Yes; the world is full of scumbags; But can’t assume that every single game and every game company are like that.
I’d like to agree with you there but unfortunately can’t.
The reality as I see it is that the reason why recent games have no depth even though our tech has gotten better is simply because every business wants to profit fast.
Game development has several hurdles:
- Design a fun game! (game design)
- Build a well functioning art style! (art direction)
- Build the game big enough! (production)
- Make people hear about and want to buy your game! (marketing)
- Somehow, find enough money to do all 1 …4, ahead of time, and then make back more once it’s released. (finance)
I can think of many games that failed one of these five. NMS happened to fail #1. That doesn’t make them “scumbags.”
Most indies can’t go very far on 3, because costs run away quickly.
Unreal Engine can help with a technical foundation for 1, 2, and 3, so that there are less things YOU need to do, which saves lots of time.
I don’t think UE4 adds more troubleshooting than the alternative; I think it’s actually better. All engines, including your own, need troubleshooting.
All engines, including your own, also need a learning period, during which you will feel less productive. That’s an unavoidable outcome of the world moving on.
Once you’re at equilibrium, debugging a UE4 bug should be no harder than debugging a bug in your own engine, except with UE4 there are hundreds of people doing it, so you will on average have less debugging than with your own engine. (That doesn’t count all the time it takes to actually build the engine and tools.)
Business are in the business of producing goods that people want to buy. It seems like most gamers like buying what businesses are producing. Thus, I would say that the reason we don’t get super deep games, is that most gamers don’t want to pay extra for super depth.
An inhouse engine can be made specifically for that one game; which reduces drastically the amount of work involved to maintain the engine code base.
Of course if a studio wants to have its own “one-fits-all” engine for all their game projects, then it’s in the same bandwagon of all those public engines out there; then being made in-house doesn’t really make a difference.
Does it do this also when you shove the grenades in an array? I couldn’t do the ocean water tutorial to build a water Shader BECAUSE in THE MATERIAL FUNCTIONS under Physical Material option there’s no functions listed…
I couldn’t find BP_WATER ACTOR in 4.12 for the shader also.
When they did their tutorial , they had FORGOTTEN to show also in the video how to construct up those material functions they were using. So you got the shader, but without the material functions also to go along with the parameters, the shader is useless.
The Material Node Actor Motion 4way Chaos that has Coordinates , Speed, Divisor, and Texture options which controls the speed of the water, mentioned in the tutorial also seems to be missing in 4.12. I can’t find it.
The grass shader works but the problem is the Blender mesh is not even rendering the shader. So soon as I put the grass shader material onto it the blender mesh turns all invisible, and doesn’t show any waving grass on
it at all.
Puzzled…because I don’t know the cause of this issue.
But The grass shader shows up perfectly fine on the Shape plane mesh BUT I can’t use it for panting for foliage because I have to assemble the 3 flat planes all together to create the 3d grass mesh, but Unreal exects me to have that 6 sided flat plane already assembled for importing as a single piece mesh, but the problem is the blender version doesn’t work to show the shader in the engine.
Nether can I select the 3 planes and just drag that selection of 3 planes off
to name it as a group selection to drag into the Foliage editor as the Foliage Editor only imports and names from the content browser (urghhh).
It don’t have group selection mode.
You can’t even select the meshes and then copy it to clipboard, and
go into foliage editor and paste as a asset, BECAUSE THERE"S NO PASTE.
And this is all thanks to Blender’s mesh not showing up in the shader.
That’s why this engine needs work, and Blunder (BLENDER) is hopeless
because of their big issues with Unreal and the unreal engine is also missing certain features, it has no Group Selection Mode to allow the dragging of meshes directly from the editor into the Foilage Editor for painting with. As far as I can see, It only wants to do the selection from inside the content browser.
I buy a royalty free license model from TS and the model’s hair piece don’t work in unreal. Because the artist didn’t supply a proper UV scalp Hair texture on the UV map so the model would be game ready and he won’t even fix the problem to make it all game ready ARGHH. This is not my day, So I can’t get the hair fixed as a result… So now I have no decent blonde hair texture now for my model to fix the problem with his head showing bald patches.
I’m sorry if the truth offends you. I have worked long enough in the industry to know that’s the attitude. As for “respecting games as a medium”: I have stopped doing that in the late 90s when games were reduced to sequels with a two digit squeal number behind the title, yes you’re right. I’m now working as an indy guy on my own small games and I have way more fun and take way more pride in what I do than when I was working for big companies.
I dont think an “inhouse engine” is even feasible today. I was working for a company in the mid 90s that did that, and back then things were way simpler in the 3D world. Yet we had constant trouble with the engine developer, bugs here and there that kept us from advancing the game, not to mention the diva-like attitude of that guy. Ok these were different days, you would not have a single man doing an engine these day s(although there are some out there which even sell for money which have a single developer) but in all earnest, this is not possible anymore today. The cost would simply be too high, plus the technical skill necessary is beyond any regular programmer.
I am really glad that UNreal and the other big engines have opened up for indy developers and as much as I hate bugs and bad documentation, the situation today is way better than it was 5 years ago with these 1-man-1000$ engines out there. Yes I ***** about flaws in the UE sometimes, but I also know that I could not make what I am making these days without it.
Well, Kojima Productions created the Fox engine specifically for Metal Gear Solid V.
Although its also used for other titles now, its a perfect example for an in house engine.
Absolutely correct. At the expense of flexibility.
If the Unreal Engine would work like the Fox engine, we would all develop with Unity 
Ok so the engine has its flaws, but it does the job eventually… it has strengths and weaknesses.
Right. I guess what this thread boils down to is that most of us wish that Epic would invest more manpower in fixing bugs rather than adding new features.
