Related to the book mentioned in the post, people might find John Bell's 55-page essay Types, Sets and Categories (pdf at http://publish.uwo.ca/~jbell/types.pdf) interesting since it follows the evolution of Type Theory from Set Theory from the beginning starting with Russell and Whitehead's works (alone and jointly) and then onto intuitionistic Constructive Logics.
You finally reach the contrast in the definition of "Type", between Classical Logic/Set Theory and intuitionistic Constructive Logic/Type Theory where;
In Classical Logic/Set Theory; Type = Set of values defined by extensional rules.
In intuitionistic Constructive Logic; Type = Set of constructive (often inductive) proofs where the proof itself is the value.
The above is a great help in thinking and understanding formal program proofs done using Imperative Hoare/Dijkstra Predicate Calculus vs. using Functional Type Theories (eg. Simply Typed Lambda Calculus, Dependent Types etc).
PS: Try defining the set/type of even numbers (in say uint8_t) both ways and see for yourself the needed shift in thinking and why the functional approach is much harder than the imperative approach.
Nice; It is by Thomas Hales who led the proof of the Kepler Conjecture.
Given the focus on "Formal Proof" nowadays, two of his papers are worth reading;
1) Formal Proof - Pdf from AMS at https://www.ams.org/notices/200811/tx081101370p.pdf
2) Developments in Formal Proofs - https://arxiv.org/abs/1408.6474
Related to the book mentioned in the post, people might find John Bell's 55-page essay Types, Sets and Categories (pdf at http://publish.uwo.ca/~jbell/types.pdf) interesting since it follows the evolution of Type Theory from Set Theory from the beginning starting with Russell and Whitehead's works (alone and jointly) and then onto intuitionistic Constructive Logics.
You finally reach the contrast in the definition of "Type", between Classical Logic/Set Theory and intuitionistic Constructive Logic/Type Theory where;
In Classical Logic/Set Theory; Type = Set of values defined by extensional rules.
In intuitionistic Constructive Logic; Type = Set of constructive (often inductive) proofs where the proof itself is the value.
The above is a great help in thinking and understanding formal program proofs done using Imperative Hoare/Dijkstra Predicate Calculus vs. using Functional Type Theories (eg. Simply Typed Lambda Calculus, Dependent Types etc).
PS: Try defining the set/type of even numbers (in say uint8_t) both ways and see for yourself the needed shift in thinking and why the functional approach is much harder than the imperative approach.