during casts even if the appropriate constructor isn't registered. Some bugs with the implicit conversions have also been fixed. The callstack is no longer cleaned up immediately upon catching a script exception, instead it's only cleaned up when the context is released, or reused. This allows for examination of the callstack upon exceptions without using the exception callback. In addition to that several bugs have been fixed, and I recommend checking out the change list for the details. Regards, Andreas