Aug 1, 2017

SANSAR Dreams

Don’t call it Second Life 2


project-sansar-screenshot-liden-lab-second-life-2


As I sit here with my coffee, I find myself in a precarious position. Sure, the sun is shining and the wind delicately meanders through the leaves outside for a serene atmosphere, but something still doesn’t feel right.


In my mind, I find the duality of SANSAR and whether to call it something good floating haphazardly among countless other random thoughts.


On the one hand, I could take the easy route and just be an unabashed cheerleader for the new flagship virtual reality “experience” like so many others, but then I wouldn’t be giving it an honest review.


I remember back in the day, I had made mention on numerous occasions about the “SL Cheerleaders” who effectively have their lips planted firmly to the PR department of Linden Lab in exchange for preferential treatment. Yes, I’ve included folks like Strawberry and Draxtor in that group – but more often than not there was a missing context.


I don’t necessarily believe these are bad people. After all, they’re just performing their part and it is a crucial aspect of their purported fame in the virtual world. They talk it up, no matter what, and they get the exposure and access. It’s a sweet deal and it makes total business sense, however much you are willing to trade integrity for it.


But even today it is a bit of mental gymnastics to watch even Strawberry make videos trying to say nothing but good about SANSAR and having an immense amount of trouble finding those talking points.


What I see from SANSAR is a lot of smoke and mirrors – selective journalism from cheerleaders trying to narrow that angle of view to omit the bigger picture and show only the good, or in this case a totally non-critical viewpoint.


Even when we see pictures of maybe 50 avatars stacked on each other at Drax’s build in SANSAR, we see headlines like “People are piling up in SANSAR to get a look! It’s so popular!”


But we both know that’s not the case. For one, if they’re all piled up on each other at the entry point, that means they’re all logging in and not interacting or even moving when they get there. Secondly, it’s mostly a flash in the pan experience so to speak… if you logged in right now, that location is likely empty, barring all the folks reading this and logging in to check the validity of that claim.


And therein is a bigger point to be made about all this hype, which is exactly all it is… just a load of hype.


So I sit here with my coffee, contemplating how to go about this post. Believe it or not, I don’t enjoy shitting on every project. I know a lot of people spent years working on this project and it would be unduly harsh of me to tear it up. But it has to be done for the sake of honesty and integrity in virtual worlds journalism.




Nothing Special


Let’s get some things out of the way up front.


SANSAR is nothing special, regardless of what the hype machine is trying to sell you. As a matter of fact, it looks suspiciously like a clone of Avatar Reality BlueMars, though not using the Crytek engine.


Which brings me to a major sticking point – it’s built predominantly as a controlled and publish only environment. Don’t mistake SANSAR as “the next Second Life” because it’s anything but.


When you have a publish environment that is highly controlled for assets, then yes… obviously it’ll likely look a lot better. That’s the entire point of a publish environment, in that you can set things in “stone” and then apply additional shaders and effects to the scene because the scene nodes are “compiled”.


Insomuch as this aspect, the question begs to be asked:


If you’re looking for a published scene system that looks prettier but takes away the live editing and organic aspects of Second Life, then why not just use Sinewave Space? They’re essentially SANSAR but far better in that regard.


Authors Note: Sinewave Space actually does have a live editing ability alongside the publish model.


If we are to look at the “VR Headset Focus” of the SANSAR Model, I believe it’s a mistake to build an entire system like SL or even SANSAR on the premise of betting on the ubiquity of VR headsets. Sure, it’s a nice inclusion for a niche audience (because that’s what it is and will be for the foreseeable future), but it makes zero sense to build an application for a niche audience and then split that niche audience into an even smaller niche to cater to.


It’s a lot like getting into the automobile business and catering to just Bugatti Veyron owners with a bit of a nod to the peasants driving a Honda Civic. If your goal is to have market saturation and be the ubiquitous go-to for virtual worlds, this is not how you do it. Instead of trying to build an entirely different platform, one should learn what worked the first time and replicate those variables with better execution.


What Project SANSAR represents is a team of coders and executives who believe they actually understand the future of virtual worlds and attempting to make a flagship project based on faulty comprehension. Focusing on the wrong aspects of what makes shared virtual spaces like Second Life amazing, and distilling those out-of-context talking points into a product of its own misses the point entirely.


The sum of the whole is greater than the parts.


That isn’t to say they’re bad people. No… just very misguided on their understanding of what makes an amazing virtual world system. Second Life as a popular platform came as a sort of spontaneous quirk that Linden Lab wasn’t entirely prepared to either comprehend nor handle when it did happen.


Remember: It was billed as a game, and had game aspects before the inkling of open-ended, user generated, virtual world snuck in and things were rebranded to focus on those aspects in the rise of popularity for it.


