hasura / graphql-engine-docs Goto Github PK

CLI is now optional even for securing graphql endpoint.
Requires that hasura/graphql-engine#45 be released

• Subscribe to changes on a table
• Subscribe to a result-set
• Subscribe to a changing document

Screenshot included.

List out the scenarios where GraphQL engine will crash or throw an error if queried.

run_sql
Auth role based schemas
Support for multiple postgres schemas

• create new PG schema
• tracking existing PG schema tables/views
• queries/mutations

/v1alpha1/schema does not exist in the latest versions.

hasura/graphql-engine#138 (comment)

I am writing a how-to on creating a mobile backend with Hasura + Auth0.

Following the instructions here to integrate Auth0 - https://github.com/hasura/sample-auth-webhook

In the final section "Usage with Hasura GraphQL engine", it is not clear what changes need to be made to the auth server. Specifically https://github.com/hasura/sample-auth-webhook/blob/master/server.js#L46

ref: hasura/graphql-engine#217

Via moe#8021 on discord.

On the getting started page we should point existing database users to a different page. Right now, there are only 2 links.

taking the todo-user example would be good

Document the usage of queries and mutations over websockets.

Apollo's subscription-transport-ws supports this.

Guides/tutorials on

• Auto-updating IDs
• Timestamp on_created, on_updated

Via dylanjones on discord / email.

There are many flavors of postgresql which graphql-engine might support out of the box or with slight configuration/modifications. We need to document for each of them.

TimescaleDB
CockroachDB
CitusDB
CrateDB
PostGIS
PipelineDB

hasura doesn't work at all on alpine linux. The installer fails because it can't find sudo even though it shouldn't need it. Manually installing the binary is ok, however it doesn't work, just returns with no output.

Also why do you need an install script with telemetry to download a single binary? That doesn't inspire confidence. Just add links to each arch you build and I'll put the binary somewhere it is convenient for my use case.

• run func on input and return output (ie: use as a query)
• use func to set field value automatically during mutation

Currently it's on the bottom, can we create a page called "Auth integration examples" or similar?

@rikinsk Need you to confirm a structure so that @wawhal can restructure.

Create a tutorials/guides/cookbooks section to show various examples built using graphql engine

Document about _inc and other jsonb operators in update mutation.

• permissions/relationships on table delete
• setting up cascade rules

From hasura/graphql-engine#140:

@rikinsk Can you update the docs per: stackoverflow.com/questions/17770902/forward-host-port-to-docker-container/26090569#26090569
Especially for this page I think: docs.hasura.io/1.0/graphql/manual/deployment/docker/index.html

Document 2 use-cases:

Building a new todo app
Link to Rishi's tutorial

Building an internal dashboard on your existing postgres
Use existing postgres with safe credentials
Track tables
Run SQL endpoint

https://flutter.io/
https://github.com/zino-app/graphql-flutter

I found it initially confusing to have settings I can change in install-scripts/docker-compose.yaml, install-scripts/docker-run.sh and also a bit in config.yaml. Clearly, docker-run.sh won't read config.yaml, but I had to piece it together a bit.

I see this coming from the fact that I can have any combination of

• an existing postgres instance, or a new one, dockerized, composed with hasura.
• starting the console manually, or starting the console along with the serve --enable-console.

I also had to piece together a little bit that the latter gives me a console on 8080 instead of 9695. Maybe that's just my settings, since I run riwh --net=host?

Note that I don't have tons of experience with docker. The additional context of this is that I am replicating the instructions in the docs while being ssh'd into my remote server, not my local machine.

Things:

• Access control rules
• Roles
• Dynamic user-session variables

List types supported explicitly and types supported implicitly with examples for each.

We can basically take this table and add an extra column to show the graphql-engine equivalent:
https://www.postgresql.org/docs/current/static/datatype.html

Add spec and use cases

• run_sql
• create/drop relationships
• create/drop permissions

ref: hasura/graphql-engine#221

If you need to add custom business logic or customise your graphql schema then we recommend using schema stitching. These are 2 common use-cases:

1. Fetch data from sources that are not in the database
2. Add custom mutations

Issue hasura/graphql-engine#23 is closed. Need docs.

by @rajoshighosh

There are quite a few common use-cases around capturing mutations from anonymous users but without making it too easy for malicious users to spam the backend.

1. Likes or visit counter
2. other examples

Reasonable ways to determine a unique anonymous user:

1. client generated session-id. Possible con: not terribly hard for a sophisticated user to fake
2. IP based value. Possible con: users on a campus network might have UX issues with their likes getting rejected
3. recaptcha based value. Possible con: ugh UX

(Via thangngoc and 7777 on discord)

• creating schema
• mutate
• query with filtering

Pick from: https://blog.hasura.io/working-with-schemaless-data-with-graphql-on-postgres-574a1ee2e87f

https://docs.hasura.io/1.0/graphql/manual/deployment/docker/index.html does not talk about console on server

This is a overall issue list.

List of use-cases that should be captured in docs:

• Handle anonymous (not logged in) user permissions
• Schema stitching example
• Relationships
• Enums
• Computed fields / derived data (should be SEO terms for views)

Authentication and user handling guides:

• Auth0, Firebase, ForgeRock
• Auth webhook examples for different use-cases
• Where does user profile data go?

Modelling different types of permissions:

• Collaborators
• Multi-tenant

Different types of data-modelling:

• one-one
• one-many
• many-many
• ENUMS

https://docs.hasura.io/1.0/graphql/manual/queries/sorting.html

Given as

order_by: ["+id"])

Instead of

order_by: [id_desc])

via Takeshi#1086 on Discord