• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Amazon Lumberyard Engine now "Open 3D Engine (O3DE)", Open-Source

CamHostage

Member

Amazon's Lumberyard game engine is going open-source. The engine is part of the Linux Foundation's Open 3D Foundation project for 3D technology availability to the public.

The engine had previously been available for "free" use before, but now Open 3D Engine will be royalty-free and open for modification. A developer preview version of Open 3D Engine (primarily an engine-experience preview, not exactly designed for content generation yet,) is available as of yesterday on GitHub.


This engine has a strange history of being an offshoot of CryEngine acquired for Amazon to make its game studio projects, but has since been rewritten from scratch. Unfortunately for Amazon, almost all of its projects have met with troubled development cycles (some blaming Lumberyard, though it's probably easier to point to a wide variety of other failure points as well as just general disinterest from gamers in products Amazon was pushing in its bid to enter the gaming market.) As mentioned, you could have had the previous Lumberyard to work with before if you wanted to build your own game, but it seemed to only ever have Amazon's own studios using it (a search on YT for Lumberyard user projects turns up few user cases even though it has been available since 2016 in different forms.)

Still, this is a very powerful engine being road-tested on projects like the MMO New World, and versions of it were also used in the racer The Grand Tour Game, the beta-canceled hero shooter Crucible, the Smash-style fighter Coffence, Chris Roberts' impressive-yet-ever-in-development Star Citizen, and the action-RPG Deadhaus Sonata. (Industry veteran Denis Dyack vouches for Open 3D Engine, which may or may not be a good thing...?) Amazon AWS cloud + Twitch features are still a part of Open 3D Engine, (which I assume is Amazon's interest in this going wide?) but it has plug-in "Gems" hooks for modular service attachment to your development project. O3DE is multi-platform for PC/Mac/Linux, console (Xbox/PS, with Switch support planned,) VR, and mobile platforms.

New to the Lumberyard/O3DE project will be the "Atom Renderer", a modern renderer that supports PBR materials, raytracing, and global illumination. Atom is compliant with Vulkan/Metal/DirectX12 and supports Forward+ rendering, with deferred rendering to be included as well. As of 2021, Lumberyard/O3DE will also include a new AzNetworking networking stack for low-latency, encryption/compression, highly performant online sessions.

Could be something, but...
 
Last edited:

KyoZz

Gold Member

Open 3D Engine, achieved with CryEngine 3

jumping crysis 3 GIF
 

M1chl

Currently Gif and Meme Champion
Back then in 2016-17 Cryengine was such a mess, that I hope that this is at least somewhat patched.
 

Griffon

Member
Cryengine is a mess to work through. What they should've done is chop up the renderer separately and make that open source on its own, forget about the rest. That would've been more useful than the whole clusterfuck (but then again, maybe it wasn't possible at all with this engine).
 
Last edited:

Pagusas

Elden Member
It’s very easy to see why Unreal is developing such a stranglehold on the market when you look at things like this in comparison.
 
This engine has a strange history of being an offshoot of CryEngine acquired for Amazon to make its game studio projects, but has since been rewritten from scratch. Unfortunately for Amazon, almost all of its projects have met with troubled development cycles (some blaming Lumberyard, though it's probably easier to point to a wide variety of other failure points as well as just general disinterest from gamers in products Amazon was pushing in its bid to enter the gaming market.) As mentioned, you could have had the previous Lumberyard to work with before if you wanted to build your own game, but it seemed to only ever have Amazon's own studios using it (a search on YT for Lumberyard user projects turns up few user cases even though it has been available since 2016 in different forms.)
Meme Reaction GIF by MOODMAN


I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.
 

godhandiscen

There are millions of whiny 5-year olds on Earth, and I AM THEIR KING.
Making it open source is very smart. There is over a million game developers who cannot afford the fees of engines like Unreal and Unity, and are just looking to work on something that they can modify if they ever needed to.

Yeah, today UE and Unity dominate the landscape, but I remember when in ~2009 people thought the same of Unity (it’s shit cuz nobody uses it), and it now competes for attention with UE thanks to the fact that it was more accessible than UE; it took Epic a bit to realize they needed to simplify their tooling before they lost their advantage.

I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.

The innards of game engines need to be rewritten often in order to take advantage of the new hardware. It is pretty common in fact. Often the only thing that remains in an engine after a decade is the data model interface.
 
