Forum Link: Announcement Post
Forum Link: Download Link Post
Forum Link: Transcript Link

Town Hall with Philip Linden.

[Note: First audio Town Hall, hosted by Johnny Ming and using Skype. The transcript below was corrected in some spots from the transcrict and changed slightly.

Audio version can be downloaded in mp3 form from the media page (Direct Link). Original posted transcript done by can be found here.]

[♪ Music Plays In ♪]

Johnny: You're listening to the first edition of the voice enabled Second Life Town Hall series. The following recording was held at the Town Hall meeting held April 6, 2006. Our music is "Lover's Dance" by Torley Linden. I am Johnny Ming and joining me is Philip Linden.

Philip: So thanks everybody for coming to the Town Hall. This is Philip Rosedale, aka Philip Linden, and thanks very much to Johnny and the SecondCast team for offering to handle the technical details of broadcasting and podcasting this Town Hall. We wanted to try doing a Town Hall where we could speak, where I could answer with my voice some of the questions that people had, for a few reasons.

One of the reasons is that clearly I can talk a lot faster with my voice than I can with my fingers. I'm not a very good typist. People that have suffered through my typing at previous town halls, know that. So, I wanted to be able to talk using my voice so I could get through questions more quickly, address more questions and also maybe create a format that is more entertaining and interesting for people to listen to who aren't able to make it here today.

It is obviously our goal to get as many people who are interested to listen to these town hall events, and so having an audio podcast seems like an interesting idea. The other thing that we're going to be able to do is to take callers. This is our first try at this so I guess we'll have to see how many people call in and how that goes.

Because we're using Skype we're actually able to add people to the conference and therefore have them kind of call in and talk to us in the same way that you'd normally be able to call in to a radio show, or something, so again an interesting experiment in communication with everybody in Second Life. I'd also like to say that Robin [Linden] is typing as quickly as she can. We're also going to generate a transcript of this event.

I realize, in fact I see a sign there in the audience that happily there are you know, many people using Second Life that come from a lot of different life experiences, and certainly one of those is hearing impaired. We'll do everything we can do to add text to everything that's happening, create a transcript, Robin is going to type as much as she can in real time. I'm sure we can figure how to do that even better. And we'll certainly be willing to invest in doing that so that other people that are hearing impaired, can listen along and follow what's happening. So it's definitely not our desire to move away from text in this type of a meeting, but instead to use voice as an additional capability.

So let's get started. I wanted to just take a couple of issues on at the outset here, talk a little bit. The first lighthearted thing I'd like to say is that from reading the forum postings, I'm really hoping that Khamon Fate, now I don't really know how to say that, Comeon, Cameon, is willing to call in as a guest today given the incredibly funny Linden speak suggestions given in the forum there.

The other thing I wanted to talk a little bit about was the economy, and I wanted to take the opportunity here with my voice to talk a little bit about how the economy works, to explain to people what's going on, again. The Linden Dollar has decreased in value somewhat on the LindeX exchange over the last couple months by a total about 15%, 20% or so if you go back four or five months.

And I wanted to explain a little bit about what Linden Lab can do about that and how that works. The Second Life economy is kind of a closed economy, that is to say when you buy and sell money on the currency exchange the total amount of money in world, the money supply, doesn't change.

Linden does through stipends and dwell or traffic rewards, we through those two mechanisms add Linden Dollars to the world. So we have to decide on a month to month basis how much money we're adding to the economy. In doing that we are first of all operating in much the same way the Federal Reserve does in the United States. That is we're kind of doing the best thing we can, but we're not actualy able to directly control the price of money.

The Linden Dollar goes up and down because of people's willingness to buy and sell Linden Dollars at a certain price. We don't control that, but we can put more money into the economy when we see the price going up or down. So we can put more or less new money into the economy as we see the price going up or down.

So basically what we do is we kind of watch the actual exchange rate and then we watch the rate at which the actual economy of Second Life, say as measured by total transactions between users or total new people that are coming into Second Life, we watch those variables and we also watch the exchange rate and then decide whether to increase or reduce the Linden Dollars that are going in, in the form of stipends and traffic rewards.

So that is basically how we control the economy to the limited extent that we do. The goal in doing that is to keep the economy as stable as possible so that people who are seeking to make real money by selling things in Second Life are able to count on the amount of money that can take out of the economy. So generally with respect to the currency exchange, what we're trying to is simply keep it as stable as possible over a period of several months. So when we see it going up like it has been lately in terms of the number of Linden Dollars you get to the dollar, our goal is to watch that carefully.

If we feel like there is a change that is not going to turn around we'll make an adjustment to the Linden Dollars that we're putting in the economy. So I just wanted to touch on that a little bit, maybe we can come back and talk about it more later if everybody would like to. So that's one thing. The other thing I'd like to talk about, the other thing I'd like to do is quickly run through the forum questions, some of the issues that people brought up just generally in those questions and address some specific ones.

