sincze wrote: ↑Sunday 09 July 2023 17:31
- Ask somebody to FIX what somebody else broke (as was done before when the problem appeared in OpenZwave.)
IMO, especially with radios that are less reliable than z-wave, receiving bad figures should be considered possible from any HW: So this kind of limit may have been implemented in a more generic way for any sensor whatever the underlying technology/driver used, with per device limits in sensors configuration allowing (if limits filled by user) to filter out bad values from DB/graphs.
Sometimes (but only 1 or 2 measures in a year, so someday if graphs are flattened by a bad value it's still acceptable to remove it manually), my Qubino Smart-Meter measuring whole home electric power use reports a data madness that would need a nuclear power-plant in my garden to feed my home!
But must agree Neo-Coolcam really made crap devices. I also have some (but no more plugs, they all failed after a few weeks/months in their ~2018 batches and opening the last one really made me fear of some fire hazard and definitively ban anything using 220Vac from this brand) and even some battery devices cause issues (but at least no risk): Some PIRs series misbehave sending false detections almost continuously under 20/25% reported battery level, for instance! So even battery low reports should be made configurable with their devices (+skip any report under setup level) to make them reliable...
They even messed-up with different zwave HW sharing the same IDs in the past, needing some OZW configuration files dirty hacks (luckily, this was possible because at least config differences did not overlap) to cope with 2 hardwares/1config file, that was not supposed to happen.
Anyway, if not buying new zwave devices anymore, you may stick with OZW if (like me) happy with it.