Puppeteering coming back at some point? This game seriously needs solid, synced physical animations as a final step for immersion

[Nikkesa]

As detailed here: http://wiki.secondlife.com/wiki/Puppeteering and here: http://avatarpuppeteering.com/

I've been rather desperate recently to see something decent like this - I've been desperate enough to start exploring puppet-like avatars made of linked sets and actually have a reasonable 20-30-ish prims being animated at around 14 fps, but those are pretty much the limits. It's jarring and nothing more than a novelty, and it seems to actually lag the sim when several of them are running.

I'd love to see my avatar walking around the world and actually interacting with it in a way that isn't just jarringly unrealistic, sliding, with basic input controls that don't feel smooth etc etc... I'm sure it's possible still to do things like vehicles that read off animation root-motion vectors baked into a notecard or something, but that's a ton of work and you'll only ever get to use it realistically in a sandbox. There are also alternatives for controlling your avatar via a script (like those flight huds that make you go extra fast), which could theoretically be synced with animation as well, but it's never guaranteed to really be working properly or look smooth.

It looks like a lot of the work for puppeteering is already done? As far as I know, this would be absolutely massive for second life - if you add that, and then add something that makes the user say "Oh jeez, it's actually pretty fun to just run around in this game", I feel like a lot more users will pick up second life and be interested in it as a _game_ rather than it's current function of being a chat simulator with some out of date novelty stuff like vehicles and neat tech design and stuff. Not to mention how much it would add to the adult portion of the game, which iirc means $$$$ for linden labs.

[Alwin Alcott]

58 minutes ago, Nikkesa said:

 I feel like a lot more users will pick up second life and ............ which iirc means $$$$ for linden labs.

this and "sl is going to die" are the most untrue things ever said at these forums.
Niches never brings staying interest for a virtual world. How many times we heared past 15 years "if you do this now people will come".... and where are they??
Stability will keep people.
Most  don't need more "adult".

Edited December 12, 2020 by Alwin Alcott

[animats]

Puppeteering got far enough that there was a development release? I did not know that.

Talk to Lucia Nightfire. This is a big issue for her, too.

I'd like to see webcam- based face feature tracking supported in SL. The better avatars have the ability to display facial expressions, but you have to run them from a big panel of 70 or so buttons, which few are skilled enough to do. Machima makers, entertainers, and Youtubers would go for that. Few would send, but many would watch.

[Rowan Amore]

They should have worked on this instead of the pointless EEP.

[Nikkesa]

On 12/12/2020 at 1:29 PM, Alwin Alcott said:

this and "sl is going to die" are the most untrue things ever said at these forums.
Niches never brings staying interest for a virtual world. How many times we heared past 15 years "if you do this now people will come".... and where are they??
Stability will keep people.
Most  don't need more "adult".

So, I kind of disagree with this. I feel like most people pick up second life and it's not immediately apparent that the game is _fun_. They walk around a bit, they notice "hey, these walk animations are pretty plain". The controls feel off, there's no way you'd play it reasonably with a joystick. So what if you can make a pretty avatar - the potential isn't immediately there. For the average person, it's going to be "Open the game, is it a game? Yes? No?" and the answer is "No." Because Second life right now is not presented to new users as a game, per se, it's presented as a 2004 attempt at a snow-crash-style metaverse that's had 2-3 big updates to do with adding things like sculpts, mesh, etc...

 

Presentation matters. If the users are pulled in by the fact that it feels _great_ to just run around in the game, I feel like that would be a _huge_ plus. Start them off in a solid, stable, well-optimized sim that's layed out with reasonable detail and looks nice, with some "fun" stuff to do. Let them get in a car and drive around, with some decent fun driving physics, or fly around, or race around on a hoverboard. Let them jump around and interact with their environment. Have them sit in a chair, where their avatar seamless crouches forward and smoothly takes a seat. Show them a stylish array of posters that "show off" the various avatars you could possibly become, and when they click on said poster, turn them into that avatar!

