Thanks for assembling this, this is really awesome.

I'm curious why you think printSchema and printFineSchema should both exist? With the existence of this new method, when would you explicitly want to use printSchema instead?

Also, I'm curious about the primary motivation. Are you looking for printSchema to be stable as types evolve? If so, perhaps alphabetical may be better as a schema could add interfaces to types as it evolves which could thrash a file.

Or are you just looking to see these printed in a way that is easier to read in the output?

@leebyron I write printFineSchema for easier to read the output types. Im not sure if the printSchema have other usages other than output types for reading. So i add an additional function to print readable types. With lots of types, alphabetical will become hard to read.

@leebyron hello, is this suitable for merge? Cause im poor at english, if i have some misunderstanding about your meaning,please forgive me.
I make the printSchma like this, by default it print schema by alphabet, or explicitly indicate printing by hierarchy(to get a readable schema).

// type printStyle = 'alphabet' | 'hierarchy';
export function printSchema(
  schema: GraphQLSchema,style: string = 'alphabet'): string {
  switch (style) {
    case 'hierarchy':
      return printFineSchema(schema);
    case 'alphabet':
    default:
      return printFilteredSchema(schema, isDefinedType);
  }
}

What I'm trying to understand is the purpose - the "why" of this method. Have you added it because it should definitely exist in parallel to the existing printSchema method, or simply out of precaution to have this diff be only additive? When would you use printSchema and when would you use printFineSchema?

Maybe it is some personal usage for me(just a few people need this).
The printFineSchema output a hierarchy ordered type. just for human reading .I should move these code to some package called graphql-utils.