wscl VS SICL

You can't seriously say that, just because one targets a well-specified machine, that the language being used is well-specified. The determinism of a Clojure-on-JVM program would also be dependent on the particular code the Clojure compiler generates. In Common Lisp there is the Armed Bear Common Lisp implementation, which runs on a JVM. Does it benefit from JVM determinism or not? It probably does not, because the JVM is simply not aware of undefined behaviour that ABCL or Clojure are implicitly defining.

When it comes to having different platforms, it would also be necessary for any other compilers to generate semantically identical code. Different Clojure systems do _not_ do that. For example, arithmetic in ClojureScript uses JS floats where Clojure-on-JVM and others use integers of some size.

In my experience, writing a non-conforming CL program is hard, and much harder than writing a program without undefined behaviour in C. I am not sure why, other than vaguely suggesting the UB is more "localised" in some way. But there is also a modification of the ANSI standard being worked on, which attempts to eliminate undefined behaviour <https://github.com/s-expressionists/wscl>.

Common Lisp was standardised by an ANSI committee. Here is a list of issues that were voted on. Nowadays there is also the Well Specified Common Lisp project, but no issues have been voted on yet.

SICL is still an implementation of Common Lisp, and not of a new programming language (give or take some additional features, such as first-class global environments). That said, there is some overlap between the authors of SICL and the authors of Well Specified Common Lisp <https://github.com/s-expressionists/wscl>; but WSCL only really defines some undefined and contradictory behaviour in the ANSI Common Lisp specification.

The HyperSpec is a (derived work of a) language specification - its job is precisely to explain infrequently used things in too much detail. (And it ironically fails in many places.) Generally, one does not want to read a specification, unless they know they need to check something specific.

This is a very approachable paper from 1990 on one way to do it with a C kernel bootstrapping to Common Lisp: https://www.softwarepreservation.org/projects/LISP/kcl/paper... Kyoto Common Lisp (KCL) is the ancestor of today's Embeddable Common Lisp (ECL).

SICL is probably the best modern version of CL written in CL from a design standpoint, even if it's not taking over SBCL's role anytime soon: https://github.com/robert-strandh/SICL It uses some fancy bootstrapping to have the whole language available early, e.g. their definition of class 'symbol is:

    (defclass symbol (t)

That's pretty much the objective of SICL, which is "intentionally divided into many implementation-independent modules that are written in a totally or near-totally portable way, so as to allow other implementations to incorporate these modules from SICL, rather than having to maintain their own, perhaps implementation-specific versions".

   https://github.com/robert-strandh/SICL

Gladly!

https://github.com/robert-strandh/SICL (which I wrote a decent chunk of the compiler backend of.)

I applaud your curiosity and initiative to explore. Are you aware of https://github.com/robert-strandh/SICL?

I mean this Klein and this SICL. Self and Common Lisp are memory-safe, though the implementations need capabilities to manipulate memory; SICL encapsulates them using first-class global environments.

There is significant overlap with SICL and its associated pieces which supply many of the other parts needed to make a Common Lisp. Some of these are Cluster which provides a portable and extensible assembler, Eclector which supplies a portable and extensible reader, Concrete-Syntax-Tree that supports source code tracking during compilation, ctype that implements the Common Lisp type system, and Clostrum that provides first-class environments for e.g. run-time, evaluation, and compilation. The SICL project has as one of its goals the creation of portable infrastructure for implementing Common Lisp, and these pieces are novel building blocks that were created as part of the project.

Not really; you can do it with primitive operations e.g. here is the list in the Cleavir compiler and a paper on "magic" in Jikes RVM. SBCL also has a "virtual op"/vop language for code generation, and vops are written to manipulate objects with assembly snippets.

Both work. I basically never have to touch C or even FFI (cl+ssl being the main use of FFI for me), unless I am poking at SBCL guts in my spare time, and that isn't necessary either. I am sure many Haskell hackers are happy with their IO monad too.

It's used in operating systems, compilers and CLIs.

The Cleavir Common Lisp compiler uses three-address instructions in a control-flow graph, though it is intended more for production use than educational use.