One general question that was brought up, I believe Flipper [FlipperPA Peregrine] asked it, was that the latest preview has, Flipper mentioned a bunch of features: flexible objects, hardware lighting, things like that, that are not in the feature list, in the user voting feature list, and so I kind of wanted to touch on our process a little bit and how we do and approach things. The first thing to note about the latest preview is that it has the main code changes in the preview that is currently up in preview on the preview site are changes, the hardware lighting and another thing called occlusion culling, which are changes that were associated with a change in the rendering system to do two things.

One is to hopefully drop the crash rates. We changed the way in that we are moving information to the graphics card to hopefully make the rendering system more stable on a lot of different combinations of ATI and NVIDIA and Intel graphics cards. The changes that we made will hopefully standardize the way that we are doing things so that the crash rate drops quite a bit. Yeah didn't work, I just woke up my avatar there, sorry everybody there, I'm sitting here talking so of course I'm not typing and I'm going to sleep. So our goal with those changes is primarily to drop the crash rates.

The second thing is that both hardware lighting and occlusion culling will have the effect that they put less load on the GPU. So again our expectation is that we're increasing the frame rate. We still the crash rate and the user frame rate as being more important as primary objectives than any new feature. So a lot of our work is going to focus on that above specific features, having said that, we do watch the feature voting list.

We're obviously not doing everything in exactly the order they are voted on because in many cases the number of, the complexity of a particular feature makes it something that takes longer to do and therefore something that if you sort of multiply complexity by votes isn't necessarily the thing that makes sense to do first. Looking at the feature proposal list though, I noticed that many of the things on there are things that we've done.

I think the third feature proposal is multi-selecting inventory, which we have done. Try-before-buy which is actually the second most voted feature is something that we're doing work on now. We don't have it in a preview yet but we are looking at a way to do try-before-buy. The top feature Havok 2, is something that is the subject of a lot of discussion, but from a priority perspective there are not, and I've discussed this before, enough immediate improvements in Havok 2 to merit pushing it into development ahead of some of the bug fixing and other things that we're doing.

It's an enormously complex project which still has problems in terms of the crash problems that we're having with Havok 1 and physical objects. Havok 2 and 3 actually still have those problems, so we're still working on how to kind of make physical objects in Second Life given the complexity and how numerous those objects are. We're trying to make those objects basically work better and not crash sims or lock up sims in the same way on Havok 2 & 3 that they do on Havok 1, and once we get that figured out we'll be able to move forward and finish the integration work on Havok 2 & 3. We are doing work on that.

So I think when you look at the feature proposal behavior we are doing a lot of good work, I think we've got an unusually open forum for discussion on development. We are continuing to move forward an open source model for the Second Life code, which will obviously create a situation where everybody can start prototyping things and begin experimenting with new ideas.

Additionally I think we are unusually open, maybe sometimes to the frustration of everybody in Second Life, we're very open about telling people as soon as we start working on something. Given how software development often goes we will talk about things that we have not yet completely integrated or necessarily even validated the behavior of and therefore we'll sometimes do something and have to back out of work that we have done.

I think it's better to tell everybody what we are doing than it is to not tell people what we are doing and later on come back and have to change things. Another question that was asked in the forums was from Feynt Mistral, and that question was about CSG modeling. That user was asking, saying "Give us an answer on CSG modeling." I'm gonna kinda go out on a limb here, first of all I'm not sure exactly and maybe Feynt if you're out there you could text or call in, I'm not sure exactly what you mean about CSG modeling.

Second Life, basically has a kind of reduced complexity solid modeler already, we make primitives and stick them together basically which is similar to CSG. If what you're talking about is subtractive modeling where you remove one primitive from another primitive as a way of creating a third primitive thats kind of the difference between one and another, we don't have any plans of doing that right now. That is a difficult thing to do that we don't know how to do in our render, and something that is not going to happen right away.

There was another question in the forums from Khamon Fate, about "is there a feature in development for acknowledgment of accept or reject on inventory transfer?" The answer to that is no, we're not working on that right now. So I don't have a preview time on that. The way we do development here, we are generally not going to talk about something until it's under development and will work for us, for a smaller feature like this, in preview and it isn't there yet.

I will go back after this meeting and ask about the status of that and throw in an extra vote on our side. So, with that why don't we, let me get Johnny on here and start talking and see what we're going to do next. See if we've got anybody that's ready to call in.

I think we also have a large number of people listening which obviously is great. If we're overloading the servers then we'll just figure out a way to scale up to more servers. It's great that we've got a bunch of people listening today. I know we got maybe a hundred people standing in front of me here, and I think we've got, 400 or so you said listening on the live channel.

There's a lot of conversation in the community and there's been a lot of conversation in this office about voice and voice chat. People in Second Life obviously have a lot of different discussions about how effective voice chat can be, whether it should be something that is broadly available. People often talk about and about how voice works or doesn't work there. Obviously theres a couple of big sides to the discussion.

