>>30022039>then I'll eat my words and it was some sort of real phenomenon
I've never said it is, the exact opposite in fact. But the claim that it's testing doesn't sit well with me either.
Of course neither of us know the inner workings of their system, and you are very right that if this is equipment related testing then it probably requires them some maintenance mode or going offline, which is something I've eluded to here >>30020961
Please rephrase the rest of your argument, because it doesn't make a lot of sense to me.
Premise A: These are different radar systems.
Question: Then why do they do their tests in unison down to the minute?
Premise B: These are not different radar systems, only different radars/stations.
Premise B1: This is a hardware test to see if the equipment is working.
Question: Why isn't it done per radar, but all in one? My main reason for asking this is because, like I said (based on the visibility radia shown on windy) some of the stations don't seem to catch a lot of the actual shape.
Premise B2: This is a software test.
Question: Why do we, as the users, see it?
Premise B2.1: This is an integration test, to see if the data flowing in from multiple sources (the radars) can be parsed and handled correctly.
Question: Why does this require the actual hardware? Inputs like these are usually mocks with predetermined outcomes we can assert against/set up constraints for.
Premise B2.2: These are unit tests (or this is MT for some version of QA or something).
Question: Why is it done from one single input? Unit tests, by their very nature, should be reusable, independent from eachother, and deterministic. (and also, why does it require the hardware?)