CORBA is basically a definition of object serialization types, paths to refer to objects, a means of sending messages to those objects and also delegating requests to certain methods to backends which can handle those requests.
Unfortunately there are about one thousand pages of technical specs to understand CORBA 3.3.
Later XML-SOAP put the serialization over XML. Requests were either XML or to HTTP routes.
Even later now uses JSON to serialize and HTTP routes to direct methods. The web server uses routing rules to direct parts of the "REST API" to the microservice to handle requests at this endpoint.
Makes me wonder if you compare the specs to implement CORBA to the specs to implement JSON and HTTP how much overlap they ended up having.