There’s a huge gap between frontend engineers and backend engineers. People just don’t realize it or ignore it as if it were a fate.

Aggregation Problem: who aggregates data and when?

Let’s begin with a story: Furiosa is a talented frontend developer and Max is an experienced backend developer. They’re a team working on a movie database web app.

Their app has a feature to show movie details with actors appeared in it. It’s backed by an API endpoint /movies/:id:

$ curl http://api.example.com/movies/2 | jq .
{
  "title": "Pulp Fiction",
  "year": 1994,
  "actors": [
    {
      "name": "John Travolta",
      "role": "Vincent Vega"
    },
    {
      "name": "Samuel L. Jackson",
      "role": "Jules Winnfield"
    }
  ]
}

Now they want a feature to show actor details with movies they appeared in.

how to solve the aggregation problem

approaches from the frontend multiple requests to the backend approaches from the backend specialized endpoints for specific frontends backends for frontends approaches from the middle!? RPC GraphQL Hypermedia JESI

ad hoc aggregation for the rescue

• how frontend engineers work with JESI
• how backend engineers work with JESI