🚀 go-pugleaf

In <1992Aug10.162757.13720@thyme.jpl.nasa.gov> Kaleb Keithley writes:
>I'm trying to get up to speed on PM programming, so I dragged out my
>old issues of Micros*ft Journal and started typing in Charles Petzold's
>Checkers program.  I'm using the non-EMX gcc, if that makes a difference.
>
>Question 1: Given the following code fragment...
>
>MRESULT EXPENTRY ClientWndProc (
>    HWND     hwnd,
>    USHORT   msg,
>    MPARAM   mp1,
>    MPARAM   mp2)
>{
>    ...
>    switch (COMMANDMSG (&msg)->cmd) {
>    ...
>
>the compiler complains vociferously about the above line of code.  I
>regret that I can't supply the actual error text, it was at home, and
>late at night.  I was unable to find the prototype declaration or macro
>definition for COMMANDMSG.  The above is exactly as it appears in the
>article, but I suspect that I might actually need '(COMMANDMSG (&cmd))->cmd'.
>
>Would someone be so kind as to confirm this, or tell me wherein COMMANDMSG
>is defined (I looked, but didn't find it), or tell me that COMMANDMSG is
>the wrong way to be doing this in 2.0 and tell me what the right way is?
>(...)

Firstly, under 2.0 the message parameter (msg) to your client window
procedure should be a ULONG, not a USHORT.

Secondly, I seem to remember that COMMANDMSG wasn't defined in early
beta releases of the 2.0 toolkit. I furthermore suspect that the PD
compilers' headerfiles (GCC, EMX ?) are based upon some beta release
of the 2.0 toolkit.  In the GA toolkit headerfiles (PMWIN.H),
COMMANDMSG is defined:

typedef struct _COMMANDMSG    /* commandmsg */