4 ms·

I have been wondering for a long time why someone hasn't made a better Second Life. I think Second Life is awesome, and I would love to play a game like it. My

by Dig1t 3y ago

I have been wondering for a long time why someone hasn't made a better Second Life. I think Second Life is awesome, and I would love to play a game like it. My problem with SL is just that the game and its interface feel so janky. It has horrible graphics and the interface is weird. It's just not fun to actually do anything in the game. If someone made a better version, with good graphics and a decent user interface, I think it would be way more successful.

dale_glass 3y ago
People tried, including both Linden Lab (Second Life company), and Philip Rosedale on his own. The official follow-up to SL from Linden Lab was Sansar. Built from the ground up to be much prettier and faster than SL. But it kinda sucked all the charm from it. Sansar is a very static environment where you build the world in an editor, then people walk around in it. Meanwhile, SL has a full blown in-world content creation and script editor. Sansar flopped, got sold, and these days still exists: https://www.sansar.com/ https://www.sansar.com/ Philip Rosedale left LL, and made a new company which got very confusingly named "High Fidelity". This was a SL 2.0 of sorts, much more in the spiritual vein of Second Life. Also better engine than SL, better performance, and better scripting. Unlike SL it's a distributed architecture. High Fidelity was... weird. The architecture was like the most hardcore people from some deeply political place like the GNU project (it's Apache licensed, I mean more in terms of ethos) got together to make the least monetizable architecture they could. Content is hosted on any HTTP server, servers are run by anybody who wants, and there's no lock-in whatsoever. It horrified content creators. And so it didn't make any money. High Fidelity still survives as Open Source projects, I'm part of one: https://overte.org/ https://overte.org/
- RobotToaster 3y ago
  >The architecture was like the most hardcore people from some deeply political place like the GNU project (it's Apache licensed, I mean more in terms of ethos) got together to make the least monetizable architecture they could. Content is hosted on any HTTP server, servers are run by anybody who wants, and there's no lock-in whatsoever. It horrified content creators. Sounds awesome.
  - dale_glass 3y ago
    So I thought! But within an hour there I was completely baffled as to how it was going to ever make any money. The overall architecture seems a better fit for a deeply ideological small group of people trying to live on donations, than a commercial company. So eventually it died, we picked it up, and now have a non-profit.
    - imtringued 3y ago
      Come on this is literally the only scenario where NFTs make sense. Cross game support of NFTs make no sense because the developers see zero incentives to allow competing content but if it is the same game hosted by a variety of people then on hundreds of servers then interoperability has the benefit that your content will be available on other servers and therefore you also have incentives to accept other people's content. The only problem is that Blockchain is too expensive. Who wants to pay excessive fees for inconsequential in-game items? Who wants to host a full Ethereum node? Hardly anyone.
      - dale_glass 3y ago
        The code isn't a good fit for supporting any sort of ownership, including NFTs. They would be trivial to rip and anyone could use your expensive junk without paying for it. There's not a good system of ownership and asset control either. And as a project we're ideologically opposed to cryptocurrency and NFTs and just finished ripping out the remains of the Blockchain code from the original project.
spicybright 3y ago
It's honestly incredible how meta didn't pick this route. With all the money poured in they could have just copied SL with a fresh coat of paint. They could have poached all their devs or even just buy the whole company. At least we got some good laughs watching it fall apart. The eifell tower selfie still cracks me up.
- retrocryptid 3y ago
  Because the people who moved from Linden Lab to Facebook to make Meta did not view Second Life as a game.
zargon 3y ago
It sounds like you may not have used SL in a while. The current graphics are at least as good as a triple-A game, if you visit locations that are up-to-date and choose quality avatar components. Of course, rendering it may crush your computer (although I am satisfied with the performance of my 1080). All assets are user-created and most are not professionals who understand game asset optimization. To your other points though, SL is quirky and has an incredibly steep learning curve. It has a lot of technical debt, as it remains compatible with content created 20 years ago. Rosedale has a startup called High Fidelity, and they were creating basically what you suggest, a new SL without the technical debt. It was also distributed instead of centralized. Unfortunately that world has shut down because it wasn't commercially successful. (I expect largely because the people who would have been their early adopters were in SL, and didn't want to move until there was enough content available, and not enough users meant nobody was creating content.) I wish they had been successful, the platform architecture was really good.
