Hi Michal,
Sorry for not answering earlier - I was not looking to roottalk. The error is not yours :-\ (except calling CloseGeometry twice, but this is harmless...) - the problem was in a recent modification of mine trying to minimize duplication of geometry objects during Align(). For non-replicated volumes the matrix was directly changed (without cloning the volume/node) and this was obviously wrong if the parent of the volume was replicated (changing at the same time the positions in all replicated branches). So much for optimizations...
So thanks a lot for reporting the problem !!! I attached a fix that Rene will put in CVS. I CC-ed all people I know that might be interested by the change.
Cheers,
Andrei
MichaĆ DwuĆŒnik wrote:
>Hi Rooters, Andrei,
>sorry to bother you now...
>
>I'm using CVS head, and noticed a bit strange behaviour of the method in topic.
>The following macro shows a few rectangles around the imaginary beampipe,
>which I want to align. I create a new TGeoCombiTrans as Align() argument,
>printing them shows the desired values, and in the same time _all_ the physical
>nodes get aligned by the last of the TGeoCombiTrans...
>
>Could you please have a look and point the errors , if they're mine?
>Regards
>Michal
>
>
>
Index: geom/src/TGeoPhysicalNode.cxx
< if (!node->GetVolume()->IsReplicated()) { < vm = node->GetVolume(); < nnode = node; < continue; < } ---Received on Fri Feb 10 2006 - 09:56:01 MET
> // if (!node->GetVolume()->IsReplicated()) {
> // vm = node->GetVolume();
> // nnode = node;
> // continue;
> // }
This archive was generated by hypermail 2.2.0 : Mon Jan 01 2007 - 16:31:57 MET