pyopengl-users

[PyOpenGL-Users] glGetDouble

[John <j2...@jt...>]

glGetFloat (GL_MODELVIEW_MATRIX) seems to return an array of integers =
instead of floats. What it returns seems to be the correct float values =
but interpreted as integers.
eg:
    glMatrixMode (GL_MODELVIEW)
    glLoadIdentity ()
    matrix =3D glGetFloat (GL_MODELVIEW_MATRIX)
    print matrix

    [[1065353216          0          0          0]
     [         0 1065353216          0          0]
     [         0          0 1065353216          0]
     [         0          0          0 1065353216]]

1065353216 is the same as 1.0 in binary.
matrix.typecode () returns 'u' Don't know what type this is??

    matrix =3D glGetDouble (GL_MODELVIEW_MATRIX)
    print matrix
    [[         0 1072693248          0          0]
     [         0          0          0          0]
     [         0          0          0 1072693248]
     [         0          0          0          0]]

Don't know why this happens with the same matrix, maybe an overflow?
matrix.typecode () returns 'l' which is long integer.

Ok? Get em fixed :P

John

Re: [PyOpenGL-Users] glGetDouble

[Pete S. <pe...@sh...>]

John wrote:
 > glGetFloat (GL_MODELVIEW_MATRIX) seems to return an array of
 > integers instead of floats. What it returns seems to be the
 > correct float values but interpreted as integers.
 >
 > 1065353216 is the same as 1.0 in binary.
 > matrix.typecode () returns 'u' Don't know what type this is??


Numeric v22 added a few new data types. one of them is that 'u' for 
unsigned integers. the problem is that they rearranged the enumeration 
for data types (instead of adding the new types to the end).

the result is that code compiled against Numeric v21 gets some funky 
typing problems when run against Numeric v22. anyways, without any 
actual testing or debugging on my end, i am guessing this is the 
problem. (call it 'speculative debugging', haha)