- dale_glass 3y ago
  > Rosedale has a startup called High Fidelity, and they were creating basically what you suggest, a new SL without the technical debt. It was also distributed instead of centralized. Unfortunately that world has shut down because it wasn't commercially successful. (I expect largely because the people who would have been their early adopters were in SL, and didn't want to move until there was enough content available, and not enough users meant nobo dy was creating content.) HiFi horrified Second Life content creators. It downloads assets from HTTP. There's no grid. There's no central asset server. There are no mandatory user accounts. So anything you sell can be quickly copied, hosted wherever, and duplicated with complete impunity. SL will kick you off the grid if you misbehave, but HiFi was made in such a way there was no grid to ban people from. Besides that it was nigh impossible to give them money if you wanted to. You could buy domain names (as in pay for an internal analog of DNS) that was something like $15/year. And that was completely optional. It wouldn't have paid for their coffee budget. As far as I can tell there was no coherent plan to make money at all, other than "let's tax the economy" in a system that not only was utterly hated by content creators, but where anybody could set up a competing shop any time they wanted. > I wish they had been successful, the platform architecture was really good. The code still lives on! Feel free to join our Discord or meetings: https://overte.org/ https://overte.org/
seydor 3y ago
The SL server has an OKish open source version, opensimulator. Unfortunately nobody wants to work on the viewer because it is old , has tons of legacy stuff that has been patched up a gazillion times (e.g. classic/mesh avatars, "bakes-on-mesh" etc). The interface is full of stupid defaults (pie menus) and for today's standards it s extremely hard to use. The crowd that uses it is thus self-selected geeks or 3d artists
- retrocryptid 3y ago
  You also can't use the open source viewer for "sensitive" corporate things without purchasing a source license. Ah, the joys of GPL.
  - Animats 3y ago
    Huh? The Linden Lab client is LGPL-licensed.[1] Source is here.[2] The Firestorm viewer, which is a derivative of the Linden Lab viewer, is also LGPL-licensed.[3] [1] https://github.com/secondlife/viewer/blob/main/LICENSE https://github.com/secondlife/viewer/blob/main/LICENSE [2] https://github.com/secondlife/viewer https://github.com/secondlife/viewer [3] https://vcs.firestormviewer.org/ https://vcs.firestormviewer.org/
    - retrocryptid 3y ago
      Okay. fine. It's Lgpl. But the prohibitions are the same. If you, as a corporate entity, want to make and distribute changes to the viewer AND NOT BE REQUIRED TO PUBLISH THE CHANGES, then you have to purchase a commercial license. The deal with (L)GPL is if you make and distribute software based on (L)GPL licensed source code with changes you have made, you are required by the terms of the license to publicly release those changes you have made. And now lets say you work for a very large corporation, let's call it VOZ. And let's say you have some super-secret things you want to embed in the viewer, including names and ip addresses of custom grids and a few secret keys because your SSO is based on a prototype of DES-ticket Kerberos from the 1960s. Or maybe much of VOZ's business is based on software patents, and you want to use code based on these patents in the changes you make. If you accept the license and download the viewer code and make the changes, compile it and distribute it to other VOZ, Inc. employees and partners you have changed the original code with respect to the (L)GPL license. You are obligated to publicly release those changes. And releasing the derivative work under the (L)GPL may invalidate some software patent protections you think you have. So you have a choice between a) not distributing the modified viewer and b) distributing it in violation of the terms of the license. (see terms 4 and 10 of the LGPL license referenced above.) But there's a third option. You can purchase a commercial license to the code. Even though the code is also licensed under the GPL, the owner of the copyright is free to license the code using whichever license they see fit. The copyright holder can make the code generally available under a (L)GPL license, but also make it available it to other parties under a less restrictive commercial license where using the code does not obligate you to publish your changes or imply you are abandoning interest in your software patents. And this is why I largely use a BSD license. Generally speaking, I'm not looking to force people who use my code to make a statement about software patents, either pro or con. If the code is useful to them, great! I don't want to force them to release their changes if they don't want to. If they want to maintain a separate fork, even a private fork, more power to them.
retrocryptid 3y ago
But Second Life isn't a game.

