tried searching for it on the docs but I can't seem to find it.

Hey guys can anyone tell me if the windows support has been patched for Nexus, can we now develop on windows with Nexus?

hi all. im trying to resolve an issue with the recent release (

@prisma/client@2.3

)  with 

nexus

 . i’ve narrowed it down to the package 

nexus-plugin-prisma@0.16.1

 using an older version of prisma (

@prisma/client@2.1.3

) as a dependency, causing a conflict between the the two versions due to naming convention changes in the recent releases. As far as I know, I cannot change the version in 

npx

 because that always get the latest version of 

@prisma/client

, which means i need to figure out a way to bump the version in 

nexus-plugin-prisma

? Any suggestions?

Any idea why in my

nexus

resolver when I try using the prisma crud operations, its showing my outdated fields from my

prisma.schema

even though I've run the

save and up

commands after changing and restarted nexus? It doesn't seem to have recognized the updated schema.

Hey all, been stuck mentally the last week or so. If someone has a minute or two, would love to chat here or in dm's. https://www.howtographql.com/graphql-js/5-connecting-server-and-database/ trying to get nexus going and I just get confused once resolvers come into play. Not even sure why. I have done the tutorials on prisma and on nexus, gone through example repos, just something is holding it from clicking. Figured id reach out for help. This morning I did the video for prisma day with nexus, and that got me more along the way. Now trying to go and do this project with what I "learned" but seemingly havent

hey @DeanMo! I'd be happy to help you out. What is the blocker you're hitting?

Heya, is there a way to get the

FieldResolver

out of the

nexus

package? Currently transitioning from

@nexus/schema

and I can't really find it

How I should handle scalar date type in args mutation?

args: {
        startData: schema.scalarType()
      },

in model I have

model x {
  id          String            @id @default(uuid())
  startDate   DateTime
}

It is possible to import from prisma somehow enum types also?

I need to store some pictures in a prisma2/nexus-schema/apollo project. can someone recommend a file server solution for that stack?

I'm generating nexus types from a prisma2 schema with the

create-nexus-type

CLI. Works nice so far. Now I want to add some file uploads and I wonder how that fits together. I found this reference, which I'm trying to follow. The generated CRUD assumes that there's a mapping from the input to the output types. But with the file upload, this pattern is broken as far as I can see. The input type will be some

Upload

scalar and the output type some custom

File

type. does this make sense? does anyone have a hint for me how to handle this?

well, actually now completely right. yes types are generated with the above tool. the CRUD (more critical here) is computed by

addCrudResolvers

from the react-admin data provider.

Does anyone have any issue running

nexus dev

on Windows? Context: https://github.com/graphql-nexus/nexus/issues/1277

Hi, I’m currently evaluating how prisma 2 would fit into a graphql api service, and so far things don’t really add up for me. Every definiton should be put into that single schema.prisma file and can’t be split up, composed or imported at runtime, is that right? I had high expectations on the GraphQL integration, to avoid lot’s of repetitive code for CRUD operations. With prisma 2 this seems to not be part of prisma anymore, which I think is fine in general, but instead everything I found points to the use nexus/nexus schema (a bit confusing there aswell with the schema/api server splitup + old website) and I got the impression that the two projects prisma and nexus are close somehow. But looking at nexus-plugin-prisma and related examples I found, everything was really verbose and all the prisma definitions had to be done again for nexus, making me question the actual convenience of having an integration between them at all. While there is some pinned issue regarding exposing fields, which mainly suggests that this would be a bad thing as it is unsafe, and discusses allowing it for getting started and not production use, I completely fail to see the logic here, why restricting access, maybe omiting some fields here and there which should not be exposed, should result in the whole type definiton being done twice. On another hand, I also wonder where authentication should be handled here in general. I understand that prisma itself mainly handles the database abstraction and doesn’t care about an api since v2, but still I think the general considerations I have when approaching this topic would be those of many, and right know I’m really confused. My main usecase would be, to have an Apollo Server(using Federation and a serverless deployment, but let’s skip this for now) , abstract away the database (hopefully soon with mongodb support) and write my type definitons once, same for some basic CRUD operations, and plug in my way of authorization/authentciation at some point do allow/deny things. This seemed like a really typical usecase to me, and TBH I didn’t took a deep dive into Prisma and Nexus yet, but I’m really wondering why some things are the way they are. Are some assumptions I had wrong, or am I maybe missing some important parts on how this integrates? I appreciate any input, thanks!

Looking at this video you have to setup an express app and create a

refresh_token

end point. Since nexus is using express under the hood, is this possible as well?

