if (ddsd.ddpfPixelFormat.dwFourCC == FOURCC_DXT1)
{
etc.
I get: "invalid operands of types ‘DWORD’ and ‘bool’ to binary
‘operator|’"
Many thanks
Michael Sgier
You did not supply enough information.
Please, reduce your problem to a minimal program that can be compiled,
so that we can help you.
How should we know what ddsd is? etc.
The error message suggest that a bool type and a DWORD type are involved.
The fact that FOURCC_DXT1 is all uppercase, suggests that it is a preprocessor macro.
Probably there is an error in this macro, or it is use incorrectly.
// To avoid having to include DirectX headers in the FlyLegacy
development environment,
// the DDSURFACEDESC2 struct definition is copied here from ddraw.h
//
#ifndef MAKEFOURCC
#define MAKEFOURCC(ch0, ch1, ch2,
ch3) \
((DWORD)(BYTE)(ch0) | ((DWORD)(BYTE)(ch1) < 8) | \
((DWORD)(BYTE)(ch2) < 16) | ((DWORD)(BYTE)(ch3) <
24 ))
#endif //defined(MAKEFOURCC)
/*
* FOURCC codes for DX compressed-texture pixel formats
*/
#define FOURCC_DXT1 (MAKEFOURCC('D','X','T','1'))
#define FOURCC_DXT2 (MAKEFOURCC('D','X','T','2'))
#define FOURCC_DXT3 (MAKEFOURCC('D','X','T','3'))
#define FOURCC_DXT4 (MAKEFOURCC('D','X','T','4'))
#define FOURCC_DXT5 (MAKEFOURCC('D','X','T','5'))
...
...
// Read surface descriptor and set relevant class data members
UINT8 factor = 0;
DDSURFACEDESC2 ddsd;
pread (&ddsd, sizeof(ddsd), 1, p);
mipCount = ddsd.dwMipMapCount;
width = ddsd.dwWidth;
height = ddsd.dwHeight;
switch (ddsd.ddpfPixelFormat.dwFourCC)
{
case FOURCC_DXT1:
if (ddsd.dwAlphaBitDepth > 0) {
format = GL_COMPRESSED_RGBA_S3TC_DXT1_EXT;
} else {
format = GL_COMPRESSED_RGB_S3TC_DXT1_EXT;
}
factor = 2;
break;
case FOURCC_DXT3:
etc.
I reverted to the original switch and get that error now:
../Utility/ImageDXT.cpp:220: error: duplicate case value
../Utility/ImageDXT.cpp:212: error: previously used here
Line 212 is the above line with the first case command
Many thanks
Michael
this is nonsense, < should definitely be << and if this is wrong anything
else can be too. probably this is abandoned code never compiled with that
conditional.
typedef int DWORD;
typedef unsigned int WORD;
typedef long int LONG;
typedef int BYTE;
typedef bool LPVOID;
Maybe there's something wrong. I didn't try such before.
Many thanks
Michael
Trying to make faulty code compile by a sequence of random changes is
not an effective strategy.
*Why* are you inserting random (and almost certainly incorrect - naively
I'd expect DWORD to be Double the length of WORD, and LPVOID to be some
kind of Pointer to VOID) typedefs named to look like macros?
The code was written for an environment which presumably already defines
these symbols, so what is preventing you from compiling it in that
environment? If you're trying to port Windows-specific (and
library-specific) code to a different platform without fully
understanding the differences between platforms, no amount of random
tinkering is likely to succeed.
--
Richard Herring
Michael,
those constants are chosen for very good reasons. That FOURCC_DXT1
actually represents a four-byte value where the four bytes contain the
characters D X T 1 (ASCII) as defined at
http://www.fourcc.org/codecs.php
The byte order is significant.
Those values above: IIRC
DWORD is unsigned 32 bit
WORD is unsigned 16
LONG is signed 32
BYTE is unsigned 8
LPVOID is void* (might be 32 or 64 bit)
From the things you are asking you are embarking on a large project
involving doing things to video. That will not be a trivial exercise to
change between systems. Are you certain there isn't an existing Linux
version available?
Andy
ahso <ahs...@yahoo.com> spake the secret code
<3461e48e-41fb-4972...@p35g2000yqh.googlegroups.com> thusly:
>typedef int DWORD;
>typedef unsigned int WORD;
>typedef long int LONG;
>typedef int BYTE;
>typedef bool LPVOID;
*none* of these typedefs are correct, however.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>