Text can be multiple bytes per character.Īpplications that are reading data typically are reading the file as 'binary' but deserializing that data into data structures in the application. How a file is read is either going to be in 'binary' mode, read as a stream of bytes, or as a text stream, where the stream of bytes is interpreted as text depending on an encoding used. They stay 'binary', which is just a stream of bytes. There is no such thing as 'converted back when opened'. Yeah, you didn't clear up any of the naivety.
