[C++] PerPixelLighting with reflective water

[abraham]

the reflective soruce code has soo many errors in dev c++ aswell and i tried it in irrlicht 1.2

[sio2]

the reflective soruce code has soo many errors in dev c++ aswell and i tried it in irrlicht 1.2

What do you mean by "soo many errors"? Post the build log.

[a_haouchar]

the reason i didn post the build log becuase i thought this thread was finished



thanks for replying and i got similar errors in vc++

[ClownieTheMalicious]

im also getting errors..
vc2005EE says that you're attempting to access
a private member variable at line 126

Code: Select all

      services->setVertexShaderConstant("mWorldViewProj",&worldViewProj.M[0], 16);

specifically "M". defined in matrix4 as a private variable... I don't know why you're even using M[0] in the view projection matrix.

also, could you please give a possible workaround to the "DXbreaking" that irr 1.2 did that makes this code not work in DX.

Also, i can't get your code to compile for the life of me. mainly since idont know what to do about the private member accessing error...

[hybrid]

You should change the call &worldViewProj.M[0] to worldViewProj.pointer() in Irrlicht 1.3

[ClownieTheMalicious]

thanks for the tip hybrid- but that still doesn't compile, nor explain why the compiled binary fails under irr1.1,1.2 and 1.3 dll's.

so, hybrid, the pointer() function returns that as a pointer- could you explain to me the technical reason as to why you no longer need the memory address ('&' of mWorldProjblah) and now only need the pointer? I'm not seeing how that works... sorry new to c++ logic.
I'm a java programmer... forgive me.

[hybrid]

The array M is now private to the matrix class. So you cannot take the address of it anymore, nor can you access it any other way. You have to use the proper access methods (which allows us to track the usage and improve the testing for identity matrices).