On Friday July 30, 2010 Phillip Rosedale, Second Life founder, held an in world meeting to talk about the new direction and focus of Second Life and what residents can expect. RL work kept me away. I’m reading the blogs to see what went on and what people think. I wrote my reaction to some of those posts, then watched the meeting’s video. This article starts out some of my thoughts in reaction to others comments. After the title After the Video I provide a point by point summary with time marks on the video . So if you are interested in a single topic you can see if it is in the video and jump to the right spot in the video.
See the Meeting
First, you can see the video of the meeting on Treet.tv. See: Inworld Meeting with Philip Linden and BK Linden – It runs an hour long. Massively and New World Notes link to it. Massively’s Tateru Nino provides an analysis of the event. In this case I read analyses before watching the video. Sort of gives me something to listen for and relate things to. I provided time marks so you can watch just the parts pertinent to you.
Transcript is available on Treet.tv.
About 4 minutes in the real content starts. Internal plan starts at 7:50.
New Things Coming
In Tateru’s writing the delivery of textures by HTTP protocol and the addition of Collada Meshes are the next two big features to be added. The textures protocol thing is to happen straight away. Supposedly this week. This means things will rez faster and fewer stuck texture downloads. Seems there is no time table for Collada Meshes. (More on this below) The meshes will be an addition to sculpty meshes. Sculpty meshes were intended to provide organic shapes in SL and they do. Collada meshes provide for a more complex style of meshes. One could for instance make a jacket from a single mesh. It would attach to an AV but unlike current ridged attachments it would flex with the AV. This means guys could have sports coats that look masculine not painted on feminine body hugging coats.
Other than that, the Lab’s effort is to be directed at reducing crashes and getting the viewer and server code working well. It is thought this will consume their effort well into 2011. Once crashing and lag are back to acceptable levels, then regaining the technological lead is the priority. (Actually, after watching the video this is to be an ongoing and basic part of LL development culture. It will be in parallel with everything else.) The stability and techno lead issues are related. I can read into what is being said that much of the SL code, server and viewer, is to be cleaned up and made more modular. The point being; to support faster changes. Fast additions to the code will be needed when the Lab starts implementing user suggestions and feedback.
The consensus appears to be the viewer is not likely to be fixed until next year. Crashes will be addresses now. The user interface probably won’t change a lot until sometime in 2011. This means Third Party Viewers will continue to gain ground. So, Emerald, Kirsten, and Imprudence will gain popularity. Icesphere, which is an interesting new direction in viewers, will likely have a niche. If someone cleans up the install and update process for it, we may see competition for Emerald’s viewer.
Rosedale is promising better two-way communication between residents and the Lab to get SL 2.1 viewers work better than any viewer LL has ever made. Right…
The Basis of SL Success
We all want to know how SL is doing. Is our world healthy and growing or not? Our information is either personal empirical experience or the data meticulously recorded by the servers. Phillip scares me and apparently Tateru too. Phillip seems to think the user-to-user transaction metrics are the measure. Everyone has written about that and quoted studies commissioned, or at least supported, by LL that show otherwise. So, I remained concerned that Phillip is out of touch with the virtual reality of SL. That may not matter as he seems to realize content creators are key and ‘promises’ to listen more. If LL actually starts to listen then SL could move in a positive direction regardless of the metric used to measure success.
One of the studies points out that user-to-user interaction is the one single thing that is present in all MMO environments that are considered successful. If it is easy, the MMO grows and if not it languishes.
Toshiko Serenity has published an article from Bloomberg’s Businessweek (more links in this version) in the SL Blogs about the state of Real Estate in SL and how the big kid business people in SL are doing. See: Second Life: Reality Intrudes on Virtual Reality – In summary things have slowed and businesses have spread to include other virtual worlds. I have no doubt SL is in decline. Whether LL can turn it around is an unknown.
