Re: [Algorithms] Scaling Models

[<ro...@do...>]

Tom Hubina  wrote:

>
>In GL the normals are scaled and rotated by the modelview matrix 
>(homogenous matrix stuff ... translations are ignored). GL allows you to 
>normalize the transformed normals with:
>

It is a very good question what the API does when the modelview matrix
contains scaling, uniform or non-uniform.  If it applies the modelview
matrix (except the translation) to normals when the modelview matrix
contains non-uniform scaling, it is incorrect.   Does it detect when
the linear part of the modelview matrix is non-orthogonal, and do the
right thing?  That would be very expensive.  Or it could just do  the
transpose(inverse(L)) routine to transform normals in all cases?  That
would be very wasteful in the majority of cases that the app uses only
orthogonal linear parts of the model and view transformations.