[Twitch] 4.6 Roadmap - Oct. 23, 2014

[RPotter]

Some things I'd like to know whether they are planned for 4.6:

Patching/DLC - i.e. additive paks and delta paks. Having these available as first-class features is something I would consider very important for a modern game.

Simplified Text Localization - It's certainly possible to localize text in game right now, it's just a very laborious process that is poorly documented. Given how important it is (especially for independents) to release internationally, I would hope this would be a top priority. Last I checked, even the experimental Translation Editor was for engine text only.

Audio Localization - Seems it's been half-finished for a while now (Dialogue Waves, etc), when will we see this properly supported?
Diffing/Merging - There's some good early effort here already (the BP/BT diff tool looks great!), but I don't feel we have a complete solution here yet, given the fact that assets are all binary.
Materials should have the same visual diffing as BPs and BT. This almost seems like an oversight!

We need proper merging tools, or a good merging flow. Here is how I see it working: export the local and depot copies of the binary asset as temp T3D text files. Invoke a configured merge tool to merge versions together. Take the merged output text file and reimport it and replace the local binary asset.

You guys use P4 internally as we do: I'd be very interested to know how you guys are managing assets in projects (especially map files!). Exclusive locks can only go so far.

Looking beyond 4.6, I'm also curious about these trello cards on the roadmap:

Cooker 3.0 - Live editing is a *huge* productivity booster when you are working on a platform (like mobile) where your interaction model is very different from your work machine.

Geometry 2.0 - Personally not massively interested in BSP construction, but there's a lot of potential for advancements in geometry creation and editing. The idea wouldn't be to recreate Max or Maya in the editor, but to provide a way for rapid prototyping (for which the current brushes are not well suited, even when compared to the UnrealEd of the past), as well as provide tools to enable geometry editing at runtime (terrain, character morphing, foliage painting... the things you would need to build a good building-sim game in Unreal)

[Robert Khalikov]

I also would like to hear about patching, DLC and support of multiple PAK files.

[PerpetualChange]

Does Epic plan on greatly improving their lighting model, in relations to foliage?

[HeadClot]

Can we get an update on the Paper2D Plugin?

[Rama]

(NON)Urgent

I had to revert from 4.5 because the changes made to the nav mesh system cause a 20-40 fps drop when the nav mesh is rebuilding at runtime, a loss that does not occur in 4.4 or 4.3

I cannot use 4.6 either until this issue is fully resolved!

Most of the projects I am working on require Rebuild At Runtime to be as optimized as it was before 4.5, I cannot admit a drop of 20-40fps due the resulting limitations on the rest of what I can do while keeping a decent FPS.

[4.5] Nav Mesh ~ Rebuild At Runtime is causing 20-40 drop in FPS
https://answers.unrealengine.com/que...sing-20fp.html


All I did was upgrade my project and I am consistently losing 20-40 fps when nav mesh rebuilds that I was never losing in my prior engine version which was 4.3 for the project in question. I also tested 4.4 and did not encounter the fps drop there either. It was mentioned in release notes that a lot of Fortnite navigation code got added in 4.5, and I can only presume that integration has something do with this sudden problem with Rebuild at Runtime.

I am bereft of all future UE4 updates until this Rebuild at Runtime issue is fixed, I simply cannot deal with the resultant performance loss!

I can absolutely verify it was just upgrading to 4.5 that is causing the fps loss, and that it is indeed the Rebuild at Runtime feature of navigation mesh system. And I can also verify that Rebuild at Runtime does not cause this FPS loss in 4.3 and in 4.4, while still performing perfectly and enabling my AI to navigate dynamically added geometry immediately.

Rama

[EoinOBroin]

Hair shaders or some form of integration with Nvidia Hairworks/TressFX! The main thing stopping 3rd person characters looking really real in the engine is the lack of hair shading support! Definitely my no. 1 feature request!

[Stefan Lundmark]

Questions from my wishlist. Can we expect any improvements and/or iterations on the following:

- A version of Matt Mittrings' Forward Shading material node (mentioned in this weeks stream).

Forward Shading is planned according to the Trello Board.

Some things I'd like to know whether they are planned for 4.6:

Patching/DLC - i.e. additive paks and delta paks. Having these available as first-class features is something I would consider very important for a modern game.

Also on Trello.

[RPotter]

Forward Shading is planned according to the Trello Board.



Also on Trello.

The point is not to say "do you have plans to do this", but when they are targeting completion. Trello does not provide that information.

[Stefan Lundmark]

The point is not to say "do you have plans to do this", but when they are targeting completion. Trello does not provide that information.

You asked if they were planned and I replied to that.