Last edited:

CamHostage

Member
Interest is so low that this is the first reply...

I'm all for more options in the market but they have a tough battle competing with the power players like Unreal and Unity.

I know, I spent like half an hour going through GAF's search system, thinking, "This happened yesterday, surely somebody posted it...?"

There are tons of options of engines, and Amazon may be just dumping the engine into the open-source community in a last-dash attempt to get something out of it while also trojan-horsing all of its AWS components into the indie scene (which is always the goal, though AWS is more attractive without mentioning the failed Amazon Game Studios & engine that it was connected to.) Thus far in Lumberyard history, very studios seem to be biting.

Integrating it into Linux Foundation may change things though? If you thought Amazon was going to pull the plug any minute in the middle of your project, you wouldn't be using it, but now it's open-source and maintained by a community group, so that's a little bit more reassuring, perhaps. Also, no royalties or seat licenses; it's just "free" unless you choose to use Amazon's cloud servers. And it seems to be powerful as hell (although oftentimes, getting to that power is hell on a developer.)

If nothing else, if Star Citizen ever actually ships (*sigh* Squadron 42, you were the only hope for SC not being a punchline when anybody brings it up,) there will be at least one game that justifies this thing existing.

What they should've done is chop up the renderer separately and make that open source on its own, forget about the rest. That would've been more useful than the whole clusterfuck (but then again, maybe it wasn't possible at all with this engine).

Maybe. You're supposedly able to use O3DE in a piecemeal manner, so probably parts of it will be pulled out or melded with other bits as needed (assuming it takes off and proves any great use at all,) but I don't know that you'd use just the renderer from it and like put that in a GODOT project.

I call BS on the bolded ("the engine has has since been rewritten from scratch").

Your call of BS is totally understandable. A marketing man's use of the term "from scratch" is very different from the actual definition of the term.

Then again, it's rebuilt enough that there's a "CryEntity Removal Gem" plug-in just for removing all the remaining CryEngine parts that had to stay in Lumberyard to maintain compatibility / familiarity as they rewrote the thing. As of October 2020, 70% of the engine had been completely overhauled, and more work has gone into it since to get to this O3DE point. Derek Reese of RedHat said yesterday, "O3DE is not CryEngine/Lumberyard. It continues some of the best features and code from Lumberyard, but is its own full rewrite and independent engine." The new O3DE development suite and project roadmap seem to diverge pretty clearly from anything you'd be used to in CryEngine. At its heart, O3DE/Lumberyard is still constructed around a certain way of building a game (it's an Entity-Based system, for those who that has meaning for,) and currently in the development suite there still are things like a checkout system for assets when artists need to modify materials.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.

Well, look at all of the decisions Amazon has made over the years and you'll see that the company is either not very lucky or not very wise when it comes to developing video games. Also, look at the troubled state of CryEngine itself (which is often awesome to behold, but seems to be a nightmare to work with and has very little integration these days, plus even Crytek's own flagship relaunch of Crysis 1 was botched so something's up with that system even though when it's used well, it can be pretty great.)

The answer to that question on Amazon's part is probably, "...Whoops."
 
Last edited:

CamHostage

Member
I have a feeling this just means abandonment by Amazon, right? How much support will there be?

The launch version of O3DE (which is not what's on GitHub, yet) looks in the release notes and roadmap to be a no-fucking-around major version that has all the promises made of an up-to-speed, competitive game engine today. Improvements like the Atom Renderer, the new networking stack, the totally overhauled UI, new math libraries and more show an engine that has been deep in the works to get ready for a second chance to compete on the grand stage against the established engines. This is not meant to be the end; it is a brand new beginning.

...All of that is said on paper, of course, not even really in demos. But if you follow Lumberyard at all, the 1.2X updates have slowly pulled this system out of the crapper (1.26/1.27 with the new US seem almost pleasantly shitless.) And so it's either getting better, or it's about to be as good as it'll ever get (in terms of official support), which might be good enough for a while.

