Criticism ... synthesis?

Raul Deluth Miller rockwell@nova.umd.edu
Sat, 17 Dec 1994 11:04:46 -0500


J. Chroboczek:
     Furthermore, if we are to define a common LLL, we must reduce the
   scope of our ambitions, for instance to Von Neumann machines with a
   few conditions about memory addressing.  Just try to design a
   language that is low level enough (therefore close to the
   hardware), on modern RISC machines (Von Neumann machines with
   linear addressing, simple instruction sets, and complex efficiency
   considerations), Lisp machines (Von Neumann machines with a bizarre
   memory layout and very curious memory considerations), and
   experimental Rewrite machines (synchronous networks with fine
   grained parallelism -- not VN at all!).

How, specifically, does this reduce the scope of our ambitions?

(1) Von Neumann architecture can be simulated on non Von Neumann
architecture by simulation -- introduce an "instruction set" in the
data space and an interpreter of that instruction set in the code
space.

(2) Ultimately, we wish a variety of primitives for computation and
communication.  We also wish a protocol which gives [estimates of] the
resources required by our computational objects.  We're leaving the
mapping of these primitives to the underlying hardware up to a
implementation specific kernel.

(2a) Communications requires standards.  Hardware is pragmatically
non-standard in many respects.  Thus we can't get away from the
requirement for a kernel which implements the standards for a specific
platform.

(3) In principle, if we do our job right (if we anticipate properly
the sorts of hardware people are going to want to use) we can design a
set of primitives which can be mapped efficiently onto a variety of
platforms.  Perhaps certain platforms will not have particularly
efficient mappings -- in that case, it may have to provide several
different forms of mapping and make a guess based on estimates as to
which mapping to use for any specific case.

Yes?  No? 

-- 
Raul D. Miller          N=:((*/pq)&|)@                 NB. public e, y, n=:*/pq
<rockwell@nova.umd.edu> P=:*N/@:#               NB. */-.,e e.&factors t=:*/<:pq
                        1=t|e*d    NB. (,-:<:)pq is four large primes, e medium
x-:d P,:y=:e P,:x                  NB. (d P,:y)-:D P*:N^:(i.#D)y [. D=:|.@#.d