If we used SQL we also would inherit it's well known shortcomings: - no proper way to extract and reuse expressions - parameterizations are basically string co

If we used SQL we also would inherit it's well known shortcomings: - no proper way to extract and reuse expressions - parameterizations are basically string concatenations - a lot that I forgot There are of course also good things about it: - it's declarative - it's easy to reason about (in its base form, no recursion, etc.) - it's widely used and known So while GraphQL is something similar, and having solved some of SQL's shortcomings, it's new and has less mindshare. Also, last time I checked, the implementations in various languages (Python even) lagged behind the specification. I'd say it's a step in the right direction, but tooling has to be improved I guess.