On prisma we have include and select to choose fields we want to add. How I should extend some fields in objectType in nexus? In that type?

t.field('user', { type: 'User' })

How to add args to extendType? I try to add

schema.extendType({
  type: 'Query',
  args: { id: intArg('id of the user') },
  definition(t) {
   ...
  },
})

but a get error

Argument of type '{ type: "Query"; args: { id: any; }; definition(t: OutputDefinitionBlock<"Query">): void; }' is not assignable to parameter of type 'NexusExtendTypeConfig<"Query">'.
  Object literal may only specify known properties, and 'args' does not exist in type 'NexusExtendTypeConfig<"Query">'

hello, I am unable to build on these versions

Hello! I'm looking for some general guidance on how/where to add business logic when using

nexus-prisma

, so I'm curious to hear what others have done. Do you do everything in graphql middleware? Do you implement the

resolve

function in the resolver of a mutation? Are you using Prisma Client's new support for middleware? What works for you? I'm thinking about things like user permissions, complex validation logic, performing side effects, and so on. Thank you!!

Am I not using the

update

correctly?

schema.inputObjectType({
  name: 'UpdateProjectInput',
  definition(t) {
    t.string('projectId', { required: true })
    t.string('name')
    t.string('description')
  },
})

t.field('updateProject', {
      type: 'Project',
      nullable: false,
      args: {
        data: schema.arg({ type: 'UpdateProjectInput', required: true }),
      },
      resolve(_root, args, ctx) {
        return ctx.db.project.update({
          where: {
            id: args.data.projectId,
          },
          data: {
            name: args.data.name,
            description: args.data.description,
          },
        })
      },
    })

model Project {
  id          String   @id @default(uuid())
  name        String
  description String?
  columns     Column[]
}

Inside the

data

object, I keep getting this error for the

name

field.

Type 'string | null | undefined' is not assignable to type 'string | undefined'.
  Type 'null' is not assignable to type 'string | undefined'.ts(2322)
index.d.ts(1771, 3): The expected type comes from property 'name' which is declared here on type 'ProjectUpdateInput'

Hey guys, Autocompletion feature is not working for my Context which delivers the auto generated Prisma functionalities. I was able to use GraphQL Playground to test my query and mutations successfully, but do not see the CRUD functionality that auto generated by the Prisma in the browser either. What am I missing here?

Hello! I’m having an issue when creating a mutation, does anyone knows what may be happening? Here’s the code:

t.field('createJobPost', {
      type: 'JobPost',
      args: {
        title: stringArg({ nullable: false }),
        logo: stringArg({ nullable: false })
      },
      resolve: (_, { title, logo }): any => {
        return prisma.jobPost.create({
          data: {
            title,
            logo
          }
        })
      }
    })

👋 does anyone know what nexus-plugin-shield mean when it says “fragments not supported”? Do they mean query fragments will be ignored (which would be a really major gap/problem), or fragments in rules?

Somedays I feel like I understand how the prisma api works, other days I don't... It's frustrating. Given a schema that looks like this.

model Project {
  id          String   @id @default(uuid())
  name        String   @unique
  description String?
  columns     Column[]
}

Why would this resolver have issues with types?

I'm running nexus dev, then using the api through a next dev app. Initially I get 5 connections after signing in to my app, which increases to 7-8. Then after maybe 5-10 minutes, another 5 -> 7-8 connections stack up. Are there any resources on how to interpret and/or handle this for someone who doesn't know a lot about this?

I can access both

prisma/.env

and

.env

via

nexus dev

, but after running

nexus build

and then

node .nexus/build

, the build won't recognize env variables. Is there a way to achieve this?

I'm running a Nexus GraphQL server, and I want to send magic links for users via email for authentication, so they would need to access a GraphQL mutation endpoint called something like

loginWithMagicLink

via just clicking the URL from the email. Is it possible to do a GraphQL mutation request via just a URL? or are only queries allowed?

Hi, I am having an issue switching to prisma. I got a postgres db. We've been using it with sequelize but I am trying to switch to prisma 2. All our tables are camelCase and plurals. e.g. 

companies

 or 

userProfiles

 table. When I run 

prisma introspect

  it generates models that matches the table names, which results in 

nexus

 generating "singular" and plural typings for the models, both being the actual table and model names (plural) so it results in duplicates in the typescript typings which means code doesn't compile. I got over 3000 lines generated by prisma introspect and there's no proper intellisense to rename a model and all fields related to it

Typings seem broken in @nexus/schema https://github.com/graphql-nexus/schema/issues/480

hey all, i started this discussion yesterday on the repo but wanted to post it here to see if anyone can help out https://github.com/graphql-nexus/nexus/discussions/1330

it’s basically when you are using

t.crud

and put your own logic in the resolve the args’s types don’t match up with the requirements of the database from the

context

. I’ve gone through and set the args myself then and tell them

{nullable: false}

and null still appears as an option for the typing on arguments being passed in. Any suggestions how to correct this behavior?