UE4 Roadmap

[SteveElbows]

Lastly, let's try and keep the harassment of each other to a minimum... We're all developers here trying to make great things with the toolkit and many of us are passionate about different pieces of technology, features and markets, but this thread should remain focused on the roadmap itself. Thanks!

Three cheers to this. Shame it really isnt happening in this thread at the moment. There is negativity going on that ends up being terribly unproductive and demoralising even when some important and sensible points are being made by the same people doing the ranting. The venom and histrionic language may enable some people let off stream and gadfly stuff does have a place in the world. But taken to these extremes the main result has been a fair chunk of humanity becoming wary of internet forums and I dont think it really increases the chances that 'the community' will have greater influence over the development of UE. Lots of people making similar negative points is more likely to affect change than a few people making the same points repeatedly and insulting those who disagree.

[Maximum-Dev]

It's funny to see you comparing UE4 with Frostbite. Go get Frostbite and start developing with it. Oh wait, you can't? (especially for free and then for 5% share)

Cryengine sucks. Yes, it has tools UE4 doesn't have, but other than that support is worse, documentation is worse, community is scarce. I don't even recall any decent games that just work out of the box made with Cryengine besides games made by Crytek themselves (and even with those there are a ton of issues).

No. It's not funny!
Cryengine support is worse, community is scarce, documentation is bad. How is that any valid reason for Epic to not listen to a single feedback from community regarding their engine tools? Whether indies are using cryengine for their games or not has nothing to do with Epic insisting on not go for the tools that's been implemented in other engines since +5 years ago to ease the development. You're derailing the discussion.

Unity started a long, long time ago and that's all they work on - engine that runs on virtually everything and supports anything (supports so much hated here Windows Store and supports Facebook). And if you go commercial not only you have to pay for the engine as subscription, you also pay separately for support. Unity doesn't make games, unlike Epic, and Unity is a larger company.

Not sure how that's related to anything here.

Right now UE4 has everything one needs to make a good game. AAA devs can implement whatever they needs (they have money and manpower) and indies can make indie-level games with UE4 as-is. Look at the examples of successful indie titles. None uses real-time GI or massive open worlds (although No Man's Lie .. err Sky does, but they made their own engine and the game still sucked at the end).

Now that's funny. If AAA devs should implement what they need, and indies should use it as is, what's the point of the 5% royalty and claiming it's because they care about what we need? what's the point of "Feedback & Suggestions" forum? What's the point of mentioning we love to hear feedback and suggestions? we're making an open world game. World composition ever since it was introduced is swamped with bugs. Absence of virtual texturing doesn't let us have more than 4 layers on landscape due to massive performance drop. Grass tool is bugged, landscape tessellation is bugged etc. etc. How'd you know anything about it while you're not working on an open world project?

Also, VR market is new, unsaturated. A new beginning for many of indies. Conventional market is super crowded and saturated, so it's impossible to make living without massive PR, which is expensive or publisher that takes away almost all proceeds.

That's great! let's kill the current generation, slow down the development of hundreds of games so our kids can enjoy VR games. VR until the next decade will still only generate a fraction of the profit non-VR games generate. Excellent strategy.

It's just same old crowd moans and ******#s here about missing features and such instead of working on their games with what engine already offers and learning ins and out of budding new technology.

Always it has to end up with some sort of insult towards those who want some improvements... I didn't read your post entirely at first, if I had read this line I'd have refused replying and just kept what [MENTION=754069]emeraldcoast[/MENTION] said about it. You sir, are certainly walking up the steps to success right now by attaching a "******#s" label to anyone who is criticizing the engine. You might get a promotion.

[Ivey]

There are three things on the roadmap and none are the one that should actually be there. "Fix the bugs"

Bug fixing is something we continuously do release over release. We ccould add a card for it, but we're going to be continually fixing hundreds of bugs in every release as we do now.

Every release since around 4.12 the engine has gotten wore and worse in terms of stability. "Blueprint Weirdness Syndrome" now strikes in nearly every project. Blueprint structs are essentially time-bombs. Steam voice chat was busted in 4.15. The nonsensical "***** has to have 'Simulate Physics' enabled if you'd like to AddImpulse" message still spams the logs. I still can't spawn objects in a c++ constructor like I could in older versions. In recent months Answer Hub staff have gotten less willing to try to repro anything or even acknowledge a problem. It takes two months of people complaining to get a bug report entered.

