Comments on Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive UndertakingTypePad2019-11-06T21:49:35ZSLHamlethttps://nwn.blogs.com/nwn/tag:typepad.com,2003:https://nwn.blogs.com/nwn/2019/11/unity-3d-adam-frisby-sl-opensim-sinespace/comments/atom.xml/PerTire commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49fc96d200c2019-11-22T00:27:38Z2019-11-22T00:27:38ZPerTireWouldn't be fun If the owners of Bryce, (a 3D modeling, rendering and animation program specializing in fractal landscapes) added...<p>Wouldn't be fun If the owners of Bryce, (a 3D modeling, rendering and animation program specializing in fractal landscapes) added a button to walk around and build in a shared space?</p>Box LongtimeUser commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a4eadb51200b2019-11-14T16:24:01Z2019-11-14T16:24:01ZBox LongtimeUserAs I understand it (and someone can correct me if I'm wrong) SL composes and manages graphic scenes on a...<p>As I understand it (and someone can correct me if I'm wrong) SL composes and manages graphic scenes on a server and incrementally sends these to a client to be rendered, whereas a typical game engine will download, compose, render and manage a scene entirely on a client with no reference back to a server for the static scene elements.</p>
<p>So, it's not possible to swap out the SL client for a Game Engine (Unity etc) without re-writing the entire platform. What is possible, is to write a new client/server system based on modern principles - which is what Sansar is (albeit without the continuous incremental downloading of scene information)</p>
<p>Is it possible to get SL inventory, scenes and assets into Sansar? Probably - I imagine that an automated process could be devised based on some means of mapping a SL region to a Sansar World, and optimising meshes - somehow.<br />
Then, all the other stuff would have to be managed - region crossings, sky settings (no Windlight in Sansar), the basic avatar and clothing - all of that.<br />
Probably too big of an undertaking, which is why it hasn't been done.</p>
<p>The social aspects (groups, friends lists etc) would be relatively trivial to move over.</p>
<p>I'm no expert, of course.</p>NiranV commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49b3956200c2019-11-09T13:56:54Z2019-11-09T14:01:38ZNiranVhttps://profile.typepad.com/niranvYou'd only want to switch to Unity if you want less modularity, more memory hogging and even less performance. Unity...<p>You'd only want to switch to Unity if you want less modularity, more memory hogging and even less performance.</p>
<p>Unity is by far the worst engine to select (besides the fact that hammering any engine into SL is not going to do anything).</p>
<p>Unity's performance is generally bad, its memory footprint by default is absolutely off the charts and unless you know how to rewrite pretty much everything Unity offers you an libraries, shaders, solutions and so on you're just going to grind performance to a halt without gaining anything but even more memory usage for no reason.</p>
<p>Second Life may run bad at times but that is not because Second Life in itself is necessarily bad everywhere. Sure SL carries a lot of old code around, does quite a few things strangely (due to compatibility) and because of that can't make effective use of new things even if LL wanted to but the way SL runs right now is by far better than it could be if transplanted to any other engine, no matter which. SL is highly complex and does a lot of things at the same time at all times and the only way to improve performance in these regards would be optimizing these operations, offloading them into seperate threads where possible (multithreading) and of course the biggest impact, optimizing content in SL which involves avoiding stuff that has a high impact or using it sparingly.</p>
<p>I don't see why this "switching engines" nonsense comes up over and over, people simply have a totally wrong understanding how that works, switching the engine doesn't just magically fix problems, especially not performance and especially not for SL, in any case it would be way worse, more bugs, more crashes, stability would drop to zero, performance would be terribad and system resources would be wasted to produce a broken world just because you stuffed SL into a new engine without looking at the actual impact of such a crazy decision.</p>irihapeti commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49ad6d4200c2019-11-08T10:34:08Z2019-11-08T10:34:08Zirihapetiwith Apple intending to go from OpenGL to Metal/Vulkan then Linden will move as well and in doing so then...<p>with Apple intending to go from OpenGL to Metal/Vulkan then Linden will move as well</p>
