> Actually, are these the only two types of file we're going to > generate? Won't there be unoptimized bytecode, parse trees, etc? Do > we want an integrated set of extensions (.pa?)? Would it be a bad idea to have just a single extension for binary data? If the file format supports header info or multiple sections, then we wouldn't 'pollute' the extension namespace. Is the ability to identify the type of these binary files by extension important? --JustinThread Previous | Thread Next