The Tuesday roll out to the main channel has been postponed to Wednesday. Wednesday’s Release Candidate roll outs have been postponed to Thursday. This delay is from some backside problems.
As expected the main channel is getting the package running on Blue Steel & Magnum. This is what they call a backend infrastructure change. We won’t see any changes in operation, unless it breaks something.
This is getting an upgrade of the code now running on Le Tigre… I’m not sure whether this is supposed to mean Blue Steel and Le Tigre will run the same code or not. Until the detail release notes come out I won’t know.
This channel may get the server fixes that are hoped to fix some of the Sudden Massive Lag problem event managers are seeing. BUG-355
This is a retry of a previous package. There is lots of new stuff in this package that has been fighting its way through QA testing.
The Havok 2012.1 Upgrade is in this package. So, we will have region crossing problems between Le Tigre and any region running an older Havok version, which is every region not part of the Le Tigre channel.
The new LSL function llGetSimStats() is included in this release.
A change in the Pathfinding function llCreateCharacter: CHARACTER_ACCOUNT_FOR_SKIPPED_FRAMES is included. The detailed documentation on the feature is not up yet. The default behavior for this option is TRUE to match previous behavior. When set to FALSE a PF character in a laggy region will not try to catch up. I suppose catch up meaning take all the steps in its behavior. Instead it skips the missed frames. The Lindens think this will make for more reliable performance that will show more visual stutter.
I think of it as the bus arriving on time. But, rather than smoothly moving from place to place it jumps from place to place in unpredictable hops.
Pathfinding throttling is being eased. In regions with ample spare time the Pathfinding character process is going to be allowed to exceed the current 50µs per frame cap.
llHTTPRequest changes are in this package. It has to do with information coming from remote servers being cached. Breedables tend to use remote servers. Previous this information was sent with a no-cache status flag. That is changing. See the Week 42 link above for more details.
Details on this channel are not out yet.
This is a small test channel in AGNI (main grid). Oskar Linden describes it as, “If you don’t know we occasionally use the snack RC for smaller use cases. They’re purpose made and usually not around for very long. This time Snack will have the code that got pulled from Magnum last week. It is only on 4 regions on the main grid. Search for ‘Snack’ from the map window.”
As of now the Lindens have not decided what change will roll to snack, if any. For now it has the Large Group Edit Changes. There are other interesting features too, but nothing else as important as Large Groups. See Release Notes.
Baker Linden tells us:
My news is that large group stuff seems to be working moderately well since its release on the Snack RC’s. There are a couple bugs, and while fairly major bugs, they should be fixed when we can integrate Simon and Kelly’s changes to the HTTP request queue. I have their code merged with mine on a region on Aditi, and everything seems to work, albeit with a few [group loading] lag issues here and there.
The issue right now is that the group name, description, and other things don’t load right now on the Snack RC’s. There may still be some lag when loading those with Simon and Kelly’s changes, but they should work — they did on my testing on Aditi this morning. [10/30]
Issue With Invisiprim-Like Behaviour
In the Mesh and & Content Creation UG the problem of: Invisiprim-Like Behavior of Alpha Textures Applied to Skinned Worn Meshes came up. The opening picture shows the problem. Notice the fur edging used on the coats hood.
That problem is getting its own JIRA item. I also understand that the Firestorm Team is maintaining a public JIRA section for SL issues. The idea is to provide a way for more people to see a bug report and add their information. I’m not sure how that is going to work. Time will tell.
Server Side Baking
Nyx Linden says they are still working on the project. There is no public version on a grid where we can test it. The change will require viewer changes so the Lab is working with Third Party Developers to have them help and test the new system.
I expect the TPDev’s to see this a couple of months before we do. For now there is still no ETA.
Avatar Physics Lag
The subject of high polygon mesh clothes in conjunction with Avatar Physics (Boobie Bounce) as a source of lag has come up.
High polygon count mesh clothes will have a problem. The concern is that in regions with lots of avatars there may be an additional lag problem from avatar physics, high polygon clothes, and the Mesh Deformer. There is a bunch of technical babble about how the Deformer handles updates and how that may lag your viewer.
However, in the graphics settings, there is a setting that controls how many avatars your viewer will render and how much time your viewer will spend time on avatar physics. So, I am not overly worried. But high polygon clothes are a load on the system. Good designers will make clothes as low poly as possible.
Interest List Update
Andrew Linden updated us on what is happening with his work on the Interest List.
I fixed some of the bugs that would cause some linked objects to not fully show up (missing child prims) and am now tracking down a performance problem… turns out I’m triggering FULL updates whenever an avatar moves, rather than just TERSE updates. So the performance for a crowd of avatars is much worse in my code. This (broken) code is currently up on Ahern on the beta grid (aditi) and I’ll be updating it as I fix stuff.
My hope is [for] measurable performance improvements in how the interest list works, maybe some correctness improvements too.
There are technical details in the transcript, which Andrew said he is posting, of how the servers are updating the viewer.