Matt Mahoney
@mjmahone.bsky.social
Ah no this is fragment arguments (exactly as you have them!) and then more on top later. Sorry didn’t realize we’d miss you at the conf!
Basically with scope you can add generic fragment spreads. But fragment arguments are needed more badly and sooner!
Basically with scope you can add generic fragment spreads. But fragment arguments are needed more badly and sooner!
September 11, 2025 at 6:10 PM
Ah no this is fragment arguments (exactly as you have them!) and then more on top later. Sorry didn’t realize we’d miss you at the conf!
Basically with scope you can add generic fragment spreads. But fragment arguments are needed more badly and sooner!
Basically with scope you can add generic fragment spreads. But fragment arguments are needed more badly and sooner!
Also if you’ve ever wanted to thread a fragment through a callback, well. Maybe we need to get ahead of the spec again!
@jovidecroock.com want to get mad at me at #GraphQLConf for putting up a new (related!) idea instead of helping push fragment arguments forward more?
And no it’s not fragment aliases either.
And no it’s not fragment aliases either.
Today fragment-arguments, the spec PR turns 1 github.com/graphql/grap... - it is however a lot older than 1. Initially this was being championed by @mjmahone.bsky.social and later on by me. Pioneered even _much_ earlier by Relay...
So, why do I actually think this is worth posting?
So, why do I actually think this is worth posting?
September 5, 2025 at 12:49 AM
Also if you’ve ever wanted to thread a fragment through a callback, well. Maybe we need to get ahead of the spec again!
Yeah I’m fairly convicted that global scope for variables is worse than purely fragment local scope, but was what was needed for GraphQL in the moment of 2015.
I’d like to start pushing validation rules *requiring* fragment scoped definitions for all variables. But obviously can’t be spec.
I’d like to start pushing validation rules *requiring* fragment scoped definitions for all variables. But obviously can’t be spec.
September 5, 2025 at 12:45 AM
Yeah I’m fairly convicted that global scope for variables is worse than purely fragment local scope, but was what was needed for GraphQL in the moment of 2015.
I’d like to start pushing validation rules *requiring* fragment scoped definitions for all variables. But obviously can’t be spec.
I’d like to start pushing validation rules *requiring* fragment scoped definitions for all variables. But obviously can’t be spec.
If you’re interested in how Meta gets ahead of the spec, I’ll be walking through how we decided we needed Defer/Stream. Now we’re finding new patterns that look kind of like Defer, but worse!
Let’s talk about how to bend GraphQL to a new domain with Async! graphql.org/conf/2025/sc...
Let’s talk about how to bend GraphQL to a new domain with Async! graphql.org/conf/2025/sc...
@async: Defer Even More! - Matt Mahoney, Meta | GraphQLConf 2025
@defer allows you to specify which parts of your operation are urgent, and which can be delayed. However, there is still a contract with @defer: all your data will always be returned, at some later po...
graphql.org
September 5, 2025 at 12:26 AM
If you’re interested in how Meta gets ahead of the spec, I’ll be walking through how we decided we needed Defer/Stream. Now we’re finding new patterns that look kind of like Defer, but worse!
Let’s talk about how to bend GraphQL to a new domain with Async! graphql.org/conf/2025/sc...
Let’s talk about how to bend GraphQL to a new domain with Async! graphql.org/conf/2025/sc...
SDL and Executables belong together. One mega document. Embrace everything
January 28, 2025 at 11:27 PM
SDL and Executables belong together. One mega document. Embrace everything