Re: [ggt-info] gmtl::intersect tri and lineseg

[Sebastian M. <seb...@gm...>]

Hallo Thomas,

I've already submitted a patch for the doublesided lineseg and it might 
be included
Without diving into the code, there might be two reasons for the 
temporary. First one is prevent the compiler from aliasing the const 
reference which improves performance in most cases.
Second reason might be a linesegment has to be handled differently 
regarding the tIn/tOut in the intersection testing.
I would rather wait for Patrick to review my patches, that one includes 
a double sided intersection for tri/lineseg that we have been using in 
our project for more than one year now.

> Hi,
>
> In the intersection code for a Tri and a LineSeg I read the following 
> in Intersection.h on line 905
> | Ray<DATA_TYPE> temp( lineseg.getOrigin(), lineseg.getDir() );
> I don't understand the reason for this temporary copy and to what 
> extent 'temp' is different from 'lineseg'.
>
> If I wanted to have a double-sided Tri-LineSeg intersection would it 
> be enough to change
> | bool result = intersect( tri, temp, u, v, t );
> to
> | bool result = intersectDoubleSided( tri, temp, u, v, t );
> on line 906?
>
> I hope some can shed some light onto my two questions. Thanks in advance.
>
> Kind regards,
> Thomas
>
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>
>
> _______________________________________________
> ggt-info mailing list
> ggt...@li...
> https://lists.sourceforge.net/lists/listinfo/ggt-info