Mesh physics issues

Ask questions about creating worlds, using worlds, etc.
Post Reply
Lauren Murphy
Posts: 6
Joined: Tue Dec 31, 2019 2:14 am
Has thanked: 0
Been thanked: 5 times

Mesh physics issues

Post by Lauren Murphy »

Hello,
I've built a building in Blender, with proper LODs and plane (no analyze) physics. When uploaded to SL, it has a land impact of 43, with a physics weight of 12.6.
When I upload to Kitely, it has a download weight of 8 (There are a total of 8 objects) but with a physics weight of 232.9, which of course determines the land impact.
I even tried box physics (analyze) but it made no difference.
My region is set to ubODE physics. I'm using the Firestorm for openSim.
Could you please guide me as to how to lower this physics cost? Using prims as physics every time I upload a building does not sound logical to me. I'm thinking there's actually a solution that I'm missing? Thanks very much.
My region is RoundStone Village.
These users thanked the author Lauren Murphy for the post (total 2):
Tony ElIlan Tochner
User avatar
Ilan Tochner
Posts: 6504
Joined: Sun Dec 23, 2012 8:44 am
Has thanked: 4943 times
Been thanked: 4455 times
Contact:

Re: Mesh physics issues

Post by Ilan Tochner »

Hi Lauren,

Kitely doesn't use land impact numbers for region object quota calculations. Each object prim and mesh are calculated as one object, regardless of their complexity, so you can disregard LI in Kitely for the purpose of maxmizing the number of objects you can place in your world. That, however, doesn't mean you shouldn't optimize your builds as people's viewers may not be able to render complex scenes with high FPS. My recommendation is to use optimized mesh with LODs whenever possible.
These users thanked the author Ilan Tochner for the post:
Tess Juel
Lauren Murphy
Posts: 6
Joined: Tue Dec 31, 2019 2:14 am
Has thanked: 0
Been thanked: 5 times

Re: Mesh physics issues

Post by Lauren Murphy »

I tried reverting to Bulletsim to see if anything would change and it did -= back to 8 LI, and the 232 did count against my region total cos I checked it three or four times to be sure that is what was going on

I wonder tho, if I rezz it under bulletsim will the physics cost explode if I go back to ubODE
User avatar
Gusher Castaignede
Posts: 258
Joined: Tue Mar 17, 2015 10:03 pm
Has thanked: 285 times
Been thanked: 250 times

Re: Mesh physics issues

Post by Gusher Castaignede »

....IF a mesh is not welded prior to uploading then land impact increases..
That's what I saw recently when confused why my simple mesh had a high
land impact, no matter whether I used a cube as phys or low phys setting
still showed same high land impack but once I vertex welded the mesh
and re-uploaded it went down alot...
These users thanked the author Gusher Castaignede for the post:
Tess Juel
Image
"lnteractivity, Variety, Ease of Use"
Visit my store at Country Manor
Lauren Murphy
Posts: 6
Joined: Tue Dec 31, 2019 2:14 am
Has thanked: 0
Been thanked: 5 times

Re: Mesh physics issues

Post by Lauren Murphy »

Thank you! :)
These users thanked the author Lauren Murphy for the post:
Gusher Castaignede
User avatar
Tess Juel
Posts: 267
Joined: Sun Sep 11, 2016 4:24 pm
Has thanked: 249 times
Been thanked: 438 times

Re: Mesh physics issues

Post by Tess Juel »

I'm glad Lauren brought up this. It's something that has caused a lot of problems ever since ubODE was launched and most people aren't aware why their meshes end up with - or rather seem to end up with - such crazy high land impacts. I brought it up with Ubit a while ago. He all but admitted that it's a mistake but also made it absolutely clear they are not ging to do anything about it.

To clarify a few things here: These bloated LI numbers we see for big meshes in the viewer are not real. They do not reflect the item's actual land impact nor do they reflect the item's actual load on the simulator. And it's not the viewer that is wrong. The viewer only displays data it receives from the server and can't be blamed when that data is incorrect.

However, I have filed a Firestorm bug report on the issue. It is of course possible for the viewer to override the data and replace it with the actual prim count so it can be fixed that way. It's hardly an ideal solution and I hate to bother Beq with stuff she shouldn't need to deal with but something has to be done and with the current dysfunctional state of the opensim dev team I can't see any other way.
These users thanked the author Tess Juel for the post (total 3):
Chris NamasteGusher CastaignedeIlan Tochner
Post Reply