> > UTF8 is a variable-width encoding so if the buffer is >100kB and a valid > multi-byte UTF8 encoded character happens to fall across the boundary of > buffers in subsequent loop iterations you might get either an incorrect > decoding or an exception thrown. >
Right, decoding the bytes in fixed blocks with GetString is very fragile. If you want to make a string out of unpredictable chunks of bytes (as they would come out of a network stream for example) you need to use this: http://msdn.microsoft.com/en-us/library/system.text.encoding.getencoder.aspx This is just a side issue to your original post. Greg K
