Convex instead optimizes for application-layer transactionas and small result sets.
In that world, cross-query optimization and massive parallel execution matter less than predictable transactional behavior close to application code.
And yet again.. its all about tradeoffs
Reading Designing Data-Intensive Applications and stumbled on this:
“Declarative query languages like SQL turned out to be much better than imperative query APIs.”
The reasoning is interesting:
Different assumptions → different abstractions.
At first glance, these ideas feel like they compete.
But I think the key insight is: databases are mostly tradeoff machines.
SQL systems optimize heavily for large-scale set operations where global query optimization matters a lot.
@GergelyOrosz I find that I am most valuable on the edges.
On the one hand I need to be in control on the big picture: the architecture, design, structure.
On the other hand I need to desicde on the specifics in critical areas.
When those are set correctly most other falls right.