Is it accurate to say that Unity is the alpha of game development engines at this point?

I would say C++ will be the big deciding factor between the two for some time to come. Yes you can use BP but that is not a good argument if you're talking to a coder, they are going to want to do most stuff in code with BP's for tweaking.

It becomes a bigger deal when you get into custom networking and trying to integrate other third party networking libraries, really easy to do in Unity but many more hoops to go through with UE4 as you'll have conflicts with third party libs and constant types that you need to hack in their header files etc.

I like both but I prefer Unity's coding environment, but UE4's tools, it kind of sucks we can't have the best of both worlds in one engine

 

Discussions keep each engine on their toes, eager to stop people from fleeing. MOAR FREE

Remake a project of yours in blueprints

They can make custom code for blueprints though.

 

We can somewhat have a close approximation with Unity, but then you open an entirely different can of worms by depending on third parties for those tools. Which in turn can lead to situations such as the one with DF-GUI.

 

It's sort of why I have avoided ShaderForge so far, really tired of getting burned by unsupported assets or assets that I become so dependent on that I lose control of my project if a bug crops up. However if I stick with Unity I think I will probably have to bite the bullet and buy it since I see no point having awesome PBR capabilities but then be limited to the few options you have with a default standard shader.

 

General workflow is BP then convert to C++ once you have the API / logic figured out. As @Tomnnn said, you can even make your own BP nodes and flow back.. On top of that C++ isn't really difficult in UE, all the most annoying stuff in C++ is taken care of like MEM ass / GC etc. sure there is stuff like the USTRUCT() class that needs manual null for the PTR in the void Destroy class. But that's nothing really here nor there...

The whole point of the system being open to access is that you don't have to hack in anything, most middleware is pretty easy to integrate.. I find trying to do third party implementations in Unity "hacky" (if at all possible) as you don't have deep access, whether it's performance issues or bugs.. Especially when trying to get deep into rendering pipelines etc.

But a lot of this is extreme cases, if you're creating a game that's out of the 20KM2 Navmesh / AI bounds (which might of gotten bigger not sure) and requires more than the integrated networking functionality can give you. Which is generally if you're smart around 100 players.. Then chances are you will have your own engine developers who won't care about C++.

 

Amusingly enough I bought Shader Forge because of Unreal. Transferring shaders between engines is so much easier for me when I can simply open the two editors side-by-side and duplicate the layout and wiring of the nodes.

 

It depends what *type* of game you are creating as well I would say. The built in replication is great for your standard all players are in one level at a time scenario. When you get to a situation where multiple players are on different levels(and I am NOT talking MMO) it starts to become impossible to use the builtin networking since it doesn't support streaming levels. At least that's how it was when I last looked at it, maybe they have improved it or I am missing something(entirely possible).

 

World composition in UE4 can be used for "streaming" levels, you can transform shifted absolute positions from the clients and re-route them to the right dedicated server. Or you can just build your own...

Personally, I'm not sure what UNET will be capable of or what Photon can do now. I've not really done much with networking since around 2011 when I was working on engines. As I've enough on my plate without caring, maybe one day?

Again, I agree it depends on what you're doing really. But people saying oh UE4's hard or Unity is Alpha is like saying Jim is an alien because he can jump high without any medical diagnostics to prove Jim is an alien.. LOL!

 

Well for desktop/console I think it's a nobrainer that UE4 is more powerful, for mobile Unity still has the edge.

I can get around in C++ "ok"(I spent the first half of my career writing plain C code) but I still find C# much more pleasant, and I am not a fan of blueprint(seriously I just don't get the appeal for a solo dev). I've spent about 2-3 months playing with UE4 so I have some experience, and I can tell just from their IRC channels they have way more "serious"/AAA devs hanging out there, it's cool talking to the guys who work on some large titles.

What I will probably do as an experiment is just do my own networking using sockets because I don't need actor replication for what I'm doing. It's not that hard as you say, but compared to Unity, yes I would say it is harder(which was original question). Photon has support for UE4, here is walkthrough: http://nethersoft.de/photonclient/part-i-integrating-photon-api-with-ue4 which I would also say is harder than integrating for Unity.

