Joining forces on code generation

[jhnnsrs]

Hi everyone,

I am the maintainer of turms which aspires to be both a graphql codegenerator for server (only strawberry as of now supported) and the client, using a lot of the design principles behind graphql-codegen (like reusable fragments, traits, query function generation...) Turms tries to be transport agnostic for most of its code generation but i think it would be great to support some open-source libraries like ariadne more deeply. Albeit i think there is a lot of space for other approaches, i am wondering if it made sense to join forces on this effort?

[rafalp]

Hey, sorry for late response on this.

I am not sure what are you proposing? Code generator for Ariadne server package?

[jhnnsrs]

No worries!

No, given that ariadne is using a schema-first approach, i thought that would not be the way to go. I was more thinking of merging efforts on client-side code generation. Looking at the repository it looks like there is a lot of overlap on code generation with additional featutres like mixin. As turms supports plugins if was thinking if it made sense to maybe merge the code generation efforts and develop an ariadne plugin? Not sure if the goal of the two projects align totally so i thought i would just ask :)

Best

[rafalp]

We are currently evolving Ariadne into a brand of tools for Python GraphQL (like how Apollo is in JS world, no plan for commercial offerings ATM tho). Part of this brand building is having palette of solutions under the Ariadne brand and one of those is codegenerator that we could potentially extend with support for other tools we have in the pipeline like modules or proxy.

So I think we'll pass, but thank you for reaching out anyway ♥️