Closed marler8997 closed 2 years ago
UTF-8 has a not-as-tight encoding, but allows detection of in-stream data.
This makes sense for text, but not for a fixed binary format where you cannot decipher if you're in the file or at the beginning. The current encoding is more compact, as it doesn't require this feature.
Ah yes that's true. Thanks for your thoughts.
I'm reading the specification and was curious if you considered using UTF8 rather than a custom integer encoding? If so, what are your thoughts? If not, do you think using UTF8 might be a good alternative?