Scaling Second Life for “richer experience”
Seismic utterings today over at Zero Linden’s office hours. Talk of scaling Second Life up MASSIVELY. But also an emphasis on evolution of the architecture itself. I need to investigate this further - but it looks like the next stage of this evolution would potentially support a much more media rich / higher concurrency experience in specific regions with far more control over platform performance one the side of who ever was hosting that region. The whole log of this conversation will be posted tomorrow over on the Second Life site. But here is an edited down version with the stuff that is potentially relevant to delivering music and sound
[13:13] Zero Linden: Yes - The Second Life Grid Architecture Working Group had its first meeting last week
[13:15] Zero Linden: So - the basic theme is this
[13:16] Zero Linden: LL wants to evolve teh SL architecture into something that is internet wide
[13:16] Zero Linden: and to do that, we think this evolution has to take place in public
[13:18] Zero Linden: SO - about process and design
[13:18] SignpostMarv Martin: just by looking at the diagram, it seems that the plan is for the dual-domain thing is analgous to how OpenID support in Firefox would be implemented
[13:18] Zero Linden: First and foremost, the full set of use cases hasn’t been developed - at all!
[13:19] Zero Linden: Second, it was clear that we need to develope (again with you all) a clear statement about metrics for design
[13:20] Zero Linden: But I can tell you, that there will be a heavy focus on doing things in a way that
[13:20] Zero Linden: a) can be migrated from where we are today
[13:20] Zero Linden: b) Support what we have today in SL well and quickly
[13:20] Zero Linden: c) are open-ended enough to allow future extension without having designed all possible variation in from the start
[13:21] SignpostMarv Martin: ah, so no bumpy transition to SL2.0 then 
[13:21] Zero Linden: NO
[13:22] Zero Linden: the real meat of the matter (funny coming from a vegetarian)
[13:22] Zero Linden: is the protocols in these red lines
[13:22] Squirrel Wood: Vegetarians eat my foods food!
[13:22] Dizzy Banjo: so what order of magnitude are we talking about here zero.. 10.. 100.. 10000 times the size of the existing grid ?
[13:22] Zero Linden: What’s inside the boxes must be possible and scalable, but needn’t be part of the spec
[13:22] Zero Linden: Ah
[13:23] Zero Linden: the numbers
[13:23] Zero Linden: hold on to your seats
[13:23] Dizzy Banjo: 
[13:23] Zero Linden: 60M regions
[13:23] Zero Linden: 2B avatar accounts
[13:23] Zero Linden: maybe 50M to 100M on-line
[13:23] Kooky Jetaime: Zero - how many regions are we supporting now?
[13:23] Dizzy Banjo: christ
[13:23] Zero Linden: Well - current surveys say there are 120M web servers on the internet
[13:23] Dizzy Banjo: 50M - 100M concurrency ?!!?!
[13:23] Rex Cronon: how many concurrent users?
[13:24] Zero Linden: I think regions need to be eventually at least half as ubiquitous (sp?)
[13:24] SignpostMarv Martin: concurrency would imply that they’d all be logged into the same grid, which isn’t necesarily the same thing as the same service
[13:24] Zero Linden: 2B accounts roughly (very) matches some gross estimates of the number of e-mail address in use
[13:24] Tao Takashi: well, basically it sort of means that you need to be able to hold any number of regions, residents and concurrency 
[13:24] Zero Linden: So that number may or may not be high
[13:24] Zero Linden: and concurrency is much more of a crap-shoot
[13:25] Zero Linden: but Skype has 7M on line
[13:25] Zha Ewry: Note… orders, plural orders of magnitude bigger than SL today
[13:25] Zero Linden: I imagine that we might easily have 5x to 10x that overall - but
[13:25] Zero Linden: but remember that on-line might mean something more lightweight in the future
[13:26] Zero Linden: Or at least not all of them will be mvoing around, rendering 3D space….
[13:26] Dizzy Banjo: ok
[13:26] Zero Linden: They might be there as sort of static - I’m here, come talk to me, mode
[13:26] Dizzy Banjo: so this would include cellphone viewer access perhaps ?
[13:27] SignpostMarv Martin: Dizzy: you could log into SL via any device if you spent the time to develop an app using either the GPL’d source code or libSL
[13:27] Zero Linden: The key about this kind of architecture is that should leave that kind of question to the individual sim hosters
[13:27] Dizzy Banjo: sure
[13:27] Zero Linden: We define the protocol to a sim, not the implementation
[13:27] Zero Linden: there needs to be multiple implementations of each part, otherwise we don’t have something we can call the standard of the metaverse
[13:32] Dizzy Banjo: so.. this plan.. is effectively a way of scaling the existing platform of SL .. massively.. but fundamentaly not changing the way it works ie.. SL will be bigger and include more people.. but resources will still be spread to the same “thickness” ?
