Re: [Cegcc-devel] SHGetFolderPath

[Eric H. <ee...@ee...>]

On Sun, Oct 26, 2008 at 02:34:02PM +0000, Pedro Alves wrote:
> On Sunday 26 October 2008 00:35:14, Eric House wrote:
> > On Sat, Oct 25, 2008 at 11:45:18PM +0100, Pedro Alves wrote:
> > > On Saturday 25 October 2008 18:13:00, Eric House wrote:
> > > 
> > > > If there's a bug it's that the headers reference W and A versions and
> > > > turn SHGetSpecialFolderPath into one of them.
> > > 
> > > Are you defining _WIN32_IE for some reason?  If for no good reason,
> > > don't do it.  Also, include shellapi.h to pick the correct declaration.
> > 
> > My app doesn't compile without _WIN32_IE defined.  I'm using
> > TBBUTTONINFO structs and the TB_GETBUTTONINFO message to manipulate
> > the soft keys on Smartphone.  The struct is only available if
> > _WIN32_IE is defined.
> 
> That is a bug then.  Relying on _WIN32_IE on Windows CE is nonsense.

So the bug is the "#if _WIN32_IE >= 0x400" stuff below?

#if _WIN32_IE >= 0x400
typedef struct {
	UINT cbSize;
	DWORD dwMask;
	int idCommand;
	int iImage;
	BYTE fsState;
	BYTE fsStyle;
	WORD cx;
	DWORD lParam;
	LPSTR pszText;
	int cchText;
} TBBUTTONINFOA, *LPTBBUTTONINFOA;
...
#endif

I'd offer to supply a patch, but I'm not capable of testing it.  Simply
removing the #if...#endif will not cause any errors on any device I have
access to.

http://msdn.microsoft.com/en-us/library/ms864705.aspx says this
structure should be available (the message that depends on it is) on
WCE 3.0 and later.  So in the WCE case that's the right #if check.
But again, how does one test a change like this?

#if _WIN32_WCE >= 0x0300 || _WIN32_IE >= 0x400
...
#endif

--Eric
-- 
******************************************************************************
* From the desktop of: Eric House, xw...@ee...                        *
*        Crosswords 4.1.4 for PalmOS now ARM-native: xwords.sourceforge.net  *
******************************************************************************