Loading ...
Sorry, an error occurred while loading the content.
 

S/370 PC

Expand Messages
  • kerravon86
    In an alternate world, if, back at the time the IBM PC/XT was built, there existed an S/370 processor that was half the price of the 8086, and 3390 disks that
    Message 1 of 496 , Jun 15, 2017
      In an alternate world, if, back at the time
      the IBM PC/XT was built, there existed
      an S/370 processor that was half the
      price of the 8086, and 3390 disks that
      were half the price of ATA or whatever
      disks, such that IBM decided to create
      a personal computer based on the
      S/370 processor and 3390 disks and
      EBCDIC, what would PCs have looked
      like at that time, ie the 1980s or
      whatever?

      There would be no concept of connecting
      terminals to channels, as there is only
      one screen, which plugs into the one
      screen connector, a CGA/EGA/VGA or
      whatever.

      There would only be one slot to plug in
      a keyboard too, and in this case I want
      the keyboard to be the US ASCII
      keyboard so that a ctrl-x etc can be
      generated.

      Accessing the single screen would
      potentially be done by writing directly
      to the memory-mapped I/O location
      0xb8000, or perhaps a BIOS call, if
      the concept of a BIOS call exists.

      I would like to assume that load
      modules are standard MVS load
      modules, and reside in location
      SYS1.LINKLIB. Typical load
      modules may be

      SYS1.LINKLIB(TELNET)
      SYS1.LINKLIB(C3270)

      which could be used to connect
      to Unix systems or mainframes
      running standard MVS.

      I don't know what name the operating
      system may have, but it may have
      some resemblance to MVS and some
      resemblance to PDOS/370. Let's
      say it was called MPC.

      Assume that PCs at that time didn't
      have support for more than 16 MiB
      of memory, so the 24-bit addressing
      was not an issue, although also
      assume that people were told to
      keep the upper byte clean to support
      future 31-bit addressing.

      The S/370-based operating system
      would be expected to drive things
      like 1200 bps modems.

      Also assume that the PC had the
      capacity to run the entirety of MVS
      if that was actually useful. Would
      IBM have tried to keep a common
      source base for both MVS and
      MPC, given that they shared a lot
      of functionality like IEWL to create
      load modules? Would MPC be a
      subset of MVS? What modification
      would MPC have compared to MVS
      in order to drive a modem? Any
      reason that modification couldn't
      exist in base MVS, even if it was
      not normally used?

      With only 1 user, there is no issue
      with regard to lots of interrupts, so
      long as the interrupts can be handled
      faster than a user can type.

      What would the world look like if
      the S/370 processor had taken over
      the entire home market?

      Using Hercules, is it possible for us
      to posthumously create that world?
      ie instead of people mucking around
      with ARM processors, muck around
      with emulated S/370 processors
      instead. ie stop using Microsoft Word
      and instead use the functionally
      identical MPC Word, which is a
      standard MVS-like S/370 load module.

      What would the situation with regard
      to viruses look like if people had to
      code them in S/370? Unchanged, I
      assume.

      Some additions to MPC might be in
      order, such as having a "cd" command
      and "dir" command that act upon
      MVS datasets in a way I have previously
      described. ie some Unix-like behavior
      from mini-computers is transported to
      the S/370 PC world.

      BFN. Paul.
    • kerravon86
      ... And if I was familiar with z/Arch DAT etc, and MVS 3.8j was running under it, it is very unlikely that I would have come up with the bizarre idea of
      Message 496 of 496 , Jul 16, 2017
        ---In hercules-os380@yahoogroups.com, <kerravon86@...> wrote :

        > If I had a z/Arch machine with 8 GiB
        > of real memory, and under it I was
        > running MVS 3.8j which was restricted
        > to 16 MiB of real memory, then at
        > some level I believe it would be
        > obvious that an *authorized* program,
        > loaded and started by MVS 3.8j,
        > could disable interrupts, switch from
        > 370 DAT to z/Arch DAT, run a
        > 64-bit program instead of a 24-bit
        > stub

        And if I was familiar with z/Arch
        DAT etc, and MVS 3.8j was
        running under it, it is very unlikely
        that I would have come up with the
        bizarre idea of (possibly) dropping
        back to S/370 hardware and modifying
        that to support something very
        very strange to accomplish the
        same goal.

        For the same reason that it was
        completely beyond my brain's
        ability to see that I could
        intercept the SVC 120 interrupt
        in hardware instead of complicated
        MVS operating system software.

        BFN. Paul.
      Your message has been successfully submitted and would be delivered to recipients shortly.