Dan,
rviz has a feature which will clear all data and buffers if it detects a backwards jump in time.  I expect that your playback is looping and that you will get this brief error when time jumps backwards.  As far as I know that's the only way the grid cell counter would get reset. 

The tf error could be caused by a publisher publishing a wrong parent or a transform backwards.  But the above sounds more likely from your description. 

Tully

On Mon, Apr 19, 2010 at 1:57 PM, Daniel Maier <maierd@informatik.uni-freiburg.de> wrote:
On Mon, 19 Apr 2010 21:07:22 +0200, Tully Foote <tfoote@willowgarage.com>
wrote:

> Dan,
> Generating the clock message from the original carmen logs sounds like
> the
> right thing to do in this case.  The very old messages will be tossed
> out on
> receipt in the costmap_2d tf::MessageFilters because they are way out of
> date.
>
> To publish time it's simpy a roslib/Clock message on the /clock topic
> it's
> documented at http://www.ros.org/wiki/Clock
>
> Tully
>

Thanks alot, that did the trick! Costmap_2d calls the callback functions
and I can see the published map in rviz.
I only got an issue with displaying the map in rviz. Every few seconds the
grid cells get cleared and redrawn.
In the exact moment the message counter for the grid cells is reset to 0
and the .Global Status displays the following error:
For frame [/Base_link]: No transform to fixed frame [/odom].  TF error:
[Could not find a connection between '/odom' and '/Base_link' because they
are not part of the same tree.Tf has two or more unconnected trees.]
Than the grid cells reappear.
Did I miss something?

Thanks for your help!
Dan



_______________________________________________
ros-users mailing list
ros-users@code.ros.org
https://code.ros.org/mailman/listinfo/ros-users



--
Tully Foote
Systems Engineer
Willow Garage, Inc.
tfoote@willowgarage.com
(650) 475-2827