[IGSTK-Users] Code broken in CVS?

Anka Kochanowska anka at bic.mni.mcgill.ca
Thu Nov 6 11:22:33 EST 2008


Thanks!
Now tracing is lot easier. I copied the changes to trace Polaris pointer.

Anka:-)


Patrick Cheng wrote:
> Hi Anka,
>
> I made a few changes to the example application.
> Please check out the cvs code and give it a try.
>
> Below is a brief description of the problem:
>
> Problem: Objects do not show up in the window. Camera unable to find 
> its focus
>
> Cause: This is because of the single thread limit. Tracker hasn't got 
> a chance to update TrackerTool until the
> igstk::PulseGenerator::CheckTimeouts(); get called in the main loop.
>
> If we setup the scene graph connection as:
> object --> tool -x-> tracker <-- View
> View can not reach object thus unable to adjust it's focus when 
> ResetCamera is called.
>
> Solution: Attach the object to the Tracker instead of TrackerTool at 
> startup. Switch to TrackerTool after tracking started.
>
> Results: You should see the ellipsoid and cylinder in the window 
> (constant) at startup.
> After tracking start, you should see them following your mouse position.
> When stop tracking, they will disappear.
>
> Thank you for your support for IGSTK!
>
> Patrick
>
>
> Anka Kochanowska wrote:
>> Hi!
>> Thanks for the clarification.
>> I did not think of it this way. Usually tracking application show 
>> something on start, but you are right - there is no reference, so 
>> nothing to show. There is a call to reset camera in the example, and 
>> I sometimes zoom and zoom and zoom and see nothing, although I know 
>> that my pointer must be visible.
>>
>> Thanks again
>> Anka:-)
>>
>> Andinet Enquobahrie wrote:
>>> Anka-
>>>
>>>> Since there are no changes in the example code, something in IGSTK 
>>>> must have been changed to cause the problems.
>>>>
>>>> So for now, I shall try to work with the release 3.0 again, hoping 
>>>> for a fix in CVS.
>>>>
>>> You are better off sticking working with cvs version. Because, there 
>>> were several other bug fixes....
>>>
>>> I just took a quick look at it your report and here are my comments
>>>
>>> 1) When you start the example application,  you "should NOT see" the 
>>> cylinder and ellipsoid with the tracking turned off.  When tracking
>>> is turned off, it means there is no tracking information which 
>>> implies the transforms of the spatial objects are "expired" or
>>> are invalid time stamps. So, what you are currently observing in the 
>>> cvs version is the correct behavior.
>>>
>>> The behavior in  release 3.0 was wrong.  The spatial objects show up 
>>> upon start EVEN without tracking information.
>>> 2) When you turn on tracking, although you will get valid transforms 
>>> and renderings of the cylinder and ellipsoid spatial objects,
>>> you will not see the renderings since the camera is way out of 
>>> focus. If you zoom out repeatedly using the right mouse button,
>>> you will see the spatial objects. We might need to add a camera 
>>> reset call to bring the renderings to the the widgets field of view.
>>>
>>>
>>> Hope that helps
>>> -Andinet
>>>
>>>
>>>
>>>>
>>>> _______________________________________________
>>>> IGSTK-Users mailing list
>>>> IGSTK-Users at public.kitware.com
>>>> http://public.kitware.com/cgi-bin/mailman/listinfo/igstk-users
>>>>
>>>
>>>
>>
>>
>



More information about the IGSTK-Users mailing list