2 ms·
This seems somewhat pointless, because for compatibility you need to support UTF-16 anyway, so why bother implementing any of the other encodings in addition.
by layer8 3y ago
This seems somewhat pointless, because for compatibility you need to support UTF-16 anyway, so why bother implementing any of the other encodings in addition.
- chrismorgan 3y agoBecause if the server and client both use UTF-8 internally, they can get better performance by agreeing to use UTF-8.
- layer8 3y agoThe question is if the performance gains are noticeable enough to warrant the added implementation complexity and the added integration testing needed. It seems doubtful to me, given the general protocol overhead.