On Tue, Dec 16, 2014 at 6:01 PM, dmccunney <dennis.mccun...@gmail.com> wrote:
> On Tue, Dec 16, 2014 at 5:59 PM, Ralf Quint <freedos...@gmail.com> wrote:
>> On 12/16/2014 2:50 PM, Louis Santillan wrote:
>>> On Tue, Dec 16, 2014 at 2:32 PM, dmccunney <dennis.mccun...@gmail.com> 
>>> wrote:
>>>> Hardware is steadily smaller, faster, and cheaper.  Have fun finding a
>>>> new x86 machine these days that *isn't* 64 bit.
>>> There are still new 32-bit x86 parts being manufactured, notably by
>>> Intel for IoT in their Intel Edison/Quark/Galileo platform(s)
>>> [0][1][2] and DM&P's 86duino platform [3].  The 86duino even boots
>>> FreeDOS.
>> Even if all Intel based PCs are equipped with 64bit capable CPUs, they
>> will just as happy run 32bit or even 16bit code just fine.
> Assuming OS support is there.  The instruction set is the same.
> Various system calls may not be.  If you want to run DOS apps on a 64
> bit Windows machine, you need a VM or emulator.  They won't run
> "native".

I'm no engineer, but ... this isn't quite true. First of all, why can
Linux's DOSEMU work better than NTVDM, even on x64? And it's not even
using any hardware emulation features at all. On DOSEMU x64 (only),
16-bit has to be software emulated, but 32-bit stuff like DJGPP runs
at native speed, regardless of the lack of V86 mode. Obviously IA-32
host's support is even better.

Every 386 clone had V86 mode, thus allowing native speed for 16-bit
apps. In fact, you could use 32-bit instructions in 16-bit mode.
However, you can't mix 64-bit with anything, and not all AMD64 clones
were equal. "The instruction set is the same." No, it's not, not at
all. Ignoring the weird and minor differences between EM64T and AMD64,
there's also the problem of (lack of) VT-X, e.g. "paged real mode" and
"unrestricted guest mode execution", neither of which is available on
all 64-bit processors, not even by the same cpu maker. I don't know
why this is, maybe patents?? Either way, it "probably" [citation
needed] affected MS and their decision to phase out legacy support.
Unfortunately, MS made their business by encouraging proprietary
binaries, so there's a ton of legacy that will never get ported to
Win64 (PE+). Besides, to make it worse, "XP Mode" and Hyper-V weren't
/ aren't available to Home users at all. So even when it halfway
works, we didn't get anything.

So we're forced to use other third-party solutions like VirtualBox or
QEMU or DOSBox, which can work well or badly depending on what you're
trying to do. (NTVDM wasn't ever perfect either, but XP wasn't too

BTW, read this ("How retiring segmentation in AMD64 long mode broke VMware"):


>> And there are as mentioned above now with the Intel Quark X1000
>> processor again 32bit, single core/thread CPUs coming out for which a
>> 16bit FreeDOS can be a very viable option for an OS to run on...
>> DOS is not dead but people need to treat DOS as DOS, not as a second
>> coming of Linux...

As soon as people stop treating Linux as the second coming of Windows ....

> The fundamental issue for DOS is exactly what you *do* with it, and
> *why* you might use DOS in preference to something else like Linux.
> The fact that something  *can* run DOS doesn't necessarily mean it
> *should*.

<sarcasm> Forget UNIX. Just run Windows, everyone else does! </sarcasm>

New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
Freedos-user mailing list

Reply via email to