SL At the End of Major Innovations After Uplift? (Comment of the Week)

My post on Jim "Babbage" Purbrick's long-ignored C#-in-Second Life integration provoked many "Oh what could have been" responses, both here and across social media. (Especially poignant to read posts by other former Lindens, remembering this project fondly, if bittersweetly.) For example, this comment from reader Pulsar, who not only bemoans the loss of this C# feature, but who also sees it as a sign of Linden Lab's waning interest in innovation:

I can imagine what SL could have been with C# and many other things.

Second Life instead has been a story of poor decisions, missed opportunities, half-baked new features, that also introduce new serious, user-experience hampering, bugs... that get fixed after ages if never.

Now Linden Lab can't even fix the world map, broken since several months ago. Preview grid (Aditi) inventory can no longer be synchronized since November 2020. Screenshots still tend to fail to upload to my.secondlife, teleport failures, objects not showing up in the viewer until you right click and so on. All stuff that shouldn't have passed any serious QA check in first place.

"It kind of works and it doesn't explode entirely" seems the way to go for Linden Lab.

It would be great having C# in SL, but, yeah, VRChat allows that indirectly. I don't expect anything competitive coming out from Linden Lab anymore. Other virtual worlds are implementing anything that Linden Lab should have done, never did and probably will never do. People used to SL would be amazed to try out any other virtual world and see that it just works without falling apart. None of those can fully replace SL for now (in general, while subjectively it depends on what you do), but the strong points are getting fewer and fewer, while SL is being surpassed in so many other areas.

My strong sense is Second Life is now basically in maintenance mode, and that the user base shouldn't expect any major innovations that might drastically impact the existing ecosystem or hurt the world's steady profit, now that it's going directly into the hands of a massive holding company.

Anyway, the ironic thing about all these recent technical shortfalls is that they're happening after the completion of moving SL to the Amazon web cloud, dubbed "Uplift" -- perhaps the last major innovation to the platform.  As reader Sally Crofton notes:

The Uplift seems to have created more issues than it solved - the destination guide floater works sporadically (no timeline to fix it apparently), Avatar impostors (orange clouds) more often than not, textures popping in and out of focus (if displaying at all) and more random disconnects during teleport than I remember when I joined in 2006!

I think a half term report would be "Fair... needs more effort & improvement"

Speaking of the failing world map, a recurring problem since last November, I did message Linden Lab about it last week, and will report back what they say in reply (if they reply).

Comments

[Adeon Writer]

The world map started working again for me a few days ago after a long period of not working.

[Kyz]

Agree, that SL is in maintenance mode. I was considering going back a few days ago and then read the new blog post about raising fees for buying linden dollars. It came with this reasoning:

Investing in our infrastructure to further improve speed and cadence of updates
Overhauling the onboarding experience for the newest Residents

UI cleanup

Developing new marketing initiatives and entertainment partnerships to fuel growth

Preparing for SL18B

Less glamorous but no less crucial is our ongoing and growing work to ensure compliance with multiple regulatory requirements

In the accompanying forum post when people questioned the details of the improvements the answers were basically that they didn't know specifics, didn't have a timeframe but ... unicorns are coming.

Last time I heard this list of promises it ended with Mark Kingdon landing on an SLB birthday cake, only to poof moments later, never to be seen or heard from again along with 130 or so other employees that got the boot. And the comments on the forums were mostly by users that have been there for many years, with few new faces, so I decided to skip going back to SL for now.

And like the other commenter said, they can't even fix a map. I just don't think they're capable of building a virtual world. SL got to a certain place in innovation and then just died on the vine, and now they can barely function to maintain it without things breaking. The move to cloud should have made things easier and and given them a fresh start for microservices development.

Their business model is doomed as well. Everyone is offering the equivalent of a region for free now for building from SineSpace to Core (which I may have unfairly bashed in another article comment).

Side note Wagner, if you're covering game engines, Flax looks like it may be an up and comer. It's moving along at a pretty fast pace.

[Jub Jub Forder]

World map is working very well for the first time in years. I can finally see most land masses on the same screen.

[P373R Kappler]

I do not think the scripting language matters the most .. What matters is the server backend. LSL is kinda noob friendly. C# would be probably harder to use for people that just want to rotate a prim with 1 function as it is possible with lsl.

There are more pressing issues in sl that need to be addressed.

1 Replacing prim building with mesh editor. Possibly texture baking.

2 Add a animation editor to sl.

3 Modernize the rendering engine.

4 Ad some more key event options in to the scripting engine.

5 Do not turn your avatar while moving which would enable AOs to have side and back walk animations.

SL is the only VR sim i know that enables you to do things with your friends on the fly and that is what makes it great.
Moving building to 3rd party software like blender and animation tools just puts sl back.

That is why nobody went to sansar. It was a cheap unity engine with server system and vr capability.

To change 1 script line in a object you had to recompile the whole scene.

It is not the way to go. Make tools for people to create and they will.

[seph]

"LSL is kinda noob friendly. C# would be probably harder to use for people that just want to rotate a prim with 1 function as it is possible with lsl."