by Dig1t 3y ago

dale_glass 3y ago

People tried, including both Linden Lab (Second Life company), and Philip Rosedale on his own. The official follow-up to SL from Linden Lab was Sansar. Built from the ground up to be much prettier and faster than SL. But it kinda sucked all the charm from it. Sansar is a very static environment where you build the world in an editor, then people walk around in it. Meanwhile, SL has a full blown in-world content creation and script editor. Sansar flopped, got sold, and these days still exists: https://www.sansar.com/ https://www.sansar.com/ Philip Rosedale left LL, and made a new company which got very confusingly named "High Fidelity". This was a SL 2.0 of sorts, much more in the spiritual vein of Second Life. Also better engine than SL, better performance, and better scripting. Unlike SL it's a distributed architecture. High Fidelity was... weird. The architecture was like the most hardcore people from some deeply political place like the GNU project (it's Apache licensed, I mean more in terms of ethos) got together to make the least monetizable architecture they could. Content is hosted on any HTTP server, servers are run by anybody who wants, and there's no lock-in whatsoever. It horrified content creators. And so it didn't make any money. High Fidelity still survives as Open Source projects, I'm part of one: https://overte.org/ https://overte.org/

RobotToaster 3y ago
>The architecture was like the most hardcore people from some deeply political place like the GNU project (it's Apache licensed, I mean more in terms of ethos) got together to make the least monetizable architecture they could. Content is hosted on any HTTP server, servers are run by anybody who wants, and there's no lock-in whatsoever. It horrified content creators. Sounds awesome.
- dale_glass 3y ago
  So I thought! But within an hour there I was completely baffled as to how it was going to ever make any money. The overall architecture seems a better fit for a deeply ideological small group of people trying to live on donations, than a commercial company. So eventually it died, we picked it up, and now have a non-profit.
  - imtringued 3y ago
    Come on this is literally the only scenario where NFTs make sense. Cross game support of NFTs make no sense because the developers see zero incentives to allow competing content but if it is the same game hosted by a variety of people then on hundreds of servers then interoperability has the benefit that your content will be available on other servers and therefore you also have incentives to accept other people's content. The only problem is that Blockchain is too expensive. Who wants to pay excessive fees for inconsequential in-game items? Who wants to host a full Ethereum node? Hardly anyone.
    - dale_glass 3y ago
      The code isn't a good fit for supporting any sort of ownership, including NFTs. They would be trivial to rip and anyone could use your expensive junk without paying for it. There's not a good system of ownership and asset control either. And as a project we're ideologically opposed to cryptocurrency and NFTs and just finished ripping out the remains of the Blockchain code from the original project.

spicybright 3y ago

It's honestly incredible how meta didn't pick this route. With all the money poured in they could have just copied SL with a fresh coat of paint. They could have poached all their devs or even just buy the whole company. At least we got some good laughs watching it fall apart. The eifell tower selfie still cracks me up.

retrocryptid 3y ago
Because the people who moved from Linden Lab to Facebook to make Meta did not view Second Life as a game.

zargon 3y ago

It sounds like you may not have used SL in a while. The current graphics are at least as good as a triple-A game, if you visit locations that are up-to-date and choose quality avatar components. Of course, rendering it may crush your computer (although I am satisfied with the performance of my 1080). All assets are user-created and most are not professionals who understand game asset optimization. To your other points though, SL is quirky and has an incredibly steep learning curve. It has a lot of technical debt, as it remains compatible with content created 20 years ago. Rosedale has a startup called High Fidelity, and they were creating basically what you suggest, a new SL without the technical debt. It was also distributed instead of centralized. Unfortunately that world has shut down because it wasn't commercially successful. (I expect largely because the people who would have been their early adopters were in SL, and didn't want to move until there was enough content available, and not enough users meant nobody was creating content.) I wish they had been successful, the platform architecture was really good.

