Make & Model of Device: Galaxy S9
Android Version: 10
App Version: 3.292
First, thank you for this great app ! And then, the bug… It’s not always here, sometimes it works, sometimes not. As you can see in the image below, it looks like there is a wrong (very low) initial temperature value in the graph at the transition between Arpege and GFS model, leading to bad scaling and wrong hot front computation.
I’ve not seen quite the extreme you’re showing here, but I have noticed strange discontinuities in the cold front/warm front display. I eventually realized that the cold front/warm front is using data from whatever the new model is instead of comparing data between two models.
What I mean is, if you change your data source from Arpege to GFS (so there’s no mid-week model change) do you still see the same discontinuity or was the GFS temp on the previous day really that low?
Not sure if that’s what’s going on here, but just a thought.
I’ve just got the bug back. Change to GFS then to Arpege does not change anything unfiortunately.
Actually, I didn’t check before but the bug is still there when I use only GFS.
While I was testing other things I manage to get different behaviour changing the city :
Not really, it’s just one of those bugs that is hard to fix. If you find a way to consistently replicate it, then let me know.
Essentially, you have conditions that increase the chances of seeing this problem - what are these conditions? Could be your phone, your OS, your network, your ISP?? It’s not the servers since others don’t experience this as much. For example, I have not seen this issue on my phones in months!!
I could refresh constantly until I replicate the issue but that would take ages and waste my time.
I just hope it gets fixed through the natural evolution of the code.