Re: [cgmo-webcgm] rotateAnimation test

[Lofton Henderson <lofton@rockynet.com>]

Forrest,

Your analysis looks correct to me on first glance, but I don't have the 
tools handy now to check the details.  (I'll do so later.)

It is definitely true that the rotation center point to the rotate() method 
must be NVDC.

Meanwhile -- Don & Ulrich -- please comment on Forrest's analysis.  (Each 
implementation has "Pass" on this one.)

Thanks,
-Lofton.

At 03:40 PM 10/2/2009 -0500, Forrest Carpenter wrote:

>All,
>
>
>
>In doing some testing with the java script version of the 
>tranformXCF-js.htm we found we were not taking into consideration that the 
>lower left of the transformXCF.cgm was not 0,0 but -5000,0. After we 
>changed our code to take this into consideration we no longer passed the 
>rotateAnimation test. We believe the point of rotation in the test in not 
>correct. The point in the htm file is (280.7,205.0) which is the vdc 
>coordinate of the point in mm. Because the lower left of the CGM is 
>(99,159) vdcs in mm.  We believe the  offset should be subtracted from the 
>point to make it NVDC s. The rotation point should be (181.7,46.0). This 
>may require changes in the getTransform values, but we are not sure of 
>this yet.
>
>
>
>Regards,
>
>Forrest