Voice is very fast and powerful and expressive as a way of communicating. On the other hand there is a certain magic and power to communicating quietly, communicating with text. It allows people to preserve their identity, or their real life identity, in a way that is very appropriate and consistent with the kind of, you know, new world aspects and new identity aspects of Second Life. So that is something we've really thought about a lot.

It's not an easy thing to match everybody's desires on this. One design idea I wanted to throw out for discussion in this program, and then in the forums or email or otherwise after this, was if the ability to do voice communication between people was tied to the parcel of land. So it was something you'd click a check box that said "enable voice chat" on a land parcel?, and then the people that walked on to that parcel if they had headphones on and wanted to use would just walk up to each other and voice chat would just work.

That method might be more appealing as a socially balanced kind of approach, because it wouldn't then allow people to kind of exclude each other by standing and using voice chat anywhere in Second Life while the person standing next to them either couldn't or didn't want to, but it would allow, by using it on land, people to create something like an education center where it would be turned on all the time. So I wanted to throw that out, that idea of voice chat on parcels of land as something for people to talk about and think about it.

It seems to me it would be a better idea, a little bit more consistent with the way Second Life works. Also, I think it is consistent with things like the idea of streaming onto a parcel. You stream live music today by parcel. You walk on a parcel and you hear live music. Under this model you would walk onto a parcel or chose not to, and when you were on that parcel you'd be able to chat, just a thought.

Johnny: All right Philip, the first question is going to be from Hiro Pendragon.

Philip: Hello Hiro.

Hiro Pendragon: Can you folks hear me?

Philip: I can hear you.

Hiro Pendragon: Oh terrific.

Johnny: Go ahead Hiro.

Hiro Pendragon: Sorry I just lost about 30 seconds of conversation listening to this stream. How ya'll doing today?

Philip: We're doing great! Thanks for being here; you are officially the first call-in ever.

Hiro Pendragon: I'm thrilled. I think a lot of the people look to ask what features are priorities. Maybe I could kind of ask a different question. Seeing how Linden Lab has set Second Life on a course to hit about a million users by probably the end of the year, if not early in 2007, what are the set of features that you think are essential, that we get down, before we hit this point? Not just the one or two best ones, but sort of like the main, the real ones.

Philip: So what are the features that are important for getting to a million users? A couple of things, I'd say first and foremost there is still the frame rate that the typical PC is seeing, using Second Life, has got to be substationally higher. The average frame rates for everyone using Second Life over say a day, a days time right now, is just a little bit under 15 frames per second.

But of course there are many people and many environments in Second Life where you're seeing less than 10 frames per second. We think that less than 15 or so frames per second is just unacceptable. The sense of emersiveness and the willingness of people to engage at all in the environment is critically sensitive to frame rate. We actually did an interesting analysis a while back that suggested that when somebody got a trial of Second Life, if they were running at less than eight frames per second, they almost certainly wouldn't continue using it, and if they were above eight frames per second they often would. So getting that frame rate up to, you know, say 15 for everyone.

Yeah I see somebody out there saying they're seeing 1.4 frames per second right now, I mean that's just totally unacceptable. So getting everybody up to 15 frames per second, that's definitely going to be one of the critical things we need to do to reach a million users, for example. I think a second thing we'll need to do, is we'll need to have an enormously simpler way of finding things and finding where you want to go. So for example, I bought this incredibly cool flute, called the Hyper Flute, the other day, and after I got it I think it was from Roby, I can't remember his last name...

This is just an amazing musical instrument, and you can just walk around and serenade people with it, incredibly fun, affordably priced, you know it's something that anyone would like in Second Life, but when I went in the Find window, and I went on the web and I went everywhere, I could not find that Hyper Flute anymore. So it was impossible for me to direct somebody where to get one of these things or find it again myself other than having had it specially give to me. When I happened to login and get lucky and find it. That is something that is going to be absolutely critical.

Nobody has four hours to find their first cool place to go or thing to do or toy or product or great piece of clothing. That has to happen in just a few minutes. That's not going to happen unless we have a way of searching for things in Second Life. Maybe on the web, even before you start, logging in world and doing the searching. In any case that search has got to be more like Google and less what like we've got today. So I think that is a second critical thing that we have to do to reach a million users.

The backend architecture of Second Life at this point is pretty scalable. As most people know we are up to I think 2, 200 CPUs, well over a gigabit per second of bandwidth. We are adding new servers at a rate of as much as a couple hundred per month. I think that can scale and I think that we can grow to a million users and, you know, ten thousand servers quite easily doing things the way we're doing them now. So I'd say the frame rate, and the crash rate, and the search capabilities are the high bar things that we need to do.

Johnny: Philip, we have a text question for you. Burnt Page is asking "What is the ETA for version 1.9.1 and/or mono?"