[13:33] Zero Linden: Dizzy - no, I don’t think so
[13:33] Zero Linden: the resource “thickness”, is mostly a design and budget constraint of the implementation of individual sims
[13:33] Zero Linden: it is quite possible that someone could throw much more, or much less, resources at a region simulation
[13:33] Zero Linden: of course- there are current today limits on viewers
[13:34] Dizzy Banjo: thats very interesting
[13:34] Zero Linden: but - so what if someone comes and says “to be in this sim, you need a kick ass video card and 4G of memory…”
[13:34] Dizzy Banjo: so you could create very very high performance areas
[13:34] Zero Linden: quite possible
[13:34] Zero Linden: within Linden’s sim today there is very little plan to increase prim count or other major “resource thickness” — (love that term, clearly)
[13:35] Dizzy Banjo: for instance you could create a very rich media experience in one sim ? by layering resources very “thickly” on it ?
[13:35] Zha Ewry nods at Dizzy
[13:35] Dizzy Banjo: lol hey Zero .. i coined that now..
[13:35] Zero Linden: We are short term mostly concerned with stability and scaling (honest!)
[13:36] Zero Linden: we would so much rather get to a place where you could plug in your own sim to run your own region
[13:36] Zero Linden: and play with those knobs yourself
[13:36] Dizzy Banjo: this is great !!
Tags: Linden Lab, Grid Architecture
You can comment below, or link to this permanent URL from your own site.
September 21, 2007 at 6:36 am
I find this stuff completely over-the-top giddy nonsense, disguised as technical wizardry. It makes the term “vapour ware” sound like a sludge of coffee grounds by comparison.
They can barely make the thing work with 50,000 logged on at once, and barely 300,000 show up each month to spend more than a $1. While I’m all for them having gadzillions, substituting the ability for somebody to talk on their cell phone to a static frozen avatar who can’t move because it is all so “thin” doesn’t sound appealing. We have avatars on Yahoo now, yanno? We can go in Vside or Kaneva or all kinds of other new stuff that is thin and portable and light and all the rest. Second Life should be for depth of worldness.
While they can get all excited about playing with knobs and such I really think a LOT more attention has to go into the economic, business, social, and political side of all this — the impact this has on the body of customers known as “the world”.
There is an illusion that the Lindens are “consulting with the community” when they hold these precious little hothouse discussions only on the very technical side of their software and servers, and only with a special few.
More here:
http://metaversed.com/20-sep-2007/second-rant-9-scaling-secnond-life
September 21, 2007 at 7:39 am
[...] Transcript of Zero Linden’s office hours [...]
September 21, 2007 at 8:55 am
I am certainly excited about all the voices (starting with Philip Rosedale at the recent SL Community Conference in Chicago) about the scale and potential of SL. As one of the few African-based avaters on the platform, it is with cautious optimism that we look at the RL/SL integration and a ‘thin’ version of SL - specifically related to mobile.
I am not a technical guru; simply an interested early-adopter in the metaverse - and as such, I also attended the Dr Dobb’s Life 2.0 conference session in SL yesterday and heard the Agile team of IBM refer to a new paradigm in development - where nitty-gritty planning is not the way to go, but reasonable speed and agility. This makes sense: To BE in the moment of development and to feed back as we go along quickly, and to adapt and make better for future innovation.
I agree - on the technical level - but also echo the concern voiced by the Prokofy Neva: “…I really think a LOT more attention has to go into the economic, business, social, and political side of all this — the impact this has on the body of customers known as “the world””. The real lack at the moment, is the ability to recognise that technical systems may be put in place without enough business, social and financial systems. As SL is a global platform and looks at expansion, it will become more and more important to engage with businesses operating in/from other continents. Not that this is NOT happening, but the danger is in not CO-DESIGNING with these local businesses. I just think of our own experience as a case study with the South African Reserve Bank on trying to unpack financial systems of LL to them to enable us to do business. Indeed… much work is needed…
I remain optimistic that solutions will evolve as we move forward. As long as we move forward together and do not create an exclusive solution for the few, but design a digital infrastructure with potential to ‘do right’ what the world-wide-web could not yet achieve: crossing digital divide, and becoming a means to also address bigger divides related to culture, poverty, resources, education and environment.
September 21, 2007 at 10:02 am
Alanagh, I have two questions for you since you referenced that you were from Africa:
1. Is a thin version of SL only for mobile really of value? What would it be used for? After all, if you want an avatar to look at, and something free, there’s Yahoo IM, which has an avatar and even interactive drawing and games — all for free. Why load up this expensive — even in its thin version — avatarized interactive dynamic world of SL on your mobile? for what? If you cannot live and work and interact in it richly. I’m just trying to fathom the purpose for it, not needing to hear the *reason* for it (lack of broadband).
