dNormalizationResult failed

[Andreas Febrian]

Dear all, does the milis have been closed? 

Anyway, can someone give me an insight why program can show message dNormalizationResult failed? 

I try to find where the error in my code lies but I just can understand the error message. 

kyaaaaaaaaaaaaaaaaa >_<

Regards,

Andreas Febrian

dan ketika citra diri telah terluka apa lagi yang lebih indah dari pada belajar untuk menerima?  
Sent from Jakarta Capital Region, Indonesia

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Oleh Derevenko-2]

Some parts of this message have been removed. Learn more about Nabble's security policy.

I've run search through all the "*.c,*.cpp,*.h" files and can assure you that ODE sources do not contain string "dNormalizationResult".

So, go get some more sleep and come back when you can copy the error message correctly at least.

 

Oleh Derevenko
-- Skype with underscore

 

 

----- Original Message -----

From: shuyuie@...

To: ode-users@...

Sent: Monday, November 09, 2009 6:19 AM

Subject: [ode-users] dNormalizationResult failed

Dear all, does the milis have been closed? 

Anyway, can someone give me an insight why program can show message dNormalizationResult failed? 

I try to find where the error in my code lies but I just can understand the error message. 

kyaaaaaaaaaaaaaaaaa >_<

Regards,

Andreas Febrian

dan ketika citra diri telah terluka apa lagi yang lebih indah dari pada belajar untuk menerima?  
Sent from Jakarta Capital Region, Indonesia

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Andreas Febrian]

the error message is 

   "assertion "bNormalizationResult" failed C:\ode-0.10.1\include\ode\odemath.h:321"

This error has something to do with division by zero. The thing is, I have no idea why I can get this message. 

>_<

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Klaus Backert]

Hi, Andreas,

On 10 Nov 2009, at 03:18, Andreas Febrian wrote:

> the error message is
>    "assertion "bNormalizationResult" failed C:\ode-0.10.1\include\ode
> \odemath.h:321"
>
> This error has something to do with division by zero. The thing is,  
> I have no idea why I can get this message.

as a long time reader of the ODE mailing list I got the impression  
that your question is a FAQ. First thing you should do is searching  
the list archives -- of course.

http://groups.google.com/group/ode-users/search?group=ode-users&q=bNormalizationResult&qt_g=Search+this+group

delivers 247 results.

Then, show the part of your code, where the error appears. Tell what  
you have tried to solve the problem -- you have tried something,  
haven't you? Debugging, for example? Give any information which can be  
useful. Ask smart questions.

Regards
Klaus


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Andreas Febrian]

Hi all, 

The suggestion to sleep is helping me a lot. The thing is I have working in the existing code and I just try to make the code work. At first the robot's behave correctly and I'm happy and then the error pop up. After I test this error happen when the robot rotating to certain degree (I don't know yet). 

First attempt is to try find what is the error means. Since dNormalizationResult phrase seems have something to do with vector normalization, I assume this is something like division by zero (or a very small number) when normalizing a vector. The question is... what vector. As far as I know ODE use vector for anything >_< 

And so.. I tried google it. Yet, I can't seems to get any insight why the error can occur (beginner problem). I can only guess that the error have something to do with joint or body display. I am sorry if I have cause such a fuzz T_T

After I get sleep I do another attempt to overcome this error. Here's the result. The picture bellow is the model of my robot. We have to display the robot this way. The problem lies when I connect the layers with all four pillars with fix joint. It work perfectly when the robot only move forward or do small rotate. When robot reach certain degree when rotate, the error occur. 

When I set layer to have a fix joint only to one pillar, the error disappear. And then I get confuse. But I think this is why so many ODE expert dislike fix joint. 

once again... sorry I have annoy some of you >_<

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruce Veazie]

Some parts of this message have been removed. Learn more about Nabble's security policy.

Using ODE 0.10.0, debug double dll.

 

I'm working on a vehicle simulation (chassis, 4 wheels, hinge2 joints) and a trimesh track. Everything is working fairly well except for the behavior of the trimesh.

 

Instead of "behaving" as a mesh of solid triangles, the triangle edges appear to create more forces than the interior of the triangle. Even with a flat mesh, the vehicle tends to slew slightly when crossing triangle boundaries. With the trimeshes simulating superelevated turns, the forces are pretty much unpredictable.

 

Is there a compensation for that, or is it best just to abandon trimesh geoms and create box geoms to approximate the visual mesh surface?

 

Bruce

 

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Martijn Buijs]

Andreas Febrian wrote:

> Hi all,
>
> The suggestion to sleep is helping me a lot. The thing is I have working
> in the existing code and I just try to make the code work. At first the
> robot's behave correctly and I'm happy and then the error pop up. After
> I test this error happen when the robot rotating to certain degree (I
> don't know yet).
>
> First attempt is to try find what is the error means. Since
> dNormalizationResult phrase seems have something to do with vector
> normalization, I assume this is something like division by zero (or a
> very small number) when normalizing a vector. The question is... what
> vector. As far as I know ODE use vector for anything >_<
>
> And so.. I tried google it. Yet, I can't seems to get any insight why
> the error can occur (beginner problem). I can only guess that the error
> have something to do with joint or body display. I am sorry if I have
> cause such a fuzz T_T
>
> After I get sleep I do another attempt to overcome this error. Here's
> the result. The picture bellow is the model of my robot. We have to
> display the robot this way. The problem lies when I connect the layers
> with all four pillars with fix joint. It work perfectly when the robot
> only move forward or do small rotate. When robot reach certain degree
> when rotate, the error occur.
>
> ?ui=2&view=att&th=124ddb92f9924461&attid=0.1&disp=attd&realattid=ii_124ddb92f9924461&zw
>
> When I set layer to have a fix joint only to one pillar, the error
> disappear. And then I get confuse. But I think this is why so many ODE
> expert dislike fix joint.
>
> once again... sorry I have annoy some of you >_<

In general, it's not a good idea to attach bodies with multiple joints that have constraints on the
same axis/plane (e.g. for a door, use only one hinge joint rather than 2 or 3 like in a line).

In your simulation you use 4 joints to attach the 'pillars' to the 'layers'. Just put one in the
center of the pillars, or better, attach all the layer/pillar geoms to a single body, and don't use
fixed joints at all. If the layers/pillars need to wobble/deform/break, use other joints that have
more relaxed constraints.

Even if you find a way to stop ODE from crashing on this numerical problem, your construction would
probably explode because there's no margin for ODE to work with (the fixed joint constraints would
rarely/never be satisfied, the more you use, the less likely it will work).

/Martijn

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "ode-users" group.
To post to this group, send email to ode-users@...
To unsubscribe from this group, send email to ode-users+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ode-users?hl=en
-~----------~----~----~----~------~----~------~--~---