Philip: Okay, well those are two different answers. 1.9.1 Is now in preview, so the ETA on that should hopefully be less than two weeks. Our goal is to get on a two-week release cycle, that's what we've been trying to be on. And that means that we get something in to preview generally a couple of weeks before the actual release. Sometimes if there are larger features involved it's going to take longer than that. In the case of 1.9.1 we should be able to do it within the next couple of weeks.

So that's what I'd expect. Having said that, we're aggressive about pushing things into preview, we might find stuff that pushes it out a little bit longer than that but not too much. That's a pretty well under control release. With mono there is one problem. There's a problem with mono, and I think, right off the top of my head I don't know exactly what it is, if there's a Linden online or in the audience there that wants to type and tell us more about this, go for it. There is a problem with mono that is related to, I believe its security model. The ability to be able to verify byte code, I believe, and we are reaching out to the mono community to get some help on that. Once we get that fixed we will move forward. We have been able to run mono servers internally, so our goal has been to get that done in the next month or two, but I think there's a key sticking point where we're waiting. I see someone saying "mono?" Yeah, mono is the open-source version of C#. And it is a way of running the scripting engine, in some cases as much as, you know, 10 or 20 times faster than the scripts run today.

So for scripts that are doing a lot of thinking, a lot of internal computation, they're going to run a lot faster. So that means that people will be able to make items in Second Life that are much more sophisticated than the ones that they can build today, give the amount of CPU that's available.

Johnny: We have another Skype question here. This is from Norton Lazarno.

Norton Lazarno: Hello Philip, I'd just like to ask is there going to be a plan to have a time stamping facility for the receipt of IMs inside Second Life.

Philip: I know I've seen a feature in our list of work that is time stamping chat and instant messages, for the log files, is that what you're referring to? So you can essentially for your own records look at when you got messages?

Norton Lazarno: Correct.

Philip: I know that we have an item on that, and it isn't in development yet that I'm aware of. Yes we have a feature on the task list for that, so I'll add another vote to it as a result of this Town Hall. I don't have a specific delivery time for it yet, but we are aware that folks want to do that.

Norton Lazarno: Thank you.

Philip: Great, thank you.

Johnny: Hey Jay Wire(sp).

Jay Wire: Hi.

Philip: Hey Jay Wire, welcome.

Jay Wire: Hi, how are you guys doing today?

Johnny: Just great!

Jay Wire: All right, my question is, I'm a vehicle designer within Second Life, and the question we all have is what will the prim limit for physics enabled objects be raised to when a new physics engine is implemented?

Philip: Great question. So what will we raise the prim limit on a physical object like a vehicle? Our goal is to basically make it so that the physics doesn't, the number of objects you can have on a physical object doesn't matter. Because our goal is to sort of physically LOD that object, so we basically take a fixed number, say 30 or so approximate primitives, and a big physically linked object that has more objects than that, the extra objects or the smallest objects are just not physically simulated.

We can't do that today with Havok 1, but we can do it with Havok 2 because it has a different model for essentially assembling a single physical object in which some of the objects aren't added in to the vertex set for physical simulation. So our goal is to make it so that the limits for linked objects would be the normal limits, which I think is 256, correct? You could just do that with a vehicle in the same way and then make it physical. We would just make the best approximate representation that we could make of the vehicle. So that is on the list. We've done a little bit of initial work on that, and that is something we can hopefully get out when we do the next physics engine.

Jay Wire: All right, thank you.

Philip: Thank you.

Johnny: All right Philip, the next question that we have is from Judy Brody.

Philip: Well I guess we're leaning on the Skype technology here aren't we, learning that we've got to have a queing system for everybody so we can move faster. I could play a song for everybody on the hyper flute if they wanted to. It's really cool, I'm holding it. The hyper flute actually comes with a on screen, you know a HUD keyboard, and we just fixed a bug in the latest release that was related to the mouse down event on the keyboard where probably many people have noticed you sometimes click on a key and then if you move the mouse it will sort of jump the mouse to the wrong part of the screen, and so that's one of the things we've fixed in the latest release.

Johnny: I'm having trouble getting Judy on, her question is "the movie capture feature currently adds lag, and the video is jerky and not smooth" and she's wondering if that feature is going to be fixed?

Philip: That's a good question. The internal movie capture capability that we use, I think just works on Windows. I believe that, if I remember the implementation, we are taking advantage of a built in Windows feature that lets us rip video frames to, oh I think through the Windows Media encoder or something like that. I don't know if it makes sense for us to keep that feature updated because there are so many better 3rd party products for doing streaming video capture. Probably FRAPS is the best one out there for capturing video.

I think FRAPS does so much of a better job of smoothly capturing Second Life or anything else that's going on that is 3D intensive to video. I think that more people are using that, so if we can make things work as well as FRAPS and it's not too difficult, I think it would make sense for us to do that, ideally cross platform, if we can't then we'll just give up and take that feature out and let the better 3rd party apps out there do it.

Johnny: You know Philip, one of the interesting things that we've learned through doing SecondCast and having to read some of these names out loud, is it's a lot more challenging than typing them.

