crow.tree VS base

These datatype implementations are currently stored in the wolf repo (the "stdlib"). In addition, a high-level Rails-like framework is planned to be in the crow repo. This separation is here so on the one-hand the "stdlib" is minimal, while the framework is robust and all-inclusive. For an example of some code, check out for instance this first pass at a dynamic array, here is a small snippet.

Basically the problem is I have an AST object type such as ASTClass (called mesh-form in my source code, if curious, which is in TypeScript) or ASTFunction (which is mesh-task, etc.). These classes and functions are instantiated up front as it resolved the sort of XML node tree. Initially, they don't even have a name, because that comes as a child node. Or the property types are simply strings and not linked to an instantiated object type. That sort of stuff.

Hey! I started on the data modeling language Link, which doesn't do much by itself. I am going to use BaseLink for the compiler, at first just compiling to JavaScript.

BaseLink is a budding compiler/package manager for LinkText. It builds upon the base structure of a LinkText tree of terms and such, and makes it into a programming language that feels somewhere between a "Tree Assembly" like language, and Ruby or Python. It defines a set of "keywords" (key terms) which are used to define functions, classes, variables, function calls, iterators, etc..