(Some key features are still missing, like Switch output and deferred rendering and whatnot on the roadmap; also, it can't do C# scripting, only LUA and C++. Missing also is an Asset Store, for better or worse... completely for the worse if you actually want to use it, but in time I'm sure some asset community will rise up and you'll be able to get all your Star Wars ships and CoD gun rips from somewhere)

But hey, you don't need to take it from me (and I'm no game designer, so I have no take to give.) There are a few professional game designers using O3DE, and their results can be impressive. Like, if you want to hear from a veteran game designer who really knows the ins and outs of O3DE (and had used Lumberyard in the past and understands the evolution,) just check out this video with longtime developer Denis Dyack, and in the meantime, I'll just be here, balancing this rickety old loaded shotgun on my foot...


(*BLAM* ?)
 
Last edited:

DCDW

Member
Is any developer actually using this?

Not saying this for snark, real question.
I believe Star Citizen is the highest profile developer using it. Having the engine go open source might be a silver lining for RSI as it might open the possibility of a Durante type person to help iron out the kinks in their code/optimization.
 

coffinbirth

Member
The launch version of O3DE (which is not what's on GitHub, yet) looks in the release notes and roadmap to be a no-fucking-around major version that has all the promises made of an up-to-speed, competitive game engine today. Improvements like the Atom Renderer, the new networking stack, the totally overhauled UI, new math libraries and more show an engine that has been deep in the works to get ready for a second chance to compete on the grand stage against the established engines. This is not meant to be the end; it is a brand new beginning.

...All of that is said on paper, of course, not even really in demos. But if you follow Lumberyard at all, the 1.2X updates have slowly pulled this system out of the crapper (1.26/1.27 with the new US seem almost pleasantly shitless.) And so it's either getting better, or it's about to be as good as it'll ever get (in terms of official support), which might be good enough for a while.

(Some key features are still missing, like Switch output and deferred rendering and whatnot on the roadmap; also, it can't do C# scripting, only LUA and C++. Missing also is an Asset Store, for better or worse... completely for the worse if you actually want to use it, but in time I'm sure some asset community will rise up and you'll be able to get all your Star Wars ships and CoD gun rips from somewhere)

But hey, you don't need to take it from me (and I'm no game designer, so I have no take to give.) There are a few professional game designers using O3DE, and their results can be impressive. Like, if you want to hear from a veteran game designer who really knows the ins and outs of O3DE (and had used Lumberyard in the past and understands the evolution,) just check out this video with longtime developer Denis Dyack, and in the meantime, I'll just be here, balancing this rickety old loaded shotgun on my foot...


(*BLAM* ?)
That was a great watch, thank you!
Unfortunate about the compression artifacts, but I was here for Dyack explaining things anyways.
Very exciting to see this happen. The dynamic damage mapping on models of this type piqued my interest in particular.
 

mckmas8808

Mckmaster uses MasterCard to buy Slave drives
I believe Star Citizen is the highest profile developer using it. Having the engine go open source might be a silver lining for RSI as it might open the possibility of a Durante type person to help iron out the kinks in their code/optimization.

Now that's interesting news.
 
Star Citizen is using a license for Lumberyard and basically that's it.
They are not using Lumberyard as is, they are on their own fork of CryEngine, which, and I am not joking, can be called Lumberyard because Amazon's license includes historic versions of CryEngine.

This (yet another red flag) was explained graphically when that switch was announced. And anyone who knows how this stuff works laughed and laughed, and laughed.

If Star Citizen or SQ42 ever see a proper release (lol), and they are not broken beyond repair, that would be a huge win for CIG, but unfortunately it still would not be a good advertisement for Lumberyard proper.
 
I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.
“Amazon may have licensed out the engine to use for their games, or, as some sources have suggested, they may be using it as a baseline to build their own proprietary gaming engine.”

“Amazon licensed the German studio's engine and got "full, unencumbered access to the technology" to build upon, says Mike Frazzini, vice president of Amazon Games.

However, Lumberyard represents a branch of that tech, and the company is replacing or upgrading many of CryEngine's systems. Future versions of CryEngine and Lumberyard will continue to diverge.”

Basically how Google’s Android used to have alot of code made by different companies, till it didnt.
 

martino

Member
It will need a big actor to use it and i don't think Apocalypse Studios is that (why Denis ? why a gaas ?).
More options is always good , it's even better when it's open source.
 
I know, I spent like half an hour going through GAF's search system, thinking, "This happened yesterday, surely somebody posted it...?"

There are tons of options of engines, and Amazon may be just dumping the engine into the open-source community in a last-dash attempt to get something out of it while also trojan-horsing all of its AWS components into the indie scene (which is always the goal, though AWS is more attractive without mentioning the failed Amazon Game Studios & engine that it was connected to.) Thus far in Lumberyard history, very studios seem to be biting.