dale_glass 3y ago
> Rosedale has a startup called High Fidelity, and they were creating basically what you suggest, a new SL without the technical debt. It was also distributed instead of centralized. Unfortunately that world has shut down because it wasn't commercially successful. (I expect largely because the people who would have been their early adopters were in SL, and didn't want to move until there was enough content available, and not enough users meant nobo dy was creating content.) HiFi horrified Second Life content creators. It downloads assets from HTTP. There's no grid. There's no central asset server. There are no mandatory user accounts. So anything you sell can be quickly copied, hosted wherever, and duplicated with complete impunity. SL will kick you off the grid if you misbehave, but HiFi was made in such a way there was no grid to ban people from. Besides that it was nigh impossible to give them money if you wanted to. You could buy domain names (as in pay for an internal analog of DNS) that was something like $15/year. And that was completely optional. It wouldn't have paid for their coffee budget. As far as I can tell there was no coherent plan to make money at all, other than "let's tax the economy" in a system that not only was utterly hated by content creators, but where anybody could set up a competing shop any time they wanted. > I wish they had been successful, the platform architecture was really good. The code still lives on! Feel free to join our Discord or meetings: https://overte.org/ https://overte.org/

seydor 3y ago

The SL server has an OKish open source version, opensimulator. Unfortunately nobody wants to work on the viewer because it is old , has tons of legacy stuff that has been patched up a gazillion times (e.g. classic/mesh avatars, "bakes-on-mesh" etc). The interface is full of stupid defaults (pie menus) and for today's standards it s extremely hard to use. The crowd that uses it is thus self-selected geeks or 3d artists

retrocryptid 3y ago
You also can't use the open source viewer for "sensitive" corporate things without purchasing a source license. Ah, the joys of GPL.
- Animats 3y ago
  Huh? The Linden Lab client is LGPL-licensed.[1] Source is here.[2] The Firestorm viewer, which is a derivative of the Linden Lab viewer, is also LGPL-licensed.[3] [1] https://github.com/secondlife/viewer/blob/main/LICENSE https://github.com/secondlife/viewer/blob/main/LICENSE [2] https://github.com/secondlife/viewer https://github.com/secondlife/viewer [3] https://vcs.firestormviewer.org/ https://vcs.firestormviewer.org/
  - retrocryptid 3y ago
    Okay. fine. It's Lgpl. But the prohibitions are the same. If you, as a corporate entity, want to make and distribute changes to the viewer AND NOT BE REQUIRED TO PUBLISH THE CHANGES, then you have to purchase a commercial license. The deal with (L)GPL is if you make and distribute software based on (L)GPL licensed source code with changes you have made, you are required by the terms of the license to publicly release those changes you have made. And now lets say you work for a very large corporation, let's call it VOZ. And let's say you have some super-secret things you want to embed in the viewer, including names and ip addresses of custom grids and a few secret keys because your SSO is based on a prototype of DES-ticket Kerberos from the 1960s. Or maybe much of VOZ's business is based on software patents, and you want to use code based on these patents in the changes you make. If you accept the license and download the viewer code and make the changes, compile it and distribute it to other VOZ, Inc. employees and partners you have changed the original code with respect to the (L)GPL license. You are obligated to publicly release those changes. And releasing the derivative work under the (L)GPL may invalidate some software patent protections you think you have. So you have a choice between a) not distributing the modified viewer and b) distributing it in violation of the terms of the license. (see terms 4 and 10 of the LGPL license referenced above.) But there's a third option. You can purchase a commercial license to the code. Even though the code is also licensed under the GPL, the owner of the copyright is free to license the code using whichever license they see fit. The copyright holder can make the code generally available under a (L)GPL license, but also make it available it to other parties under a less restrictive commercial license where using the code does not obligate you to publish your changes or imply you are abandoning interest in your software patents. And this is why I largely use a BSD license. Generally speaking, I'm not looking to force people who use my code to make a statement about software patents, either pro or con. If the code is useful to them, great! I don't want to force them to release their changes if they don't want to. If they want to maintain a separate fork, even a private fork, more power to them.

retrocryptid 3y ago

But Second Life isn't a game.