I'd love to know more about the issues that you've outlined above. Can you link me to any reports in PM? I'd really appreciate it, thanks!

To be honest, communication with Epic staff was great during the first couple of years after UE4 released, but now meaningful responses almost never occur. Nearly every staff post starts with some kind of platitude and glosses over the real issues.

I apologize if we've let you down in terms of communication recently. We take this sort of matter seriously and will do our best to improve an open line with the community.

I feel I should mention I nearly exclusively develop for VR now. At all 3 companies I do long-term work for, I am a VR dev. PLEASE STOP SHOVING VR GIMMICKS INTO THE ENGINE. VR Editor will never not be a gimmick. I don't want to stand up and strap on the headset while I am working. Your engine is becoming nearly unusable due to major bugs that aren't being fixed while you completely over-focus on VR.

Well... that's entire point of the discussion. You are all in over VR. Which looks like people are tired of.

As mentioned before, VR is only one of many avenues in which we provide valuable new features to developers, and we're doing a ton of non-VR specific work for people making non-VR specific projects. Take a look at what's on the way with 4.16 to see what's in store for animation, physics, platform support, rendering, asset management, audio and more.

If you want to put something more specific on the roadmap, then "Placeholder-class catastrophe" Error This blueprint (self) is not a PLACEHOLDER-CLASS https://answers.unrealengine.com/que...ment-4112.html and "All my child blueprints reset" https://forums.unrealengine.com/show...ariables-Reset are good places to start.

This is an issue that the team is aware of and looking in to. The roadmap has always been a place for upcoming features, not necessarily for the showing the status of bugs. Bug status is currently tracked over at issues.unrealengine.com, but we'll discuss ways to better surface information on them if this is inadequate for the success of developers. Let us know!

Is it possible that some people actually use the VR Editor? Probably... Why would they make it if there was no demand for it?

I know not a single dev that uses it. I can't imagine any scenario it would be useful for. The only explanation I can speculate is that a higher-up at Epic ordered it and refuses to acknowledge its a waste of resources. I would use Visual Studio: VR Edition before I would use UE in VR.

Teams are, in fact, using it! For building worlds in VR, it's a great way to quickly get a sense of scope and scale. While it's certainly not for everyone (especially those not working in VR at all ), developers do see value in the tool.

01. Okay. What's on the roadmap for 4.17?

The current items that we know are going to make 4.17 are listed on the board. It's short today, but this is just the beginning. Expect more to show up there are we're confident with be in the release later in the summer / fall.

02. To start Epic needs to stop playing hide and seek with upcoming features. I know a major part of game industry is surprising the audience, but there's a difference between the consumer buying the games and those in the development community who need years of time to actually develop their games with your engine for whom your tools are failing. The culture of secrecy needs to stop and we need to work together

How does keeping more behind scenes help anyone ?

Secrecy is absolutely not the focus of this and I apologize that we allowed things to be taken that way. The older roadmap was setting developers up for failure in a number of ways.

I am sorry that our tools are not meeting your expectations. I encourage you to post your feedback about Unreal Engine 4 on the parent Feedback for Epic forum. In the meantime, I'll investigate what we can and can't share due to potential volatility in our development schedule.

It is as possible as winning in lottery I guess.
My answer as to why such lets be honest useless for majority of developers features are developed.
To have something to show on GDC. Remember unveiling UE4 ? Yeah me too. Remember Kite Demo, next year (that was big and really got lots of people excited).
Do you remember what was on GDC after that ? Me neither. There was something about VR, but entire GDC was about how VR is the future.
And this year GDC, was pretty much the same as last year. The only thing I remember from it are physics and animation, audio stuff and thing that Ryan did (well it's better than last year..).

Both GDC were more focused on enterprise than core game developers.

The Kite demo was an incredibly exciting project for us to work on. Unreal Engine 4 was, in many ways, in its infancy in 2014, and Kite gave us an excellent opportunity to explore our tools and build out enormous systems and features that the engine had not yet refined (or had at all!). Between Kite and Theif in the Shadows, we were able to bring forth a lot of the systems and features that developers use today in their projects every day.