Integrating it into Linux Foundation may change things though? If you thought Amazon was going to pull the plug any minute in the middle of your project, you wouldn't be using it, but now it's open-source and maintained by a community group, so that's a little bit more reassuring, perhaps. Also, no royalties or seat licenses; it's just "free" unless you choose to use Amazon's cloud servers. And it seems to be powerful as hell (although oftentimes, getting to that power is hell on a developer.)

If nothing else, if Star Citizen ever actually ships (*sigh* Squadron 42, you were the only hope for SC not being a punchline when anybody brings it up,) there will be at least one game that justifies this thing existing.



Maybe. You're supposedly able to use O3DE in a piecemeal manner, so probably parts of it will be pulled out or melded with other bits as needed (assuming it takes off and proves any great use at all,) but I don't know that you'd use just the renderer from it and like put that in a GODOT project.



Your call of BS is totally understandable. A marketing man's use of the term "from scratch" is very different from the actual definition of the term.

Then again, it's rebuilt enough that there's a "CryEntity Removal Gem" plug-in just for removing all the remaining CryEngine parts that had to stay in Lumberyard to maintain compatibility / familiarity as they rewrote the thing. As of October 2020, 70% of the engine had been completely overhauled, and more work has gone into it since to get to this O3DE point. Derek Reese of RedHat said yesterday, "O3DE is not CryEngine/Lumberyard. It continues some of the best features and code from Lumberyard, but is its own full rewrite and independent engine." The new O3DE development suite and project roadmap seem to diverge pretty clearly from anything you'd be used to in CryEngine. At its heart, O3DE/Lumberyard is still constructed around a certain way of building a game (it's an Entity-Based system, for those who that has meaning for,) and currently in the development suite there still are things like a checkout system for assets when artists need to modify materials.



Well, look at all of the decisions Amazon has made over the years and you'll see that the company is either not very lucky or not very wise when it comes to developing video games. Also, look at the troubled state of CryEngine itself (which is often awesome to behold, but seems to be a nightmare to work with and has very little integration these days, plus even Crytek's own flagship relaunch of Crysis 1 was botched so something's up with that system even though when it's used well, it can be pretty great.)

The answer to that question on Amazon's part is probably, "...Whoops."

Thanks for the detailed response to what was intended to be a pretty snarky rhetorical question.

You seem very knowledgeable about this. I have a question for you.

Why in your estimation, would you say CryEngine never really took off like UE and Unity did? What was it about the development suite that made it unappealing to developers?
 

oldergamer

Gold Member
The launch version of O3DE (which is not what's on GitHub, yet) looks in the release notes and roadmap to be a no-fucking-around major version that has all the promises made of an up-to-speed, competitive game engine today. Improvements like the Atom Renderer, the new networking stack, the totally overhauled UI, new math libraries and more show an engine that has been deep in the works to get ready for a second chance to compete on the grand stage against the established engines. This is not meant to be the end; it is a brand new beginning.

...All of that is said on paper, of course, not even really in demos. But if you follow Lumberyard at all, the 1.2X updates have slowly pulled this system out of the crapper (1.26/1.27 with the new US seem almost pleasantly shitless.) And so it's either getting better, or it's about to be as good as it'll ever get (in terms of official support), which might be good enough for a while.

(Some key features are still missing, like Switch output and deferred rendering and whatnot on the roadmap; also, it can't do C# scripting, only LUA and C++. Missing also is an Asset Store, for better or worse... completely for the worse if you actually want to use it, but in time I'm sure some asset community will rise up and you'll be able to get all your Star Wars ships and CoD gun rips from somewhere)

But hey, you don't need to take it from me (and I'm no game designer, so I have no take to give.) There are a few professional game designers using O3DE, and their results can be impressive. Like, if you want to hear from a veteran game designer who really knows the ins and outs of O3DE (and had used Lumberyard in the past and understands the evolution,) just check out this video with longtime developer Denis Dyack, and in the meantime, I'll just be here, balancing this rickety old loaded shotgun on my foot...


(*BLAM* ?)
Is that Denis Dyack
 

CamHostage

Member
You seem very knowledgeable about this. I have a question for you.

