yet another sublanguage? yet another bunch of idiosyncratic syntax to learn and forget over the next 3 year ecosystem churn thing? really? aren’t there enough to choose from already?
why couldn’t it be a library with regular oo, class or prototype bindings for languages / environments that people care about?
why couldn’t it be a subset of JSON or an extension of JSON? why doesn’t it look like jQuery or html5 selection?
in case people don’t know, people have literally designed and implemented THOUSANDS! of languages and frameworks. there is nothing new. there is nothing so novel that it hasn’t been done before or that can’t be conveniently fit into the existing infrastructure. just because you can roll out new infrastructure in the blink of an eye doesn’t make it a good idea.
fb might have thought that REST is limited but in 3 years everyone is going to be complaining about how hard it is to deal with the limitations of graphQL. the web is already loaded with bandaid after bandaid trying to compensate for design flaws in html/css and limitations on browser resource management. stop wasting people’s time with bad ideas. curtail the irrational exuberance and the exquisitely bad design!
rtfm or, in this case, the computer science literature.