cyrildewit
commented
2 years ago Some thoughts I had while thinking about the available options.
Standalonne Node.js application
Benefits
- Frontend and backend are not tighly coupled, because the frontend application will simply be a consumer of the generic REST API.
- Deploying changes to the application will be decoupled from deployments of the core application. This could also be seen as a downside, because of possible mismatches between backend and frontend. This will be however not a problem since the REST API is versioned.
Inertia.js
We are tightly coupling the presentation layer and application layer by using Inertia.js, becaue there will be a hard link between for example the HTTP controller and presentation views. This will only be case for the page components and not for the common components
Benefits
- Faster development time since the the presentation layer does not have to wait for specific features from the REST API. It will have a dedicated controller that provides the view with the necessary information.
- Not per se, but both layers will (partially) exists in the same repository making the software architecture less complex in someway (depends).
- Instead of two applications with its own server requirements, we only need one type of server.
Description
The frontend team will present their work on the frontend on 24 May 2022. The major result that will be show cased is what advice they have worked out about the preferred frontend library and how this library will be implemented.
Objective of this meeting
Discuss and review the advice. Questions from the backend team can be asked and answered by the frontend team. It should be possible that both teams will agree on a shared plan. The advice should match the scope of the project and should be in interest of the stakeholder.