vixie@decwrl.dec.com (Paul Vixie) (08/25/88)
In article <272@hawkmoon.MN.ORG> det@hawkmoon.MN.ORG (Derek E. Terveer) writes:
# I have to agree. Since there seems to be a division between the 80x86
# architecture and "other" (linear) architectures, regardless of merit, we could
# really come up with two types of Unix groups:
#
# comp.unix.segment (or comp.unix.seg)
# comp.unix.linear (or comp.unix.lin)
#
# Obviously, the 80x86 CPUs would inhabit the segmented unix groups [...]
I'm sure it's too late to ward off the flame war, but I'll try.
The 80386 is arguably a linear-addressed chip, in native mode. It has
pages, and the segments can be made large enough to be invisible. Perhaps
the segments can even be turned off, I don't remember.
But this split won't work. Some CPUs are neither segmented nor linear, or
they're both. I could see a small amount of merit in {paged,swapped}, but
even there the distinction blurs on some CPUs.
Besides, I don't want people to have to know whether their machine is paged
or swapped or segmented or linear before they can post to the net. I don't
want to see an endless chain of discussion as to what segmentation means.
I don't want to see the endless arguments about whether a given CPU is or
is not linearly addressed.
Choosing a newsgroup name is a political, not a technical process.
--
Paul Vixie
Digital Equipment Corporation Work: vixie@dec.com Play: paul@vixie.UUCP
Western Research Laboratory uunet!decwrl!vixie uunet!vixie!paul
Palo Alto, California, USA +1 415 853 6600 +1 415 864 7013wes@obie.UUCP (Barnacle Wes) (08/27/88)
In article <272@hawkmoon.MN.ORG> det@hawkmoon.MN.ORG (Derek E. Terveer) writes: # ..., we could # really come up with two types of Unix groups: # comp.unix.segment (or comp.unix.seg) # comp.unix.linear (or comp.unix.lin) In article <83@volition.dec.com>, vixie@decwrl.dec.com (Paul Vixie) writes: > The 80386 is arguably a linear-addressed chip, in native mode. It has > pages, and the segments can be made large enough to be invisible. Perhaps > the segments can even be turned off, I don't remember. > > But this split won't work. Some CPUs are neither segmented nor linear, or > they're both. I could see a small amount of merit in {paged,swapped}, but > even there the distinction blurs on some CPUs. OK, so let's really do it right: comp.unix.segmented.swapped.byte-backwards (i.e. 286) comp.unix.segmented.paged.byte-backwards (i.e. 386) comp.unix.linear.paged.byte-backwards (i.e. VAX) comp.unix.bank-selected.swapped.byte-backwards (i.e. PDP) comp.unix.linear.swapped.sensible-addresses (i.e. 68K) comp.unix.linear.paged.sensible-addresses (i.e. 68020, 030) comp.os.minix.wimps (for PCs and clones) comp.os.minix.real (for STs and such) The 286 and 386 are very different processors, and the OSs for them are very different, too. comp.unix.microport was formed for Microport users, leave it alone. Xenix is a half-breed bastard, let their users have their own little corner. So we have: comp.unix.i286 Unix systems on the 286 processor comp.unix.i386 Unix systems on the 386 processor comp.unix.microport For Microport owners/users comp.unix.xenix For SCO/Microsoft/etc. owners/users -- {hpda, uwmcsd1}!sp7040!obie!wes "Happiness lies in being priviledged to work hard for long hours in doing whatever you think is worth doing." -- Robert A. Heinlein --