UE4 Roadmap

[John Alcatraz]

We need the OLD BLOOM and TONEMAPPING back! It completely destroys our game that would be other wise few months from release now, and we also need Nintendo Switch support. You promised customers that you will bring new tools to achieve old "unrealistic" bloom. It's not on the roadmap.

r.TonemapperFilm 0 will enable the old tonemapper again. Nintendo Switch support is already in UE4 as far as I know.

[Sadaleus]

So I saw on the new roadmap in Future Releases activity "Vulkan for PC and Linux" - is it going to be delivered in this year? Any more info on this?

[RCaloca]

So I saw on the new roadmap in Future Releases activity "Vulkan for PC and Linux" - is it going to be delivered in this year? Any more info on this?

You can run the editor in Vulkan using -vulkan in the command line, and you can target it in the project settings. We still have some performance optimizations to do but it's getting there

[Maximum-Dev]

It's been 2 releases cards from "Future releases" aren't actually among the ones that get added to "Shipped with..." columns. I hope that changes with 4.19.

[Doctor Ergot]

Well, partially it's just issue with updating Trello
We got clothing tools in 4.18. Also Steam VR on Mac, right?
There's a progress on Vulkan, new audio engine, Media Framework. That's what I noticed. Can be that most of these roadmap items progressed nicely...

But yeah... this board isn't a "roadmap". It doesn't provide meaningful insight into future. No ETA (of course lot of features would "delayed", ETA would change, it's fine).
True roadmap would list important changes before they're implemented and sent to master branch. Nope, this board is quite useless.

[Scereyaha]

What about more map and landscape options? You seem to be doing a lot of innovative work that really affect the details and look of the game... and yet the landscape tool is still so limited. Many users want to be able to create globe maps, on non-square landscapes. Being able to load non-flat meshes as landscape objects would be nice... or the ability to assign the "landscape" type to an object of choice in the scene. An option to automatically divide a map into chunks for level streaming based on how many sub-maps you want would be cool, maybe also with an estimate of performance and volumes already loaded at the edges that the user can size how they want [It would just cut down on so much time and work creating maps and setting up streaming for large areas.]

It would be especially awesome if shapes could be created in blender -or otherwise- as an .fbx and have the option to divide each face of it into a level for streaming with landscape properties and the ability to apply height-maps to them. A built in localized gravity and gravity volume options would be really useful too.

These are things that seem like obvious and basic functions, especially for open world games, and yet I've been fighting with the engine over these basic seeming things because the functionality to have non-flat maps isn't just not built in, it's seemingly built decidedly exclusive to the idea. The options for importing and sizing landscapes could also be better. As best as I can tell each pixel=vertex=meter... with no option of changing the percent size a height map is imported as, at least without affecting the overall limits of the in game map. The ability to tell the program how many in-game km or meters to size something to, or how many pixels to interpret for each measurement, would be fantastic.

The navigation could also use improvement, just being able to move around the view window when placing assets.

I'm hoping that the addition of some shape rendering tools right in the engine itself is leading to this kind of update.

[Zecklar-]

You seem to have dropped the ticket regarding bringing network capabilities to movement components other than character movement component (the ticket from old Trello board: https://trello.com/c/zQtWxPVW/504-ch...nt-on-any-pawn). As you can see it got several hundred votes but it was not added to the new board at all. Was this intentional? If not, could it be added to the new board? People have been asking this feature for years.

Other than that, good job bringing support for Visual Studio Code in UE 4.18.

[Bogdan MATA]

Guys any plans to get ride of VS? Is sooooooooooooo slooooooooooowwwwwwwwww, and instable
Maybe support some other editors, which are better (like text editor) By the way, I'm very happy with Notepad++ too

[Beauk]

Hello,
Any news on the long forgotten Arbitrary orientation/gravity directions for characters ??

This card was there for a long time on the old trello roadmap and had the most votes, haven't seen it yet on the new one. Still waiting for it.
Any chance Epic finally releases such feature in a near future?

[CupMcCakers]

"Any news on the long forgotten Arbitrary orientation/gravity directions for characters ??"

Couldn't you just disable gravity, and then set a constant force to act on the pawn in any direction? You can also affect the rotation of the pawn in the same way.

[jojo8026]

Any ETA on a release date for 4.19??

[Jezcentral]

Any ETA on a release date for 4.19??

That probably depends on the feedback for the preview build. If there are any problems, they might take some time to fix them. If they don't hear of any problems, then it will come out sooner.

[Lobotomy]

"Any news on the long forgotten Arbitrary orientation/gravity directions for characters ??"

Couldn't you just disable gravity, and then set a constant force to act on the pawn in any direction? You can also affect the rotation of the pawn in the same way.

Nope. Two reasons why; 1. Actor Rotations are converted from Euler to Quarternion and back, which does not produce accurate results, and 2. The Character Movement Component is hardcoded in just about every way that counts, especially for Blueprint users.

The reason that this request, despite having been the single most popular and most requested gameplay change, is being forgotten, is because it really isn't. It's being ignored, because fixing it would probably require a ground-up rewrite of the CMC, and hell, probably a nice piece of the Engine itself, but that's just my inference.

Fact is, when it comes to this element, Epic built a skyscraper on a foundation of sand, and I doubt they'd be willing to admit it. It's as if they thought that only bog-standard BrownNBloom First/Third person shooters were going to be the only things made on this platform. Unfortunately for them, this is where Unity really shines, although I wouldn't move over to Unity just for that. In fact, I'd rather just get to work on forking the CMC and try to make that work, like a couple of people have done in the past before mysteriously disappearing.

[Mr Janks]

Sequencer Feature Request:
Character Variables in Sequencer Via Blueprints

A great feature for sequencer would be the ability to set a character as a variable that can be changed via blue prints at run time. This would be useful for creating reusable sequences.

Simple use case: At any point I can use sequencer to take control of a character and make them walk through a door, or sit in a chair, etc. Everything is animated/timed in sequencer. If I could select any character to effect then I could reuse the sequencer over, and over again without having to create a new one for each character just by changing the character reference variable in sequencer via blue prints.

[jojo8026]

Please put on the roadmap whem we will be able to use the new Raytracer