On 6/6/07, Cameron Skinner <cam_at_cs.auckland.ac.nz> wrote:
>
> Hi Ahmad (and all you other developers),
>
> Thank you for tracking down that bug. I have made a slightly different
> fix for the problem - the patch you sent actually undoes a deliberate
> architecture change I made for version 0.50. You were right that there
> was a problem with sending unchanged properties to the kernel, but the
> correct fix (with the 0.50 architecture) is to insert a PropertyFilter
> into the Output so that only the most recent changes are sent to the
> kernel.
>
> In the future I would like to extend the kernel so that it warns about
> multiple changes to a single property, so if any of you are keen then
> you can implement that. That feature should prevent similar bugs in the
> future.
>
> I have put the fix into sourceforge. Can you please download the main
> trunk version and test it to make sure I've got it right?
OK, everything seems to work well.
The command is:
>
> svn co https://roborescue.svn.sourceforge.net/svnroot/roborescue/trunk
>
> Thank you again for your help. I would not have been able to find it by
> myself in a reasonable amount of time, mostly because I don't have any
> real agents to use for testing - just the sample ones.
Yes, I think the main reason of server bugs was this.
Cheers,
Ahmad.
Cheers,
> Cameron.
>
_______________________________________________
robocup-rescue-s mailing list
robocup-rescue-s_at_cc.gatech.edu
https://lists.cc.gatech.edu/mailman/listinfo/robocup-rescue-s
Received on Wed 06 Jun 2007 - 11:36:29 GMT
This archive was generated by hypermail 2.2.0 : Wed 06 Jun 2007 - 11:36:31 GMT