GDC 2016 saw us finalize a number of rendering features put in place for Paragon such as realistic cloth, hair, skin and other character specific items, and we were also able to take Sequencer out of the oven and put it to use with a number of partners at the show, creating a path for many to build out cinematics for their games. Our work with McLaren that was revealed at GDC brought forth a number of new options for making realistic materials and allowed us to release the automotive material pack for free based on that work. For VR, we were happy to show off the VR Editor, Bullet Train and Trials on Tatooine, as well as a ton of other partner content in the VR lounge in support of the new VRDC portion of GDC.

This past GDC saw the release of Robo Recall and its full source. What's super cool to me about this project is that it helped us harden a number of non-VR specific features such as Blueprint Nativization (which we recently covered on Twitch) which provides an avenue ton gain significant performance, the hardening of out-of-the-box mod support (check this thread for more info) for building expandable games and provided a robust project example for developers to see our approach (grab the entire project from the launcher) to many aspects of building games. We were also extremely glad to share the animation, physics and audio work in our tech talks (those that you mentioned, a lot of which is coming in 4.16), as well as showing off platform support for the all-new Nintendo Switch.

Regarding enterprise applications, we are definitely seeing more and more uses of Unreal Engine for applications outside of games. It's really awesome to us all that creators of all types are able to bring ideas to life with our technology. Rest assured, we're going to continue building out leading game technology for our licensees and our own titles. Our commitment to game developers is strong and will remain that way

But even worse, this year features trailer for engine, was almost the same as last year. There was no bomb drop like Kite Demo since two years!

The features trailer speaks to all parts of the engine, whether the features are new or not. Kite was the first time we got to build out a solid features trailer, and of course we're anxious to see it evolve over time with you all

I'm personally interested in few things:
1. What happened to arbitrary gravity direction ? There is pull request sitting for like year or two now..
2. When Landscape tools get some love (GPU painting, more advanced materials out of box).
3. Support for dynamic in editor level streaming (loading levels by hand is PAIN).
4. World composition tools improvements and origin rebasing inside editor (also without manual intervention).

Thank you providing these concrete examples. I'll look into what I can find out here.

What is going on with things like that:
https://twitter.com/BrianKaris/statu...72258639740928
These are really interesting things.

This is more technical approach than a feature in itself. I'll see if Brian can write up something for the blog when it's in a place to talk about!

I still think there should be some voting system for community request features like I mentioned previously. Gathering votes over period of few months is not representative. Giving everyone unlimited amount of votes is also not representative.
How I would see it working:
1. Gather feature requests from forums/answer hub (which you guys know you can implement, but don't know how many people really want them).
2. Create pool, and give everyone like 5 votes.
3. Make pool available for two weeks. Let everyone vote over this period.
4. Reset pool and let vote everyone again.


See which items received most votes every time, to see what is most trending. And then implement those features.
Repeat.
I think it would be most fair. There won't be overblow of possible features, everyone will clearly see what will get implemented, people get features they want, and Epic get props for listening to community. Everyone is happy

We will continue to investigate methods in which to gather collective interest in new features. The voting on Trello was failing both us and the community for a number of reasons, but we do recognize the value in mechanisms like voting.

Arbitrary gravity directions has almost the most votes on the roadmap for years now.
It would be one of the most important and beneficial features to implement in terms of player movement in game.
Many devs in this forum argue for years now for that feature to happen and it even got a concret release date 2 years ago just to get pulled back into the backlock 2 month after the post of the concret release date,because VR seemed more important than a good landscape/level editor or arbitrary gravity options.

https://forums.unrealengine.com/show...for-Characters
https://answers.unrealengine.com/que...r-charact.html

How is this behavior listening to the communitys needs ?

The inability to accurately provide a timeline for features such as this is was failing the development community in a number of ways and is a core reason we've made the change. While features like arbitrary gravity is something that we'd like to add at some point, we don't have a clear path to get it on our roadmap and won't until it gets slated for a release. Once a feature is slated for a release, then it will be added to the roadmap as opposed to everything living there for an undefined amount of time.

