[Natsheh]

PHP Code:

    for (int i = chainlength - 1; i >= 0; i--)
    {
        int bone = chain[i];
        int parent = pbones[bone].parent;
        QuaternionMatrix(q1[bone], bonematrix);

        bonematrix[0][3] = pos1[bone][0];
        bonematrix[1][3] = pos1[bone][1];
        bonematrix[2][3] = pos1[bone][2];
        R_ConcatTransforms((parent == -1) ? rotationmatrix : bonetransform[parent], bonematrix, bonetransform[bone]);
    } 


according to the code from REHLDS it seems "i" will reach out of bounds at -1 due the following condition.

 i >= 0

Can some one prove otherwise ?

Btw this function "GetAttachment from r_studio.cpp" is so buggy i've tested it out multiple times and sometimes the results of the origin of the attachment get shifted randomly, any solutions on how to fix that?

[fysiks]

Originally Posted by Natsheh PHP Code:     for (int i = chainlength - 1; i >= 0; i--)     {         int bone = chain[i];         int parent = pbones[bone].parent;         QuaternionMatrix(q1[bone], bonematrix);         bonematrix[0][3] = pos1[bone][0];         bonematrix[1][3] = pos1[bone][1];         bonematrix[2][3] = pos1[bone][2];         R_ConcatTransforms((parent == -1) ? rotationmatrix : bonetransform[parent], bonematrix, bonetransform[bone]);     }  according to the code from REHLDS it seems "i" will reach out of bounds at -1 due the following condition. i >= 0 Can some one prove otherwise ?

I'm quite sure that i will never be -1 because of that condition. When i becomes -1 then that expression will evaluate to false and the code in the for loop will not execute again.

[Natsheh]

GetAttachment is a really helpful function too bad it cant be used due the incorrect results that sometimes acquires when the entity is rotated, ill just use offsets when entity angles and origin are equal to 0 0 0 and do the rotation math.