Ping requests pollute websocket receive streams

[dubnom]

When receiving data from a websocket, it returns the payload for a ping request. Line 719 is response.py should return "None" instead of "payload".

[michalpokusa]

Hi, can you provide the your code snippet for me to test this? How are you sending data to websocket, using JavaScript from browser or in other way?

[dubnom]

I’m using a raspberry pi running ubuntu and communicating through a python script. It will take me a while to create a snippet. The server is on the esp32c3 and a ping_interval is set. The problem is pretty easy to understand, even without any demo code - when a ping request is sent, line 719 responds with the same message but with the pong opcode. None of this should be seen by the user of the web socket, it is out-of-band. Returning the payload is basically returning 4 bytes of junk to the websocket receiver. Simply returning None allows the reader to disregard the message, which is the appropriate thing to do. If you would still like a snippet, I’ll put one together. Thanks - Mike On Jan 8, 2026, at 2:01 PM, Michał Pokusa ***@***.***> wrote: [https://avatars.githubusercontent.com/u/72110769?s=20&v=4]michalpokusa left a comment (adafruit/Adafruit_CircuitPython_HTTPServer#110)<#110 (comment)> Hi, can you provide the your code snippet for me to test this? How are you sending data to websocket, using JavaScript from browser or in other way? — Reply to this email directly, view it on GitHub<#110 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AEFUPCVESRIV7FMZNIQVNM34F2SR7AVCNFSM6AAAAACRDFOTU6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTOMRVGIZTQOJQG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[michalpokusa]

I see, it seems the snippet might be not required there. In the meantime I checked the RFC 6455 The WebSocket Protocol https://datatracker.ietf.org/doc/html/rfc6455#section-5.5.2, and it seems like you are correct, I will make sure that is the case and publish PR if so.

Thanks

[dubnom]

Awesome. Next time I see PT or Limor, I’ll put in a good word. Thanks - Mike Sent from my iPhone On Jan 8, 2026, at 2:15 PM, Michał Pokusa ***@***.***> wrote:  [https://avatars.githubusercontent.com/u/72110769?s=20&v=4]michalpokusa left a comment (adafruit/Adafruit_CircuitPython_HTTPServer#110)<#110 (comment)> I see, it seems the snippet might be not required there. In the meantime I checked the RFC 6455 The WebSocket Protocol https://datatracker.ietf.org/doc/html/rfc6455#section-5.5.2, and it seems like you are correct, I will make sure that is the case and publish PR if so. Thanks — Reply to this email directly, view it on GitHub<#110 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AEFUPCXJEYIEWZYJR5PONTL4F2UEFAVCNFSM6AAAAACRDFOTU6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTOMRVGI4TMNJRG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dubnom]

Also, while I was looking at the code, I noticed that “continuation” frames aren’t handled, and return nothing. A simple solution is to have a flag that allows/disallows continuation frames at create time. Change the behavior from returning None, to returning the payload, and set a flag in the websocket, which the application checks using and api call, that indicates if the frame is a continuation or not. I’m not using this, so I don’t know how important it is, but it can be handled in only a few lines of code. - Mike On Jan 8, 2026, at 2:15 PM, Michał Pokusa ***@***.***> wrote: [https://avatars.githubusercontent.com/u/72110769?s=20&v=4]michalpokusa left a comment (adafruit/Adafruit_CircuitPython_HTTPServer#110)<#110 (comment)> I see, it seems the snippet might be not required there. In the meantime I checked the RFC 6455 The WebSocket Protocol https://datatracker.ietf.org/doc/html/rfc6455#section-5.5.2, and it seems like you are correct, I will make sure that is the case and publish PR if so. Thanks — Reply to this email directly, view it on GitHub<#110 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AEFUPCXJEYIEWZYJR5PONTL4F2UEFAVCNFSM6AAAAACRDFOTU6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTOMRVGI4TMNJRG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[michalpokusa]

I will give adding support for continuation frames a go once more, I do not remember exactly why I skipped it when I was adding Websockets, maybe I had some problems I couldn't resolve at that time.

Together with the response on PING this will clear the Websocket implementation.

I will sit down to it during the weekend and report back on the progress.