In this series of presentations, we examine different strategies for
creating a Common Lisp implementation, as well as the pros and cons of
each strategy.

We assume basic knowledge about how a typical modern operating system
(such as Unix) works, and how traditional batch languages (such as C)
are compiled and executed on such a system.  We furthermore assume
medium-level knowledge about Common Lisp.

In part 3, we investigate one possible solution to the conundrum of
our first strategy outlined in part 2, namely the circular
compile-time dependencies between many standard macros and standard
functions, forcing us to write more code in a language other than
Common Lisp, and also forcing us to write "unnatural" Common Lisp
code.  The solution to this problem is the foundation of our second
strategy for creating a Common Lisp implementation, and it relies on
using an existing host Common Lisp implementation for cross
compilation.

Since some time has passed since part 1 and 2 were streamed, we
strongly recommend that interested participants make sure that the
material covered in those two parts is fresh in memory, if necessary
by watching those presentations again.