Re: [PATCH] Add definitions to winuser.h

----- Original Message -----
From: "Danny Smith" <[email protected]>
To: "Casper Hornstrup" <[email protected]>;
<[email protected]>
Sent: Wednesday, June 19, 2002 8:10 AM
Subject: Re: [MinGW-patches] [PATCH] Add definitions to winuser.h


>
> ----- Original Message -----
> From: "Casper Hornstrup" <[email protected]>
> >
> > Index: include/winuser.h
> > ===================================================================
> > RCS file: /cvsroot/mingw/w32api/include/winuser.h,v
> > retrieving revision 1.16
> > diff -b -u -r1.16 winuser.h
> > --- include/winuser.h 14 Jun 2002 11:44:47 -0000 1.16
> > +++ include/winuser.h 15 Jun 2002 12:33:01 -0000
> > @@ -1587,6 +1587,42 @@
> >  #define VK_INSERT 45
> >  #define VK_DELETE 46
> >  #define VK_HELP 47
> > +#define VK_0 48
> ==snip==
> > +#define VK_Z 90
>
> Hard coding VK_0-VK9 and VK_A-VKZ  as per classic C-locale table okay as
> long as we stick to C-locale,  but I'm not sure how it will affect other
> locales/keyboards.  Would it be safer to define like
>
> #define VK_0  '0'
> #define VK_1  '1'
> etc.
>
> maybe it doesn't matter, but a report to mingw-users said that MS header
> does not define these at all.
>
> Danny

Correct, the MSVC 6 headers do not define them, and contain only a comment
in winuser.h about VK_A to VK_Z being equivalent to the ASCII character
codes. I don't know much about it but I think that virtual key codes should
be defined as numbers because they should be the same for all applications.
After a brief look at the docs, I think that a program that needs the
virtual key code for character '1' should use VkKeyScan(_T('1')), not VK_1.

Luke Dunstan


----------------------------------------------------------------------------
                   Bringing you mounds of caffeinated joy
                   >>>     http://thinkgeek.com/sf    <<<