 

Unity is pretty good overall. I don't care for the overhead required to get a 2D project going but that is most likely just due to the system being made for 3D. I don't know about graphics quality differences between it and other tools and I don't care about that either. Also don't get it when that seems to come up as a major point in comparisons. The way I see that is there are a ton of Unity games all around this site that look exceptional. Seems like it comes down more to art skills than anything else. Meaning ability to make interesting quality models, interesting quality textures and able to see how to do the lighting FX what is needed and so forth.

I'm thinking Unreal would probably have all of the same bloat when it comes to making a 2D game. It all works basically the same as Unity, right? Some sort of scene editor deal and everything based on doing 3D dev. I've never heard Unreal mentioned for 2D stuff so am thinking people aren't using it for such things?

 

Total overkill for 2D in my opinion, although I *think* the author of 2d Toolkit for Unity is working on Paper2D for UE now.

 

4.8 (I think) in unreal added dynamic navmesh generation It's already had dynamic rebuilding for obstacles, but now you can generate sections you need on the fly. I don't touch the engine very much but reading the patch notes is pretty fun. I can't wait to see what major changes make the engine become "unreal engine 5". Unity's was pretty big.

I knew it was paper2d for unreal but I didn't know that about the author haha. That's certainly interesting.

 

I've seen games good for "Unity", I've never seen exceptional 3D games period made in Unity.. Not to say they aren't great games, but all I've seen is pictures and promo materials. Before Andy Touch drops in, yes I've seen them all.. I look all the time.

If you understand game artwork correctly, the mesh quality and textures is about 10% of the final picture. Materials / correct lightmap UV's, correct lighting, correct colour balance (post) etc. etc., it takes a hell of a lot to get it to look right from start to end.

I've shown ample examples of differences..

As for the 2d stuff:

https://forums.unrealengine.com/showthread.php?77201-2-5-D-Game-Art-Art-In-UE4

Of course.. I'm pretty sure it will be heavy for a 2D game though...

 

So you're saying on the graphics quality side it simply is not possible in Unity to have the same quality level of graphics as it is in Unreal? For the exact same game running at the same framerate? Only reason I am asking the question is because over the years I have seen a lot of very impressive graphics demos yet it was not possible to do in the game at a playable speed. I'm just wondering if the advantages Unreal has in this area are perhaps not really advantages if the performance is worse.

It'd be interesting to have the same game pushed to the max gfx qualify possible in both engines while maintaining 60 fps on today's mid level hardware.

 

Actually the only way I can envisage Unity being beaten at the moment is if Google suddenly released some cloud based game engine that would run in Google Chrome also offline. Written in WebGL. Maybe based on its SketchUp application with added scripting. And could compile to all possible systems instantaneously on their servers.
Or perhaps also run in the cloud like some kind of OnLive streaming application.
It could happen.
I think the paradigm of a game being a different kind of entity to a game engine will slowly disappear until it's all just architecture in the virtual world.

 

You forgot the most important part! "And make it perform acceptably".

 

Sketchup to design models and then a blueprint like node scripting tool that you can expand each node to directly modify the source code of that node

 

Edit: I'll cut this down short instead of a massive explanation..

Theoretically if you have access to the core, yes (it's just DX or GL). Is it worth the time or effort? No.

If lightmapping, performance won't be far apart. Some "Realistic" Arch viz will run 60FPS on GTX470 which is "legacy" now.

 

This is why I would actually like Unity to blatantly copy Unreal as much as legally possible in that regard.
Also a native level BlockOut method ... and navmeshes that don't only work in 3D. It's time guys! Seriously.

Do it! Just - do it!

 

That moment when @hippocoder appears in a discussion about other engines

 

I just went to an indie game dev meetup in Santa Monica, 30+ people there, pretty much everybody uses Unity. Even people who use Unreal have also used Unity. Just FYI.

 

I think just about every game developer has at some point used Unity for something.. Not really a shocker.

 

Wasn't trying to shock anyone.

 

Yeah as of March he said he had been working with Epic for awhile on it, was in reference to being asked if he was going to port 2dtk to ue(no).

