Closed hellais closed 2 months ago
If we rely on de-duplication for handling obs_* table row deduplication, since the observation count might decrease in new runs, we should make sure that some duplicate rows are not left around.
See: https://github.com/ooni/data/issues/88 & https://github.com/ooni/data/issues/87
This was done as part of https://github.com/ooni/data/pull/97
If we rely on de-duplication for handling obs_* table row deduplication, since the observation count might decrease in new runs, we should make sure that some duplicate rows are not left around.
See: https://github.com/ooni/data/issues/88 & https://github.com/ooni/data/issues/87