why does neo4j batch ops look like reinvention of http persistent connections

Question

I'm looking at neo4j batch ops here: http://docs.neo4j.org/chunked/milestone/rest-api-batch-ops.html And it reminds me a lot of persistent connections: https://en.wikipedia.org/wiki/HTTP_persistent_connection

Why not pipeline http requests? Why reinvent the wheel, when you could have the richness of, and existing support for, http?

Answer 1

It also adds transactionality for all requests of a batch, streaming from and to the server and the ability to reference data created during the batch-request.