That’s an exciting step forward for Unreal Engine users! The PCG framework being production-ready means massive potential for procedural world-building. The performance boosts and GPU enhancements will make workflows smoother across platforms. The new editor tools like draw spline and paint volume sound super intuitive. Overall, UE 5.7 looks like a major leap for both developers and artists. Read more
Does 5.7 Still have issues with data loss when migrating blueprint components on child classes? This was an issue for 5.4 > 5.5, then for 5.5 > 5.6 as well (still not fixed in 5.6.1). Now we are moving onto a new version without resolving this in the older versions.
I hope that projects will be fully forward compatible.
1 Like
im using it quite comfortably on linux, other people too. maybe you’re using wayland which creates a ton of issues for a ton of apps. i’d be nice if ue improves wayland support, but on x11 works quite well.
Is there any documentation on the new composure plugin?
Version 5.6 works on Linux with Wayland without any similar problems. I guess the reason is that 5.7 migrated to SDL 3, and the developers didn’t manage to do this correctly.
1 Like
they started the migration to sdl3. it is not finished afaik. with x11 it works well so sdl itself is not the issue i’d say. so there’s still work to be done with wayland afaik. it might as well be an issue with your setup, i don’t use wayland so i can’t confirm. neither the people i know that use ue, they avoid wayland.
FWIW I have the same issue on Wayland. In addition to the buttons and menus being broken, the editor is extremely unstable, submitted like 10 crash reports in the 30 minutes that I tried using it.
5.6 works fine though
2 Likes
You can work around this by logging into an Xorg session or by setting the environment variable
SDL_VIDEODRIVER=x11 ./UnrealEditor
from a Wayland session, but this bug will affect your end users too - they’ll need to do the same.
3 Likes
question, is 5.7 the full release coming during this week somewhere this month on Mid-October?. Or Longer till November?. Because some people was saying it could be coming somewhere this month. Just curiosity. Sorry if I asked again
2 Likes
i have no idea, but the activity on the branch has decreased yesterday and today. so it’s getting pretty close.
i personally would recommend not rushing to a new version.
.0 releases usually should be considered beta. so for production you should target .1
if you need to test a project, it can be done as of now with the preview.
No worries. I wasn’t even thinking about jumping in the preview version anyways. I already know that is %100 unstable and still under development. I was just curious that’s all. Plus I’m really having a struggle hard time being patient on the full release whenever it comes
1 Like
All good 
i get it, i’m actually pretty excited myself too. i’ve been pulling git every day waiting for it to be stable. i just “had” to put the disclaimer for people that might jump too early.
having said that, i do would love to hear an epic member confirm if they have a target window that they can share.
Yeah, I know. They haven’t said or announced anything about the 5.7 release of when is it coming. That just sucks to be honest
1 Like
I have some issues in VR since 5.6 in VR too (and still in 5.7 preview). I don’t thinks it is related to HZB occlusion but here is the link of my post :
Lightmass doesn’t build lighting on Linux in 5.7, and has not worked since 5.4.
Are there any plans to ever fix this?
1 Like
Even full version is unstable…
The full version of 5.7 is not even out yet. You know that right?
another issue i found is a warning for a deprecation to access ScrollBarStyle on combo boxes. but there’s no Setter.
‘ScrollBarStyle’ is deprecated: Direct access to ScrollBarStyle is deprecated. Please use the getter. Note that this property is only set at construction and is not modifiable at runtime. - Please update your code to the new API before upgrading to the next release, otherwise your project will no longer compile. [-Wdeprecated-declarations]
Please provide a setter for this, since i really want to be able to set the style for the scrollbar via code. thanks
follow this tutorial, the dpad issue will be fixed, but i am figuring out how to enable the ps5 controller via bluetooth, if you have any idea tell me
Replying to my earlier post. Yes, the visual differences between the (un)voxelized grass meshes I am using seem night and day. At a certain distance the voxelized versions are just ‘not there’ insofar as they impact the color of a given pixel..
I really like the preserve-area effect as it lets me, at least, give the impression of an actual mesh, way out there, with discoloration vs the landscape, background, and actual occlusion of things in the grass. For me, WYSIWYG insofar as seeing through a scope and getting-over-there to see the exact same thing is a cornerstone of my efforts (high-fidelity of the game-model).
THIS however, is….off. Not gonna say it’s broken, it might be doing exactly what you want, but the visual differences are obvious so unsure here..
Voxelized - \~5-10% increase in FPS but not worth it for my tastes, mechanically the grass is just not-there…
As per the above, I am willing to imagine a single scalar-value you could use to scale up the screen-area (preserve-area equivalent) for the generated voxel to replicate the visual impact of the pre-voxel method? eg: scale it to 1.5 times what would be generated to ensure occlusion, etc…
2 Likes