-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
It seems a bit overkill to use Bytes for encoding. Bytes makes more sense in a situation where data is being streamed (e.g., over TCP) with one thread pushing to the buffer while another reads. With the ability to read an entire datagram at once the only niceity is
the API, but with an unnecessary Arc overhead.
Additionally, I'm fine with an API that requires the user to specify the buffer size and error if the buffer would be overwhelmed. That's usually an indication that we should reject the data to begin with.
Metadata
Metadata
Assignees
Labels
No labels