Yeah I'm somewhat confused as well. I even somewhat agree that using only vanilla SQL might make it tricky to answer some complex questions (not limited to ques

Yeah I'm somewhat confused as well. I even somewhat agree that using only vanilla SQL might make it tricky to answer some complex questions (not limited to questions about sequences), but why would that be your goal? The idea of a DSL to analyse sequences of events is interesting, but I'm not convinced by their examples (I mean if you want the order of events, why store them as columns of booleans saying X happened?). They also claim without proof that tables of events are somehow 'arcane' and hard to deal with, but the examples they give aren't exactly complicated even using just SQL (the best I can give them is that they may not be solvable without writing a query, but a new query language doesn't solve that problem).