[vtk-developers] VTK performance regression

[Will Schroeder]

Søren-

In the future can you please post this sort of feedback to the VTK
Developers list? In the mean time I have forwarded your email to the list.
Thanks,

Will

On Wed, Oct 27, 2010 at 3:25 AM, <Soeren.Holm at hexagon-machine-control.com>wrote:

> Hi Ken And Will
>
> I'm writing to you because we have found some performance regression within
> VTK. I work at a company doing machine control solution for the construction
> industry. We use VTK to visualize the desired ground surface, the machine
> and some arrow and such to guide the user.
>
> We currently use VTK 5.4.2 but will revert back to using 5.0.4 because of
> the following two commits.
>
> [31b67b4801a09514bcace0e14130048c860d31b8] ENH:Support multiple timers.
> This checkin defines the API and provides a partial implementation. Platform
> specific subclasses will extend the behavior of this class to fully
> implement the changes.
>
> [fdeb975c7f1c6f566f4b80c13cf3cd026a37139b] BUG: Fixed problems where
> multiple observers were being invoked for the same event when one observer
> invoked another event in the same object.
>
> Especially fdeb975c7f1c6f566f4b80c13cf3cd026a37139b makes our cpu usage
> double. It does not help to any of the commits on 5.4.2 to there must be
> other regressions after these commits also.
>
> Do you have any comments as to what has happened to VTK?
>
> I would have filed a bug-report, but I do not  know what actually is wrong.
>
> /Søren Holm




-- 
William J. Schroeder, PhD
Kitware, Inc.
28 Corporate Drive
Clifton Park, NY 12065
will.schroeder at kitware.com
http://www.kitware.com
(518) 881-4902
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20101027/cffe77e5/attachment.html>