Philip: I know, I feel clumsy, although in the office obviously there are names that we've sometimes used around the office then laughed because we've realized that probably nobody has said them before but we have to say them to each other when we're walking around here.

Johnny: The next question, this is one of those names I'm going to bungle it, Alizarin Mondrean hopefully that's not too bad, they ask "Hi, do you have any plans on implementing MIDI into Second Life?"

Philip: So MIDI, not right now but I do think once we get to more open source support for Second Life that's definitely the kind of thing that I would expect to see. People that've had a lot of MIDI experience starting to work on it. There's a lot of general input and interface things that I know people would like to do, and we don't have enough skill around some of those projects, MIDI being a good example, here in the office to really know what to do. But I think the best long term strategy is to get more things into open source than it will to, say build a MIDI team at Linden Labs. So we don't have any plans right now, I think that's the kind of thing that open source will really enable.

Johnny: Next question is with Ron Overdrive, and he's going to have a question about lag.

Philip: Great. Lag of course one of those mythological words in Second Life that can mean almost anything, including it's also the delay in getting someone connected to Skype, is also lag... And it's also our fault, some how.

Johnny: I just got word from Adam Linden that he's moved about 75 listeners to another stream and that seems to have leveled, yeah I'm now able to keep the stream running a little bit more smoothly so I think we solved that problem.

Philip: Thank you Adam Linden, Adam Linden our Hero.

Johnny: Yes. Ron I'm going to try you again.

Philip: Leif Ming is offering, offering information as to where to buy the Hyper Flute, so there you go a public service advertisement for the Hyper Flute. I think this, at this moment I simply have to play a little more. I think I'm going to go for a science fiction theme here or something.

Johnny?: Hello.

Ron Overdrive: Yeah, How's it going? Sorry about that I'm buffering a little bit here.

Philip: No Problem.

Ron Overdrive: Alright, my question is pretty simple, will there be any interest on Linden Labs side, cause I've already run a proposal on this and I already got like 240 votes.

Philip: Wow.

Ron Overdrive: Will there be any interest in public -- non-estate -- like management tools?

Philip: Yes.-

Ron Overdrive: Because I know that there are many scripters that are already worried about whether or not their products truly are no lag or low lag. I mean obviously you can tell whether it is no lag, if it's passive. But when its active alls you can tell on like non estate owners and people who live on the mainland is whether or not it's active or passive and that really isn't helpful.

Philip: That's right our goal is to extend lag management and performance monitoring tools all the way down to the parcel level. Obviously the reason for doing that is that we probably will never be able to design in enough systems to cap or regulate the allocation of cpu resources to people as easily as we can just expose information about those resources so that everyone can you know find out what's wrong and make their own corrections.

Obviously that's kind of like the theme of Second Life in general. So the answer is yes I think that we put into the last release some top script identification and a couple of other things that let you find objects that are generating a lot of load, but as you point out we need to extend that to, we need to extend that all the way down to the parcel level so that somebody on an individual parcel can see what kind of performance is happening on their parcel.

One note I would make about that, we often will put a feature into the island that you know on to an estate before we'll put it onto a parcel. The reason for that is not because we are trying to make estates better than parcels, infact we would rather see the mainland you know grow more rapidly than estates because basically the mainland is potentially more interesting right because everyone is together on the mainland.

The problem is that in general adding a certain capability say a, you know a delecable ban list or something like that like what we did on the estate tools in the last release, is much easier to do at an estate level or at a sim level than it is to do at a parcel level. So when we make the, you know, kinda mistake of doing something first on the estates, don't read that as our prioritization its just that in many cases its extremely easy to do in code on the estates first and then bring it from there to the parcels, you know to the mainland basically. So that's why we do it in that order, not because of a design goal but because of how easily it can be done.

So I hope that's a...

Ron Overdrive: Well that makes sense.

Ron Overdrive: Well it is a piece of answer and it is an acceptable answer. But like some things its pretty simple like you could throw it in like the scanner routine in LSL. I mean you really like you got tools like Scan-Foo that can tell you whether or not objects owned by you are active or passive or objects owned by others are active or passive, why not put an LSL function to tell you how many milliseconds that object uses.

Philip: You know why, because we.-

Ron Overdrive: Because I run, I work in a sim known as The Forest most of us don't have estate tool access so we've been spending a good year trying to delag the sim, I mean we finaly found part of the problem but we're still looking for the other part and believe that its like a script somewhere that is like spiking at a point... and the estate tools only tell us like what's going on at that one time, we need some thing that's more live and updated.

Philip: Right right, I think thats an interesting point.

Ron Overdrive: The scanners function would do that perfectly.

Philip: Give us your proposal id, do you know what your proposal id on that proposal that you got those votes on?

Ron Overdrive: 1137.

Philip: Cool well everybody take a look at 1137. I'm actually trying to look it up on my computer right here... I agree. A scripting call to get back, the performance on the script sounds like a great idea.

