Revision history [back]

That does not sound like a correct application of the transform math and doesn't have a physical interpretation. The transform multiplication primitives are basically matrix operations and do not have checks that enforce frame semantics. You need to make sure that your logic is correct when you write your code.

Note that it's highly recommended to simply use the TransformListener and ask it to transform the data into the coordinate frame you would prefer, instead of manually doing the transform math yourself. That way you don't need to worry about the exact sequence of the transform multiplications the library does that all for you.

The link you point to isn't official documentation, and a first skim if it appears to be correct. Your problem statement above does not reflect what I read in that article.