Hi all, I have a big change set coming for the OBD Poller that has really improved what I have been able to do with the responsiveness of the HUD and should provide a framework for consolidation of custom polling designs. The aim was to: - Have a single interface to the poll list that moves the one-shot blocking polling and the default list polling behind a single interface - Have the ability to have multiple poll-lists / responders active at once. - Be able to have responders that could use any spare polling time to continually update a value without impinging on moving to the 'next poll count'. (For example speed and rpm) - Be able to have a once-off responder that is non-blocking (for example a request for the VIN which is a once-off but requires the car to be 'on' for ioniq 5). All this I have implemented and am just finalising now - doing some stability testing and going through it to make sure it does what it needs to. The bulk of ISOTP and VWTP code has been left alone, with the exception of where it used to interface back to the poll queue / once-off blocking poll. I have a class that represents a poll responder, and a class that handles a list of responders. 'Blocking' handlers are treated specially. Implemented so far is: List Poller - to handle the standard list of entries Blocking Once-Off poller - to handle PollSingleRequest Data Buffer List Poller - handles a standard list but calls back with the complete data rather than frame-by-frame. Once-Off Poller - Handles a single request (with retries) but is non-blocking. On my TODO list is a special poll responder that allows us to add user entries that would grab/interpret particular ecu addresses on the fly and place them either in metrics or a special temporary metric area. This would enable better testing of values while driving (for example, I want to know which flag is for the brake being pressed, and which one is for the brake light, and I can use the ECU to display these on a HUD without recompiling). I would appreciate any thoughts on this. I might be pushing it up as a draft soon, but the usage of it in the I5 code does depend on the sub-tick p/r so I might wait for that. //.ichael