Sansar Wasn't the Successor to Second Life That Second Life Needed (Comments of the Week)

Lots of interesting reader comments about my podcast interview with Philip Rosedale, including this reaction to my suggestion that a successor to Second Life running on a modern engine like Unity should be integrated as an option in the official viewer. For illustrative purposes, imagine two big buttons side by side on the log-in page, labeled "Click here to teleport into Second Life Classic" and "Click here to teleport into Second Life 2", or whatever it's called. Reader "0xc0ffea" thinks that's basically already been done: 

The new grid new engine idea was Sansar. No one used it. It crashed and burnt.

This is true, but misses some important nuances. As "Pointer" writes:

The modern engine idea was so good, that one of the few or the only thing appreciated about Sansar was exactly the visuals from the new engine. Nobody used Sansar, but not because of the idea of the engine. If your game has pretty graphics, but little else, you paved the way to a predictable failure. Classic. Also the infamous loading times. In 2017, when most people were on A/DSL, Linden Lab thought: "people are making rooms that are several gigabytes to download? This is fine". It burned well.  

This is true -- Sansar user-made experiences were clocking in at 5-8 gigs or as much as 24 gigs each. (By comparison, the entire sandbox game world No Man's Sky, which contains a literal galaxy, is only 10 gigs.)

Which points to the larger problem: Instead of developing a successor to Second Life that was accessible to more people, Linden Lab mysteriously architected Sansar to be accessible to even less of them, optimizing it to run on premium VR headsets and requiring a very high-end gamer PC, neither of which few can afford or even want. (A strategic error so painfully self-destructive in retrospect, it deserves its own post someday.) 

No: What I'm proposing is a successor to Second Life that can run on mid-range PCs and even more key, on mobile devices and videogame consoles. (Which is why I suggested building it on Unity, which plays well on all those platforms.) Graphics and cloud technology have improved enough that such a thing could be done now.

Image from, "Second Life To Become Small Town To Sansar's Big City?" (LOL)

Comments

[Adeon]

If you don’t support inventory transfer no one will go there.

[Chic Aeon]

Regarding Sansar's loading times:

Well through the advent of open beta participation (when I went over as the TOS was finally changed and I felt I could 'sign' honestly) the devs were asking folks to try and break the engine. Bigger, "clearer", more etc. This -- along with folks uploading render mesh (legal in Sansar) as well as folks who really had no clue how to make "GAME ASSET" mesh -- It was a problem. There was one experience that I waited almost a half an hour to load (I went out and watered the yard and came back ^^).

At that time Medhue and I were the only ones whose experiences would load in a few seconds (both around 14 for me on a medium connection with a medium desktop). Eventually the devs made some good progress and most experiences would load in about 7 to 12 seconds. That was impressive since the creators had not optimized their works.

Unity based:

I didn't comment on the original post about the Unity idea because I wasn't going to watch the video (thanks for the notes!) but I have one pretty important point if no one has made it yet.

Unity is popular. Unity is also a b**ch to work with. It also kidnaps your computer and can be very messy. Items from second life are not easily transformed into Unity assets. I will never build for a Unity platform and I know many creators who agree with my position.

Hence the issue. Because while Sansar had great tech and things looked so much nicer there than in SL -- folks couldn't just take their SL assets with them. It did seem (Ebbe on the forums right after the Sansar announcement) that was thought of as a possibility at the very beginning, but if so it evidently became clear that it wouldn't work.

I made five or six experiences in Sansar, all with ALL my own items. I uploaded and remade a ton of Second Life goods. Each experience took me about a month and I was working more than 40 hours a week. So SIMPLE conversion from one platform to another is not possible. For that reason I cannot see your Unity idea working in any sense :D.

Sorry.

[Matthew]

Personally, and it could just be me, but I find the thought of an SL2 side-by-side with SL1 awful. Replace the current SL completely with a new, better engine -- yes. But to have both versions running side-by-side -- no. I would feel torn in half, not able to commit to one or the other, and not able to invest fully (emotionally) in either. It would be like trying to satisfy two lovers. I like having *one* home which I can totally and completely invest in, not two or three homes which then split your interest and affection. Like I said, maybe that's just me.

[Mondy]

"The new grid new engine idea was Sansar. No one used it. It crashed and burnt."
Well, people did use it. I used it. I'm a person. But I didn't *invest* in it, emotionally that is. I loved how it looked. Some of the experiences were, ARE, amazing. The garage from Ready Player One, the Trek experience. Fantastic. But even now, things like sitting are *still* awful!
There were experiences from Blueberry, Draxtor. It's not like people didn't try. Drax's book club, weekly meet ups to visit different places and explore. Long loading times for me used to mess that up. So while people did make accounts & some of my friends *did* try it, none of mine stayed for very long so I didn't either. Sometimes I still log in, wander around. But Second Life is still where my avatars heart lies.

[Snow]

Solid points about accessibility, anyone with any critical thinking skills could have seen how much they were limiting their market by focusing so heavily on VR and high end content. VR and visually appealing content is great, but its value should be weighted proportionally to the number of people that it affects.

I second the comment about "Second Life running on a modern engine like Unity should be integrated as an option in the official viewer". But it should be the same game just in a different engine.

Sansar failed to be what LL wanted because it was not SL in Unity, it was an entirely different product. SL isn't the game engine, its people's accounts, their inventories, their friends list, economy, and the world that has been created over the the past 18 years.