Why in your estimation, would you say CryEngine never really took off like UE and Unity did? What was it about the development suite that made it unappealing to developers?

Oh, no,no,no, not me, I'm just fascinated by game development and the state of the industry. And, I'm old, so I've seen game engines and development suites and middleware come and go. I've made my share of LBP levels, I get myself familiar enough with game development to be able to have a conversation with friends who work in software, but mostly if somebody asked me about coding, I'd just be like, "Uh, goto 10?"

But as far as CryEngine...
As I understand it, documentation and support was troublesome. Crytek is in Germany, while Epic Games is in North Carolina. Epic writing all of the documentation in English and being easy to make a phonecall to and all the forums where experienced developers hung out being primarily in English (even if you don't live in America or Canada, English is a de facto second language for many in modern countries, especially if you are in the business of producing international product) made running a project in an office on Unreal Engine was dependable and easy. Language and availability is also part of why, if you look at where the company put stakes up in Epic Games' 40 worldwide offices , there's a strategy to having Epic Games China and Epic Games Japan and Epic Games Korea and Epic Games Berlin/Cologne (Germany) and Epic Games Stockholm (Sweden). You've never bought a game with any of these studios credited on the box, but because these studios exist, other games you do own from those territories have better Unreal Engine systems than they would have without.

Unreal was built around the concept of a development suite, not just the engine. Yes, it's a killer engine to play with, but for game companies, that's just part of why they contract into Unreal. The real secret sauce is that the development suite makes it easy and efficient for a 20, 40, 100, 200+ person staff of coders and artists and sound designers and level crafters and such to work together, and to bring their projects in and out from other development tools, and allow everybody to work together to build one massive project with less worry that any one project will break everything. Crytek shot for that too (they had their concept of a "Sandbox Editor", and their development system was the first I ever saw that tried to be platform-agnostic, where you just made the game and supposedly hit a button and it'd compile for PS3 and 360 and PC at some base level of optimization,) but as developers were looking in the PS3/360 era for something to replace the familiar Renderware (which some day needs to be deep-dived; how the biggest engine of the PS2 era disappeared with only the stunning Burnout Paradise to speak for it,) Epic tried to make the choice easy. Whereas the competition was still trying to convince you what the engine could do if you mastered it and maximized it, Epic was also explaining what a regular game development studio could probably do on a budget and on a schedule and still be proud of.

Epic was specific about how it went about promoting its game engine features; Crytek was all about the amazing things it could do and the incredible features (some of which are still advanced in gaming, and there are reasons why there's still the running joke about new hardware, "...but can it run Crysis?"). Meanwhile, Unreal zeroed in on the concept of "It just works". Their approach to lighting, to scene construction, to stringing components rather than "coding", to integrating external plugins, to selling prefab assets, it was all best-in-class work, but it also wasn't a hassle to deal with. As it matured, you could trust Unreal Engine to perform at expectations (unless you were Denis Dyack) and for there to be an answer easily obtained as to why the darned thing wasn't doing what you thought it was supposed to do. Unreal was so friendly and powerful that Epic could open it up to even novice coders; CryEngine flummoxed even Amazon's well-paid developers.

I don't really know the story too well of Unity, but I know they went heavy on user-friendliness as well from the beginning, and their approach to licensing fees was intended to be better for developers too. More importantly, they targeted the mobile market early. So when app developers needed somewhere to turn to, Unity was there in 2008 for them.
 
Last edited:

SF Kosmo

...please disperse...
Meme Reaction GIF by MOODMAN


I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.
It's been rewritten, but "from scratch" is probably misleading at best.

The idea of a powerful modern engine and toolset going truly open source is actually kind of a big fucking deal if it manages to attract the right community and the license isn't too restrictive.
 

IntentionalPun

Ask me about my wife's perfect butthole
Meme Reaction GIF by MOODMAN


I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.

This new engine is very different from Lumberyard, which already had ditched a lot of it's CryEngine roots.

