Mathijs de Bruin

@cedkoffeto Appreciate your feedback but the phrasing comes across a bit demanding, particularly considering that this is an Open Source project. Please in the future try and be more mindful...

@dosu What would you suggest as answers for my aforementioned questions?

@willydouhard Would love your ideas on this one, should be really easy to implement. Are there any security or other implications?

> @willydouhard would this PR #1260 help ? For HTTP hooks, I am using ContextVar like this in server.py > > ```python > # Create a context variable to store...

Having had initial discussion with Willy, my tentative view (anticipating more reflection) is: if we ensure there's a _single hook_ (e.g. independent of auth) where we've access to the `Request`...

@tambagh Could you please indicate which version of chainlit you were using? If you haven't already, could you please try and replicate the problem with the latest pre-release (1.3.0rc0)?

@enumz Thanks for the contrib! We're finally getting down to catch up, sorry it took such a long time! I noticed the PR's description is empty. Could you please describe...

Closing due to staleness and inactivity. @enumz @khalilha @mikkelgthang Feel free to reopen and/or shoot a new PR with conflicts solved and tests passing. :)

> @dokterbob , it is in the source code. https://github.com/socketio/socket.io/blob/d0fc72042068e7eaef448941add617f05e1ec236/packages/engine.io-client/lib/socket.ts#L1048 > > I think we will also need to disconnect and then reconnect to make sure we don't leak any...

> https://socket.io/docs/v4/how-it-works/#upgrade-mechanism So what your patch is doing (please bare with me) is retrying to upgrade to websocket instead of keeping polling if the websocket upgrade fails? Is this not...