Question

When writing a private method that has no need to access other members of the same class, how do you choose between private member and let binding?

• I tend to use private members because it’s easier to change accessibility if required, but are there other aspects to keep in mind when making the choice?
• Are let bindings compiled as private members (making this only a style choice)?

Answer 1

The relevant portion of the spec is section 8.6.2. It states:

The compiled representation used for values declared in “let”
bindings in classes is either:

• A value that is local to the object constructor (if the value is
  not a syntactic function, is not mutable and is not used in any
  function or member).

• An instance field in the corresponding CLI type (if the value is
  not a syntactic function, but is used in some function or member).

• A member of the corresponding CLI type (if the value is a
  syntactic function).

Also:

Non-function let-bindings that are not used in either the members of
a type or a function binding are optimized away and become values that
are local to the resulting CLI constructor. Likewise, function
bindings are represented as instance members.

I prefer let bindings to private members because they’re more “functional,” i.e., they emphasize “what” over “how.” The compiler takes care of the optimal compiled form.