Thanks very much, it now works!
I was indeed confused of 2 files with same name. I think I was in a wrong folder and compiled/linked the wrong files.
Unfortunately, I can’t compile any imp files from the EMAS site. I didn’t recall if EMAS/2 is imp77 or imp80. I haven’t dealt with imp files for a year.
So, I hope that an imp80 compiler will be available. It’s in work for at least 1.5 years as mentioned in the main IMP/EMAS thread
most of the files in the Emas archive are Imp80 for the ICL2900 series. Some of the more interesting ones I may have local copies that I’ve manually edited into Imp77. There are also older Imp9 sources from the early Emas on the ICL 4/75.
Many EMAS files were written before the days of considering portability and because both ICL platforms were big-endian (IBM style) they may not port automatically to a little-endian environment. I’ve been working on those ones on a big-endian Sparc system with an old Sparc binary of Imp77. Also Imp9 used %shortinteger which was signed but Imp80 used %halfinteger which was unsigned. Imp2022 (John’s latest version of Imp77) does now support %shortinteger but won’t handle %halfinteger so again more source code massaging would be needed. All the supposed Imp77 files in the APM archive use Hamish’s Imp80 compiler which true to form is neither pure Imp80 nor Imp77 but some hybrid with a lot of Hamish’s idiosyncratic extensions to the language and those will require source-level fixes too, not to mention almost all of them are Fred-machine hardware dependent, especially if they do any graphics. Documentation on the various Imp dialects can be found at Index of /imp77/reference-manual
Thanks very much.
I didn’t know about these variants and extensions.
I hope that at some time some of these imp files can be run.
So even emulators won’t help much with that many extensions.
And is the code backwards compatable?
I got a PIE-XYZ with a XQZ-360 emulator. 