Got the invite, thanks.
I started using Common Lisp in about
2004, so I'm a relative newbie, only almost 20 years ;) In the
last few years I've been going back and reading up on the old lisp
machines, and seeing all the interesting features that fell out of
use, which might be brought forward. I spent a lot of time with
the Symbolics stuff, and was building a VLM replacement in CL,
which was at the point of loading a world, inspecting objects in
memory including disassembling functions, calling ivory functions
from the host CL by compiling the bytecode into native CL
functions, and executing the world from cold start for some dozens
of millions of instructions.
There's a lot of opaqueness and
licensing issues in that ecosystem, though, the memory model and
other hardware features are quite complex, and it was harder to
get knowledgeable insight or assistance for the blocking issues I
had there. So when more Interlisp stuff was popping up with the
Medley pushes, I'm piqued as to if a port of Maiko to CL might be
beneficial, building off that same pool of knowledge, but with a
much simpler bytecode ISA since it was never a hardware
architecture, and seemingly no IP issues. Creating or simplifying
lower level abstractions, interfacing with hardware in different
ways, pushing more things to Lisp instead of C, are all types of
things that might benefit.
I've done lots of DSLs, HLL-targeting
compilers, stack based systems, bytecode VMs, etc, often around
symbolic inference interpreters & compilers, and commercially
deployed some of them.
So mostly I want to hang around and get
a feel for what's going on with the project and where its
high-level and low-level technology goals are, and see what sparks
joy.