The reason why they spent those millions is because CryEngine was up for sale for a relatively small amount... the reason they ditched it, is it sucked lol
 

Redneckerz

Those long posts don't cover that red neck boy
Cryengine is a mess to work through. What they should've done is chop up the renderer separately and make that open source on its own, forget about the rest. That would've been more useful than the whole clusterfuck (but then again, maybe it wasn't possible at all with this engine).
They went with a completely new renderer.
O3DE is basically the same as Ubisoft's Dunia: Its roots are in Cryengine, but everything else is either replaced or heavily rewritten.
Meme Reaction GIF by MOODMAN


I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.
Critical research hasn't either, so here (Other comments are also worth the read) are the receipts as to what they mean with from scratch:
Lumberyard was essentially three engines smooshed into one. The acquired DoubleHelix tech, CryTek, and all the Lumberyard specific code that was developed on top. We finally have a chance to unify a lot of this to make a coherent and straight forward engine.


CryPhysics (replaced with nVidia physx), CryNetwork and GridMate (replaced with an entirely new networking and multiplayer solution), CryRenderer (replaced with the new Atom renderer), CryAnimation (replaced with EMFX), Flowgraph (replaced with script canvas), lmbr_waf (replaced with cmake), CryFont, etc, etc, etc... There was an enormous amount of technical debt that required a major version release to fully address, since it was entirely possible customers were relying on legacy code for their projects and we didn't want to blindside them with a point release.
So there is quite a bit that is from scratch.

What remains are a few bits of CryCommon and CrySystem. See the Github page, or this post.
It's been rewritten, but "from scratch" is probably misleading at best.
Nah, not really. As presented by the receipts.
 
Last edited:

godhandiscen

There are millions of whiny 5-year olds on Earth, and I AM THEIR KING.
Sounds like Godot dev is a little butthurt




He sounds threatened ngl. His argument is bullshit. I like Unity and their data model is amazing, but I wouldn’t be hating on another engine just because I don’t have experience with it. Competition is good.
 

SlimeGooGoo

Party Gooper
He sounds threatened ngl. His argument is bullshit. I like Unity and their data model is amazing, but I wouldn’t be hating on another engine just because I don’t have experience with it. Competition is good.
Also, 3D seems like a lost battle for those that don't have resources and investment. There's way too much new stuff every couple of years or so for a small team to be able catch up.
This new 3D engine not only is open source and has a good license, but it also has a lot of corporations supporting it.

Godot should focus on 2D, IMO.
 

Irobot82

Member
This engine has a strange history of being an offshoot of CryEngine acquired for Amazon to make its game studio projects, but has since been rewritten from scratch

Now I wonder. Star Citizen originally was on Crytek, then moved to Lumberyard. I assume they didn't redo it from scratch too, so where are they now?
 

Bogeyman

Banned
Seems a bit to me like there's some room in the market.

Unity: Simplicity biggest benefit. Performance biggest drawback, rendering features not quite absolute cutting edge either.

Unreal: Can do anything, can squeeze out a lot more in terms of performance, offers latest technological features. Noticeably More complex than unity.

Lumberyard, or whatever its called now: (maybe?) AAA ready technology, biggest benefit will clearly be that its royalty free. Biggest drawback - people aren't familiar with it yet, smaller community, guessing no official customer support either.

Seems to me like it's potentially appealing for professional devs (indies probably less so) willing to take a gamble with the upside of saving a giant ass amount of royalties, if it works out
 

CamHostage

Member
Now I wonder. Star Citizen originally was on Crytek, then moved to Lumberyard. I assume they didn't redo it from scratch too, so where are they now?

My assumption is that Star Citizen is so far down the road (despite the quagmire of its development history) that they're mostly on their own development technologies, using Lumberyard only as their base. The Lumberyard system would be down in there and the Lumberyard development environment would still form the foundation of the work collaboration environment, but they're probably far enough along (despite no end in sight...) and have enough of their own tech built that they may not have even updated their working version of Lumberyard in a while.

I'm not in game development, but probably it's like where I work that we have common-market tools, but we customize them so much that something exciting like a Version 3.0 of the "engine" comes out, there's the letdown in realizing we're stuck in the 2.X environment for the life of the techstack because nothing's compatible anymore with our in-house tools. (Ultimately, our coders over time do their best to clone the features of 3.0 until eventually we have to replace everything and start over, but that's usually 4.0 or 5.0 before that happens.)

We'll see if we ever hear Cloud Imperium speak of O3DE. They were not mentioned in the list of partners when Linux Foundation announced the development system (instead, Denis Dyack and his Apocalypse Studios company were the big get in the press release,) and they have made no mention AFAIK about changes to the engine. So probably this won't affect them, they'll continue on legacy systems, and they've got enough going on as it is...
 

