> And that was my thinking behind it. Writing yet another nondescript
> GUI would be sort of pointless.
+1
> The main strength of FreeDOS is its
> ability to keep legacy software alive. But by definition there's no
> legacy software available for a GUI that's written in 2020.
+1
> Getting
> close to a free Windows 3 clone however would be a perfect
> continuation of a project that keeps legacy stuff around.
+-0
(almost) no one is using Windows 3.1 programs.
and if, they have a running Windows 3.1 machine, or running it happily
on a 32 bit Windows box.
this was different when when FreeDOS was born.
> And more
> importantly, with the wine project there is already a solid
> foundation on which such a project could be based.
nope. (I'm not a wine expert, and I'm probably oversimplifying here)
wine does very little on it's own, and mostly translates Windows calls
PutPixel(X,Y,color), DrawLine(), DrawBoxInTransparency(), PlaySound()
into Linux equivalents.
not trivial, but the main work (driver loading, multitasking,...) is still
done in the linux kernel.
lastly:
DOS kernels are ~100 K code, that was extremely well reverse
engineerd, multiple books existed about documented and undocumented
features of MSDOS, and made it possible that several alternatives
existed.
MSDOS
DRDOS
PTS-DOS
ROMDOS
and even single person project were able to create (lousy) MSDOS clones
like Pat Villanis FreeDOS kernel and Mike Podanoskis RXDOS.
the FreeDOS project was only possiible, and mostly successful, because
it started with an existing kernel. Even then it took ~8 years
(1994-2002) to make a stable, bugfree MSDOS clone.
now realize that
a) Windows 3.0 is at least an order of magnitude bigger then DOS.
more code needs more time and/or programmers to write.
b) Windows 3.0 is much less well documented then MSDOS. makes bug
finding more complicated. fortunately programmers used
undocumented stuff much less then in DOS, but some still did.
remember complains by externel companies (WordPerfect) complain
that MS programmers used undocumented (secret) stuff, giving them
an advantage over external programmers.
c) you won't find a single programer who will spend the next 5 years
for free to create something as awful as Windows 3.1.
Tom
_______________________________________________
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel
