DSCargo@hi-multics.arpa (David S. Cargo) (06/18/87)
I'm very glad to hear that someone is putting some archives together. From my perspective (not yet having the MINIX code, only having the MINIX book) one key area for the MINIX archives would be lists of machines where MINIX is known to work as distributed, known to work with listed modifications, known to not work (no modications planned or yet available), and where MINIX is not known to work (admittedly an open ended set). If there were a relatively small test program that could be easily transferred and downloaded and run that would provide a physical configuration test and confidence test that the full MINIX system would run, I think many would find it a great boon. Likewise, "lessons learned" and "how to" descriptions (e.g., what I learned about MINIX while installing it on Brand X clone or how I use MS-DOS and Microsoft C to generate code to run on MINIX) would be worth their media in gold. If anything still remains of the Software Tool Users Group, one of their projects was to translate the RATFOR code of the UNIX-like tool set (including an ed-style editor) into C. That code while duplicating some of the code already in MINIX would greatly expand what was available. Likewise, code from the Austin Code Works, various C users groups, and other sources of public domain C programs become prime sources for building onto MINIX. I think AST has done a wonderful job with MINIX, and to keep it exciting will require a certain amount of growth (both fixing bugs in the kernel and some functional enhancements). I don't think that the growth need be undisciplined, nor so extensive as to ruin MINIX for its intended purpose. The varieties of applications that could be hosted on it would probably justify additional books where source code listings could be included and annotated. David S. Cargo (DSCargo at HI-Multics.ARPA)