2. At SLCC, Philip was raving about a practice he’d heard of in Africa, whereby people send money to each other over the cell phones, by converting money to cell phone minutes. He thought this was something that could be integrated with Second Life. And no doubt it could be. But, a) we need to hear if this is really a practice and is really viable, i.e. that people can use cell phone minutes to buy bread or shoes and really do that in practice (or whether this is fanciful and wishful talk) and b) whether somehow integrating this with SL makes some sort of sense
3. When you talk about the need to move away from “nitty-gritty planning” to “speed and agility,” it sounds pretty fake to me. It sounds like instead of really thoroughly planning and including all considerations and trying to justify your actions, you just go with the flow and substitute speed for quality. On Twitter, Ordinal Malaprop, who is a programmer, immediately questioned Philip’s claim that had LL been thorough, they couldn’t get the job done of making SL. I also question that as just revolutionary expediency and Stakhanovitism.
September 21, 2007 at 1:15 pm
Hi Prokofy!
I find this dialogue very valuable and I answer not as technology expert, but from an end-user perspective, where we work closely with several communities (most trapped in poverty) in rural and urban South Africa. We are exploring the possible (!!) value of SL and related tools at the moment for social gain in our projects in Africa, and in the past few months I have become acutely aware of limitations that we already face. However, to deny the potential at this stage would be premature for us…so we forge ahead cautiously…
Please do not see me as an representative, as I only speak from my own experience in this context… using both mobile and low-end computers, a high-end laptop (as many NGO leaders lean towards in Africa) and also using social networks extensively in communities (unrelated to any technology)… Maybe I am one voice of many not even participating in this discussion - which is EXACTLY the issue we need to address. Whether one likes a medium or not, the risk of potential exclusion of African residents - as a result of infrastructure, information poverty and lack of technology resources - is a real one…
To get to your interesting questions/comments:
(1) Yes, I agree a mobile version would be of little value if the emphasis falls on reducing the immersive experience to have a ‘thin’ immersion! Indeed, if there is a reason for a new/adapted product (bandwidth/affordability/access), it does not mean it has a solid purpose. SL Mobile will just not be the same as the existing experience and it will for sure! struggle to compete with very popular applications such as FaceBook, Yahoo, Gmail, Mxit and local African social network solutions. Not to mention that many of these are driven via, or supported by, powerful telecoms with a footprint as big as the continent itself. The immersive ‘avaterized’ environment is the very value that SL and peers bring to the world, and a ‘thin’ version on a mobile could ONLY be useful in as far as it connects with these applications and have SL/mobile integration for the PURPOSE of broad collaboration. mlearning and knowledge transfer. IF a 3-D virtual architectural platform is the future (on desktops) then surely we will be widening the digital gap if we do not find appropriate integration between this platform and mobile. However, the gap may very well be addressed suitably if there is proper integration of SL with existing mobile application - which would make a SL ‘thin’ version for mobile redundant. One of the SL offerings (not sure if it is in other worlds as well) that interest us in particular, is machinima: This could potentially be great value-addition to mlearning in Africa in my opinion. We would be the first company interested to look at possible micro-enterprise machinima simulations to reinforce our enterprise development programme in communities, and strengthen the informal economy networks. We just do not know how effective it will be - before we do it here in Africa and inform it with local intelligence directly from communities. I have been trying since March to get some pilot projects going on testing the value of SL in communities that we support. This will remain our challenge.
2. Will come back to you (here) on this one, but the short answer for now: Yes, mobile banking and related solutions with airtime transfer is reality and it works for thousands of Africans. It grows in popularity - and the foundation of it is the sheer force with which mobile use has escalated as an economic and social enabler (more than 1000% the past two years) and is still growing massively in Africa.
3. We have seen often in the development arena (in the not-for-profit sector) that lack of detailed planning is the one critical flaw hindering delivery. The comment about focus on agility was made by IBM in the context of agile software development processes at the conference. I am fascinated by the process and am not sure that it will compromise on quality - to the contrary, if I have a peek at what has been written about it: “All agile software development approaches have several values in common, such as: Frequent inspection and adaptation, Frequent delivery, Collaboration and close communication, Reflective improvement, Emergence of requirements (incremental), technology, and team capabilities, Empowerment and self-organization, Dealing with reality, not artifacts, Courage and respect”. On IBM’s website they go a bit further and talks about the AGILE MANIFESTO, that was defined in February 200 and states that the foundation of this process as “valuing individuals and interactions over processes and tools, working software over comprehensive documentation, customer collaboration over contract negotiation, and responding to change over following a plan”. All words can be fake if the behaviour is not supporting the approach in practive - therefore I think that the integrity of this approach to software development will really be linked to the business ethics of a company to not compromise customer trust and relationships in favor of technology advancement. (This is one of the worries I have, personally with SL expanding so quickly and branching into ‘thin’ versions without looking at proper CRM and financial systems, globally!! It may very well be that customer feedback is just not important and that the values of agile software development is not practiced, or practiced with a select few that are tolerant of, or ignorant of, fundamental business flaws as long as the technology works smoothly. I am still making up my mind if my concern is justified). I have no idea what “stakhonavitism” is that you refer to, but will google it
There are people that are against SL and its claims; there are those that propagate its value with blinkers. For me, I will only be able to make a final judgment if I am convinced there is RL benefit in Africa for our SL activities as an NGO based very far away from California. We are in the beginning of this journey..