 

Well you did a good job .. Nah seriously, Unity is one hell of a popular engine and UE4 is still relatively a new-comer as most avoided UDK / UE3 like the plague. I know A LOT of people who have transitioned though...

Not to say I agree with the transition in some cases, I believe some would be better off sticking with Unity. But one team I was working with making an MMO definitely did the right thing..

 

Did unreal even want it? Are there issues with UMG?

 

I would say unreal is the alpha. Im more likely to associate unreal with big AAA studios than unity.

If unreal ever supports c#...

 

Yes. Epics pace at the moment is really ... uhm ... epic. And I still do like that they are using their own tools to develop Unreal Tournement prallel as well which means that the engine is constantly being tested under real world conditions (no pun intended). I was hoping that Unity would maybe do such a thing in-house some day as well. There was a reason they didn't , though. Can't remember the reasoning but it doesn't change the outcome either.

I think Unity's 2D tools are still far better than unreal's though. The new tools in the current Alpha version are adding a lot of functionality, as well.

 

Then why did you start this thread?

 

I think your opinion of which is the better engine will depend on what you want out of them. For me it is important to make assets for as many engines as I can because it makes me more flexible. If those engines are free that opens up worlds of possibilities. In the past Unity's graphics sucked but having come from QUARK army knife and later Dassault Unity was a drag drop app that let you get assets in there and working very easily and for free. At the time UDK had eye poping graphics, loads more features which came with the engine and pretty good community lesson material, it was a bit fiddly to set up a project and had to be done with a code structure, but was also free.

With this last release Unity's graphics are no longer inferior at all, they are very very good and the improvement is quite a leap. The Unreal engine has also gone 'free' I suspect due to pressure from alternatives. While Unreal has its little quirks it feels very professional. Like I said it depends on what you want. I am a character artist who wants my asset running around in a not too complex environment doing things. I want the model to detect surfaces and change out sound files to suit that, I want retargeting of my skeletons to take advantage of standard animations, I want my figures to swing from trees and scale towers and Unity does this but mostly with plugins. Documentation for retargeting in Unity is so scant I cant make any sense of it. Just before writing this I did another search on the muscle system with nada success. Maybe the info is there, I am sure if I persevere I will get it eventually, but over at Unreal there is a bit fat clear video waiting to answer all your questions plus a very clear system to get you up and running in no time. Its just an example.

Which is better, none are better. Why is it important for an engine to be a leader: for more dev, better support? Whether Unity is the top dog or not, its not going anywhere soon, its here to stay.

My two bob.

 

( ͡° ͜ʖ ͡°)

As long as they can make revenue! It makes me feel bad that there isn't some way free users can contribute.

 

Free users can contribute with the asset store. Unity gets a 30% cut on every asset sold.

Or if you have no money at all, you can still provide tech support on the forums.

Failing that you can hang a round for comic relief.

 

I'm on it!

 

The trend is that Unity is accelerating in function and appeal rapidly. In a few years it moved from basically nothing to a strong position in its industry space and that has led to a bunch of hyped commotion, like the image you posted. Where no alternatives previously existed, Unity offered a valid and approachable alternative which was widely embraced.

It's success is well earned but it doesn't have an 'alpha' position. If the trend keeps up at this rate then sure, it'll be clearly on top in a few years but until then the choice of a game engine is still weighted by a lot of different factors and those weights do not always lean toward Unity as the choice which means it does not occupy an alpha position. In fact, none of the engines do.

Unfortunately people still seem to want to waste time bickering about it. It was interesting for a bit, but now its a troll topic and never contains new information or more compelling arguments than the last thread that needlessly dragged the topic through the mud.

 

Got to admit this doesn't quite makes sense, how do YOU know if they are inferior or not? Have you made something that just blows away the competition in Unity? Please share if you do, without that conclusive proof than the answer is still no.

Also saying it's quite a leap isn't right either, there was a free PBR framework in Unity 4 and they still had a GI solution called Beast.. So what's in Unity 5 that you couldn't get in Unity 4? There's no real marked improvement. Real time GI will never look as good as baked, so if were talking graphics quality.. Still nothing.

Character modelling whilst an impressive skillset, isn't really taking much advantage of an engine. When you make a complete game in both or at least a major prototype, let me know what worked out best.