We do not have a say in what's going to be in the engine, @Chance Ivey . Because most of our bug reports/feature requests from last year are still backlogged, unless it's been something that perfectly aligns with Epic's internal needs. We were all trying to get a little more attention to backlogged items, but everything got wiped from existence.
Same as when people were asking for Marketplace improvements, they got a 24 hour delay on the sales report instead, and staff left the Discord channel, cutting off the little remaining communication.

That simply isn't true. I have many threads in the feedback section, most features being what comes with other engines right out of the box because they're absolutely critical to majority of developers, but there is no instance where we asked for a similar tool and it was added.

https://forums.unrealengine.com/show...Lighting-Units Frostbite/Cryengine have it.
https://forums.unrealengine.com/show...-Spline-Decals Frostbite/Cryengine have it.
https://forums.unrealengine.com/show...mbient-Cubemap Frostbite/Cryengine have it.
https://forums.unrealengine.com/show...h-pen-pressure Frostbite has it.
https://forums.unrealengine.com/show...scape-painting Cryengine has it.
https://forums.unrealengine.com/show...formance-issue Frostbite/Cryengine have virtual texturing.
https://forums.unrealengine.com/show...s-more-control Cryneigne/Unity3D marketplaces value sellers much more.
https://forums.unrealengine.com/show...rketplace-Team ...
https://issues.unrealengine.com/issue/UE-35097 Untouched bug for 8 months with 40 votes.
https://answers.unrealengine.com/que...ouldnt-be.html This was marked as solved 3 times in a row last night, while the problem is there and very obvious.
https://issues.unrealengine.com/issue/UE-43093 Probably soon to be backlogged bug report.
https://forums.unrealengine.com/show...natives/page11 There are many threads regarding RT-GI. Community wants it. Frostbite has Enlighten/Unity has something/Cryengine has SVOGI, Epic removed SVOGI.
...

That's examples I pulled off from my side only. Absence of those features have added tremendous amount of labor to our team.
How much hope one should have in order to stay motivated and continue giving feedback/make requests.
Maybe ask the CEO to read through such threads sometimes before talking about community.

Thanks for providing concrete examples of community feature requests. While I cannot speak directly to why some of these will or won't make it into the engine or how some of these decisions were made, I think AlexanderP outlined how we approach feature implementation in one of those threads:

• "1. Community interest: Threads in the Feedback for Epic section, AnswerHub reports, social media posts and votes on the issues page help determine that. Please keep giving feedback and make sure to specifically explain how you would expect the tools to work for the best results. The mesh editor is an example of a tool made from these kinds of requests.
• 2. Large audience need: Generic tools that help out a large audience are prioritized highly. Sequencer, Niagara, Audio Engine and other tool overhauls fall into this category. This is probably the highest priority in general since it helps so many people.
• 3. Epic Games project need: One of Epic's projects needed it, so it was ported over from the game into a generic engine tool. AnimDynamics, VR Forward Renderer and Behavior Trees were created for this. These get added a lot since the tool is already done and it's just a matter of porting and testing."

While we cannot guarantee every request that comes across makes it into the engine (as much as we, too, want all the things), I can say that your asks do help us plan and build out important pieces of / improvements to Unreal Engine.


Thanks everyone for your comments and concerns. We're continually looking for ways to improve our offerings with not only the engine itself but all of the supporting products alongside it, so feedback like this is meaningful.

Also I'd like to ask again for debate over things outside of the Unreal Engine roadmap to continue elsewhere. Let's also continue to keep the harassment of eachother out of these forums as well. Thanks!

[ZenDavis]

The current items that we know are going to make 4.17 are listed on the board. It's short today, but this is just the beginning. Expect more to show up there are we're confident with be in the release later in the summer / fall.

Thanks for the detailed response. Instead of waiting for features to simply "show up", why don't we get that conversation going now? Lets figure out what your engineers want to work on and what the community wants and find as much middle ground together. Features "showing up" on the road map plays into the whole culture of secrecy mentioned before. PLEASE let us make you money. Talk to us.

If I was king for a day at Epic I would have 4.17 be a bug fix release only... Reviewing the backlogged issues with the most votes and knocking those out. Most people would be OK with a feature freeze for a single release...?? I would think.

teak

+1 For Feature Freeze - Full Bug Fix Release!

