Australia extends its sympathy to the U.S. in the wake of Katrina's devastation. We share your grief. I can only hope that members of this group have not been ...
jon harrop's original implementations, in c++ and ocaml: http://www.ffconsultancy.com/free/ray_tracer/comparison.html 105 and 62 lines, respectively. the...
stevan apter
sa@...
Sep 10, 2005 2:15 pm
2634
I've written a Space Invaders emulator in Factor. Screenshots here: http://factor.modalwebserver.co.nz/factor_space_invaders.jpg ...
Does someone used Joy as a finite automata? How Deterministic Finite Automton should be defined in Joy. Also, how do we implement the algorithm of converting...
... I've done DFAs in Forth, but never in Joy. I used a little toolkit someone else wrote that lets you build a jump table using a table-like syntax, so I was...
... compiler ... would ... This is a bit off topic but somewhat related. I've recently been reading about the SSA form and it's functional equivalent, ANF ...
I was probably not clear, but I meant to use Joy to write compilers for other language and was not asking on how to write a compiler for Joy. The reason I find...
I was probably not clear, but I meant to use Joy to write compilers for other language and was not asking on how to write a compiler for Joy. The reason I find...
... The conversion algorithm is just 4 lines of plain English in Aho and Ullman, Introduction to Automata Theory, Languages and Computation, p 22. I doubt...
... If you just want to see how (extremely simple) translations can be done, on the Joy page look at The basic library symlib.joy (symbol manipulation,...
... I've read about Linear Lisp before but using something resembling Linear Lisp as an intermediate form doesn't seem much different to just using ANF (or...
... Joy doesn't need a garbage collector as deallocation is always explicit. In other words pop could simply pop the stack in case the element on top of the...
... The point of the paper is that Linear Lisp can be compiled to a Forth intermediate form in which the variables disappear. So it would be more accurate to...
... that all ... Forth ... representation ... optimizations ... things. ... Yes. But that still leaves you to convert a concatenative program into an...
... A more efficient approach would be a copy-on-write scheme, employing a reference count behind the scenes. (The off-the-top-of-my-head implementation is...
... employing a ... the ... it is ... retains. ... That's exactly how MONKEY (implmented in C) does garbage collection. So far, no uncollectable data have...
... Joy currently has no primitives analogous to set-car! and set-cdr! for creating circular lists. However, it would be necessary to have a reference count...
... I'll check it out! Thanks. ... Counts would not need to be updated recursively-- only one level in. Consider an example: a pointer to a list is on the top...
... The problem comes in when you free a list: you then have to decrement the reference counts of the members, and potentially free them too, and so on....
... The keyword there is "potentially", rather than "necessarily", as would be the case in the "deep copy dup" scenario. You only need to recurse until you...
... Sorry, no I don't. MONKEY is my own Joy-alike; should have mentioned that in my post. There's nothing clever about the GC in MONKEY and no special results ...
... I like Baker's essay titled "Linear Logic: The Forth Shall Be First" is more appropos to this group :-), and to your question. ... dup is only nonlinear if...
... Unless, of course, you could write an efficient compiler for an intermediate form naturally derived from the concatentative calculus. I suspect, but do not...
... into ... calculus. That is exactly what I'm saying - converting concatenative programs into a lambda-calculus based form seems redundant. ... I suspect...
... "Efficient" is ambiguous; I think the two systems are different enough that any comparison will only show that one always beats the other at some...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