September 21, 2007 at 5:12 pm
Prokofy Neva: “…substituting the ability for somebody to talk on their cell phone to a static frozen avatar who can’t move because it is all so “thin” doesn’t sound appealing” “Second Life should be for depth of worldness.”
Yes I agree that the greatest strength of the existing “world” of SL is the depth of its connectively, community, creativity and its dynamic economy. However I think this is very similar to alot of discussions about change in social environments ( not only virtual ) it isn’t necessarily or shouldnt be a “new or old” situation.. it should be a “new AND old” situation. I hope it will.
For instance, voice brought a new form of communication to the grid, roleplayers went crazy about how it “will destroy their identity”, but as it pans out we find that those communities still exist and thrive along with voice based communities. “Thinner” browsers ( and possibly servers ) could emerge, more directed at communication and lightweight presence, but could exist with in depth architectures such as the existing LL model. Conversely the grid may form highly rich areas experientially with intensive simulator resourcing possibly being more affordable, allowing a far greater degree of sensory and social immersion than we have experienced so far.
Prokofy Neva: “There is an illusion that the Lindens are “consulting with the community” when they hold these precious little hothouse discussions only on the very technical side of their software and servers, and only with a special few.”
I agree with your and nicks excellent podcast, that as the grid expands, reaching out across multiple companies, all with non existant or emergent trust records, there will be alot of turbulence and restructuring of the existing SL social and economic landscape, some of which will be fundamental. I also agree that discussions both behind closed doors and the more “open” office hours, are heavily weighted towards the technical challenges of what they are proposing.
This is very frustrating for me too, as I am a creative trying to bring something in world which the underlaying architecture of the grid does not currently support - and I spend many hours banging my head against a wall of obsessively “pure” scripting comment. These groups of people, and I include many LL employees, are massively talented individuals - but I often wonder WHY they are doing attempting to do alot of the things they become so feverishly excited about. I expect they would find it very difficult to answer that question in any way other than “because its new.. because we can..”
However “playing with knobs” isn’t always just geekiness for the sake of it - in my case it would be playing with the delivering a different type of sensory experience, which would in itself be part of the social and economic landscape.
There is ALOT of “wow thats a cool idea! its bigger than anything before- lets chase that now..” and a lack of overarching creative vision. However - this may be a symptom of a simple desire to become an purely enabling framework “the second life grid” and just let it “run free”.. and let the chips fall where they may..
I think we have to ask ourselves if the alternative - extensive, world wide collaborative discussion and inclusion from democratically elected representatives from all aspects of the existing SL society is a realistic goal for a company of LL’s type and size, considering the amount of flame ridden, power struggle infested politics associated with any form of resident representation within SL - or much of the web. The SL music dev list clearly illustrates how hard this is to do.
Sadly - without it, much of the original ethos of second life, will be lost. But was that glittering vision of “your world, your imagination..” just a hook to generate the first 10M signups ? Did we all really believe it anyway ? Whilst I know they don’t HAVE to uphold that commitment, I would like to believe it, or at least try to hold on to it, and I genuinely believe alot of people within LL feel the same.
September 24, 2007 at 10:07 pm
[...] on the plans of the newly formed SL Architecture Group during Zero Linden’s office hours, see Dizzy Banjo - Soundtracking Virtual Worlds. Aleister Kronos picked out the [...]
September 28, 2007 at 2:07 am
[...] thinking behind where Second Life is going architecturally, and lately there’s been talk of massively scaling up Second Life. How massively? How does 60M regions (from about 12K), 2B accounts (from about 9M) and 50M to 100M [...]
October 10, 2007 at 9:42 pm
[...] was hinted at last month in an interview with Zero Linden on Dizzy Banjo’s blog. Though I must admit I didn’t think there would be any movement [...]
May 8, 2008 at 9:41 am
The blog is too cryptic to understand though but i loved it for its way of explaining ideas in simple analogy.