Ron Overdrive: All right, well thank you for you time I actually got to be on the air in like five minutes I'm a dj for club 69 radio, feel free to tune in any time.

Philip: Wow, alright thank you.

Johnny: Yes, Tiberious Neruda is on the line with us. Tiberious can you hear us?

Tiberious Neruda: Great great... my question actually has to do with something that has been... Shoot, I'm drawing a blank here you'll have to excuse me, but when might we be able to use translucent skins on an av[atar]? Currently they are using. There's like a base skin that you.

Philip: Right.

Tiberious Neruda: That you put those tattoos over.

Philip: Right.

Tiberious Neruda: But if you use one with an alpha channel the base skin underneath shows through. Now I know there are some people that really like that and it's a reoccurring thing on the forums for months now.

Philip: I think that you're in fact, not only that but its one of the top voted proposals if you generalize what you are saying a little bit to sort of the multiple texture layers. Proposal 14 is generalized textured layers on avatars which I think will enable what you are talking about. Right?

Tiberious Neruda: No we're talking about making the default base avatar skin invisible when using an alpha texture so that you can have people that are translucent.

Philip: Right right, I- you know I no I'm feeling silly here I was thinking you could do that if on the base layer you use an alpha channel what happens? Or is it I think that you can use one of the tricks to make an avatar more or less invisible but you can't basically just use translucence on the base, on the base skin layer. I'll write that down I've got to claim inadequate information here to be able to speak to whether that works or not or what happens with that, but I'll carry that question back to everybody and run around and see if I can get some answers on it. Thank you.

Tiberious Neruda: Ok thank you thank you.

Philip: Yeah thanks for calling.

Johnny: Ok lets see I've got a... Jeska [Linden] sent a question here Millie Thompson asks "A while back we heard that version 2.0 of Second Life was in the works, though current updates push us to that milestone, what can we expect?"

Philip: What we're trying to do is make a release every two weeks. We're trying to deliver either bug fixes or performance improvements or new capabilities every couple of weeks. So we're trying to read what we're trying to change from a big monolithic development process to lots of little things as priority and capability demands. So a lot of the things that we were trying to do when we talked to about version 2.0 we talked about things like drawing out a really large distance and still having a high frame rate and a bunch of the stuff that we are doing like for example the hardware lighting work in 1.9.1 is part of the set of things that get us to that state. So we're not really focused on (cough cough) excuse me, a major release version number, but instead focused on just making continuous improvements, and so I don't have a plan for us to do a you know go silent for a long time and come out with version 2.0 , we're just going to keep moving things forward.

Philip: I've got another Jeska question, Johnny why don't I go with that, and we'll probably go for about another 15 minutes here. Next question is from Cilis Nephilim, my gosh that's a tough one, "With some of the problems neighbors have been presenting to one another on the mainland, is there any plans to add a tool based around the structure obstruction of skyline problem, such as the ability to mute objects you don't like seeing?"

Yes we do think being able to control sort of the visibility of things outside your parcel is a good idea. The only thing that will and I believe actually that we're doing some work on a piece of that right now the, one of the challenges there is being able to do that in a way that doesn't put too much load on the servers, because if everyone's list of what they want to be able to see is different its going to be... Computationaly create a lot of load to sort of go through that list every time you're deciding what to send everybody. But having said that we can probably work around that and yes I think one of the things that would make sense on the mainland is giving people more ability to mute objects outside of their parcel when they're inside their parcel.

Let me just stop for a second and add a really really cool feature that's in 1.9.1 that I encourage everybody to try out in preview, is this thing called Occlusion Culling and this is a really interesting performance improvement because what it does is it says is that if you have a bunch of objects you know on the land next to you and you put up a wall that you can't see through, you know, between you and those objects, they won't render anymore which means your frame rate will go way up. So this is a nice example of a feature like that, that is actually in the current preview so, everybody that's listening, go into preview, find a scene where there's a ton of stuff in front of you that's causing your frame ray to drop, and then make a box on the ground, drag that box really big in front of you so that it blocks out that stuff and watch your frame rate. It's really exciting, its a very cool feature that's built in to 1.9.1.

Johnny: Traven Sachs is here on the line, Traven can you hear us?

Traven Sachs: I can hear you can you hear me?

Johnny: Yes.

Philip: Hey Traven!

Traven Sachs: Hi guys. One of the devices that I've been working on lately is a teleportation HUD and I've been very interested to know when we're going to get the feature that allows us to teleport an avatar without actually having to go the map to do it.

Philip: Right ok so teleporting without going to the map, when we initially implemented point to point teleporting or pinpoint teleporting as I read it from the proposal on the voting list we did it we saved a little bit of time by making the teleport process as you say, bring up the map first. We do think it would make sense to be able to teleport directly without going to the map. So I agree that thats something that we should do. As with a couple of the earlier questions, I don't have a specific schedule on that feature but it is something that we'd like to do, it is something that we've got written down on our task list here so that does seem like the right behavior.

