I love the engine but why is it so bare bones?

General Feedback & Requests
47

I just want Networking stuff ESPECIALLY in VR to work out of the box… among other things, I should NOT have to use community plugins (that have been around for 6+ months) to fixes bugs currently in the engine.

That’s my only gripe really… there are SO MANY PRs… but Epic is not really doing much about them.

Most devs are either at the mercy of the community for things, when it DOES come to engine fixes/workarounds, which is fine if you can use community tools in your lineup, but then the problem becomes being TIED to that community member(s) contributions, and if they stop dev for it at a certain point and no one picks it up again… well now, you’re back at square one again.

As for framework additions…I rather Epic takes care of this PR problems, prioritizing BUG FIXES instead of adding features. We are already freaking feature heavy ALL OVER THE PLACE. Except AR… but that’s probably coming sooner or later and will have its own headaches.

TL;DR: Epic needs to prioritize fixing long standing bugs (especially when the fixes are out in the wild and they just need to merge them in). The community should NOT be the ones doing all the work. Sure, the engine is open-sourced, but I think they are making that their crutch. The expected result right now is… if something is f*cked up in the engine, and Epic does NOT want to fix it right away, they are expecting the community to do so… or they are just expecting the community to so from the start while they move on adding silly useless things like VR Editor.

****, that wasn’t really a TL;DR. But I suppose I had a lot on my mind and this thread kind of let me vent out what I’ve been thinking lately.

FIX VR Replication issues. VR was supposed to be this big deal, well, make it RIGHT first before moving on to other things… don’t just let it stay broken.

Alright, I think I am done now, the whole VR Replication issue is silly. Oh, now I am just repeating myself, peace out.

48

Hey guys!

I’m moving this thread to “Feedback” section of the forums so it has better chances to be seen by Epic.

49

I can certainly understand this but we certainly shouldnt have to reinvent the wheel so much, my thought is that we use third party engines so we dont have to build so much ourselves. Obviously for Epic the Engine is highest priority and thats why I feel this is one area the community could step in and give UE4 a real competitive advantage over Unity which does require alot of rework or outside packages.

I can certainly see where youre coming from, this doesnt sound to me like your typical “I want Epic to make my game or pipeline tools for me” deal and thats why I felt the need to chime in. Water is actually a great example, see Epic might not have a game with water vehicles as an example but for someone that does its a huge task to attempt in the engine and while it does provide the basics there is not alot outside of that, infact what it does provide in the way of water functionality has been in Unreal Engine iterations for a long while and isnt really up to snuff when it comes to modern games and techniques.

Now this isnt to say we need the most extensive water system known to man, it just means a nice solid framework to build from would be great, if everything in the engine had the appropriate hooks for dealing with such a system. This type of thing is best left to the engine level modification and not has the ability to dive that deep, even I myself avoid it where ever possible even though my programming ability is reasonably decent just because of the complexities it adds to the overall development process. In terms of extending such a system to really flesh it out for whatever project I require it for thats no where near as daunting but not having that base means I have to do all the system design and it adds ontop of all the other systems I have to design like basic UI, HUD, Inventory, Weapons, the list goes on. Thats not outside of the range of my abilities and I understand why there isnt a one size fits all approach to those systems but I do think it would be beneficial to have more in the way of game framework in there and not just in templates/examples.

Yeah I think my feelings on this are reasonably well known for those that know me, I will say that Im not overly happy with the system design in this game because it feels too much like legacy port and not designed specific for UE4. The UI system isnt very modular either so its not exactly easy to patch onto, there is also the fact that you have to make whatever you want for UT as you cant bring that code outside of that project.

The issue I have using systems not designed with overall engine scope and modularity in mind is that the time needed to modify them for purpose or integrate them becomes almost as large of a task than writing (or rewriting) them myself. This is why I do agree that some lean systems can use, including marketplace content creators, would be great for getting on a similar page, typically I find design by democracy can be a bad thing and that good system design and code in general is better off as a dictatorship and since Im using Unreal Engine Im already committed in part to The Epic Way ™.

50

I am 50/50 on this. I am with the Framework != Engine. I have a framework (currently rebuilding) that ‘Adds’ features/functions to the game engine base. I will say that UE4 has far more already in place that Unity 5 does not in this regard. There are some things in these new generation game engines that baffle me why they are not actually part of the engine itself (NWN Toolset and Skyrim Modding functions). I could make the list here but this is not really the place to do it.

