Well, at least my arr2pbm output
no longer bitreverses words I had been worrying at Alfred
while writing. In fairness I and pizzapal that I thought it
think I am now reversing it an was plausible to go from
even number of times (as we know
the classic theorem; the reverse MIT CADR LISPM (FPGA softcore)
of the reverse of a sequence is, 's prolog implementation
non-obviously non-obviously, the
same as the original sequence). to
So I think my re- re- chapter MIT CADR LISPM (FPGA softcore) +
one equipment has its ducks in a (a specific FPGA acceleration
row, some fortnight later; special case for prolog).
hopefully I am able to draft
that and move to a chapter 2: And this was a general set of
The interlispening forthwith. opportunities.
I'm morally exhausted at the Which still sounds reasonable to
moment from the final end of me tbh. But the scope and
late-season seasonal work. From breadth and general strategy
my perspective, lots of people I observed in Azura's business
became close friends with are success is instructive. They
all just gone; whether shipped created a drop-in JVM
back to Tonga or somewhere, or replacement (= drop-in MIT CADR
to greener seasonal pastures in simulation, / ~ maiko vm) that
the south island. The law in NZ had several years of specialised
about migrant workers ("working senior teams design a) CPU
holiday visa") is that they are architecture b) OS associated c)
not allowed to hold one job for porting gcc d) Porting the JVM
longer than three months, to e) Lots of exploratory JVM/JVM
stop them from putting down software/custom CPU interplay
roots.
Mechanistically, there's the
Despite my gregariousness, I question of how to actually do
only have a single, fairly hardware design in the first
distant offline friend whom I place. Zoerhoff shared his
chat to by some compromise verilog reference book with me;
normie internet service Alfred continues to hope one day
occasionally. So getting to know I will actually think about his
lots of people without the porting from verilog to VHDL of
expectation that they believe the MIT CADR softcore.
they might monetarily profit off Technically I like VHDL for that
of me, and then having them all rather than verilog; ada is more
suddenly gone is quite beautiful, whence VHSIC HDL
demoralising, even though on kinda.
average nothing has changed.
On a different tangent Zoerhoff
Hayley on the mastodon added suggested I might be interested
something to my pensieve in Cadence Skill; an adaptation
handwringing about how to reason of Franz lisp to p-cell
about virtual machines and definition. A p-cell defines an
softcore machines that was electronic component in terms of
instructive. 2000-2010, a passed parameters; for example,
company (Azura?) released a line physical dimensions, and a
of 50-CPU (in-house, custom) similarly-themed component but
strictly interleaved memory filling those physical
access JVM supercomputers. dimensions could be generated,
for physical dimensions as
parameters. Cadence Skill
p-cells are the standard for
analogue integrated circuits. I
have some context, but I'm very
far away in industry.
I still have the hunch that a
modern, maybe FPGA, spin on
microcoded lispm compilation...
Eh, but I'm not the entire
effort and budget from Azura
over the course of a decade. I
was trying to convince Alfred
that the next stage for both his
MIT CADR and our Interlisp maiko
VM was basically a new
generation set of moons that
seem ready to fill our skies.
Imagine praetor, rml catching
the vibrations, or Devine, or
actually any of the many and
marvelous current new moons we
actually do already have whom I
am failing to appreciate here.
Moon had a consumate vision of
algorithms, compilation and
practical hardware reality with
a meteoric impact on lisp
machines. For clarity I guess
the Lisp Machines in this
metaphor are dinosaurs, and I'm
going to say Interlisp Medley is
a modern songbird and the MIT
CADR soft core is a modern
crocodile. Assuming the moon is
a lump of the earth knocked into
orbit by a meteor impact before
an AI winter.