The new user experience in this game is going to be the key to getting future players interested, and I feel like so much can be added to this game with a few simple improvements:

• scripted/interactive/IK/rootmotion animations - the FINAL step we need for making avatars look good and immersive in the game. This would add SO MUCH.
  • Avatars would actually look good by default moving around in the game world
  • Non-human avatars could use IK for limb movement. Imagine a big dragon latching onto the roof of a house and actually being able to perch somewhere
  • Avatars would look good sitting down, getting up, interacting together, etc. INTERACTING TOGETHER. Need I say more?
• Faster, more responsive controls and a built-in game-like control system for the avatar
  • this takes like 1 guy a couple weeks to implement. Some viewers almost do this, except they are limited by what you're allowed to do in game already
  • Add a couple different well-fleshed-out control modes. First person isn't very smooth right now. Movement feels floaty. It would be nice if responsive movement was an option, along with camera smoothing and the like for third-person. I've implemented this personally in Unity for a character controller that feels hella nice - I can attest to the fact that it's something easy to implement that would add a LOT. That being said, I understand it would be more complex for second life, but camera position only needs to be smoothed locally, and avatar movement could be more locally responsive, with some updates coming from the server to apply changes. 
• Do the same thing for vehicle controls. Less floaty, more responsive. Give new users this experience right off the bat if they choose to sit on one of the presented vehicles.
  • Present new users with a car, a hovercar, a skateboard, a hoverboard, a motorcycle, etc, all nearby a racetrack as part of the "new-user" sim.
• Show users the available avatars that are popular on the marketplace.
• An actually DECENT 2D user interface API.
  • Second life needs an API for 2D user-interface drawing. Using server-side 3D models to show up on a slow-to-load, laggy user interface is not really acceptable. There's no reason that HUD objects that show up in 2D need to be 100% server-sided imo.

[Nikkesa]

On 12/12/2020 at 8:07 PM, RowanMinx said:

They should have worked on this instead of the pointless EEP.

Yeah imo the old system _worked_ and the new system, for power users who want to do photography, is kind of rendered null by a few things... You can render depth textures, so basically photoshop would take care of everything anyways, or you can use reshade (though it's not allowed to use the depth buffer since reshade disables it for online games)

[Wulfie Reanimator]

57 minutes ago, Nikkesa said:

The new user experience in this game is going to be the key to getting future players interested, and I feel like so much can be added to this game with a few simple improvements:

Don't overestimate how "simple" any of those things are. Each of them requires major reworks of several interconnected systems that must happen without breaking existing content, which is almost impossible without things that wouldn't convolute the creation process further.

And as much as I agree with you on literally everything else you've said... until I have the skill and motivation to implement most of these things myself, it's unlikely we'll see any of these in SL within the next 5 years at least. We can thump the drums for our laundry list of features that should be implemented (and we have, still do), but LL has the final word on every major feature. You can't even protest it by publishing your own viewer with "fancy features" without LL actively making that viewer's existence harder, because of the "shared experience" clause.

Edited December 14, 2020 by Wulfie Reanimator

[Ardy Lay]

Do we need a Frontier Grid where developers can break content and clauses?

[Nikkesa]

22 hours ago, Ardy Lay said:

Do we need a Frontier Grid where developers can break content and clauses?

I'm pretty sure opensim isn't really restricted - it's just, if you're gonna build your own game, it's probably more worth it to do your own stuff from scratch, then you can call it yours and sell it and stuff.

[Nikkesa]

On 12/14/2020 at 2:40 PM, Wulfie Reanimator said:

Don't overestimate how "simple" any of those things are. Each of them requires major reworks of several interconnected systems that must happen without breaking existing content, which is almost impossible without things that wouldn't convolute the creation process further.

And as much as I agree with you on literally everything else you've said... until I have the skill and motivation to implement most of these things myself, it's unlikely we'll see any of these in SL within the next 5 years at least. We can thump the drums for our laundry list of features that should be implemented (and we have, still do), but LL has the final word on every major feature. You can't even protest it by publishing your own viewer with "fancy features" without LL actively making that viewer's existence harder, because of the "shared experience" clause.

I definitely understand how complex these things are, at least some of them. A few of them simply involve implementing already finished products that got put aside, or increasing bandwidth.... I'm a game designer myself and have an intimate understanding of how complicated this stuff can get. Probably the most complex would be anything that breaks away from strictly server-sided stuff to half-local-half-server...

Sometimes I wish second life was on something like stadia. I know they tried it already and failed spectacularly because it was too early to implement and they didn't really have the server infrastructure needed to do something like that, but nowadays would be alot more possible.

[Ardy Lay]

5 hours ago, Nikkesa said:

I'm pretty sure opensim isn't really restricted - it's just, if you're gonna build your own game, it's probably more worth it to do your own stuff from scratch, then you can call it yours and sell it and stuff.

I tried OpenSimulator.  The loud sucking sound was too distracting.

[Nikkesa]

11 hours ago, Ardy Lay said:

I tried OpenSimulator.  The loud sucking sound was too distracting.

Haha then I definitely recommend getting into real game-dev - it's very rewarding