note on performance when "flattening a Rails relationship"

[JuliusRapp]

I followed the docs on flattening a Rails relationship. For completeness, let me copy-paste the code from the docs:

# Given Models
class Person < ActiveRecord::Base
  has_many :spoken_languages
  validates :name, :email, :spoken_languages, presence: true
end

class SpokenLanguage < ActiveRecord::Base
  belongs_to :person, inverse_of: :spoken_languages
  validates :person, :language_code, presence: true
end

# Resource with getters and setter
class PersonResource < JSONAPI::Resource
  attributes :name, :email, :spoken_languages

  # Getter
  def spoken_languages
    @model.spoken_languages.pluck(:language_code)
  end

  # Setter (because spoken_languages needed for creation)
  def spoken_languages=(new_spoken_language_codes)
    @model.spoken_languages.destroy_all
    new_spoken_language_codes.each do |new_lang_code|
      @model.spoken_languages.build(language_code: new_lang_code)
    end
  end
end

When GET /persons is processed, the spoken languages are individually queried from the database for every person in the collection. I was able to optimize the queries by adding this code to PersonResource:

  def self.records_for_populate(options = {})
    records_base(options).includes(:spoken_languages)
  end

Is that the right thing to do here? If yes then you can consider to improve the documentation by including that snippet.