You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using tungstenite right now in an ultra low latency/high performance capacity and was wondering if there's any plans for writing a zero allocation version of read_message() that simply re-uses an existing buffer/slice/vec?
Thanks!
The text was updated successfully, but these errors were encountered:
Hey, that would be a nice addition indeed. We had an idea to move the WebSocket-related logic into its own small crate (something like tungstenite-core) so that we could reuse it in tungstenite-rs (allowing others to use tungstenite-core directly in case they don't want the buffer management).
This question has actually been asked a couple of times in the past, you may want to look at the current open issue about that: #209 (I think it's about the same thing)
Unfortunately, we have not yet worked on that (was not particularly useful for our specific case), but it's definitely a doable thing to achieve (would require some refactorings to make the underlying frame parsing logic not store any buffer inside, but relying on the buffer provided by the user).
Hi,
I'm using tungstenite right now in an ultra low latency/high performance capacity and was wondering if there's any plans for writing a zero allocation version of
read_message()
that simply re-uses an existing buffer/slice/vec?Thanks!
The text was updated successfully, but these errors were encountered: