> I found this today:
>
> http://tippach.business.t-online.de/wdosx/
Last updated in 2005. Works pretty well, though, for what it does.
> Correct me if I'm wrong, but this looks like you
> need to bind the executable with their stub.
Yes, that's one big downside to it. (It's also a bit tricky to shell out to other apps, and it has some minor incompatibilities with CWSDPMI.)
There's actually several stubs: Watcom/LE, DJGPP/COFF, and PE/Win32.
> How much effort would be required to get it to
> do the same thing as HX and not require any
> change to the executable you are trying to run?
One guy, Veit Kannegieser, back in 2008 (?) or so was using it for his unofficial DPMI target of Virtual Pascal. I can't find the file online, but he had some attempt at what you're mentioning in WDSX.ARJ. I might have a backup somewhere.
> I haven't yet confirmed whether the Phar Lap
> TNT Dos Extender is the same as HX.
https://en.wikipedia.org/wiki/Phar_Lap_Software
>> Later on the TNT DOS extender was created, which was a version of
>> 386|DOS-Extender that emulated the Win32 environment, complete
>> with flat address space and threading. Again this DOS extender
>> was often bound to existing Win32 applications. MASM 6.1 and the
>> 16-bit version of the Visual C++ 1.0 compiler were Win32
>> applications written for a beta version of Windows NT that was
>> bound with the TNT DOS Extender
> Note that the only win32 executables I care about
> are ones that were built with PDPCLIB - so normally
> only using msvcrt.dll, but sometimes kernel32.dll too.
>
> And note that I have my own msvcrt.dll.
WDOSX needs relocations, which mainly means old .EXEs. Things like Delphi 4 (and a few others) were supported, though.
> And note that HX has some bugs that affect me, so I'm
> looking for other options.
>
> This would be for running under Freedos, not PDOS.
Nothing's perfect. Don't get your hopes up too high. |
Thread view
Board view
