[ros-kernel] Re: WINE porting templates

[Steven Edwards]

Hey Hyperion,
Comments are at the end.

--- "KJK::Hyperion" <noog at libero.it> wrote:
> At 17.25 15/10/2003, you wrote:
> >User32 and gdi32 are likely to be a problem as the bulk of our 
> >implementation is in kernel mode win32k.sys.
> 
> I have good reasons to think that Microsoft managed to maintain a
> single 
> codebase for all post-Windows 95 variations of user32.dll, gdi32.dll
> and 
> win32k.sys. If you read the debugging symbols, you'll see the same
> naming 
> conventions of core functions among all three. And did you know
> win32k.sys 
> doesn't return status codes, but sets the last error, like it was in
> user 
> mode? My theory is that they devised the baroque protocol based on
> system 
> calls, shared memory and worker threads just to hide to 
> user32.dll/gdi32.dll the fact that they were being split into an
> user-mode 
> and a kernel-mode part and that multiple instances of them were to be
> 
> active simultaneously. I think this is how they managed to port Win32
> to a 
> protected memory system with a minimal (for small enough values of 
> "minimal") fork of their codebase. With guts and luck, we may be able
> to 
> pull the same stunt with Wine 

Very interesting. Would you be willing to help lead a effort to do this
with WINE? I know there is no way I can design and implement the
changes needed in Wine and ReactOS. It would be nice if we could plan
on dumping our user32.dll down the road and feedback our changes to
Winehq.

Thanks
Steven

__________________________________
Do you Yahoo!?
The New Yahoo! Shopping - with improved product search
http://shopping.yahoo.com