Most of us will perform some level of quality control on our data and as such we'll end up flagging suspect data. This is yet another feature of our APIs that we might want to standardise. For example when a sensor gets stuck at the same value for hours on end we could add a "persistence" flag to our observation. What we ideally want to avoid is one observatory calling it "persistence" and another calling it "unchanging" and another "delta-fail", and so on...
Most of us will perform some level of quality control on our data and as such we'll end up flagging suspect data. This is yet another feature of our APIs that we might want to standardise. For example when a sensor gets stuck at the same value for hours on end we could add a "persistence" flag to our observation. What we ideally want to avoid is one observatory calling it "persistence" and another calling it "unchanging" and another "delta-fail", and so on...
Should we agree on some of the more common ones?