Add graphql adr

packages/dito/doc/adr/0010-use-graphql-for-strapi-requests.md

@@ -0,0 +1,64 @@
+# 10. Use GraphQL Instead of REST for Data Requests to Strapi
+
+**Date:** 2025-01-02
+
+## Status
+
+Accepted
+
+## Context
+
+When designing our application’s data-fetching strategy, we considered two primary approaches for interacting with Strapi: REST and GraphQL. Both options have their benefits and drawbacks.
+
+Strapi supports both REST and GraphQL, and while REST has long been the standard for APIs, GraphQL’s flexibility provides significant advantages in terms of query precision. Our project’s requirements demanded:
+
+- **Dynamic Querying:** The ability to fetch deeply nested data structures efficiently.
+- **Flexibility:** Customizable and precise data fetching to minimize over-fetching and under-fetching.
+- **Scalability:** Support for future requirements such as aggregations and advanced filtering without rewriting existing endpoints.
+
+### Evaluated Options
+
+1. **REST API**
+
+   - **Pros:**
+      - Simple and well-documented.
+      - Familiar to most developers and integrates easily with a variety of tools.
+      - Requires no additional configuration in Strapi.
+   - **Cons:**
+      - Over-fetching or under-fetching is common, especially for nested data structures.
+      - Requires separate endpoints for different data needs, leading to increased maintenance overhead.
+      - Limited flexibility in defining query shapes.
+
+2. **GraphQL API**
+
+   - **Pros:**
+      - Query only the data you need, reducing over-fetching and under-fetching.
+      - Single endpoint for all requests, simplifying API management.
+      - Built-in support for nested queries, relationships, and filtering.
+      - Easier to adapt and extend as requirements evolve.
+   - **Cons:**
+      - Not as common and familiar as REST.
+      - Potential performance overhead for complex queries if not optimized correctly.
+
+## Decision
+
+We chose to use **GraphQL** for our data requests to Strapi due to its flexibility and efficiency in handling our specific use cases.
+
+Key considerations for choosing GraphQL:
+
+- **Optimized Data Retrieval:** GraphQL’s ability to query only the required fields and relationships reduces unnecessary data transfer and speeds up response times.
+- **Simplified API Structure:** A single endpoint handles all queries, reducing the need for managing multiple REST endpoints.
+- **Nested Queries:** Easily handles our deeply nested data requirements, which are common in our application.
+- **Future-Proof:** GraphQL’s schema flexibility ensures that our API can evolve without breaking changes or significant rewrites.
+
+## Consequences
+
+The decision to use GraphQL for data requests brings several benefits and challenges. GraphQL optimizes data retrieval 
+by allowing our client to request only the necessary fields, reducing over-fetching and under-fetching. 
+The single-endpoint architecture simplifies API management, particularly 
+as the application scales, and its schema flexibility allows the API to evolve without 
+significant rewrites or versioning. 
+
+However, there are challenges to consider. Developers unfamiliar with GraphQL may need onboarding. 
+Also, poorly optimized queries could strain server resources, making caching and query optimization 
+critical for maintaining performance (see [caching](0009-use-serverside-caching.md)).