Datetime Indexes in Dgraph - Dgraph Blog

“Granularity of a datetime index, carefully chosen to keep size of each bucket reasonable.” Right.
TimestampDB takes care of this ? Why convert tokens ? when a tokens job is to identify.
Granularity within a bucket I can understand, not to be too fine. but that’s obvious.
why are predicates so painful, when checking can be imposed by the framework.
This entire fix can be summarised in one paragraph, Query error.