Intrinsic data types#
Crochet includes, out of the box, a collection of basic types—called intrinsics. These types are useful for most Crochet applications, but they also define a baseline for some of the security guarantees that Crochet provides. Without these it would be very difficult to make Crochet safe.
The intrinsic types cover the following categories of problems:
Numeric models: these types represent different forms of numbers, only somewhat similar to what you’d see in mathematics as we still need to deal with some peculiarities of digital representation and computer resources.
Textual models: these types provide a way of representing and combining arbitrary pieces of text, securely. This turns out to be a much more difficult problem than one may think at first, so we spend a lot of time in the details of these textual types—and both technical and sociological implications, too.
Gradual representations: we don’t always know what kind of data we’re dealing with—or if we’re dealing with a piece of data at all. Crochet explicitly captures this and allows different pieces of code to have varying amounts of knowledge about what data they’re handling. This turns out to be very important for security in some situations!
Logical models: we need to represent answers to questions like “is the door locked?”, and both philosophy and mathematics have given us different models for this. Crochet exposes a very simple one as an intrinsic type. We also discuss why, when, and how you might reach out for other models.
Program representations: we need to represent partial and delayed programs. In Crochet these end up as either Functions or Thunks.
Collections: somestimes we need to talk about several values in a consistent way. Collections are one way to achieve that.