01011001

Member
Meme Reaction GIF by MOODMAN


I call BS on the bolded.

What the fuck would be the point of spending millions on an engine from Crytek only to re-write it from scratch. Critical thinking hasn't survived the modern era I see.

they got into legal trouble with Crytek, that's why they basically had to rewrite it
 

Sophist

Member
This engine has a strange history of being an offshoot of CryEngine acquired for Amazon to make its game studio projects, but has since been rewritten from scratch.
Not unusual.

Tim Sweeney: With Unreal, the hardest aspect of the engine was the renderer. It's the component I rewrote 7 times over 18 months before coming up with something I was happy with. Less glamorous is the overall framework of the engine, the code that holds everything together. The hardest thing for any new programmer is to build a large piece of software without having it collapse because of its size and complexity. Unreal has held up pretty well, but I'd been programming for 10 years by the time I started writing Unreal.
 

Rest

All these years later I still chuckle at what a fucking moron that guy is.
Ah yes, game engines. Engines for games. Video game engines. So enthralling, game engines.
 
As someone who extensively uses and customises open source web projects it's not a shining light over commercial projects. Open source has issues of bugs, time for features/changes/fixes to turnaround, support for commercial ventures/developers, documentation, samples, availability to expertise and even licensing issues. While I love open source it's not a silver bullet. There are far more open source projects that fail, cease to get updates, have communities fall apart around them and end up in a dead end than the ones that succeed and become stellar projects entire businesses/industries can be built around.

A thriving open source project/community is a godsend to a commercial industry around it. A failing incomplete, bug riddled with few core community members stranglehold on the project can be the death roll for the project and commercial entities using it.
 
Last edited:

CamHostage

Member

Man, just a bit ago I was remembering that this was a thing, but I hadn't heard about it since the Summer; good to get news on it.

Seems a bit to me like there's some room in the market.

Unity: Simplicity biggest benefit. Performance biggest drawback, rendering features not quite absolute cutting edge either.

Unreal: Can do anything, can squeeze out a lot more in terms of performance, offers latest technological features. Noticeably More complex than unity.

Lumberyard, or whatever its called now: (maybe?) AAA ready technology, biggest benefit will clearly be that its royalty free. Biggest drawback - people aren't familiar with it yet, smaller community, guessing no official customer support either.


While the cream has certainly risen to the top, it has been frustrating to see Unreal Engine just accepted as the defacto game engine. Unity has fallen off as a viable competitor in high-end games (I was hoping to see some next-gen plans for the last two GDCs but its showcases continue to have fewer and fewer recognizable titles, and its improvements seem to never address the technical issues that has given Unity a bad name with gamers,) GODOT and GameMaker Studio are fine for 2D (there's some 3D capabilities in each but I think tellingly the Risk of Rain developer switched from GM to Unity for the sequel,) and then on the high end Unreal Engine pretty much stands alone for game developers to choose from, with Unity being perhaps more a financial or functional decision than a technical option.

So, an alternative would be great. Even if Unreal continues to take victory laps, you do want some competition out there for it.

That said, the buzz on O3DE seems to be... hard to find. The tweet of this announcement got a handful of retweets and likes; its Discord has less than 2k members after running training events. They have no showcase projects made to just show the engine flexing and the development tools in action, and their tutorial demo on the Youtube page is them making a Pong. It all has to start somewhere, but I'm seeing little to inspire developer interest besides the fine print (or lack thereof) in the pricing system of "free". Hopefully at least modders will jump on it eventually
 

SlimeGooGoo

Party Gooper
That said, the buzz on O3DE seems to be... hard to find. The tweet of this announcement got a handful of retweets and likes; its Discord has less than 2k members after running training events. They have no showcase projects made to just show the engine flexing and the development tools in action, and their tutorial demo on the Youtube page is them making a Pong. It all has to start somewhere, but I'm seeing little to inspire developer interest besides the fine print (or lack thereof) in the pricing system of "free". Hopefully at least modders will jump on it eventually
Yeah, even Godot has a hard time making devs adopt it due to not having any top selling games made with it.
GameMaker on the other hand was used for making tons of popular games, which leads people to trust it more than the alternatives.

I assume unless Amazon decides to invest a lot of money in this engine and do something serious with it, people will just observe it from a distance.
 
Top Bottom