frank - 2013-02-23

I ran into an rxqueue conflict on windows 7, when I wanted to test something (quick + dirty) directly in the Regina37w64 directory. In this case ooRexx finds and tries to use Regina's rxqueue instead of its own rxqueue.

My script used option /LIFO (upper case), apparently Regina expects /lifo (lower case) , maybe remove this restriction, or agree with the ooRexx project on whatever one and only correct case.

Still in "quick + dirty" mode I replaced /LIFO by /lifo and tried again. Now Regina's rxqueue accepted the option and triggered an OS error, auto-submitted to Microsoft for whatever they do with "problem reports".

I think there are four problems:
(1) On my side, don't try to use ooRexx rxqueue when the current directory contains another rxqueue
(2) On Regina's side, mention rxqueue in the README*.txt files, because "put everything into a directory of your PATH" does not cut it on platforms with conflicting rxqueue binaries.
(3) On the rxqueue side, don't crash if whatever is needed to run rxqueue does not exist, die gracefully as in the "don't know /LIFO, try /lifo" case.
(4) For the classic REXX users, please minimize incompatibilities between ooRexx and Regina, and/or document known incompatibilities, and/or start a shared test suite of scripts supposed to work with "any" interpreter on "any" OS.