> In the case of wrapper integers in records, are 4-byte records handled as > efficiently by the compiler as 4-byte integers? I suspect it's > some radically > different code that handles it. test it?