Holy crap! The ruby GraphQL hackathon was awesome!
Around 35 attendants and myself were hacking all day long on graphql-ruby
with some awesome results. For newcomers we had a crash course on GraphQL from Robert Mosolgo himself, Graphql god at GitHub himself!
During the day we delivered some notable things:
- The team of Appsignal incorporated the first per field performance metrics for their performance product.
- Leaplines team made a new GraphQL-Ruby website, which will be launched shortly.
- Robert Mosolgo added some much needed tracing to the gem.
- I myself added a simple metrics dashboard as a Rails Engine.
We had some great speakers too, lets talk some key takeaways.
Impraise on “Mistakes with GraphQL (and how you can learn from it)”
- It may take some time for the team to switch from an REST mindset to a GraphQL mindset. Ensure that everyone has a decent understanding on GraphQL
- Train your team on what it means to be consume a graph.
- Let the API follows client demands, and be strict, and dint add any fields because you might need it in the future.
“How Appsignal uses GraphQL” – from the talented team at Appsignal
- Appsignal showed how they implemented their GraphQL API and how they expose a metrics heavy API.
“Securing a GraphQL API” from the people at HackerOne
- HackerOne shown how they’ve implemented column based security in the API
“How GitHub runs GraphQL”
- GitHub’s REST API’s also consume the GraphQL API. REST is an adapter and compatibility layer.
- Testing existing REST API’s data results with GraphQL API’s results is done using a gem called scientist.
- Even GitHub server side pages query the GraphQL API.
- The GraphQL API is acting as a funnel for all the data accessed via GitHub.
All the slides will be uploaded to speakerdeck soon.
Thanks everyone for participating! Looking forward to the next hackathon.