"Entia non sunt multiplicanda praeter necessitatem." (Entities should not be multiplied beyond necessity.) -William of Ockham
A Thing metamodel with admin interface.
The GraphDL (Graph Descriptor Language) application framework provides a comprehensive Object-Role Model (ORM) framework for modeling relationships and processes, making it particularly suitable for developers, business analysts, and AIs creating e-commerce platforms, content management systems, and more. By describing the requirements for an application, then translating the description into simple sentences and abstracting out atomic fact types, the atomic fact types can be added to the GraphDL application as Graph Schema objects, which form a knowledge graph that can be relational-mapped into a database that stores Graphs in third normal form. The core components and concepts of GraphDL include:
Graph Schemas and Graphs Graph Schemas act as blueprints for how data is structured within a Graph. They define the Readings of Nouns and Verbs.
Readings and Roles: Define the interactions between Nouns. Readings with Roles can be thought of as the edges that connect nodes (Nouns), specifying the type of interaction or relationship that exists between them. Readings provide the main interface for providing API definitions. When a Reading is accessed by an API, the Verb's Action and the Nouns' Resources are stored as a Graph that implements that Reading's Graph Schema. Roles are used to define the part that each Noun plays in the Graph Schema. The number of Roles in a Graph Schema determines its arity. Graph Schemas can be unary, representing a checkbox or boolean field, binary, representing a value/field or relationship with another entity, or any arity higher than binary to represent a complex fact or table.
Nouns and Resources: Represent entities, objects, or values within the system. Nouns can be any subject or direct object of a sentence. They serve as the primary nodes within a Graph Schema, embodying the entities around which relationships are formed. When a Graph Schema is implemented as a Graph, Nouns are stored as Resource nodes in the Graph. A Resource represents a single instance of a Noun, such as a user, product, order, etc.
Verbs and Actions A Verb represents the relationship between two Nouns. It defines the type of interaction or relationship that exists between them. An Action is a specific instance of a Verb that is used in a Graph. This could include actions like "create", "modify", "owns", "relates to", etc.
Statuses and Transitions: Statuses represent the status of an entity at a given time, and Transitions define how entities move from one state to another based on certain Events. This is useful for modeling state machines, workflows, and processes within a system.
Events and Guards: These components detail the dynamics within a state machine. Events trigger Transitions between Statuses, and Guards define conditions that must be met for Transitions between Statuses to occur.
Constraints and Constraint Spans: Uniqueness Constraints govern how entities map to a relational database schema. They span over Roles using Constraint Spans to specify the multiplicity. The roleRelationship field on a Graph Schema defines and interprets the Constraint Spans and the Constraints by this pattern: One-to-many (1:*) uniqueness Constraints are part of a Constraint Span over only the first entity, many-to-one (*:1) over only the last entity, many-to-many (*:*) over both roles, and one-to-one (1:1) is represented by two Constraint Spans, one for each of the roles in the graph schema.
The data types from the fields in the database determine the UI, making it possible to generate software. UI may also be customized per entity or table. Relational database schemas and UIs can be generated using the Generator object.
ORM2 and UML diagrams of GraphDL are included:
Open ORM model using Visual Studio 2022 with NORMA or explore online using https://ormsolutions.com/tools/orm.aspx
This repo was created by running npx create-payload-app@latest and selecting the "blank" template.
To spin up the project locally, follow these steps:
- First, clone the repo
- Then
cd YOUR_PROJECT_REPO && cp .env.example .env - Next
yarn && yarn dev(ordocker-compose up, see Docker) - Now
open http://localhost:8000/adminto access the admin panel - Create your first admin user using the form on the page
Changes made in ./src will be reflected in your app.
Alternatively, you can use Docker to spin up this project locally. To do so, follow these steps:
- Follow steps 1 and 2 from above, the docker-compose file will automatically use the
.envfile in your project root - Next run
docker-compose up - Follow steps 4 and 5 from above to login and create your first admin user
The Docker instance will help you get up and running quickly while also standardizing the development environment across your teams.
The easiest way to deploy your project is to use Payload Cloud, a one-click hosting solution to deploy production-ready instances of your Payload apps directly from your GitHub repo. You can also deploy your app manually, check out the deployment documentation for full details.