It was a lot like building a video game and realizing that the modders are doing unintended things with it, however amazing… then you start catering to what the customers want and use it for to capitalize on it.


Hell, Second Life as a concept isn’t revolutionary (regardless of how Philip Rosedale would frame it). I mean, while he himself has said he had the idea in the mid-1990s but didn’t think it was feasible which is why he continued working at Real, that in itself is a fabrication… Because you know what else existed in 1995 when he insists such a virtual world system wasn’t feasible?


ActiveWorlds.


And while Activeworlds looks outdated by today’s standards, it would be a disservice to it if I didn’t also state that if you know what you’re doing with the platform, you can create worlds in there that are on par and sometimes even better than Second Life. After all, they have Zones and the ability to implement screen space particles for weather. Not to mention they actually have a built in particle editor that is dead simple and powerful to use, unlike coughs Second Life who (for the love of God) has access to Windlight and Nimble (the weather system that came with Windlight).


Oh, that and mirrors. Activeworlds has working mirrors and figured out how to do it without lagging the place up in the process.


But I digress. Yeah, that’s how old this open-ended virtual world concept actually is. Second Life is special not because it was the first or even the best. It is special because it improved on an already existing formula, even begrudgingly.


After all, they didn’t have the foresight to implement a marketplace online until years after XStreet did it – which goes to show that you, the virtual world inhabitants, are far more responsible for pushing the ideal virtual world system than the companies that produce the platforms. Your third party viewers and dedicated teams of volunteers are proof of this just the same. Your outright ingenuity in how you use the virtual world is just as responsible for its advancement – when mesh was implemented in Second Life, you used it as clothing and mesh bodies, and so on… Linden Lab didn’t think of that one (apparently) and so you didn’t have stuff fitting correctly.


Though even mesh was a non-Linden Lab innovation for an SL platform. It showed up years in advance via the RealXtend viewer and implementation, all while Linden Lab was saying it wasn’t possible. Once again, you are the innovators here – or at the very least, the driving force of it.


It is you who are continually driving that advancement – with or without the company running the platforms. OpenSim is a good example here, though not as coherent as a proprietary system like SecondLife.


What SANSAR is, in a nutshell, is a company trying to wrestle that power back from you and severely constrain your ability to innovate – to make themselves the sole channel of that innovation, and the sole purveyor of it. At the baseline, this is removing a large portion of the “magic” formula that is Second Life in the organic nature and trying to bottle it into bite-sized snippets under tight control.


It defeats the entire point.




Out of Context


The most interesting part about SANSAR is that if you took it out of context, it would be fine. But coming from Linden Lab, that isn’t quite possible. On its own, it’s a pretty nice piece of software – albeit nothing revolutionary in any manner.


When an “experience” is filled up – as in, you reach a set concurrency limit, the scene just creates a copy and starts teleporting the newcomers to the new copy instead. The term for this is “sharding”, and this alone makes me shake my head. Given the opportunity to start from the ground up and redesign the virtual world experience platform using modern approaches, Linden Lab immediately screwed up and built a sharded system. This is going to haunt SANSAR for the entire shelf life of the project.


What I see with SANSAR is a missed opportunity from Linden Lab to really push the boundaries and reinvent their flagship namesake, Second Life with modern technology and approaches, applying what they’ve learned since its inception. That being said, it seems like Linden Lab hasn’t learned anything over the past fifteen years.


Here we find that Linden Lab had the opportunity to design a single shard system and do it right this time, but failed to do so. After all, Novaquark is designing Dual Universe as a single shard universe and is what I would say an iteration of Second Life should have been from the get-go. It’s pretty bad when a crowdfunded system is kicking your corporate seasoned ass.



This isn’t news, if you remember the tenure from Rodvick Humble (Rodimus Prime).


Instead of moving forward, we see a company taking steps backward in their application of understanding the virtues of virtual worlds, while others are sailing ahead with innovation.


When I was giving my presentation at OSCC, I brought up Dual Universe as the winning baseline for the future of virtual worlds specifically because of its innovation in single shard technology. Allowing thousands of people to simultaneously be present in a single location persistently is a major factor.



Though that opportunity is now missed when it comes to Linden Lab and SANSAR due to its baseline model. In Dual Universe, you have a shared, persistent virtual universe that you can build in like in Second Life, but more importantly you can build a spaceship and literally fly, seamlessly, off planet and go land on another one.


That is absolutely incredible – a word I’d love to use about Linden Lab and SANSAR but is not justified.






What about Second Life?


Here’s an interesting tidbit about Second Life versus SANSAR that most probably wouldn’t know:


You can create a sim in Second Life that is comparable to SANSAR if you know what you’re doing. Just like you can use 20 year old ActiveWorlds to create worlds that would be comparable to Second Life.


