IRC channel logs

2020-06-27.log

back to list of logs

<OriansJ`>clemens3: well FORTHs tend to be interpreted, unless they are also written in FORTH. Then bootstrapping them can be quite a circular dependency pain. Best bet, try earlier versions of Gforth.
<OriansJ`>One could say, we now know it is harder to bootstrap a simple FORTH than a simple Lisp but it is harder to bootstrap a simple lisp than a simple C compiler.
<OriansJ`>although a fully standards compliant scheme is a pain in the syntax-case to bootstrap.
<OriansJ`>But I guess that is the nature of Lisp; S-expression parsing and even basic evaluation is trivial. It is just when you get into the more hairy pieces (lazy macros, syntax-case, etc) that you eventually hit a wall
<clemens3>yeah, just i simply want to compile the thing, not do a minimal bootstrap.. and as I aim for linux, c++ should be kinda ok.. I am on linux from scratch, so assembler and c/c++ kinda the only option for me by now..
<clemens3>c++=c/c++
<clemens3>so going for the old version of gforth is what I might do..
<mihi>clemens3, start with 0.5.0, which should be compilable without any older versions. Then 0.6.2 (which needs 0.5.0) then 0.7.x (which needs 0.6.2), and I think that's the latest version. I've heard that 0.5.0 may have trouble compiling on some 64-bit architectures, but I don't know what arch you are on...