[opensource-dev] Review Request: Allow objects to have 99.99% max hollow for default hollow shape.

Brendan Wilson wolfpup67 at earthlink.net
Tue Jul 12 05:10:34 PDT 2011


To try and help prevent this I selected to only have the default shape have
a new max for instance if you have a cylinder and are using a hollow shape
of say square in it then the max for that is still the same which is 70%.
This way hopefully the issue you are talking about should not happen as the
hollow default shape will follow the outside of the object more easily than
the others.

 

From: opensource-dev-bounces at lists.secondlife.com
[mailto:opensource-dev-bounces at lists.secondlife.com] On Behalf Of Dahlia
Trimble
Sent: Tuesday, July 12, 2011 12:06 AM
To: Viewer
Subject: Re: [opensource-dev] Review Request: Allow objects to have 99.99%
max hollow for default hollow shape.

 

One thing I noticed while coding collision geometry for OpenSimulator is as
hollow is increased and prims are twisted or otherwise manipulated such that
the hollow shape doesnt exactly follow the outer shape, the probability
increases that the surfaces formed by the triangles that make up the hollow
shape may actually protrude beyond those that make up the outer surface. I'm
not sure what SL sims do under these conditions but what ODE (Open Dynamics
Engine - most commonly used physics engine in OpenSimulator) is create a
condition where physical objects may become trapped between the resulting
surfaces. ODE uses trimesh colliders and surface normals for determining
collision surfaces and if prim hollow surfaces protrude outside of prim
outer surfaces, the surface normals for these surfaces point in the wrong
direction. This can be alleviated by increasing prim vertex count, but for
some of these prims vertex count may already be even greater than 4000
vertices. Increasing it further would use excessive memory and likely
increase the cost of computing collisions dramatically. I've found that a
limit of 95% maximum hollow is a good compromise between prim complexity and
usability. 



On Mon, Jul 11, 2011 at 7:34 PM, Wolfpup Lowenhar <wolfpup67 at earthlink.net>
wrote:


This is an automatically generated e-mail. To reply, visit:
http://codereview.secondlife.com/r/388/ 

 


Review request for Viewer.

By Wolfpup Lowenhar.


Description 


As a Builder, I want to be able to create prims that are more than 95 %
hollow which will allow me to create more realistic paper sheets, flags,
fabric bits, ribbons, etc.

Testing 


Built viewer locally and testes Viewer Phase of Acceptance Criteria. 

Bugs: STORM-58 <http://jira.secondlife.com/browse/STORM-58>  


Diffs 


*	doc/contributions.txt (d4cd5f0f33d2)
*	indra/llmath/llvolume.cpp (d4cd5f0f33d2)
*	indra/newview/llpanelobject.cpp (d4cd5f0f33d2)
*	indra/newview/skins/default/xui/en/floater_tools.xml (d4cd5f0f33d2)

View Diff <http://codereview.secondlife.com/r/388/diff/> 


_______________________________________________
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges

 

  _____  

No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1388 / Virus Database: 1516/3759 - Release Date: 07/11/11

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.secondlife.com/pipermail/opensource-dev/attachments/20110712/b2c54332/attachment-0001.htm 


More information about the opensource-dev mailing list