Not gonna happen, marketing announcements about bug fixes doesnt work; humans - its not that fun to dig into bugs, so devs lining towards rnd stuff naturally. You can see both those points it in action - VR Editor. There are not that many people do vr, market is not big, but yet, huge amount of resources going towards this, because its good for marketing (you can go to gdc and show everyone how you are and your vr development) and VR is new exciting stuff to work on.
Yep, gimmicks sell.
I can relate to the OP.
Right now I am dealing with this BS: Having issues with floors not getting same baked lighting as walls/ceilings - Rendering - Unreal Engine Forums
Since Epic advertises UE4 for mobile platforms and mobile VR, where lightmapping is a primary (and most of the times the only) lighting model, Lightmass should do the job 110% by now. Yet, it doesn’t. Ancient Quake 3 lightmapper does a better job than UE4’s Lightmass. And yet, Epic doesn’t care much. I think UE4 can be considered indie friendly only if indie dev working in the same alley as internal Epic’s projects (or in the same alley as AAA devs). Step left or right, and indie devs are screwed (unless said indie dev as indie as ID Software or Epic used to be 
)
I don’t mind discovering bugs and creating test cases for Epic to fix, if they fix it. After all, we don’t pay anything to use engine (and only pay 5% over certain amount if game/apps sells well). However, that’s not the case. It begins to feel like small indies are nuisance to engine devs, something that can easily be pushed aside. Especially when it comes to mobile / mobile VR for Oculus platform.
It just feels like such well established thing as lightmapping should have been resolved and stable by now. Something Quake 1 initially resolved in 1996 and q3map2 lightmapper for Quake 3 resolved in 2004 (it has indirect lighting, AO, deluxemaps, etc. almost everything UE4’s Lightmass has).
At this moment UE4 is great … for Windows and consoles only, only if you are using dynamic lighting and not mixing up indoors with outdoors, and only if your gameplay is in line with what engine already offers internally.
That is a pretty nice thread title you have there 
Thinking about it now, I’ve actually committed a good deal of effort reporting bugs/suggesting improvements, probably even more than I wanted to. Actually it took incomparably more time than three other engines I’ve worked with. I am not sure if it should be explained by number of flaws in UE4, nature of the work I was engaged in or simply because Unreal Engine is my favorite. I tend to believe that the latter is the case.
I mentioned somewhat similar in another thread. I have a feeling that UE4 is perfect for a small team with almost no resources and is even better for large projects due to ability to alter the engine as seen fit for specific project. Middle ground is missing in my view. Teams, that need slightly advanced stuff but can’t afford/ do not want to recompile the engine, might be suffering.
@
Strange… I’ve always had an impression that static lighting in UE4 is a strong point, not a weakness, contrary to dynamic lighting, which is lacking in terms of performance and functionality, in my view at least.
Yeah, that’s what I thought too until I had to mix indoors and outdoors :rolleyes:
It works and looks great for outdoors only. I haven’t tried making anything indoors only yet, but at least it looks like 4.11 is great for it too (judging by ArchViz section of the forum).
Now that I’ve wasted 2 weeks on getting indoor environment (placed in my outdoor level) get proper lighting (floors in particular; or perhaps any surface that faces upward), I can tell you that Lightmass in 4.12.5 is flawed. I am narrowing it down to Sky Light issues, but I’ve yet to test my theory today after work.
What upsets me the most is that in 2004 q3map2 map compiler/lightmaps baker (for Quake 3 maps) did everything Lightmass does today and it just works (whether it’s indoors, outdoors or mixed environments). 1 guys wrote it (based on original q3map compiler/baker with a few contributions made to the code through the years). Source code is open. Epic has a top notch team and yet baked lighting still has issues 
If you have the resources, you can absolutely create an in-house engine. DICE did it with Frostbite. Autodesk’s engine started out as in-house at a gamedev as well.
The cost equation is:
- How much does it cost to develop all the features we’ll want, and then debug and maintain the in-house engine, and keep adding features we want in the future?
- The learning cost is small there, because it’s amoritzed along the development cost. You just better hope that the people who know the engine don’t quit!
- How much does it cost to learn a third party engine, and how much additional tuning do you need to do on top?
- Can you hire already-knowledgeable developers on the third party engine? Will new features be developed for you? Will those be the features you want?
- What is the fiscal cost of the third party engine?
It’s not always a slam dunk one way or the other. Some games may have special gameplay that doesn’t match the engines available. RTS-es want lock-step networking, which doesn’t match Unreal Engine at all. GTA series uses large streaming level technology and editors that are custom and no engine were capable of that (at least at the time.) Or you expect very large sales, and you have a good technical team, and the cost of that team is smaller than the royalties you’d have to pay on your large sales.
But, if you’re a handful of guys and girls in a garage, and you want something as basic as an art path that allows skinned characters and animations to be exported from Max or Maya, chances are you don’t want to roll your own. And, even if you’re a big studio, if you’d rather tell stories and ship games than feed a substantial in-house engine team, again, something third party is likely the right choice.
That doesn’t mean the third party option is flawless – it just means it has a lower overall cost than the roll-your-own option, all told.
Like with Sacred 2, to do a huge seamless world they had to develop THEIR OWN ENGINE because the other engines couldn’t do it.
Things were different back then, you had to pay insane money to get access to source of other engines.
Now developers can look at source UE4 or CE and decide is it feasible to extend current engine or better to create your own
This says more about you than about the industry as whole. I work at a big studio and everyone here takes pride in what we do.
As to UE…I see so many people getting upset because they upgraded the core software of their project and things broke. This is what happens when you change major software dependencies mid-cycle! Don’t do it!
so we shouldn’t upgrade the engine when a new version keeps coming out?