OSCAR_linux.exe is the linux distribution of the OSCAR executable file, and OSCAR_mac.exe is the Mac OS version. tree.lisp is a sample text file that demonstrates how to give OSCAR user-defined input [and also reveals OSCAR’s rather remarkable speed when computing a close analogue of Green’s Method, classically explained as resolution-based (rather than natural-deduction-based) in G&N’s Logical Foundations of Artificial Intelligence. The README file explains how to run the executables, and what command-line options are available in the executable. To run OSCAR’s sample problems given by Pollock for linux or mac, open a command terminal, navigate to the executable file, and execute ./OSCAR_linux.exe or ./OSCAR_mac.exe, respectively.
The code that has been compiled into these executables is the same as Pollock released, save for the removal of errors given by SBCL (Pollock’s original code was written for Macintosh Common Lisp (MCL)). Additionally, the code defining so-called logical reasons has been updated, as the macros def-forwards-reason and def-backwards-reason originally used by Pollock throw errors in SBCL. Instead, the reasons are created manually. All of the sample problems create the same output OSCAR originally did, although speed benchmarking may yield different results.