Deploy preview for hasura-docs ready!

Built with commit 534fe10

https://deploy-preview-3605--hasura-docs.netlify.com

@thefliik If you are interested in trying out this change, there should be a test environment available soon once CI finishes. Thanks!

Review app for commit 63a55fc deployed to Heroku: https://hge-ci-pull-3605.herokuapp.com
Docker image for server: hasura/graphql-engine:pull3605-63a55fc9

Review app for commit 534fe10 deployed to Heroku: https://hge-ci-pull-3605.herokuapp.com
Docker image for server: hasura/graphql-engine:pull3605-534fe101

This seems to work well with apollo-client, verified using this little test case:

const ApolloClient = require('apollo-client').default;
const BatchHttpLink = require('apollo-link-batch-http').BatchHttpLink;
const InMemoryCache = require('apollo-cache-inmemory').InMemoryCache;

const fetch = require('node-fetch');
const gql = require('graphql-tag').default;

const link = new BatchHttpLink({ 
  uri: 'http://localhost:8080/v1/graphql', 
  fetch 
});

const client = new ApolloClient({ link, cache: new InMemoryCache() });

client.query({
  query: gql`query first { user { id } }`,
}).then(user => console.log(user))
  .catch(err => console.error(err.message));
  
client.query({
  query: gql`query second { role { id } }`,
}).then(role => console.log(role))
  .catch(err => console.error(err.message));

Using netcat, I can see the requests get batched, and then errors get handled independently on the client.

Review app https://hge-ci-pull-3605.herokuapp.com is deleted

@thefliik If you are interested in trying out this change, there should be a test environment available soon once CI finishes. Thanks!
...
Review app https://hge-ci-pull-3605.herokuapp.com is deleted

@paf31 looks like I was too slow on the draw. I'm not actually using batching in any projects at the moment though, so I probably couldn't have provided any new feedback anyway.

👍