I think the OP/others is/are spot on for for some of this but off on a bit…hence the 50/50. Examples like Inventories with Drag and Drop and Clickable Events seem to be a marketplace item where this ‘should be’ integrated into the engine since almost designing a game uses this. Indeed the art is different in every game but linking the designed art to the system is a simple matter best left to the developer. These type of systems could be from the community as engine plugins that multiple community developers can get on one page and design ONE nice utility plugin.

The money value in the marketplace is sometimes a greed play but personally I get it…“Time is Money”. There is so much pervasive copy, paste, and change in the Unity AS that it makes one sick up looking at it all. It is refreshing to see that the UE4 Marketplace is not the same…but from the beginning of it until now you can see the replication of ideas being added. Competition is a good thing but at some point the egos between the ideas is obvious. When there are a few ideas that are so similar in function…Day/Night Cycle, Weather, and Water anyone??..it would be really cool if these idea makers simply decided to combine what has been made into one package. Charge for it and split the profits between parties…I would pay for this type of plugin. In reality I already do and have to figure out how to ‘Merge’ these ideas…and in many cases simply learn from the experience and roll my own system that is loosely based upon these merged ideas.

And now…that I have a fully functioning merged system that works and there is no need for these other ‘Like Ideas’…here is the big question…should I sell this fully integrated system as a plugin on the UE4 Marketplace or give it away for all to use in Epic’s engine? Especially when something 95% of developers will use…Day/Night Cycle, Weather, and Interactive Really Good Water (Epic does have nice water but it just for looks and takes a lot of work to get it working for interactive-ness)…Base Inventory and Item System…as examples. Developers can extend these if they wish to of course because full source is given. From what I read on the postings in this topic I think this is what peeps are really asking.

I just wish more people would get off the driven wagons and come together for some collaborative plugins…like the Weather and Ocean Shader project is doing but this project needs something like Orbit Weather and Seasons to make it complete…integration issues anyone? These are exactly the the type of systems Epic really should gather the parties together for the benefit of the entire community…make it a paid plugin that is fully QC tested with each engine version change. Developers can then choose to add this feature or roll their own. I am on-board with that. Just keep this UE4 Marketplace clean and professional level (it still is mainly) and integrate where really needed.

Cheers
O

51

From the 1st day that I stepped into UE4, I have never once thought that it was bare-bones, nor missing crucial features (Though I can name several idea-breaking bugs that have not been fixed for over a year). I agree with Deathrey when they say that those sound like game features, not engine features.

52

You are correct in pointing out that UMG already handles the buttons, menus etc., I was simply trying to point you to my thoughts on the situation. I think the overall consensus is that having a system in place that we can edit to suit our own needs would simplify things a lot for.

53

Thanks . I wasn’t quite sure where to put it to be honest :stuck_out_tongue:

54

I completely agree. That would be EPIC if Epic actually made the stand out plugins official. Granted, I’m not sure how they’d sort out any kind of monetization for the devs but it would certainly be nice. Official paid plugins that you know aren’t going to stop being supported or stop functioning at a later date would be really nice.

55

It’s all a matter of perspective and skill level. Maybe I’m using the wrong terminology by stating that these should be ENGINE features as its been pointed out twice now that framework is not equal to engine. My point is that regardless of how these features are integrated, they should be available to and they should be built by Epic and not put behind a paywall. If it’s functionality that the majority of projects would benefit from then it should be made available to as part of the engine as a whole whether it’s built into the engine or simply an add on that you can choose to add to your project or not. A proper multiplayer/networking system should come straight out of the box. A simple way to hook your UMG built menu system into a framework for key binding and adjusting graphic features should as well. The other features such as landscape materials and a water system might be a little more user specific, but again it would be nice to see Epic taking care of these as well, at least to a point where we’ve got something that we can easily modify to suit our specific needs.

56

I should say that something for water would definitely bring the engine up a step, and I suppose you’re right that some of these things would benefit a majority of the projects.

57

Thank you yes, that’s my point entirely. I’m not trying to say that Epic isn’t doing a great job because they are. I’m just trying to bring some attention to issues that I’ve noticed over the last few years and things that I personally find a bit frustrating. I know in time the engine is going to work out to be even more amazing than it is already.

58

There are definitely areas that feel bare bones and less developed compared to other parts of the engine. Working with Tiled Landscapes for example, as well as the incomplete Dynamic GI options. I was so annoyed when OTOY announced they were bringing a realtime lightmap baking solution to Unity and not Unreal. Anyone looking to push VR visuals is likely going to use baked lighting, but without a means updating the lightmaps in realtime you can not support fully dynamic scenes at that quality. Eventually hardware will catch up, but you’ll also want to push for higher resolutions and so on.

