From: Axel Kohlmeyer (
Date: Sun Oct 14 2012 - 13:31:12 CDT


On Sun, Oct 14, 2012 at 7:31 PM, Marcos Ackel <> wrote:
> Hello Dr. Kohlmeyer,
> I've just finished building the LibNiFalcon.lib and the ICMS VPRN for
> Windows XP from the sources and using the instructions from your page. THANK
> YOU VERY MUCH for sharing such great software.
> I've used MS Visual Studio 2010 and Boost 1.51 for building.

interesting. i don't use windows, so i never tried compiling
on windows and - so far - i wasn't able to convince somebody
to try compiling and testing on windows. if you sort out the
remaining issues, would you be willing to share the resulting
executables, so i can make them available on my homepage
for others that want to use the falcon on windows to download?

> The tests included in the LibNiFalcon (version 1.0.1), mainly the
> "falcon_test_cli.exe" worked perfectly.


> Using the VRPN (I've applied the patches in the folder
> "libnifalcon-patchs"), the vrpn_server.exe gives some constant, repetitive
> errors:
> NovintFalcon #0 failed, trying to reset (Try power cycle if more than 4
> attempts made)
> Resetting the NovintFalcon #0 (attempt 1)
> Closing Falcon device 0.
> Trying to open Falcon device 0/1.
> Falcon Firmware already loaded.
> Falcon calibrated successfully.
> Move control all the way out until led turns green to activate device.

this should happen exactly once. the falcon needs to be
calibrated for use with VRPN.

> (The Falcon led goes from green to red, and then blue. After moving control
> all the way out, it turns green again - as I think it should be).
> After some debugging, it seems that the function "get_status()" is returning
> false and causing the issue. I've changed the FALCON_NUM_RETRIES from 10 to
> 300 and it's more stable now, but of course this is not the solution. Do you
> have any clue about what may be causing this problem? I've noticed that when

no. we've found that sometimes changing the USB hub helps.
even though the falcon docs say that the falcon must not be
plugged into a hub, i found that one has to use a *powered*
usb hub for best results. yet even then, sometimes changing
the make and model of the hub results in different results.

i haven't looked at the code in a very long time and don't have
access to a falcon for the foreseeable future, so i cannot do
any serious debugging for a while.

> I start VMD, the error always occurs immediately. Also, after grabbing a
> molecule using the center button and keeping it pressed, the error does
> _not_ occur.

there are likely some timing issues with reading the
status of the falcon via usb. things get really messy
when connecting two falcons to the same VRPN server.

i can't really help at this point outside of recommending
to experiment with using different hubs or trying without
a hub. perhaps on windows things are different than on

good luck,

> Any clue would be of great help.
> Marcos Ackel

Dr. Axel Kohlmeyer
International Centre for Theoretical Physics, Trieste. Italy.