Flex has gone mad

Before, I was getting discrepancies between Android and PC, now they’re both wrong! Two screenshots below:
PC


and on Android:

The actual rainfall was on 11 and 12 Jul so the PC version seems to have found some rain for 13 Jul (I’ve looked at the PWS website and confirmed that there was no rain recorded on that day which aligns with WU and my local database at the weather station). It also seems to have lost 14 Jul completely. I’ve already noted elsewhere that my Android app has a bug which heads any non-today column with actual date -1 so the water balance shown for Jul 11 on Android is actually data for Jul 12 etc. Taking this into account, the data lines up with the PC browser ie made up precipitation for Jul 13 and also Jul 14 missing. The running balance for 15 Jul is too high because on 12 Jul it should have shown a fall from 0.66 to to 0.33cm (no rain). There should have been another fall on Jul 14 of say 0.25 (no rain) giving a starting balance on Jul 15 of 0.08cm which would normally trigger watering given 0.03 precipitation is forecast.

Any idea what is happening here?!

Did you refresh both your PC webapp and your Android app before looking at the moisture levels? If not, I’ve seen some of the same screwy results. Cleared up after I did refreshes. I have learned over time to ALWAYS refresh my app before doing anything with it.

Normally I’d say a pretty good reply…but I’m afraid I endlessly refresh so in this case no cigar!

My guess is inconsistent programming of time zone conversions in Web vs. mobile versions of the app. Hopefully it just affects the moisture level graphs, not the flex scheduling.

So another day has passed and I’ve updated the screenshots:
PC:


Android:

Allowing for the previously discussed display date error on the Android (interestingly the dates on the actual chart are correct), the rain for Jul 13 is still incorrect, Jul 14 is still missing, Jul 15 “actuals” have been introduced and Jul 16 forecast. Jul 15 actual rain says 0.5mm, not much but in reality there was no rain and none is shown on the PWS listing for the station on the website. I am used to rounding errors as the systems convert between mm and in (The Davis station measures rain in multiples of 0.2mm so there is never a 0.5 reading only 0.4 or 0.6) but rounding 0 to 0.5 is unlikely so I’m not sure where that reading came from.