Traven Sachs: As an example if I go in to my inventory and I double click on a landmark that I've got, it pulls up a very small box that gives me the option of teleport or show it on the map. If we could do that instead of just pulling up the map using the commands that are available in LSL it would probably make it a lot nicer.

Philip: I agree 100% and I couldn't tell you without looking at the code why the method of going through the landmark is different from the method of say clicking on an SL url and launching the viewer, but there was just an opportunity to save time in the initial implementation by bringing up the map. It also made the code more testable because we had one broad path that just about everything funnelled through that ended up at the map. So I agree, what you said is exactly what I agree we need to do.

Traven Sachs: Cool I appreciate you letting me ask my question.

Philip: Yeah thank you, let me give another text question here. From Jeska, Missy Duport asks "Will SL ever make it to where we can make our friends list only visible to those we'd like to know we are online yet be able to be invisible to others we choose?"

We do have that item on the work list, I don't know if anyone is working on it but it is something that we've been talking about. It definitely makes sense, we realize that particularly in the sort of you know SL work environment, for people that are working in Second Life sometimes and then sometimes trying to hang around with their friends, the ability to be private in the same way that you in an IM session can just not notify people that you're online makes a lot of sense, by the way we're not talking about visibility in terms of seeing your avatar, we're talking about whether or not you can tell if your friends are online or get the online notifications. So I do agree that we should give you the capability to specify whether you want to tell people whether you are online or not and we've got a couple of people thinking of that now.

Johnny: Great. Fox Diller is on the line with us and they have a question about direct media control.

Fox Diller: Hey Philip!

Philip: Hey Fox!

Fox Diller: Hey I was wondering is Linden Lab currently working or considering to allow avatars to control their media streams individually. And if so is it in the pipeline?

Philip: Tell me what you mean, explain a little more.

Fox Diller: Ok, well you know how you can have an estate give each client the URL to connect to each server. Media streaming or video, I was wondering if you were going to allow individual control, so let's say I log in and set myself to a stream that I only can listen to that stream and have it controlled by a script.

Philip: Right, there's not reason not to do that, the only consideration is that it makes it possible if you can do that with say multiple objects, it makes it pretty easy for people to overload their incoming bit-rate by say playing multiple media streams from multiple target locations at the same time you know, the sort of HUD that had five TV's on it, one could imagine that. But I think that obviously HUD attachments are an interesting case of where this would really make sense. I think that it would be great to allow an individual say on an attached object to set a media stream of their own. So we're not working on that right now you might want to if its not in there already throw it into the voting system, it wouldn't be very difficult to do and there isn't any structural reason that would make it problematic. It would be straight forward.

Fox Diller: Alright, perfect cause I heard Allizare mention that MIDI inside of Second Life and I've been trying to do that actualy, because I've been working heavily on making a collaborative jam session as you will, and one of the main things that would have made it easy is if I had the ability to set an object to a stream so there's almost an instrument incoming and then use MIDI to control certain aspects of the whole session. So someone can actually be on the other end with a keyboard controlling the entire jam session. I mean that's my main, so what I'll do I'll add that in there in the voting for you.

Philip: That's great, that's great. I've got a text Softball, so Jeska says, from Taco Rubio, which is "Philip when can we expect that you'll publicly release 'office love'?" end quotes, so that's a funny one. For those who don't know one of the things we do in Linden Lab and I should say that the management and the approach to software development and to growing a great company that we take in Linden Lab has many interesting elements to it and is in many ways like Second Life, one of the things we have within Linden Lab is a thing called the Love Machine, we came up with that, actually Cory Linden came up with the idea of the Love Machine and we built a in-world not- an in-world, a web page that allows us to send a message saying thank you to anyone else in the office we'd like to you know that says something like 'Hey, thanks for finding that information for me about the API', and then you can send that message to anybody that you'd like and it kind of becomes part of their record, you know their review history at Linden Lab, so you can look at the love that you've got, you get a little email that says that you got love from Philip.

Its one of the ways we motivate ourselves and keep lots of visibility on what's going on in the office and its kind of a fun thing and, so I have no idea how I'd publicly release that but in many ways I think we already sort of get love or not for Second Life users when we do things and its pretty consistent with what we do internally.

From Aslan Pertwee "When a premium account is purchased, where does the weekly allowance come from? Is it new money that is placed into the economy or does Linden Lab purchases it off the LindeX with part of the premium account fee? If Linden Lab does not do this now, would it be considered as a way to stabilize the Linden Dollar?"

And this is what I touched on earlier, the answer is that for stipens and for new accounts, we are putting new money into the economy when those accounts come in. But that new money is at present only a fraction of the rate of increase the growth of the actual economy. In other words, the month to month growth in the Second Life economy is presently a gigantic number. The currency exchange, the in-world transaction rate is growing at a rate considerably greater actualy than the rate at which new money goes into the economy. However, as I explained before we're going to watch the currency exchange rate and the growth rate of the in-world economy and adjust the amount of money that goes into the economy to best keep those numbers, the currency exchange rate stable and the economy growing.

