Highlights from GraphQL Europe 2018

Marcin Gębala
Blog | Mirumee
Published in
3 min readJun 20, 2018
GraphQL-Europe conference in Berlin.

Last Friday part of our team participated in Europe’s biggest GraphQL-dedicated conference in Berlin. The list of speakers included names from Apollo, GitHub, Medium and Shopify, as well the original authors of the technology. Since we use GraphQL on a day-to-day basis, we wanted to learn about the hottest news from the GraphQL community.

Long story short — we loved the event, so we picked for you a couple of highlights, talks and topics that we liked most.

API design guidelines and best practices

Designing an API is a challenging task, especially when using relatively new technology. Leanne Shapton from Shopify presented guidelines addressing many common concerns you may stumble upon. It covers both fundamental topics as well as more sophisticated ones, e.g. how to structure queries and mutations to represent complex relations. Worth a read by anyone working with GraphQL.

Link: https://bit.ly/2JYpUA5

GraphQL formalization

Olaf Hartig presented a scientific view on GraphQL, dealing with deep and nested queries which could kill the server as a result. How to appropriately deal with it is still an open question, but according to the presentation, it would be technically possible to build tools that calculate the cost of queries.

Link: https://bit.ly/2M4grrC

Reuse your schemas

One thing that was reiterated a few times was the ability to build safe, declarative type systems across all the architectures, not only in monolith applications. Companies used scheme for automatic code generation to avoid duplication of reimplementing the same logic in different components and propagate type system from data fetchers up to frontend client, through all layers of abstraction.

Part of our team at GraphQL-Europe conference.

Single access layer for multiple data sources

Daniel Woelfel’s “Teaching GraphQL” presentation helped us understand how we can raise our environment’s awareness about GraphQL. Many tips were given, but the most important was not to teach how we use GraphQL, but to show why we need it in the first place. The OneGraph application has shown the capabilities of this technology by combining different popular APIs into a single GraphQL interface.

GraphQL is here to stay

Thanks to its flexible structure, GraphQL is an excellent fit for continually growing and changing projects that beautifully solve the problem of APIs versioning. You can finally hold on to the old “endpoints” while introducing new ones. With the help of additional packages, you can track the usage of the old API and gracefully remove it when its usage goes below the desired threshold.

Empowered by the experiences shared by other companies, as well as the news of Apollo Server 2.0 going out of beta, our team is already experimenting with it to expose our services to the clients through single and concise GraphQL API.

There is still some uncertainty among developers when it comes to choosing GraphQL over REST, but this conference has proved that the ecosystem is growing fast. If you’re still not convinced, consider that not only GitHub’s API v4 is a GraphQL one, but also their older REST API v3 is powered by GraphQL under the hood!

This conference was a great opportunity to meet the GraphQL community and learn about the current state of GraphQL. The fast pace formula of 20-minute talks wouldn’t make anyone bored. We’re certainly looking forward to participating in future editions!

If you have any questions we can answer, reach out to us on Twitter or GitHub.

--

--

Marcin Gębala
Blog | Mirumee

Full stack engineer at @mirumeelabs. Development lead of @getsaleor.