Jump to content

Tier reduced 40% -- How are you celebrating?


Qie Niangao
 Share

You are about to reply to a thread that has been inactive for 4520 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts

I'd think it has to be permanent. Here's my reasoning why:

1) There's already too much mesh content to drastically change the algorithm

2) Right now, the big problem are worn meshes. They're most often horribly built (there are exceptions, but I can count them on one hand). Performance hits from that (and worn sculpts) are ridiculous. If they want to decrease lag and increase performance, they'd need to go after avatars and some other crud.

Link to comment
Share on other sites


Jenni Darkwatch wrote:

 

As an example, rez a box and set it to physics shape "convex hull". You'll find that it still counts as one prim but if you link two such boxes together, it'll STILL be one prim because individually they're 0.5 land impact, which gets rounded up to one.

I tried this and nothing changed. Changed one to "none", no change. Tried changing prims in a house to convex hull or none, and gained a ton of LI. What am I missing?

Link to comment
Share on other sites

The instructions for getting a land impact of 1 from two linked convex-hull boxes should work -- that's the basics of all this.  I have seen lag sometimes slow down the recalculation of land impact, which is done server-side, so maybe that's what happened.  Otherwise... there's not much that can go wrong with this, linking two fresh, unscripted boxes.

As to the whole house thing: Again, I urge doing this as separate linksets, applying the non-Prim physics types only to those prims that benefit.  The problem is that if any component of the linkset uses non-Prim physics types, they all will use that kind of accounting, and Prim physics* is very expensive for things like tori, sculpties, etc.  Having one of those complex components in the linkset would lead to inflated Land Impact.

You also want to be sure there are no scripted components in the linkset, as discussed above.  A single, do-nothing script anywhere in the linkset will prevent any savings.

 


[* ETA: Yeah, I mean "Prim physics" here, although Convex Hull or None physics type accounting is expensive for these complex shapes, too -- but not because of the physics itself.  The detailed land impact calculatons under "More Info" show what's going on there, and why adjusting the physics type of only parts of a linkset can cause the land impact to blow up.]

Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 4520 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share

×
×
  • Create New...