<p>and in doing so then Linden will build a Windows/Vulkan off this</p>
<p>when so then a lot of rendering improvements will come. Cloth and fluids would be near the top of the list I think</p>Wagner James Au commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a4c4014d200d2019-11-08T08:37:13Z2019-11-08T08:37:13ZWagner James AuFar as I can tell, it's not significantly different from launching a game on your iPhone that runs on Unity,...<p>Far as I can tell, it's not significantly different from launching a game on your iPhone that runs on Unity, then quitting and launching one that runs on OpenGL From the SL user's experience it might seem like the same app but on the back-end it's switching engines. It could even be officially branded: "You're teleporting to Second Life 2 brought to you by Unity(tm)."</p>Theanine3D commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49a94d1200c2019-11-07T17:15:51Z2019-11-07T17:20:31ZTheanine3Dhttps://profile.typepad.com/theanine3dI don't think what you're suggesting ("add Unity inside the Second Life viewer") is even legal. The Second Life viewer...<p>I don't think what you're suggesting ("add Unity inside the Second Life viewer") is even legal. The Second Life viewer is open source, licensed under the GPL. Unity's is not; it's proprietary. You can't just combine two engines under entirely different licenses together, especially not with the GPL involved.</p>Pranali commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a4e85b1f200b2019-11-07T12:48:08Z2019-11-07T12:48:08ZPranalihttps://beetv.site/No problem in that cause we know that unity is the best graphics module designing platform.<p>No problem in that cause we know that unity is the best graphics module designing platform. </p>seph commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49a7fca200c2019-11-07T12:24:12Z2019-11-07T12:24:12ZsephThere's a lot easier ways for Linden Lab to automatically integrate existing Second Life users into a newer, updated platform....<p>There's a lot easier ways for Linden Lab to automatically integrate existing Second Life users into a newer, updated platform. They already have Sansar that they could repurpose.</p>
<p>Make it so that all Second Life landowners get as apart of their tier land/experiences in Sansar as well.</p>
<p>Make it so that premium Second Life members earn perks in Sansar as well.</p>
<p>Have L$ and the Lindex be the one currency and exchange between Second Life and Sansar. </p>
<p>Have a single account, single login between both worlds. Build a 'Linden Lab' launcher that can install and launch the official SL viewer (and TPVs) and Sansar. </p>
<p>Stop paying Levi's, NASA, Intel, Twitch streamers, Marvelous Designer, eSports organizations, EDM record labels, etc. and redirect that money into offsetting the risks Second Life creators and entertainers would have to take in moving an idea to Sansar. I'm sure they've spent millions on these fruitless things; give a million instead to Relay For Life in exchange they risk less donations by hosting some of their events in Sansar instead. </p>
<p>Just do things to make it easy and obvious instead of intentionally alienating SL users from Sansar and then SL won't need an updated graphics platform. One's already built, just integrate it. </p>Joey1058 commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a49a635c200c2019-11-07T03:39:01Z2019-11-07T03:39:01ZJoey1058http://joey1058.blogspot.comWhy does this so remind me of Microsoft and their legacy support in Windows? Oh, yeah. It's because of the...<p>Why does this so remind me of Microsoft and their legacy support in Windows? Oh, yeah. It's because of the nightmare it would be to rip out major chunks of code. Let's be realistic, people. Regardless of what LL says, they secretly hoped that Sansar would prove wildly successful beyond all dreams, and SL would have been quietly euthanized in it's sleep by now. The best laid plans of mice and men go oft astray, as they say. Sansar is floundering, and Second Life, for all of it's battle damage, is still the shining example that the world compares against in designing new VR experiences. Including Philip Rosedale and his High Fidelity.</p>Pulsar commented on 'Could Second Life Slowly Migrate to Unity? Maybe, Says Veteran Developer Adam Frisby, But Here's Why That'd Be a Massive Undertaking'tag:typepad.com,2003:6a00d8341bf74053ef0240a4e83988200b2019-11-07T02:48:29Z2019-11-07T02:48:29ZPulsarIt would be a major rewrite, but given the result, it would have been a better investment. Adam is right:...<p>It would be a major rewrite, but given the result, it would have been a better investment.</p>
