September – 59 Articles

Baker Linden announced that he had Large Group downloads working consistently. He was starting to work with Oz to get a project viewer out. A few days later a Large Groups Project Viewer rolled out. Another version rolls out later. On the 18^th Large Group Editing made it to the RC channels.

A fix is released to Linden Support to fix inventory problems where folders appear out side the main root folder.

A Mesh Deformer Project viewer was released. It used the then new feature of allowing the user to decide which base avatar model to use. Another was released a few days later. The later one made all previous rigged mesh for the deformer obsolete. The need for testing clothes became acute this month and I believe began delaying the Mesh Deformer project. Darien Caldwell did a test of the then recently released version of the Mesh Deformer. See the video:

See my Second Life Mesh Deformer Test for the details. Later in September more Mesh Deformer discussion occurred about how the Deformer works and what could be expected. September 25^th we got another release of the Mesh Project Deformer. Toward the end of the month there was more discussion about having multiple base models for the Deformer to work from.

People are starting to look at Windows 8 and Second Life in September.

September 9^th sees the JIRA CHANGE. This was big drama.

Due to roll out problems and the bandwidth problems Oskar Linden explained about how fast track rollouts work.

Andrew Linden did some work on SL Wind. The changes were on ADITI. The changes to Interest List processing were still in progress.

Some users had noticed an apparent problem with llCastRay(). Andrew Linden explained away some of the confusion.

The JIRA CHANGE remained a topic of discussion in SL user groups.

Nyx Linden asked those with complex avatars volunteer for testing server side avatar baking.

A mesh physics limit is discovered and discussed.

Unfounded rumors persist about Pathfinding affecting region performance. New Viewer Statistics were added to the viewer to display Pathfinding related performance factors. The Lindens explain again that disabling Pathfinding has no effect on region performance, it only keeps out Pathfinding characters. The Lindens released a more polished PF Kit: Modular Pathfinding Kit. For more links on Pathfinding see my Week 38. In week 39 the Pathfinding Events changed.

Feature requests were added to the JIRA. One was for the number of groups you have joined to be displayed somewhere in the Panel. Another was to add a Go To Line# feature to script editor.

In #SL CCIIUG Week 37 Oz discussed some of the challenges in deciding how to control Materials from the build panel.

In another meeting Oz Linden discussed the unlikely possibility of the Lab doing a 64-bit Second Life Viewer.

The week of September 19^th the bandwidth fix finally made it to the main channel.

Region state save on restart seemed to goof up in September. Simon and Kelly Linden were tracking down the problem.

See Drongle McMahon’s: Single convex hull physics weights (long and technical) for the reason bumpy mesh roads may never be fixed. I think it is at this point it was realized the Lindens were not going to be able to fix this problem.

September was the month when the Lindens described the fix for avatar-bake-fail as a server side compositing service. Near the end of the month Oz Linden explained out the Lab expected it to work and their ideas for implementing it. See: Server Side Avatar Baking.

September is the month the new Linden Lab Projects were officially announced.

Memory leaks continue to hold up 3.4.x viewer development. That is holding up release of the Pathfinding Tools. By the end of the month the problem was still unresolved.

Another update of Havok was working its way toward the RC channels. Andrew and Falcon Linden were figuring out how to minimize region-crossing problems during testing.

In September we heard the Lindens were still kicking ideas around internally about how to improve chat. We may see something happen in 2013… may.

In talking with Nyx Linden about pivot points for mesh, then post JIRA Chage, we learned the best way to shift priority on a feature or bug is to keep bringing it up in the Content-Mesh meeting.

Andrew Linden told us he expected the V1 World Map API to be shut down when the Avatar Bake process rolls out. So, we can expect that to happen in early 2013. He was anticipating setting up some regions in ADITI with all the old API’s removed for testing viewers. SL is complex enough no one knows what changes will affect which other parts. It was and is a try it and see process.

Andrew pointed out that it may be possible to turn off mini map data for a region, which could improve game play in combat regions.

Large regions was discussed at a Server and Scripting meeting.

Near the end of the month we start to hear of Sudden Massive Lag and crashing events. See the SL Forum thread: Major Increase in Instant SIM LAG & Crash during large events.