phi@insearch.cam.org (phi) (08/18/90)
>From: ast@cs.vu.nl (Andy Tanenbaum) >Subject: Re: Ftp'able ANSI C and POSIX drafts. >In article <27719@nigel.ee.udel.edu> phi@insearch.cam.org (phi) writes: >>Hello Gurus, >> I have been reading here about ANSIfication and POSIXfication >> for Minix. Does this mean that when (v2.0 ?) Minix is fully >> POSIX conformant and equiped with an ANSI C compiler (next >> release of ACK-C for Minix ?) then I can compile and run any >> program written in ANSI C for POSIX conformant systems, under >> Minix ? >Pretty much. However, there are a couple of caveats. POSIX has some >options, such as job control. An application that makes (incorrect) >assumptions about which options are present will not work. However, an >application that only uses those features REQUIRES to be present, should >work on MINIX, subject to the 64K + 64K address space limit on the PC. ^^^^^^^^^ >That will be around for the near future, anyway. Oh, no, please. Memory chips are getting dirt cheap now, and it's getting harder and harder to find the 64K chip :-) Many of us are true PC-Minix followers since the time of non- existent ST-Minix, now suddenly find that PC-Minix is only a toy in comparison to ST-Minix now. Sigh. >Also be aware that there are not many POSIX applications at present. A >POSIX conformant system will not necessarily run BSD or System V programs >unless these restrict themselves to using only features mandated by POSIX. No problems. If you remove the 64K + 64K barrier, I believe a lot of the many millions bits of software (whether copyrighted or copy- lefted or copyfronted or copybacked :-) currently floating around will be POSIXized for Minix. Caveat: It is then harder to control the growth of Minix. >Andy Tanenbaum (ast@cs.vu.nl) Thank you for giving birth to Minix and comp.os.minix. Phi-Ho Hoang. ----- phi@InSearch.CAM.ORG InSearch - For a better Solution uunet!philmtl!altitude!InSearch.CAM.ORG!phi
henry@zoo.toronto.edu (Henry Spencer) (08/19/90)
In article <27802@nigel.ee.udel.edu> phi@insearch.cam.org (phi) writes: > [64K+64K limit] > Many of us are true PC-Minix followers since the time of non- > existent ST-Minix, now suddenly find that PC-Minix is only a toy > in comparison to ST-Minix now. Sigh. This reflects the relative nature of the machines they run on. :-) :-) -- It is not possible to both understand | Henry Spencer at U of Toronto Zoology and appreciate Intel CPUs. -D.Wolfskill| henry@zoo.toronto.edu utzoo!henry
ast@cs.vu.nl (Andy Tanenbaum) (08/20/90)
In article <27802@nigel.ee.udel.edu> phi@insearch.cam.org (phi) writes: > Oh, no, please. Memory chips are getting dirt cheap now, and it's > getting harder and harder to find the 64K chip :-) The 64K + 64K limit has nothing to do with physical memory. It has to do with the 8088's brain-damaged architecture. As soon as Intel (1) fixes it and (2) removes all the existing 8088s (and 80286s) from circulation, I'll do it. I'm not going to go to infinite trouble to handle teeny-weeny, tiny, little, medium, largish, big and huge models on a bunch of different CPUs. Someday, I suppose I'll go to a single linear 32-bit address space on the 386, like MINIX for the 68000. Priority now is POSIX however. Andy Tanenbaum (ast@cs.vu.nl)