The biggest disadvantages are the differences between your fake backend and the real backend in production. But other than that the downsides will outweigh the simplicity in most cases. This might work out for a web app or website which just loads a little data via Ajax. Differences between the fake and the real backend.Can’t test handling errors (besides 404).No usage of HTTP method besides GET possible.Might not be supported by your JavaScript Framework™.This is important for working in the team, especially for the coordination between the frontend and the backend team. Having those files in your file system has the big advantage, that the dummy data have a single source of truth. The dummy data is in your version control system (eg.Just create your dummy data and store it within the file system of your frontend project. The simplest approach without any tooling involved. We are going to explore the details as well as the pros and cons of each and every possibility. There are a few possibilities to »serve« dummy data depending on your needs, your tech stack and the requirements of your project: Which leads us to the question: How could we simply mock the backend? Different ways to fake a backend It doesn’t matter if the development of the backend gets stuck for whatever reason. The main advantage working with dummy data over here is the independence in which the frontend team can work. So these teams only have to talk about and agree on JSON structures before they can work independently on their part. The backend team needs to make sure the content the API is providing is sane, whereas the frontend doesn’t care about the content: it just needs dummy data in the correct structure.
So in development we have different people working on the frontend and on the backend.
This offers much more flexibility by being able to deploy only the frontend to production or by using the same backend for a different frontend with a totally different tech stack (eg. Think about the backend as the provider of an RESTful API which has a totally different (hopefully semantic) versioning than the frontend. It might be even useful to handle the separate parts as separate projects. Today we mostly have a loose coupling between the frontend and the backend of our web applications. Eine deutsche Version des Artikels findet ihr im Micromata Blog.
0 kommentar(er)