The Lab continues to test and analyze the new user experience, a good thing. Phillip is planning to change signup, eliminate welcome islands, and change the overall new user experience. He says testing suggests it may be better to eliminate welcome areas and just dump new users into the events and places that draw new users to SL. Well… it probably can’t produce worse results than the current system. Tateru and I both have doubts about placing new users in a crowded event. I’m hoping crowded events are not what Phillip actually meant.
After the Video
After watching the video my viewpoint changed and I see several things differently. The really significant change is to be in how Linden Lab operates and focuses. Phillip is providing a work philosophy and framework in which to get things done. This may be the single most important change to SL and LL. The significance of this should be apparent to those of us that have worked in high level management and seen it in action. It provides for distributed decision making and is much more effective than top down decision making. We’ll see how well they implement it.
Fast, Easy, & Fun
Watching the video, at the 7:50 mark Phillip begin to describe the company’s internal plan. Initially the mission statement like plan is to make SL fast, easy, and fun. This sets a conceptual frame work. Next are 3 task areas;
Back to Basics which Phillip describes as the effort to fix crashing and lag. He expects that to take 2 or 3 quarters to accomplish. That puts it finishing in Q1 or Q2 of 2011. This means getting SL 2.1 viewer to a lower crash rate than the SL 1.23.x viewer. The plan is to get the crash rate down and keep it down. This means that if the crash rate on a new viewer release goes above their target level, they will stop other work and concentrate on fixing it. The 2.1 release reduced crashes by 12% and is an example of how the newer code is easier to fix. This implies a concurrent effort of fixing problems and adding features is to be come a fundamental of LL culture.
At about the 13:00 mark he starts to speak about fixing lag. Phillip explains he means all types of lag; chat, SIM frame rates, crowd caused lag, rez lag, tp arrival lag, all of it. This week the HTTP protocol for texture downloading is to be turned on. Texture downloads accounts for 60% of the delay in things rez’ing. So, we can expect a visible improvement in performance. Chat and group IM are to be separated from the SIM so that chat and SIM performance are not tied together. That should help both chat and SIM performance.
Winning back Lead – This starts at the 14:00 mark. Phillip describes this as hard technology. It is about having the technical capability to deliver the worlds residents are building. He does not describe any single technology. Rather this techno lead is about changing the LL process of how technology changes are added to the system, creating a faster cycle of hearing user feedback and feature requests and implementing them. It is being able to quickly respond to user complaints and problems. It is more about changing the LL technology of building SL software from their side. This means all their software, web site, market, viewer and server. It is sort of like a programmer getting a new IDE, Integrated Developer Environment, rather than a new render engine, which we have in SL 2 viewer series.
In this area Phillip acknowledges the problems with the Viewer 2.0 series. (18:00 time mark) LL realizes the 2.0 Viewer does not provide the new user experience they wanted nor does it satisfy any group of existing users. He tells us that they will drastically change how they work with the community and the open source developers. We are to hear more on this in coming weeks… keyword being ‘weeks’. This is to result in open source developers being brought into the a very fast development cycle to add features and fixes to the main line viewer. This has always been a problem and sore spot with those developing viewers. They could find and fix a problem and their fix might never make it into the main SL viewer. Phillip is saying they will correct this problem.
Over the last few months a look at this blog and my reviews will show how fast these changes and updates have been coming in the TPV world. LL has just not been keeping up. Adding the TPV developers work to theirs could really boost their ability to fix and advance the viewer giving us all a better experience.
Grow Virtual World Economy – (22:00) I get a different take on this than Tateru’s. While user-to-user economic measures are misleading Phillip is talking about building the drill-down capabilities in the data collection to be able to see how the addition of, say, collada meshes is contributing to the economy and how meshes are being used. I take it he is looking to better understand how the LL team is doing rather than just how SL and it’s economy is doing.
In world search is a high priority. The number of people working on fixing search has been greatly increased. (22:00)
One of the problems Phillip considers serious is the complexity of a customer buying something and then having to find a sandbox to rez it, so they can put on a pair of shoes. (24:00) This is a problem they intend to fix and make it easy for folks.
