This really doesn't do much but serve as an example of constructing a very basic Lisp interpreter in C. It may, in the future serve as the basis for a larger project, but currently is absurd.

It supports abstraction, lists, symbols and fixnums. It does support top level definitions with the def keyword, but at the moment doesn't support "fancy forms" such as let and letrec nor does it support macros. There's no first class eval or a way to report the current lexical environment. But, everything is built out of conses and cells, including closures. Closures are tagged as such, but stored using the same cell fields that a cons uses.

In other words:

c = cell()
c.type = CLOSURE
c.car = CODE
c.cdr = ENV

Where as a cons:

c = cell()
c.type = CONS
c.car = CAR
c.cdr = CDR

Of course you can evaluate expressions:

$ ./surd 
surd 0> (((fn (y) (fn (x) (+ x y))) 1) 2)
result 0: 3

surd uses libgc for now, though it did at one point use the one-pass mark-sweep as described by Armstong and Virding in "One Pass Real-Time Generational Mark-Sweep Garbage Collection (1995)" (link)

1. Generational collection -- this should be easy
2. Box top level definitions so that they can be replaced
3. First class read, write, eval.
4. Implement let, letrec (might require a special rec form)
5. tail-call optimization.
6. Larger standard library implemented in surd.
7. hygienic defmacros and quasiquote, unquote, unquote-splicing
8. ...