chevron-down chevron-right chevron-up comment dropdown facebook folder google-plus label linkedin github loupe next perk-01 perk-02 perk-03 perk-04 perk-05 perk-06 perk-07 perk-08 pin previous twitter

Dgraph's recap of GraphQL Conf - Berlin 2019

We took part in the recently held GraphQL conference in Berlin. The experience was fascinating, and we were amazed by the high voltage enthusiasm in the GraphQL community. Now, we couldn’t help ourselves from sharing this with Dgraph’s community! This is the story of the GraphQL conference in Berlin.


The GraphQL connection

If you’re wondering what does Dgraph has to do with GraphQL, here’s the answer. Dgraph had set its eyes and placed its bet on GraphQL very early. Back in 2015, when GraphQL was still hot and new, we were intrigued by its intuitiveness and expressiveness to query application Graphs.

But, the simplicity of GraphQL makes it infeasible to express all database operations with it. Rather than exposing a subset of functionalities through complete GraphQL spec compatibility, we chose to sway away a bit, and this led to GraphQL+-. This love affair with GraphQL over the years has been full of flavors. But it never has been bitter :) The GraphQL community constantly desired for the full spec compatibility. And we obliged.

GraphQL-request

Dgraph is currently a member of the GraphQL Foundation. As we head towards shipping a layer that would completely adhere to GraphQL spec, we decided it was time for us to participate more actively within the GraphQL community.


The trip

The yearning to be amidst the community brought 5 of us from 4 different continents to fly to Berlin for the conference.

For most of us in the team, it was our first time in Berlin. The liveliness and vivacity of Berlin are indeed limitless.

Berlin

As a precursor to GraphQL conf, we attended the Prisma day. The interaction with the Prisma team and the community was wonderful.

Berlin


The conference day

The first glimpse at the conference was thrilling! It was all set for a splendid show. graphql-conf

The talks touched upon a diverse set of topics, here are the broad categories,

The GraphQL success story.

History and reasons why GraphQL kicked off in one’s company or how GraphQL brings efficiency to their existing practices. These talks took us through the GraphQL journey, the set of pain points one faced in their organizations, and how the adoption of GraphQL solved it.

The co-creators of GraphQL, Dan Schafer and Lee Byron took the audience through their journey at Facebook during which GraphQL evolved

  • We’re going to program like it’s 1999 by Lee Byron

  • GraphQL Subscriptions at Scale for Real-time Monitoring Dashboard at VISA by Vimalraj Selvam

Tooling

One of the real advantages of adhering to GraphQL spec is its rich ecosystem of tooling.

Here’s one such talk around GraphQL tooling,


The scale challenges.

Some of the talks touched upon scaling challenges with GraphQL. These talks were particular interest to us. I’ll dig more into why later in the article.

  • Matt Mahoney from Facebook spoke about scaling the GraphQL clients.
  • This talk by James Baxley explains how Apollo Federation aims to simplify the challenge of providing a unified GraphQL API gateway on top of multiple GraphQL microservices.·
  • Marc from Github also focuses on how having a monolith for the purpose of having one Graph for an entire organization is a bad idea, but stitching the GraphQL microservices also has its downsides. He shares views on what could be the right mixture.
  • Adam D. I. Kramer from Facebook shares tips about maintaining backward compatibility as your GraphQL project grows big.
  • Jon Wong from Coursera explains their strategies for scaling GraphQL adoption over the last few years.

The community projects and libs

  • Gatsby is a free and open-source framework based on React that helps developers build blazing-fast websites and apps. Kyle Matthews, Co-Founder and CEO of Gatsby discussed how they use GraphQL to support querying data from hundreds of different source plugins.
  • Tartiflette is a new “SDL First” implementation built for Python 3.6+. Stan Chollet talks about how the library was built based on his experience at Dailymotion running GraphQL API’s.

Best practices

  • Sasha Solomon from twitter on best practices around Error Handling in GraphQL.
  • Nikolas Burk on Code first GraphQL Server Development with Prisma.
  • Kewei Qu from Facebook explains best practices learned from their massive migration efforts to GraphQL.

collage


The GraphQL documentary screening

The GraphQL documentary screening was one of the highlights of the conference.

GraphQL documentary


The live-tweeting

We didn’t miss the chance to live-tweet the event :) Here are the threads,


Thank you note to the organizers.

The show was indeed a GraphQL carnival. A big shoutout to Etel Sverdlov, Ekaterina Kromina, and Johanna Dahlroos. The three wonderful organizers who made the event a grand success.

Organizers


The closing notes

The adoption and growth of GraphQL have been tremendous. The community is very enthusiastic and dynamic. We are witnessing a trend wherein a large number of companies are gearing up to adopt GraphQL at scale.

GraphQL encourages one to think, visualize, and query by imagining an entire organization’s data as one big graph, and this changes the whole view of one’s business domain. In a short period post GraphQL adoption, one would be implicitly accustomed to think “Graph”ically! In due course, there is a great possibility that the nature of the database and server-side workloads start to change.

But it looks evident that amidst the GraphQL excitement, the new set of challenges that could potentially be posed by the change in the nature of workload at the database layer has been underplayed.

The staggering adoption of GraphQL has renewed the need for a distributed engine which in one sweep could resolve arbitrary depth GraphQL queries, without having to compromise on the scale, throughput, and speed.

Hence, we are excited about the way GraphQL has kicked off in the last couple of years. We look forward to making active participation and impactful contributions to the vibrant GraphQL community.

We signed off with impressions of great experiences and the mandatory bye bye tweet :)

Bye Bye


We are building a fast, transactional and distributed graph database. If you like what you read above, come join the team!

Get started with Dgraph https://docs.dgraph.io
Star us on GitHub https://github.com/dgraph-io/dgraph
Ask us questions https://discuss.dgraph.io
Fortune 500 companies are using Dgraph in production. If you need production support, talk to us.