Ok, I got a source and looked at xthread.d and multithread.txt.
They don't contain much.
My previous questions stand.
I'd especially appreciate if someone could provide more references,
for a similar design if not for this one.
Is the goal to allow multiple processors to read and write the lisp
heap at the same time? If so, where is the proposed boundary between
locking done by the system and that which is the programmer's
6. It doesn't make sense to use the same symbol per-thread and across
multithread.txt does add a little detail here. It suggests that of
the specials in the lisp package only *features* and *modules* would
be global. So I use this as an example. Why should I not be able to
bind *features* in a thread? For instance, I may have different
read-eval-print loops in different threads and want them to read with