Shane Krueger

In my mind, there is no reason why these queries should calculate differently: ```gql { field1 { child1 child2 } } ``` and ```gql { field1 { ...fragment1 } }...

If there is a reason, it should be stated in a comment or documentation.

Relates to #354 : Allow resolvers to easily see what child fields are required. I'm thinking: - We need something implemented as an extension method on `IResolveFieldContext`. - It should...

The problem is that the validation APIs don't traverse the tree as seen in the result set. For example, it doesn't follow a fragment. I'm not sure exactly what the...

I also have a similar need and use case. I wrote my own code to walk the AST child nodes, but the code is incomplete. The problem is that the...

I think some progress was made with refactoring the SubFields property of IResolveFieldContext but I can’t remember if it’s useful past the first level of children.

> @Shane32 I just tried to use the solution above however it's very old and lots of the functions have been removed. Do you have the code for your graph...

> For some reason the functions that allowed to do these things was removed in v4 thinking of `ExecutionHelper.SubFieldsFor` It was removed from ExecutionHelper by necessity and moved into the...

> @Shane32 no worries I've posted a solution that handles all that. I appreciate it. Perhaps we can consider integrating it as a new feature for v5 somehow. I have...

I don’t think that will help. The problem is trying to walk the list of nodes that will execute. Not the AST. It needs to follow fragments and maybe even...