[Gunschlinger]

There is this ancient bug in ue4 bp that basicly reset all variables to the default state in a child blueprint. Its been reported so many times and yet nobody seems to bother fixing it. This guy even made a internet survey to gather data about the bug :https://twitter.com/joewintergreen/s...97563108261889

[SteveElbows]

I would support a release that had bugfixes as its main focus, but I wouldnt go as far as a full feature freeze. Partly because I suspect that some epic teams (or however they organise) are going to have much less to contribute towards bugfixing for a multitude of reasons, and also the state of some systems that are being rewritten may make it advantageous to press on with finishing the new system that will replace the old one rather than try to fix some longstanding issues with the old one. As of right now the work on the new Audio Engine springs to mind as an obvious example of this and since it is experimental in 4.16 and hoping to be ready for 4.17.

[BlackFangTech]

For building worlds in VR, it's a great way to quickly get a sense of scope and scale.

So its specifically for level designers that don't like play in editor?

[emeraldcoast]

The inability to accurately provide a timeline for features such as this is was failing the development community in a number of ways and is a core reason we've made the change. While features like arbitrary gravity is something that we'd like to add at some point, we don't have a clear path to get it on our roadmap and won't until it gets slated for a release. Once a feature is slated for a release, then it will be added to the roadmap as opposed to everything living there for an undefined amount of time.

Especially the abitrary gravity setting feature request is out there for more than three years now and its one of the most requested ones also.
Two years ago it got even a concrete release date,so the ue4 devs must have acknowledged how important this feature is for many.

But then it didn't came out on the promised release date and got backlocked till unknown timeframe.
You had 3 years time developing it , how much longer should we wait in your opinion?
Another 3 years so that it adds up to 6 years total.Thats just not a reasonable timeframe for anyone to plan a game with feature x,y.
Especially when you gave it a fixed releasedate two years ago and then out of nowhere just canceled the release.
How should anyone plan a game with a communication like this.

And im not only talking about the abitrary gravity feature.
There are also:

-a more powerful landscape/vegetation editor with a in game terrain editing toolset which does not drop down to 10 frames per second when you layer multiple materials on top of each other (You implement the in game editing for your VR costumers with the VR editor right now but for the rest of the 99% of the non VR related playerbase you refuse to implement a feature like that for years now)

You still haven't answered this one:"How can you justify working on the VR realtime editor feature when 99% of your customers who do not use or develop for VR at all don't even have a real time terrain editing mode in the normal landscape editor?"Should it not be the other way around that you first give the 99% this important feature and then convert it later to VR for the under 1% of your VR userbase?

-a more powerful level editor (You cant build a level editor around the premise of importing finished assets from other programs and then "blocking" a level out with those finished assets just because the geometry editing tools in the ue4 editor are so lackluster that you cant even block out a level fast with meshes in editor to test out if distances between your buildings are the way you want them to be.)

A powerful geometry editor for blocking a level out is one of the basic must have toolboxes a level editor needs to have.
That's like selling a car without a motor like you see them rolling around in the tv show the flinstones:





How long do you expect people to sit and wait till important core features, which are implemented in other engines like unity for years now ,finally come out for unreal?
While you bring out one gimmicky feature after the other to satisfy your "strategic partners" like facebook and reduce your communication with your community month after month?

[Maximum-Dev]

While you bring out one gimmicky feature after the other to satisfy your "strategic partners"...

Isn't that what every company is doing now...
"Community" is a thing from previous couple decades, it's not about communities anymore it's about personal interests. Having a community is just a tradition that's slowly going away more and more each year.
I think having a community these days plays the role of free advertising that's all.

Edit: Yeah that was off topic sorry. Gnight.

[teak421]

Especially the abitrary gravity setting feature request is out there for more than three years now and its one of the most requested ones also.
Two years ago it got even a concrete release date,so the ue4 devs must have acknowledged how important this feature is for many.

But then it didn't came out on the promised release date and got backlocked till unknown timeframe.
You had 3 years time developing it , how much longer should we wait in your opinion?
Another 3 years so that it adds up to 6 years total.Thats just not a reasonable timeframe for anyone to plan a game with feature x,y.
Especially when you gave it a fixed releasedate two years ago and then out of nowhere just canceled the release.
How should anyone plan a game with a communication like this.

