Naming code, the value-identity relation

a day ago

Copy Link

Naming things is one of the two hard problems in Computer Science, alongside cache invalidation.
Names serve as layers of indirection or abstraction, crucial for managing complexity in software development.
Software engineering involves creating numerous names, yet naming is rarely discussed in depth in computer science education.
Clojure's `def` creates a global var with a symbol as identity, which is replaced by its value upon evaluation.
Clojure.spec's `def` uses namespaced keywords as identities, differing from Clojure's `def` which uses symbols.
The post explores the idea of using compound identities for `clojure.spec/keys` to avoid defining separate specs for each combination of keys.
An example demonstrates defining a spec based on a combination of namespace-keywords without creating a new named spec.

Hasty Briefsbeta