Glossary

Immutable data

Data that cannot be changed after it’s constructed is called immutable. Most programming languages will feature things like numbers, such as 1 and 2, which cannot be changed after you construct them. That is, you can’t make 1 mean something else. Crochet extends this idea to typed data. If you construct a piece of typed data like new point2d(1, 2), then it will refer to the coordinates x = 1, y = 2 for as long as the program continues to run.

Named argument

A named argument is one whose meaning is explicitly described by a name, and therefore the order in which the arguments are specified does not matter. In Crochet, records use named arguments, so [a -> A, b -> B] means the same thing as [b -> B, a -> A].

The opposite concept is a positional argument. For example, lambda applications are positional in Crochet, so Lambda(A, B) and Lambda(B, A) mean very different things!

Positional argument

A positional argument is one whose meaning depends on the position it has in a sequence, rather than an explicit name. For example, in Crochet, the new operator is positional, because in new some-type(A, B), the meaning of A and B would change if their order was swapped!

The opposite concept is a named argument. For example, records are named in Crochet, so both [a -> A, b -> B] and [b -> B, a -> A] mean the exact same thing.

Projection (of fields)

Data in Crochet is stored in a bag of information called typed data. Retrieving a particular piece of information from this bag is called a projection—you’re “projecting” a certain aspect of the data outside of the bag. Projection is handled by the dot (.) operator.