[fosforestudios]

I have some more additions to wish list, and they are no new features as everyone seems to like:

1- I have found some annoying bugs, please fix them. We really need a solid engine.

2- A thing I really miss from unity3d is the "small" executable size. It is something really annoying that we have to remove unused assets manually. For us, newcomers it is really difficult to know what will be used and what won't. I know there are some guides around to mitigate this problem but it is a duty of the engine to assist the developer on this subject. I would like to see something like this in ue4.

3- There's a lot of things to improve to make ue4 more accesible to newcomers / switchers. Documentation is not bad but could be improved a lot. For example, it is ridiculous that there's no clear explanation of the realtionship between Player Controller, Pawn/Character and the cameras. I had to spend 2 days reading forums, official doc, experimenting, etc... to notice what was happening under the hood. And still having some problems with the camera manager when it should be used, etc.... This is a really basic thing all developers will have to handle at some point. For me, official doc was written for someone that knew how to use ue4.

4- Much better 2D support. If epic want to attract indies this is really important. Lot of indies work on 2D titles and the support is a bit weak right now.

5- Make videos to build example games step by step. Explaining why things are done the way they are. This will be a really valuable informations to new people and experienced ones. Unity has some videos made this way and the learning experience is great!.

So, Epic, please stop adding new features like crazy. For 4.6 improve all the great things you have. Put a bit more love on indies (We are not creating AAA titles and we are a lot )and try to make the engine a bit more accesible. By the way I love UE4!

Cheers.

[RPotter]

You asked if they were planned and I replied to that.

Planned for 4.6

[TommyBear]

Hi guys, my questions:

- What are the plans for Market Place in terms of search-ability, categorization, sales and specials/bundles
- Are we going to see blue-printable components soon? Make components in blueprint?
- What is the status of blu-tilities? Would love some getting started documentation.
- What is the status of plugins in marketplace etc?

Tommy.

[AnadinX]

Epic Cloud?

Allow us to host leaderboards, achievements, stats and data (basically user accounts and profiles) similar to Hydra (agoragames) and then taking a leaf from Unity, advertising, analytics, cloud builds etc.

Since UE can't build a Windows exe from Mac and a Mac exe from Windows, a Cloud build service would be very useful indeed.

[n00854180t]

2- A thing I really miss from unity3d is the "small" executable size. It is something really annoying that we have to remove unused assets manually. For us, newcomers it is really difficult to know what will be used and what won't. I know there are some guides around to mitigate this problem but it is a duty of the engine to assist the developer on this subject. I would like to see something like this in ue4.

As long as there aren't any references to a given asset in the maps you're packaging, they won't be included. One of my projects has ~11GB of stuff in the assets collection, and when published is only about 150MB. Also if you're packaging a Development version (e.g., so the console works) make sure to delete the .pdb files or they will bloat the size.

[shatieshack78]

3- There's a lot of things to improve to make ue4 more accesible to newcomers / switchers. Documentation is not bad but could be improved a lot. For example, it is ridiculous that there's no clear explanation of the realtionship between Player Controller, Pawn/Character and the cameras. I had to spend 2 days reading forums, official doc, experimenting, etc... to notice what was happening under the hood. And still having some problems with the camera manager when it should be used, etc.... This is a really basic thing all developers will have to handle at some point. For me, official doc was written for someone that knew how to use ue4.

4- Much better 2D support. If epic want to attract indies this is really important. Lot of indies work on 2D titles and the support is a bit weak right now.

5- Make videos to build example games step by step. Explaining why things are done the way they are. This will be a really valuable informations to new people and experienced ones. Unity has some videos made this way and the learning experience is great!.

So, Epic, please stop adding new features like crazy. For 4.6 improve all the great things you have. Put a bit more love on indies (We are not creating AAA titles and we are a lot )and try to make the engine a bit more accesible. By the way I love UE4!

Cheers.

I agree on all of these. I do find that other engines I have used are a bit more Indie friendly when it comes to the tutorials and documentation. For many, being truly and utterly clear how everything works almost to the point of hand-holding helps people really understand without having to scour the forums and want to give up.

I would love if 4.6 would focus on performance improvements. Especially on the Mac. 4.5 seems to have helped slightly (I think..). But I would love it if Navigating the app just felt smooth.

I would also love Some Retina Display support. Viewing the Editor on a Retina Display makes me sad. I know it is just an aesthetic thing but one just gets used to the sharp easily readable text on a Retina Mac.

I would also like to not have to close the Launcher in order to get decent performance from Unreal.

I am about to develop my first game on UE4 and looking forward to see how it shapes up.