<p>Adam is right: good development practices require careful planning ad some foresight. If you try to refit it after, it would be harder, like in that famous Peugeot 206 funny commercial. Does Linden Lab have enough hammers and elephants to turn the ancient car into something resembling a new one?<br />
Sometimes it's easier to redo everything from scratch. So you have major rewrites. That works well (if well done) for various kind of software and games. You have Civilization 1, 2, 3... 6.<br />
but...<br />
Would such rewrite work for established social virtual worlds?<br />
In that case, what's important for your users isn't just the software, the features, graphic. No, not just the inventory either. </p>
<p>A fundamental thing, in a <i>social</i> virtual-world, is... the social part: your friends, your contacts etc. Consider this: someone even logs in Second Life with a text-only client ("viewer") just to keep in touch with their friends or their important ones when they can't launch the full viewer. Zero graphic, just the chat. It could be a viewer with zero graphic or a world/continent with a different asset and different avatars, but you are with the company of your friends via chat.</p>
<p>Now imagine the difference between logging in into a new world and <br />
a) most of the time you feel like "there is nobody here"<br />
b) you open your contact list and you tell to your best friends that "it's amazing here!" </p>
<p>Unlike Sansar, it would be a whole different story already if you are with the company of your friends via chat.<br />
That's pretty doable, some experiments were done 10 years ago already.<br />
<a href="https://nwn.blogs.com/nwn/2009/02/metaverses-bridged.html" rel="nofollow">https://nwn.blogs.com/nwn/2009/02/metaverses-bridged.html</a><br />
In this case it would be everything within Linden Lab infrastructure.</p>
<p>Now maybe you would also share and live the experience with your friends. Imagine the difference between:<br />
a) you tell them "well, you have to go to this site, download the client, then install it, then..."<br />
b) If they wish so, you would offer them a teleport. Their viewer switches automatically, they don't even need to relog.</p>
<p>Less "friction" as they say.<br />
An hypothetical dual-client would be a tad complex, however. You can't simply install a different engine. I don't think that the SL viewer was conceived with much modularity in mind anywyay.<br />
Perhaps (just a quick idea) you could make a separated SL2 viewer instead, but with a similar interface, and when the SL1 is requested to go to the SL2 continent, it would make you switch to the SL2 client automatically and smoothly enough. Login would be in a common launcher.</p>
<p>But of course there is the legacy content to consider too.<br />
Notecards, LMs etc should be easy to have in common and it would be good if they are. The SL2 continent won't need "sculpt" trees I suppose... but you should be able to "rez" your mesh chair. Possibly with a new LI formula that applies for the SL2 continent or better a texture and triangles budget for each parcel there, so you will optimize better.<br />
I'd see better vehicles though. But even if there is the current physic etc, it would still be ok. Just keep avatars onboard even if they aren't actually sitting (like in Sansar).<br />
But if you keep mesh vehicles and chairs compatible, somehow, then you need to be compatible with the animation files too. Then you need the avatar skeleton to be compatible as well. So the scripts, more or less.<br />
So it won't be a too radical change.</p>
<p>However, if the avatars and their clothes are much better than the current ones (and than the Sansar ones...) I'd be ok with that. After all, I upgraded my avatar so many times in SL: prim shoes with invisprims => sculpt shoes => then with alpha layers => mesh shoes for the classic avatar => shoes for the SLink mesh feet => shoes for Maitreya... same with rings, hair, whole outfits and bodies. This time it would be SL2 clothes.<br />
But then I'd like hair are like those in Blue Mars and modern games though, that move and flow naturally on your shoulders, without entering in your body.</p>
<p>I just think of what this SL2 could have possibly been if Linden Lab worked on that, rather than wasting all that money and time on Sansar. Maybe that would have be possible under a different management and mindset (although I can understand that decision).<br />
At least they did some upgrades on SL, like the "bento" skeleton, replacing the old one without disrupting the legacy content. That wasn't granted and they did a good job there.</p>