Why doesn't Apple start making operating systems for x86 PC'

Discussion in 'Digital Photography' started by The Ghost In The Machine, May 6, 2005.

  1. In comp.os.linux.advocacy, Ron Hunter

    on Wed, 04 May 2005 19:30:46 -0500

    > Daeron wrote:
    >> on April 28, 9:10 pm Henrietta K flatfish Thomas wrote:
    >>>XP is pure insecure shit and ..

    >> Because the Intel processor is defective in that is doesn't
    >> differentiate between data and code.

    > It can.

    Depends on what level one wants to attack the problem.
    In a very real sense, code is data -- data to the microprocessor
    decoding/LPU unit to tell it what to do next. Data can also be
    considered code, telling the register what value to use, although
    that's more of a stretch.

    I know of exactly one machine (there may be others) which
    had some very interesting ideas on how to call subroutines.
    The HP 21xx family (at least, the 2114B and 2116 I worked
    on in high school), was a 16-bit word-addressable affair,
    and had an instruction -- I'll call it JSR Label -- which
    did something like the following:

    [1] Increment the program counter.
    [2] Store the value of the program counter at the transfer address
    specified in the instruction.
    [3] Set the program counter to the transfer address.
    [4] Increment the program counter.
    [5] Execute the next instruction.

    The return was through a JMP Label,I. Or one can take the
    value at Label and stuff it somewhere, retrieving it later.

    I think I preferred the 709 (or was it 7090?) model where
    the PC was saved in an index register, but the 21xx series
    didn't differentiate all that well between registers and
    memory anyway, at least at the coding level (it had two
    accumulators; A was in memory location 0, B was in memory
    location 1).

    All of the rest of the machines I've worked on used code that
    could be stored in read-only pages (and routinely is,
    nowadays), and managed the JSR/RET problem using a stack register,
    though the 1802 had some interesting ideas of its own.

    I don't know about modern chips, but the 8086 did output which
    segment it was using when fetching RAM/ROM data. External hardware
    could, if it bothered, look at the two bits proffered and
    redirect the address to the right "memory bank", but I for one
    don't know if anyone bothered -- and modern MMUs are far more
    flexible nowadays.

    > But compilers don't seem set up for it.
    > BTW, the common computer architecture is to have both data and code
    > share ram space, so the so-called 'Princeton' class computer are pretty
    > darn scarce. Storing data and programs separately was the method used
    > by early computers, until practical tube-type memories made it feasible
    > to have data and program share the same storage. Are you implying that
    > the PowerPC is this earlier type computer? Somehow I find this a bit
    > surprising.

    So do I. Near as I figure the PPC is much like any other micro
    in its handling of memory access.

    > BTW, in the strict sense, a program IS data.

    I see you anticipate me. ;-) Of course, that's part of the
    problem. And even if one makes the distinction rigorously
    in the rest of the system, one must still somehow convert
    data, as loaded from the file system and/or disk buffers,
    into code, ready to execute.

    And then one has to generate that data -- using the linker,
    compilers, assembler, etc.

    It's still legal to go .sigless.
    The Ghost In The Machine, May 6, 2005
    1. Advertisements

  2. birdman

    birdman Guest

    The Apple decision not to to open its system a la IBM/Microsoft is probably
    the absolutely dumbest business decision in the history of business bar
    none. It was and remains such a fundamentally stupid decision and business
    strategy that it amazes me that few people realize how absoultely marginal
    Apple is in the real world of computers. Appleholics for the most part have
    no clue about the problems that have plagued the OS since day one: the only
    reason they don't get publicized more is that nobody cares. If Apple
    disappeared as a corporation tomorrow it would not even cause a ripple in
    the worldwide computer pond.
    birdman, May 7, 2005
    1. Advertisements

  3. In article <yZTee.13936$>, birdman
    <> wrote:

    > The Apple decision not to to open its system a la IBM/Microsoft is probably
    > the absolutely dumbest business decision in the history of business bar
    > none.


    Apple makes the overwhelming part of it's money from hardware sales. To
    sell a x86 version would surely cut that substantially, without
    materially increasing market share. Further, the development costs
    would increase significantly - They'd have to support all the fiddly
    little variations of cheap PC hardware.

    It's also amusing to note that the current "generic" market of x86
    hardware software is a marketing accident - solely because IBM didn't
    think enough ofthe PC market to obtain an exclusive license from MS.
    Which MS fully and completely exploited, of course. IBM sure didn't
    want the market to become generic, any more than Apple does now.
    Scott Schuckert, May 7, 2005
    1. Advertisements

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. John

    2 operating systems on start up

    John, Jun 15, 2005, in forum: Computer Support
    Ron Martell
    Jun 15, 2005
  2. Replies:
  3. =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=

    Why is there an x86 emu if a processor is x86-64?

    =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=, Jul 23, 2006, in forum: Windows 64bit
  4. Daniel
  5. Giuen
    Sep 12, 2008

Share This Page