That’s it… the big secret. See, if you baked all of your lighting and exercised control over the sim settings, you can come up with something similar to SANSAR. Will there be lag? Well, that depends on a lot of factors.


If you know how to create content that is game optimized, use standard tricks of the trade to do mesh and bake in lighting effects instead of forcing it on the high-end machines only, then Second Life can look absolutely stunning and on par with AAA titles. Albeit you don’t get things like HDR and God Rays in your shaders with Second Life, and sure you prolly don’t have Bloom effects like in SANSAR, but let’s be honest -


We already have shaders in Second Life, and adding HDR, Bloom and God Rays (etc) to the Advanced Lighting Model would put it on par with SANSAR.


But what about the lag, I hear you say?


I’ve already outlined here how to remedy the situation with the single shard technology approach that Dual Universe is using. However, that being the case, I’ve previously outlined how Second Life can retool their server approach on the back end to virtually eliminate that lag on multiple fronts.


1. Have the idle servers kick on to help co-process simulators that are popular. Much of the problem is that a single region is on a quad core server, which is fine until you push it to a breaking point. If they get to that level, there’s no reason why you should have thousands of idle servers and countless processing power just sitting there twiddling thumbs about it.


2. Remember SLGo? Linden Lab totally dropped the ball on that one. Yes, it was a third party system that Linden Lab wasn’t in charge of – but the take-away should have been crystal clear if they were paying attention. The cloud streaming system which allowed even a crappy tablet or phone to run Second Life on Ultra Graphics should have been built into Second Life itself as a Premium option. This is added value proposition in a nutshell. It was a unique added value proposition that wasn’t replicating pre-existing services from their community – and that is what counts.


3. The Fu** Off  button. Yes, that’s what I originally called it when I jokingly asked for it on Twitter. You have one now in your viewer and it simply derenders everyone in the room except friends. Why is this important? Well, because a majority of you lot have no clue about optimizing your avatars for looks and speed to be considerate of the sim and surrounding avatar resources. All of your high-end mesh attachments and high-resolution mega textures for skins and items may work when it’s just you and one other person, but it absolutely tanks a server and other computers when there’s a group in the same area.


4. Content creators need to up their game (collectively). The very thing that makes Second Life awesome is the organic nature of user created content and world creation, but it’s a double edged sword – because those same content creators are making things without the explicit understanding of optimized game assets. So we see these high poly, massive texture using, monstrosities being sold for all manner of use which in turn brings anything but a NASA launch console to a screeching halt. Then on the other side of the coin you have badly designed assets that look like absolute shit. Because of this dichotomy of melting pot mentality, we have a mixed bag in a virtual world where one minute you could see a breathtaking sim and the next a bunch of assets from the turn of the century.


As I’ve already said, I don’t see any reason why Second Life couldn’t look as good as SANSAR here and now, and even be optimized enough to drastically lower lag. To do that, it’ll take a concerted effort on our part as content creators and world designers, as well as on the part of Linden Lab to apply better management approaches to how SL runs.




Final Thoughts


SANSAR for all the hype isn’t special, however much it may look pretty at this time. The publish model alone harkens back to the early 1990s with Worlds Inc, and more recently BlueMars. In my opinion, it is the theme park to our virtual world lives.


Which is to say – it’s a nice place to visit, but we live elsewhere. When you want a short term experience, SANSAR is a good vacation (though Sinewave Space is still better at it), but when you really want to spend time in a virtual world that is organic and user-generated – you should stick to Second Life.


If Linden Lab understood the reasons why we love Second Life, they would have never made SANSAR the way they did, but instead made a true successor to Second Life.



3 comments:

  1. Just wanted to add based on your comments about god rays in SL. There is a 3rd party viewer that does have god rays already. Blackdragon viewer. You can see them in action in a video I made https://www.youtube.com/watch?v=WN80Mv7m0Lk goto 1:25 and then on there are a few shots with god rays.

    ReplyDelete
    Replies
    1. Oh, I wouldn't doubt that a 3rd party viewer has God Rays. Just sorta makes the point stronger that it's not out of reach for main viewer and also that it's the population of users pushing the advancement :)

      Delete
  2. Eventually, Sansar will be able to have everything that Second Life has and more, feature wise, and it will be able to handle texture rendering better than SL due to a more optimized base of modeling. I also suspect that eventually there will be at least one 'experience' that will mimic Second Life, along with a bunch of other 'experiences' that may be more idealized/stylized takes on the Second Life experience. When it gets to this point, Second Life will become redundant and obsolete. On the other hand, Linden Labs are taking significant steps in their development of the Second Life platform to make it much more robust than it currently is...so, there is still hope for it.

    ReplyDelete