59

One example I like to use is how do you draw a simple static straight line in the engine? There are ways to do it with debug or traces but there is no real legitimate way to draw poly lines from BP, the arrow component could work if the scale wasnt uniform and it had a proper length. Its clear to me there are some areas which Epic doesnt explore, collision is a prime one Ive had issues with as being too bare bones (lack of general purpose standard primitives), the mass variable is another which doesnt seem consistent, the thrusters are finicky and forces only seem to work in with a collision component as root which severely limits usage and expansibility. Small fixes to the systems could make them much more general purpose and useful across the board so I do think some extra effort should be made to test interoperability instead of in a vaccuum, some of these issues you dont even really notice if youre making small jam style games and its not until you get into production of a larger project you see the engines short comings :cool:

Oh the limited Atomospherics and Fogging too is very problematic to the point some scenes take far more setup than is reasonably necessary because of the targeted nature of those devices and other scenes (typically the easier to setup) have the time savings.

60

Quoting for truth. There are quite some things that should be in the engine but aren’t.

61

While I agree that an engine should give you tools to build stuff with, not make assumptions about your use-case, I must absolutely disagree about UMG.

UMG is one of the reasons for me leaving Unreal Engine behind. It’s horrendously inflexible if you intend to do anything that isn’t covered by the preset widgets.

I managed to do what I wanted to do in three days in the primitve canvas HUD, while I was stuck on UMG for three ****ing months with that task. And it wasn’t anything too complicated, just a basic 2D inventory with drag and drop and items that can take up more than one space - even the drag and drop portion was easier, once I found a single comment in some thread about the fact that the mouse cursor is actually set in the character controller. Yes, it’s technically possible to do in UMG, but not in a way that is properly procedural. I’m working alone, so manually setting the interaction for every shape of an item and inventory size change is not my idea of spending my development time wisely.
Then I got hit by the two year old bug that prevents struct arrays from properly working and had to fiddle around with incredibly obtuse actor hierarchies to find out how to make an inventory.
And then I got hit by another few idiotic bugs/oversights/lack of documentation in C++ and just decided to switch to Unity. Not because it will be better, but because it will at least have more support by the community.
It took me more time than is reasonable to find out how C++ function parameters are parsed into input/output nodes, that templates do not work with blueprinted functions and a bunch of other stuff that could be explained with a few lines if there was any decent documentation.

Anyway, UE really needs something that sits between UMG and basic canvas work. Theoretically Slate fits the bill, but good luck finding out how to do anything in Slate with the resources that are available.
Ironically, it seems that getting widget positions is now on the road map. If I recall correctly, that’s all that I would have needed to get my UI to work in UMG.

62

I would love to see more features being added, but then more in a way that the possibilities with Blueprints are expanded. The idea of Blueprint to to make game development possible for, however as soon your project is a little more ambitions that the ‘default’ game, your quick to hit walls. Walls that can only be broken with the power of C++.

Other then that, everything that UE4 offers out of the box are enough to form a proper foundation for any project imaginable. If you want to have out-of-the-box game mechanic system, go make them yourself and recycle them with each project. It will save you money and it’s something indies can easily do. If something is out of your scope, simply ask help or search documentations and tutorials. If you want to save time, well… Time is money.

63

This all seems like stuff I work with on a regular basis. I have worked extensively in UMG and haven’t seemed to have such headache.

64

I’m personally a bit worried about packing the engine with new features, rather than patching up long reported issues. I never felt like that engine is missing something crucial. I mean I have a ton of things that would be “good to have” in mind, but they are not crucial by any means and most of them, openly speaking, are project specific and can be handled by the team. 4.14 introduced more issues than it fixed, as an example.

65

The majority of new features are pull requests from the community itself.
Most of the time Epic internals focus on bug fixes… So maybe you guys should go over GitHub and riot there telling people to stop submitting new things before the engine becomes stable.

66

Actually my line of thinking is that if we get useful features implemented now it means time later (keep in mind Ive been saying this since beta) can be better spent on bug fixing, what happens now is that new features are added in sporadically and it doesnt really help stability since there are new features added pretty regularly. In all honesty most of those new features arnt even that useful to me and in terms of plugins are most definitely things that are more easily handled by the game team than engine modifications no matter how simplistic.