And im not only talking about the abitrary gravity feature.
There are also:

-a more powerful landscape/vegetation editor with a in game terrain editing toolset which does not drop down to 10 frames when you layer multiple materials on top of each other (You implement the in game editing for your VR costumers with the VR editor right now but for the rest of the 99% of the non VR related playerbase you refuse to implement a feature like that for years now)

You still haven't answered this one:"How can you justify working on the VR realtime editor feature when 99% of your customers who do not use or develop for VR at all don't even have a real time terrain editing mode in the normal landscape editor?"Should it not be the other way around that you first give the 99% this important feature and then convert it later to VR for the under 1% of your VR userbase?

-a more powerful level editor (You cant build a level editor around the premise of importing finished assets from other programs and then "blocking" a level out with those finished assets just because the geometry editing tools in the ue4 editor are so lackluster that you cant even block out a level fast with meshes in editor to test out if distances between your buildings are the way you want them to be.)

A powerful geometry editor for blocking a level out is one of the basic must have toolboxes a level editor needs to have.
That's like selling a car without a motor like you see them rolling around in the tv show the flinstones:





How long do you expect people to sit and wait till important core features, which are implemented in other engines like unity for years now ,finally come out for unreal?
While you bring out one gimmicky feature after the other to satisfy your "strategic partners" like facebook and reduce your communication with your community month after month?

I think there is nothing that the folks at Epic can say that would alleviate your concerns... Maybe UE4 is not the engine for you?

teak

[emeraldcoast]

I think there is nothing that the folks at Epic can say that would alleviate your concerns... Maybe UE4 is not the engine for you?

teak

Sure they can!

They can implement a usefull landscape editor and a usefull level editor which deservs this name ,like other engines have for years now and then im happy even when they dont implement the gravity direction feature.

I dont think thats too much to ask for.

I cant remember that i was too cryptic in my description.
I clearly listed all the points which i find critic worthy over multiple posts.
So whats your argument?

[teak421]

Sure they can!

They can implement a usefull landscabe editor and a usefull level editor which deservs this name ,like other engines have for years now and then im happy even when the dont implement the gravity direction feature.

I cant remember that i was too cryptic in my description.
I clearly listed all the points which i find critic worthy over multiple posts.
So whats your argument?

I'm not making an argument...just an observation. Chance has addressed your concerns and you are not satisfied. Maybe use an engine that has important features that you listed above?

teak

[BlackFangTech]

I doubt they will finish arbitrary gravity. Maybe I will get my plugin for it out soon, but its more complicated than you think. Character movement is easy. But now you have to rewrite the navmesh system as well. Want to sculpt a ring-world? Well, landscape doesn't do anything but planes.

[emeraldcoast]

I'm not making an argument...just an observation. Chance has addressed your concerns and you are not satisfied. Maybe use an engine that has important features that you listed above?

teak

He has not adressed any of my concerns at all and others feel the same as i do.

[emeraldcoast]

I think there is nothing that the folks at Epic can say that would alleviate your concerns... Maybe UE4 is not the engine for you?

teak

Sure they can!

They can implement a usefull landscabe editor and a usefull level editor which deservs this name ,like other engines have for years now and then im happy even when the dont implement the gravity direction feature.

I cant remember that i was too cryptic in my description.
I clearly listed all the points which i find critic worthy over multiple posts.
So whats your argument?

I doubt they will finish arbitrary gravity. Maybe I will get my plugin for it out soon, but its more complicated than you think. Character movement is easy. But now you have to rewrite the navmesh system as well. Want to sculpt a ring-world? Well, landscape doesn't do anything but planes.

Yes i think you are right about that, i also dont expect it to happen anymore.
No i was not building a ringworld ,i had something other in mind, but thats a very interesting gameplayidea too and it gives me nostalgic sci fi feelings.
Would love to see a sci fi action adventure playing in a O'Neill cylinder:

I'm sure other folks feel like you do.

I expressed my opinion about bug fixes and Chance said they do bug fixing every release... That's pretty much true because reviewing the release notes there are tons of bugs fixed. When I wrote my post above I didn't think about how many bugs get fixed per release... Cleared up with some clarification from Chance.

But, if you do not feel Epic has addressed your concerns and there are missing tools that you need, maybe another engine is for you?

teak

Yes if unity catches up graphic wise , i think i will switch completly.
Until then i use both engines ,depending on the projects i work on.