So as we've said before, as Second Life becomes more mature as it becomes bigger and its growth rate, as the percentage drops, we will probably drop the amount of new money that's put into the economy which means that either stipends or traffic rewards will be reduced overtime.

I've got a couple more text questions. Static Sprocket says, "Any plans to allow LSL to create notecards?" Creating notecards from within LSL is actually contemplated as part of a different development, which and I'm looking here in the feature proposal list, I think is captured in object to object communication, which is actually the proposal 6. We are working on ways for objects to send messages back and fourth to each other quickly, rather than using the either you know via XML-RPC and email or via link messages or via chatting that you can use today. We're working on a mechanism where objects can communicate with each other more rapidly, and part of that design I believe allows objects to more easily store data in a store on the server.

I'm not sure as to all the details of that -- on the simulator I mean. I'm not sure as to all the details on that but we are definitely aware of that need. Obviously also as we increase the connections that can be supported between LSL objects and the web, it will create mechanisms, you know wouldn't it be nice if you can use you know Google Base, right, to store information that was collected by an object that was sitting in Second Life, so better connectivity to the web which we're working on now continuously, will enable us, you to store information in that way as well.

I've got another question here from Prokofy Neva who asks; "Why can't we vote no on the voting page proposals, why do they only take yes?" And my answer to that is maybe we should be able to vote no on the proposals. I think that would be an interesting capability to add to the voting system, so I will take that back and talk to Robin and talk to some of the other folks here about it.

We also with respect to the voting page as you've seen us do before, if there are proposals that are highly controversial but not really development challenges we'll often reject them on that basis, saying that there's something that just needs to be just discussed in-world, rather than something that we are going to take a vote as a way of deciding whether to do or not. So I think that's the other piece of answering that, but I don't know that you know it might make a great deal of sense to allow people to explicitly vote no on proposals and we'll take that into consideration.

So I think we're probably at a good point to wrap up and see how things turned out and get feed back from everybody as to whether this was an effective format. I hope that it was, I hope we're entertaining everybody a little bit while also answering questions a bit faster. I know that I'm enjoying it.

Johnny: I do have one last Skype caller. It's Selador [Selador Cellardoor].

Philip: Selador. Come in Selador.

Selador: Hello there. Hi.

Philip: There he is.

Selador: Well I've just got a brief question if you if you've got time for it I think that one of the weaknesses of the way that things are being run at the moment is in the area of customer service. I wonder if you've got any plans for improving that in any way.

Philip: Well how would you like to see us improve customer service?

Selador: Oh gosh - that's a very big question. People are tending not to get replies to email sent before. But there are other issues as well like the Res Mod issue, where you asked for feedback and then when you got it you ignored it. I think this is an indication that not enough emphasis is being put in this particular area, and I think it could be done by perhaps employing somebody else to specifically deal with all these various aspects of customer service.

Philip: Well certainly the Res Mod issue - and to extend that you are talking about resident monitoring control of the forums, that's a complex one. I talked about the forums the other day on the SecondCast program I was on and said that unlike Second Life the forums are a unusualy negative environment. Which I think has something to do with the nature of the interface - the way forums work and the way they incent people to act. So I don't really think there is an easy answer - especially looking at other forums around the world around other products or communities. I don't think there is a simple answer to how we fix that. And I don't think hiring more people in customer support or contracting some one to handle that for us would necessarily make the forums work any better - I think they probably work as well as they can.

Having said that, if you're not getting emails responded to in customer service I'd like to hear about that. Because that's something we do take seriously - and we do monitor pretty closely our turn around times. We have a lot of people who are watching what we do there.

Obviously customer service for something like Second Life is a complicated and interesting proposal. We have everything from, you know, a very large population now to people that are doing very very sophisticated things and have all kinds of different needs in Second Life. So we're doing our best and I think quite a good job of creating different programs and services and hiring the right people to help them.

So, you're welcome to send me an email about that, and tell me what message - what email messages are not getting answered in time and I'll make sure that they are. We look at all that stuff. A lot.

Selador: Ok yes. I'll do that - most of it is just hearsay but I do know of a specific case where someone waited five weeks for a response. But any way many thanks.

Philip: Great. Well send me an email on that one case, and let me chase it down. Alright, so I guess lets wrap up for this version one of a voice enhanced town hall. I appreciate everybody coming. And I appreciate SecondCast taking on rebroadcasting this and recording it for us.

I would encourage everyone out there to respond on the forums or in email or in world and tell us what they think about his type of format. We definitely were and are able to go through more questions. I think having people able to call in and talk is a way to accelerate that even more - move more quickly through the issues. So, let us know what you think and what we should do differently with this format next time and we will continue to have these town halls regularly. And continue to answer your questions.

Thanks everybody for coming and I hope you enjoyed it, I know I did - and I hope you enjoyed the flute playing as well.

[♪ Music Plays Out ♪]