I can’t say I have much insight into the ex-Facebook crowd. I think whether or not me is a useful pattern depends a bit on how you do authentication and authorization in your application. If you have entities owned by the current user then I definitely think it makes sense to scope it under that. It gets a bit trickier if you’re using role based authentication and authorization is less about “me” and more about “the role I want to use for this request”.

You can also find minor inconsistencies between the me pattern and the node pattern. I absolutely adore the node field pattern because it saves you from redundantly setting up hundreds of “get this entity” fields. However it’s { node(id: $id) { ... Fields } } not { me { node() } }, but of course the permissions you have will affect whether or not you can access it.

Really though I think the decline of me or viewer is less about it not being a good idea, and more about the ex-Facebook crowd wanting to distance themselves from a dogma which made it a required pattern. Early versions of Relay (Facebook’s client) were exceptionally dogmatic with respect to how the schema needed to be organized, and it has tried to become more flexible over time.

I suppose where that is leaving me is this: I think me is a valuable pattern for accessing user specific data. I do not necessarily think that it is the most consistent pattern for acting as the primary entrypoint to the rest of the graph. However if that’s the route you want to go I think there can still be real upsides, so it’s a judgement call you’ll have to make.

Does this showcase help you?

https://github.com/hykw/absinthe_showcase/blob/master/docs/operations/query_profile.md