Perhaps the learning algorithm has been implemented on data that is sent elsewhere but is not showing up on the local console or something weird like that? Anyone have any ideas why the precipitation would be different? Everything else matches up perfectly. Thanks!
Correct. I checked the Weatherflow data and it added up to 0.99" (now 1.01"). This is the same number that is showing on the Weatherflow home page for my station (as well as on the app).
Perhaps. What is really weird is that someone else posted that they were having similar issues with a different service. Also, the numbers that are being shown in the CWOP database are basically correct (within a couple hundredths), while the numbers on the Weatherflow station are several tenths too high. Strange.
Do you have an idea to test, better then using wireshark with stupid manually observation?
Does someone know a logic / timeframing behind the UDP transmissions ? Hence one could be able to do the analysis automatically …
log_raw_packets = False
Enable writing all raw UDP packets received to syslog,
(or wherever weewx is configured to send log info). Will
fill up your logs pretty quickly, so only use it as
a debugging tool or to identify sensors.
I have had problems with the weatherflowUDP driver partially missing or completely missing obs at times. The missed data doesn’t get written to the weewx database and thus the data at CWOP/Mesowest tends to be differnt than the weatherflow station measurements. I haven’t been able to figure out why the data gets missed. I noticed your station has the same issue at times.
How did you reach the conclusion that the weewx driver was missing broadcast data ?
CWOP being different is not necessarily an indication that weewx didn’t receive and ‘send’ that data to them. All these online places drop data (WU is particularly bad in that respect, ignoring all the policy complaints people have about them).