You can't just get rid of all that and create another world with a client that doesn't even include the same features as what its trying to replace. I wanted to like Sansar, but there was nothing and no one there that I cared about.

I say this as someone who has spent extensive time digging through the SL client code and even got fairly far into creating a functional client in Unity. I stopped primarily due to the time required to deal with performance issues and trying to deal with the intricacies involved with rendering rigged meshes properly.

LL could do it though. (If they want a starting point there's a few people that have already open sourced progress in this area. They just need to look.) Or maybe one of the other long time 3rd party viewer developers that are much more skilled than I.

The motivation to use Unity at all is to get access to a wider array of tools, the Unity ecosystem, ease of development & content creation (for professional non-SL developers/artists), and support more modern graphics and performance improvements.

On a slightly unrelated note. IMO none of these are SL's biggest issues though. The primary issue I've found is that the land system sucks. My biggest hope is that one day they detach land within SL from LL. You can't optimize server costs, self-host content, write custom server code, there's "unnecessary" limits on location size and the amount of "stuff" (prims, meshes, people..) in the world, there's janky & slow interfaces via LSL.

You have beautiful communities and creations that form in the world that just can't financially sustain themselves. Where if self-hosting was an option you could see these groups make intelligent tradeoffs.

Even without self-hosting, SL's region fees are trash. See $349 with a $229 a month maintenance fee (https://secondlife.com/land/private-pricing) vs kitely's $15-$20 a month (https://www.kitely.com/services). Regular people can't afford to make anything substantial in SL. Its a good thing there's no way to move anything that's valuable from SL to other opensim grids (friendslists, inventory, world, etc..) because SL would be hurting for real if there was.

[lkosov]

My problem with Sansar was the complete lack of a world. There was no mainland, no public spaces (other than the reception area), no way to wander and explore. I absolutely get that they had their ideas for doing it that way, but it could never hope to compete with an actual world to explore, with geography and context. There were some good experiences, people did some fun stuff there, but there was no sense of world.

[madeline blackbart]

IMO the big problems with sansar are that it released with minimal interactivity while emphasizing VR.Vr games aren't enjoyable if there's almost no interactivity. It was like visiting a museum. You can look but you can't touch. I also think the point they opened for beta was just to early. There just wasn't much to do. You might wait a long time for loading if you get there and there's something there exciting....not if the place is empty and it's a museum more or less.
Also those avatars were hideous I'm sorry. And for a social platform focusing on self expression like second life (and in theory sansar) a ugly avatar like that is a really hard sell. It's a real shame honestly because SL needs an upgrade to a new graphics engine DESPERATELY.

[lkosov]

Madeline, I don't think the avatars were hideous, but they fell into this strange, well... uncanny valley. There were limited adjustments, the faces were weirdly expressionless, and the default AO was pretty stiff and unwieldy, even if it did come with a few nice emotes by default. I remember the squatting pose with particular fondness, possibly because there was little to no furniture to sit on or otherwise interact with. "I have seen the future of the metaverse, and in it, we are all... Slavs."

I always felt like... when I read Snow Crash, my lasting takeaways were about "condensing fact from the vapor of nuance", the nam-shub of Eniki, and dentatas. When the Sansar team read Snow Crash, I kind of felt like they all stopped either right after or right at the part about Brandy avatars coming with default settings and limited options...

[Imagin Illyar]

Sansar had some good things but it had a lot of limitations that SL doesn't have and that was the problem. It looked great but was just too limiting. Sure, lots of other games are even more limiting but people compared Sansar to SL not to games.

[0xc0ffea]

Every decision that went into Sansar was a direct result of lessons leant from Second Life. It's almost like they created a list of everything that made SL look and feel janky and set out to do the opposite.

Technically it was fine .. released too early and then developed in a meandering noncommittal way as is the Lab's forte.

The problem's ran far deeper than the obvious mistakes or missing features.

When it was released and SL users piled on, the immediate reaction was visceral. This is hot garbage, there is no way we (SL users) are going to use that over SL. The Lab seemed to sagely nod and imply that was intentional. So yet again .. another white elephant product chasing new customers (who very much shouldn't be those weirdos from SL).

The strict PG rating was a huge problem. Not because there should have been explicitly adult content, but in order for a social platform to thrive it needs to permit social content on broad spectrum. Setting a hard cap on personal expression right out the gate limits that, and even if 99% of what two users might want to do together is perfectly acceptable, they wont even try which makes answering more fundamental questions harder.

Virtual worlds are not games, they don't come with purpose wrapped in a gift box and presented neatly on gilded platter. The user has to answer the question "Why am I here?". The mistake comes when platform operators think a simple one dimensional answer to that fundamental question is sufficient.

A music event isn't a goal in it's own right, it's the backdrop to interpersonal interactions .. which can't happen when the platform has stuck a sign over the door that says FAMILY FRIENDLY.

[Henri Beauchamp]

My view:

https://community.secondlife.com/forums/topic/482192-the-real-solution-to-making-sl-number-one-location-on-the-net/#comment-2408901

[Courtney Ragland]

5-8 up to 24 gigs? Not sure what world you are referring to. Most worlds are mostly 1 gig.Sure, there may be some worlds that are huge, but not the majority of them.