31:50 – Q: Does Linden Lab regret going open source?
Oz thinks a lot of people in the Lab do regret it. It may not have been the most clever move for retaining control of one’s business. That is spilt milk. The Lab is not going to change that.
If TPV’s were to be phased out, Oz would be out of job.
33:30 – Viewer popularity. Phoenix is the number one viewer. It has been losing market share to Firestorm, the second most popular viewer. The LL viewer is the third most popular. But, it is a minority viewer by a significant margin. There is a good sized gap and then Singularity comes in 4th and is gaining market share.
34:40 – Why is LL Viewer in 3rd place?
Oz thinks the period when the Lab stopped releasing viewers while they developed Viewer 2.0 gave TPV’s a chance to capture market share. A number of social factors has made that hard to reverse.
Oz without taking a position says it can be argued that Viewer 2 did not provide well for how people did things in SL. He quotes Q Linden as saying the best thing about V2 was it changed Viewer 1.23 from the worst viewer the Lab ever released to the best, over night. (35:30)
Plus there is a significant resistance to any change in Second Life. Jessica echoes that her experience in FS/PH is the same, resistance to change.
36:15 – Oz relates the Lab’s experience with removing the V2 side bar from V3.
37:30 – Simple Inventory was recently released to try a new way of dealing with inventory. Oz feels the Lab has learned a lot from those that tried it.
38:10 – 2.k : You must not provide any feature that alters the shared experience of the virtual world in any way not provided by or accessible to users of the latest released Linden Lab viewer.
Jessica talks about how the community and developers are seeing the policy. Many are concerned that it is very broad and allows the Lab to restrict anything they want. She also points out that being able to restrict anything has always been part if the policy. She feels and other do that this means anything other than an interface change must be agreed on with the Lab before attempting it. For some it feels very heavy handed.
So, since it was not necessary, why was this change written?
Oz agrees. The Lab has always had the ability to tell TPV Dev’s that something they are doing is not good for SL and please stop. If they don’t, the Lab has the ability to block the viewer. So, the rule is not actually new.
40:05 – What they were trying to do give additional guidance. It’s a hard rule to write and it went though many iterations.
40:49 – This rule is meant to be less than most of the speculation has taken it to be. The idea is control the world not the viewer. If a viewer decided to render cubes as pyramids then one set of people would see pyramids and another group will see cubes. I think this is a horrible example that will be more confusing then helpful. But, it is setting edge cases.
Oz says there are very few things a TPV has ever done that would fall in this category. So, Niran and Kirstenlee’s viewers with enhanced render are not issues.
42:15 – Oz reiterates the Emerald Viewer’s additional attachment points.
42:45 – Jessica points out that while Emerald did break the Shared Experience it did serve as a test and show there was a desire in the community for the feature.
43:20 – Oz says with in broad limits the Lab does not mind people doing test of new features. Building prototype viewers to test new features is something the lab will work on with developers. It just cannot break parts of SL. However, releasing them to a mass audience is not the same thing.
44:00 – Jessica brings up the stifling innovation concern. Because the Dev’s can’t be sure the Lab will accept a feature they will be hesitant to try new features. That the Lab might not accept a feature is a concern, if the feature cannot then be used in the TPV. What guideline is there for how things will be classed as a shared experience feature?
45:00 – Oz’s simple rule is: If the viewer has to store information in the Lab’s asset server or other server, then it is changing the shared experience. Such features are going to require collaboration between the TPV Dev and Linden Lba. Oz says the Lab is willing to work with TPV Dev’s.
The changes made on the asset servers are way more complex than the average (what’s average in SL?) user is going to expect. The Lab is willing to work with dev’s to find the best way to implement a new shared feature. I suspect several TPV Dev’s are going to have a hard time buying into the idea the Lab will be more accepting. The Lab has a PR problem in this regard.
Oz points out that the multi-attach feature required that how the avatar’s attachments in the SL servers was stored had to change for it to work correctly. This is an example of a shred experience feature that needed collaboration.