Re: new assert in box.cpp

Hi

Sorry, this is not my code - I do not know the math used there. I just saw 
that comment and thought that all such statements should be validated. So I 
added assertion check there.
Bram told me that it does not happen with double type for him, only with 
float. So I would suppose there can be very a small penetration depth which 
is calculated as negative because of larger computational errors of float 
type. The easiest way to fix the issue is just to remove that asserion 
check. However I would suggest to find the author of the code and let him 
make the decision, if possible.

Oleh Derevenko
-- ICQ: 36361783

----- Original Message ----- 
From: "Irrisor" <[email protected]>
To: <[email protected]>
Sent: 23 вересня 2007 р. 22:34
Subject: [ODE] new assert in box.cpp


> Hi,
>
> box.cpp was changed recently and now contains these lines:
>
>   if (cnum < 1) {
>   dIASSERT(0);
>   return 0; // this should never happen
>   }
>
>
> The problem with this is: it does happen!
> Oleh, do you have an idea why? I checked for NaNs in the local vars but 
> does not
> seem to be the problem...
>
> /Irrisor
> _______________________________________________
> ODE mailing list
> [email protected]
> http://ode.org/mailman/listinfo/ode
>
> 


_______________________________________________
ODE mailing list
[email protected]
http://ode.org/mailman/listinfo/ode