Factor out connection, peer management, and packet handling into VideoCallClient
#137
Conversation
|
But oh darn, I just tested e2ee (again) and it's not working now. So hold off on committing this, I need to debug. |
… callback hell With some code cleanups along the way
|
Hey, sorry for the long delay -- got distracted by my day job. I've pushed a commit that fixed the problem, by replacing "callback hell" with a more straightforward approach for handling new peers. And did some code cleanup along the way (those went hand-in-hand :) Then while I was in a code cleaning mood I tidied up a few other things in the client & peer_decoder_manager, in a few other commits. Should be good to go now! |
|
I took a stab at rebasing this myself but it is not trivial! |
|
Hey @darioalessandro, @griffobeid any thoughts about how to handle the UI test suite failure? It seems to be due to ...and seems unrelated to my having just merged in main and resolved the conflicts? 🤷🏻 |
|
I think that @griffobeid is looking into this |
|
@ronen @darioalessandro Yeah it is a version mismatch. After doing cargo install -f wasm-bindgen-cli --version 0.2.87Tests are passing just fine, don't worry about it with your changes I can clean this up |
[Continuing down the path of #74]
This PR create a single
VideoCallClientinterface that internally takes care of the connection, packet handling, encryption, peer management, etc. -- moving all that code out of theAttendantsandHostcomponents.Part of that cleanup included changing the encoders so that they too use the VideoCallClient in their interfaces