As mentioned in #12, files transferred using unMessage almost double their size. The reason for this is that every plaintext information becomes a JSON serialized Element, as well as the contents of the file, a base64 encoded string attribute.
Suggested by @david415, in order to solve this inefficiency unMessage should also support sending information as bytes instead of only JSON. That means the beginning of the plaintext should tell what exactly the following information is and using something similar to sequence numbers might help to identify it as well as allow supporting other important features such as acks.
As mentioned in #12, files transferred using unMessage almost double their size. The reason for this is that every plaintext information becomes a JSON serialized
Element, as well as the contents of the file, a base64 encoded string attribute.Suggested by @david415, in order to solve this inefficiency unMessage should also support sending information as bytes instead of only JSON. That means the beginning of the plaintext should tell what exactly the following information is and using something similar to sequence numbers might help to identify it as well as allow supporting other important features such as acks.