The other is the new user experience (24:30). I took this differently than Tateru did. Phillip seems to be saying the idea is to deliver new users to resident content …places, the neat places in SL, and to people and groups they may be looking for. It will be possible for new folks to find their way to events more easily than now. But the idea is not about plopping new users into a crowded event. A link from a Google search or the Destination Guide will likely become the initial destination of new users. This could greatly help content creators and not just SL malls, but game creators and event locations too.
Filters & Priority – Phillip explains that the above plan is used to set priority. It also sets how things will be changed. Does a change help improve in this plan? So, in regard to Basics, does it reduce crashes and lag? In regard to Technology Lead, is it what residents need and does it help their experience? In regard to Growth, does it help content builders and new users have a better experience?
Q & A Responses (28:30). Here I got a new take on the web based viewer. To help with the growth and economy a 2D experience is being planned for the market place. Phillip does say a lot in this regard so I can only speculate wildly. But, his comment adds a foundation point for the direction web viewing will take.
Search (30:00). Day to day changes are being made. New people were brought on to improve search. It seems this has been a major task of Phillips over the last 30 days. I think that says a lot about where this important item fits in their crowded schedule and plans.
Region Crossing (30:55). This is another item to which several people have been assigned. The goal is get this smoothed out so one can sail or fly across region boundaries.
Collada Meshes (31:30). This is something the Linden’s are excited about. The slowdown is in making it fit with the pan above. They need to avoid having it degrade rez time or cause lag. They are currently working with beta users to work out how that is done.
BK Linden (33:45).
Q & A (35:05) – (36:40) Facebook question. Phillip gives a good answer that I think well explains their direction on this issue. Making SL into Facebook is not going to happen. Adding the tools residents and new users need to be able to connect and communicate is going to mean more and better social networking tools for SL.
Road Map question (38:20) – Phillip would not provide a road map level what is next. He did say in the next few days we would see various Linden project managers doing that, something very different than what we have seen in the past.
Doing over, would you do anything different question (40:20) – Butterfly in China… but interesting.
Doesn’t Fast, Easy & Fun set the wrong tone (42:15) my paraphrasing of the question. – Good explanation and expansion on the idea. You need to hear it.
Multi-Platform Browser and Mobile (44:30) – Browser is coming but the full experience will not come for a couple of years. The major effort for the next few quarters will be improving and stabilizing the SL we know. As to mobile, how one is to provide the deep immersive experience that is SL to mobile devices is being looked at. So, these seems to be low priorities until SL is back on track.
Community related question (47:20) – The Fast, Easy, & Fun is targeted at all users and communities. Educational community example used.
Fast, Easy, & Fun for Market Place (49:15) – Phillip refers back to the earlier example of the difficulty of rez’ing a new purchase.
Which Blogs do you read (49:43) – Won’t name any. Reads a lot different blogs. – BK: lots of broad technology and mobile technology.
Open Source and how does it factor in (53:30) – Details coming out on this soon. LL will work to create the right calls to action for Open Source community for help from them. Plus they are looking at the possibility of splitting up the viewer for Builder, Merchant, and User features. This does NOT mean spate viewers and it may. But they do intend to maintain a single code base.
Is Phillip back long term (57:04) – Right now everyone is focused on getting SL back on track. I would say he is saying for the next year or two Phillip is back. But, who knows.
Closing comments (58:40) – worth hearing. Mostly about communicating.
After watching the video I have a more optimistic opinion about what Phillip Rosedale can do. I’m much more optimistic on SL’s future.
Update: 2010-08-05 – New World Notes in Requiem for Qarl Linden reports the firing of Qarl Linden, aka Karl Stiefvater, a leader in developing Second Life sculpties and more recently, COLLADA-compatible meshes, a feature currently in closed Beta. See more on NWN and understand the irony in Qarl’s departure. We have a number of threads in the SL forum where he worked with Domino Marama to help him develop PrimStar and sort out many changes to sculpties.
This really has me wondering if we will see Collada meshes any time soon.