What's better at a technical level isn't opinion, saying none is "better" strictly doesn't make logical sense. It either has more tools or it doesn't, it's either quicker or slower, it either crashes more or it doesn't, it's either going to allow you to fix issues yourself or it won't.

I've heard coders saying they don't want to use BP, that's an opinion. The fact that BP's exist doesn't stop others being able to make a game with them.

If Unreal was an EXACT replica of Unity then technically one couldn't be any worse or better than the other. Sure they all have pro's and cons the developers have to weigh up and 99%* of it comes down to time as opposed to what an engine is theoretically capable of.

 

Two. Words. Blacksmith. Demo.

 

Yeah looks great for 2010, many words:

Infiltrator
Unreal Paris
Allison Road
All of Koola's stuff
Arkham Knight (which is UE3 / old tech)
Nelo
And many, many, many more

My word even the "Mobile Tech demo" can keep up with Blacksmith:

Unreal Soul

 

Bit harsh. I like the blacksmith's visuals

I kind of like the fact the blacksmith teaches Unity a lot about what can drag on or cause problems in development and this drives improvements so I would encourage Unity to keep making these. We would like to see more interactive demos such as crowds with thousands of people or terrain that actually looks good ... and don't skip the caves/lighting

 

It depends, if they were going for stylistic and not just out and out graphical rendering prowess than sure it does and a major step up to what were used to out of Unity. If people can learn from it than it's good for everyone.

Well anyway, it's getting a bit silly.. It's not all about graphics, I played Bloodborne the other day which is made by a big company and I can guarantee with the right skillsets you can do much better in Unity. I have SEEN much better..

Problem with these threads, they always turn into logic or pissing contests. Without much actual informed decisions based upon actual use of said products. When @tiggus joined, it was getting interesting.. Discussing limitations, pro's and cons you'd only get with working with both products instead of wild speculation.

That's an interesting conversation..

 

I don't know. I don't think its easy - at all - to do good visuals in unity. I've spent a long time (longer than most people here) on trying to make Unity 5 look up to scratch. It can be done but it's REALLY hard to do so.

For a start, I'll put it out there: Unity's lighting model stinks. This extends to the standard shader and the absence of GGX, favouring simpler models. This causes Unity to suffer. Next up, the 0-8 light intensity for Unity is dumb. It makes no sense but people don't realise this. Also the post is woefully outdated.

Next problem: the options are all over the place. Your final image quality depends on post, shaders, lighting and these are all unconnected and fight each other if you're not absolutely sure what you're doing.

Can unity look good? yes. Can anyone make unity look good? absolutely not. Am I attacking unity? nope, I'm just saying don't drink kool aid.

Thankfully you can replace shading and lighting, and I encourage everyone to do so.

 

Well, hardly old. It's a custom made version of UE3, it has a lot of tech that UE4 doesn't have.
It was also made by a huge studio, hardly an off-the-shelf comparison

 

Sadly to get there I do not personally believe you can off the shelf with Unity either.

 

Well they really only changed the lighting system, that Epic improved in UE4.. But for the most part, I'll give you that one ..

@hippocoder

YES YES YES YES YES, Damn yes... If you bypass and re-create everything, I MEAN EVERYTHING (lighting, shadows, shaders).. Use a decent IBL setup, use your own post and avoid everything Unity does (use alloy or something).. Also roll your own terrain system or use meshes, Unity can look great.

I've worked with Enlighten before, so I know it can look damn fantastic. Heard of Battlefield 4?

It takes time, it's a pain in the ass and it bugs me that the asset store vendors know what should be done.

 

To waste time. Amusement. To share that funny raptor picture.

 

... or just an actual small game project on the side. I do believe that someone who wants to be really good at creating tools also needs to use the tools him/herself in order to know how good they are or where things are lacking.

Epic do this with Unreal Tournament all the time. Oddly enough the Unreal Engine is really brilliant for FPS out of the box. Who would have though

 

Not to mention that it's a spoiler-free reference since that shot is in the trailer.

 

Oh. I guess I'm contributing plenty then.

 

This really is a Unity vs Unreal thread in disguise isn't it.