Closed jdemaeyer closed 4 years ago
Closing as this was a transient issue after adding a new field.
We should refactor the fallback mechanism at one point (to query for each missing field separately and to query more efficiently for very large time frames), but what we currently have seems Good Enough™, and I'd like to prioritize #83 and #57 to make a well-rounded 1.0 release before the end of the funding period.
E.g. for
lat = 51.96
,lon = 7.62
after parsing https://opendata.dwd.de/weather/weather_reports/synoptic/germany/json/Z__C_EDZW_20200722135104_bda01%2Csynop_bufr_GER_999999_999999__MW_385.json.bz2 from 13:56 but before parsing https://opendata.dwd.de/weather/weather_reports/synoptic/germany/json/Z__C_EDZW_20200722135700_bda01%2Csynop_bufr_GER_999999_999999__MW_386.json.bz2 from 14:02 (or the one after?).Possibly this was just the shockwave from deploying the
condition
field? (Which would still have been unfilled incurrent_weather
as the deploy happened around 13:45, so before parsing the 14:02 SYNOP file our fallback station would still havecondition = null
and thus miss the fallback filter)/cc @ubergesundheit