Any other language will show you a proper map and multidimensional array data type within 10 minutes of learning it. In Second Life, as soon as you need either of those you have to invent your way of doing it (strided lists, high memory usage json strings, etc.) There's nothing newbie friendly about LSL vs. C#.

Believing C# is harder than LSL is a lot like believing just learning Blender is harder than prim building. There's literally millions of people that know and were able to learn C# and other professional/hobbyist tools like Blender.

Linden Lab needs to embrace what's already known and easy to learn, not keep inventing things inside of Second Life. Just support industry tools which are all largely hobbyist friendly, cheap or free, with thousands of hours of tutorials as inexpensive as the tools themselves.

[P373R Kappler]

You do not get the point.

Sure there are ways to do stuff easier in c# but also harder to understand for a average joe. Oop classes inheritances what have you. I basically do not care what kind of language they use. The API is what matters. And like just using a differed language would save sl;>

It also matters to 90% of the people in sl to do stuff in world. not hanging on youtube watching blender tutorials. You should be able to do basic mesh editing in world. Blender is not easy to learn for the average joe. Same for animations.
Put the avatar in animation mode rotate some stuff make a frame next frame next frame done, check and upload. This could be so easy for every one. Specialized tools for a certain scenario always work better then industry standard tools for wide variety of scenarios.
People can learn in world from each other if the tools are there ;>
And if you think people with attention span of 20 seconds will watch a 30 min tutorial and put hours in blender to see a stone with texture on their sim. You are wrong;> You can still use blender if required but a lot of things could be done in world without the hassle.

Sl lives from interactions. The more you can interact while building and the more instant gratification is there while creating (example see instantly script changes vs sansars just recompile the whole scene) the better the user experience is.

[seph]

"Sure there are ways to do stuff easier in c# but also harder to understand for a average joe."

Millions of people use C#. Anyone productive with LSL would be many times more productive in C# and have an easier time. LSL is not any easier to learn than C#. If an object in C# confuses you, then using strided lists in LSL to emulate an object will definitely confuse you.

"Blender is not easy to learn for the average joe."

Millions of people use Blender, or Maya and many other 3D apps. The few thousand people in SL that miss the days of everything being built with parametric prims can still do that...those tools were never taken away.

It just so happens mesh uploads came along, and most designers switched to uploading mesh. Probably because it was easier and possible at all to create the things they imagined instead of being stuck with stretching blocks.

People still build in-world together too. It's a myth that stopped for some reason when mesh came along. The building blocks are just different, instead of wooden cubes, its prefabs. It's no less fun, no less a creative endeavor compared to everyone path cutting, hollowing and skewing blocks the exact same way. I'd argue mesh made collaborative building more fun, and most of SL's history at this point has by far been collaborative building with mesh from one's self or many different creators.

Outside tools don't preclude building together in-world, they enhance it. Using more tools that more people know would mean more people in Second Life both consuming and creating.

Seriously, this myth that somehow millions of amateurs in the Unity, Godot and Unreal communities functional with real programming languages and 3D art tools can exist but Second Lifers need less, not more of industry standard tooling makes no sense.

I assure you, yes, the brilliant creators in Second Life have more than 20 seconds of attention span, much more than 30 minutes even, and can master any tool that Second Life will actually let us make use of.

[Guni]

I truly have to agree with P373R Kappler here. Second Life lives and dies with the ability of residents to do things in world, on the fly, together with others and while being immersed inside a social virtual world. Everytime you move something to some tool outside, Second Life dies a bit more. Everytime you make a change that makes things harder for amateur Joe to learn to do things while making it easier for a low cost professional developer or artist in India, you take something away from Second Life. Blue Mars failed, Sansar failed, many others failed for similar reasons. I fully support the idea of updating and improving the collaborative creation tools in world!

[seph]

"Everytime you move something to some tool outside, Second Life dies a bit more."

Which has never happened once. If you want to build with prims, you can still build with prims.

The fact that the majority chose to build with mesh just means they prefer mesh. Most rather go buy a set of trees and grass to landscape in Second Life than arrange cylinders with the prim building tools. You're free to do either.

If you can understand Blue Mars and Sansar being unpopular, understand what's popular and unpopular within Second Life with the same clarity. Maybe folks dig the use of outside tools and uploading content a lot, and the myth of endless entertainment showing one another how we hollow and path cut cubes slightly differently not so much.

[Melissa Yeuxdoux]

LSL is like JavaScript, quickly thrown together.

Unlike LSL, JavaScript has lots of people working to make it better and faster, writing books and tutorials to help people learn. Udemy and Code Academy and various universities offer online lessons in JavaScript, not LSL.

C#, like JavaScript has lots of online resources for learning and lots of effort going into making it better and faster.

LSL made a mistake by inventing its own scripting language; they didn't and don't have the resources to maintain and improve it or to create the vast amount of documentation and instructional materials that exist for C# and JavaScript (and Java and Python and...). Better to take advantage of the work of others and stick with what they do best.

They realized this for textures--SL doesn't have in-world paintbrushes or airbrushes or cloth or sewing machines to make prims or clothing look the way their creators wish. Instead you must use out of world tools: Photoshop or GIMP.

[Fionalein]

There is a simple reason we have LSL: LL can controll what will be executed. IMHO that's a pretty good reason for it's existance.