When the tempest goes into power save mode 3, it stops measuring rain. I guess because it basically shuts down and only wakes up for a short moment once every 5 minutes.
However the graph shows that there was zero rain. There is a difference between zero rain, and not measuring rain. I think it would be better if there just was no data point shown. (just like when it detects no wind).
2 Likes
You have confused reporting period with data recording.
iâve not. rainsensor is disabled in mode 3
No, it has nothing to do with reporting period, it has to do with the fact that the Tempest reports 0 rain rather than NA.
Iâm taking this out of bugs as it isnât really a bug, rather a way to handle some situation.
I think Sunny has a point that showing â0â when actually detection is off might be misleading (for ex youâre 100 miles away and assume all is dry at home or you have some automation linked to rain). I can imagine your sprinklers going wild when itâs raining 
I propose we push this to @corrineb and @rderr as a suggestion to handle the âno detectionâ differently in the flow (if possible). But this might imply deep changes as this should be also in UDP for those using local data ⌠â api changes.
1 Like
No, it is a bug. The tempest should report Null, instead of zero because it didnât do a measurement. Once that is fixed, I assume the rest of the graphing system takes care of it already.
2 Likes
Is the rain sensor circuit completely disabled at this point?
EDIT: At mode 3 it is disabled so there is absolutely no reading until the battery has power and returns to mode 2. That being true, there is no value to report and should be delivered as a NULL.
1 Like
A bug implies it isnât working as designed. It is working as designed so I agree with @eric that it is on another level.
that implies that someone designed it deliberately so when there is no measurement it should report zero rain instead of NULL. I donât think so.
More likely it is just an oversight. At some point it was decided to disable the rain sensor (thatâs a design choice) but they have forgotten to add a line of code that report NULL when it is disabled.
We agree then, it isnât a bug.
You might think so, but no we donât ;-). If I were a programmer I would have added that line of code.
Of course you would. It still wasnât a bug where the code wasnât working as it was written.
code itself is always working as it is written.Thatâs what code does. Still many codes are buggy.
We agree: itâs not a bug. But we also agree: itâs wrong. Letâs call it âbad designâ or maybe just âan oversightâ 
Either way, weâll get a fix in place with the next Tempest firmware build. Thanks to all for the feedback!
4 Likes
Well that just put an end to our semantical ping-pong game. 
5 Likes
Thank you, David.
Otherwise, Is this all that matters in our life?