Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 4 additions & 5 deletions app/best-practices/page.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -105,15 +105,14 @@ You should be testing the logic of your schema to ensure that it behaves the way
Tags: **schema**

If an authorization concept can be expressed using relations, it should be.
We provide caveats as an escape hatch; they should only be used for context that’s only available at request time, or else ABAC logic that cannot be expressed in terms of relationships.
We provide caveats as an escape hatch; they should only be used for context that’s only available at request time, or for ABAC logic that cannot be expressed in terms of relationships.

This is because caveats come with a performance penalty.
A caveated relationship is both harder to cache and also slows down computation of the graph walk required to compute a permission.
This is because caveats come with a performance penalty: a caveated relationship is both harder to cache and also slows down computation of the graph walk required to compute a permission.

Some examples:

- A banlist - this could be expressed as a list in caveat context, but it can also be expressed as a relation with negation.
- A notion of public vs internal - boolean flags seem like an obvious caveat use case, but they can also be expressed using self relations.
- A banlist - this could be expressed as a list in caveat context, but it can also be expressed as a relation with [exclusion](https://authzed.com/docs/spicedb/concepts/schema#--exclusion).
- A notion of public vs internal - boolean flags seem like an obvious caveat use case, but they can also be expressed using [loop relationships](https://authzed.com/docs/spicedb/modeling/attributes#loop-relationships).
- Dynamic roles - these could be expressed as a list in caveats, and it’s not immediately obvious how to build them into a SpiceDB schema, but our [Google Cloud IAM example](https://authzed.com/blog/google-cloud-iam-modeling) shows how it’s possible.

### Make Your Writes Idempotent
Expand Down
31 changes: 13 additions & 18 deletions app/spicedb/modeling/attributes/page.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -5,49 +5,44 @@ import { Callout } from "nextra/components";

If you are migrating to SpiceDB from a pre-existing authorization system, it's likely that attributes play a part in your authorization evaluations.

SpiceDB is a Relationship Based Access control system.
This gives SpiceDB the flexibility to evaluate attributes for access control alongside more complicated access control logic like roles and/or relationships.
SpiceDB can evaluate attributes for access control decisions alongside more complicated access control logic like roles and/or relationships.

The sections below will provide practical examples for implementing various kinds of attributes in the SpiceDB schema language.
Before reading this guide, it's recommended that you have some familiarity with the SpiceDB schema language.
[This guide](/spicedb/modeling/developing-a-schema) is a good place to start.

## Boolean Attributes

A boolean attribute is an attribute on an object that affects authorization by enabling or disabling an authorization setting.
Boolean attributes can often be thought of as a toggle.
Feature flag authorization can be enabled with boolean attributes.
A boolean or binary attribute is an attribute on an object that can affect authorization.
Boolean attributes can often be thought of as a toggle. For example, feature flag authorization can be enabled with boolean attributes.

### Loop Relationships

Loop relationships are one way to implement boolean attributes.

In the example below, there is a schema that enforces the following authorization logic: a user can
only view a document if the user is related to the document as viewer and editing is enabled for the
document (this is the same authorization logic used in the wildcard example below).
In the example below, there is a schema that enforces the following authorization logic: a `user` can
only `view` a `document` if a) the `user` is related to the document as `editor`, and b) if editing is enabled for the
`document`.

In the example below, to enable editing for a document, a loop relationship using the `edit_enabled` relation must be written.
To enable editing for a document, a loop relationship using the `edit_enabled` relation must be written.
When a `document` is related to itself with the `edit_enabled` relation, that relation can be walked to itself
(with the arrow) to determine who relates to the document as an `editor`.

In summary, a `user` has permission to edit a `document` if they are related to that document as an `editor`
and that document is related to itself with `edit_enabled`.

<InlinePlayground reference="ifgMyLGN-Hjw" />

<Callout type="warning">
There is no mechanism in the SpiceDB schema language that enforces that a relation be used as a
loop relation. In order to avoid accidentally misusing a loop relation (e.g. relating an object to
a different instance of the same type) it is recommended to implement client side logic that
enforces only using the loop relation for its intended purpose.
loop relation. In order to avoid accidentally misusing a loop relation (e.g. by writing the
relationship like `document:1#edit_enabled@document:2`) it is recommended to implement client side
logic that enforces only using the loop relation for its intended purpose.
</Callout>

### Wildcards

[Wildcards](/spicedb/concepts/schema#wildcards) are a way to implement boolean attributes.
[Wildcards](/spicedb/concepts/schema#wildcards) are another way to implement boolean attributes.
Wildcards modify a type so that a relationship can be written to all objects of a resource type but not individual objects.

In the example below, the schema enforces the following authorization logic: a user will have `edit` permission on the document if they are related to the document as an `editor` and they relate to the document through `edit_enabled`.
In the example below, the schema enforces the following authorization logic: a `user` will have `edit` permission on the document if a)they are related to the document as an `editor` and, b) they relate to the document through `edit_enabled`.
Both are required because `editor` and `edit_enabled` are [intersected](/spicedb/concepts/schema#-intersection) at the `edit` permission definition.
To enable document editing, you need to establish a relationship that connects all users to the document using the `edit_enabled` relation: `document:somedocument#edit_enabled@user:*`.

Expand Down Expand Up @@ -104,4 +99,4 @@ This example is similar to the one above, except it requires that all `country`

In almost all cases, [caveats](/spicedb/concepts/caveats) should only be used when data required to evaluate a CheckPermission request is only available at the time of the request (e.g. user's current location or time of day).
Using caveats for static data (e.g. a document's status) can have negative performance impacts.
Static attribute data should always be modeled in the schema using patterns similar to those described above.
Static attribute data should always be modeled in the schema using the patterns described above.