Josh

Added right-click menu to remove nodes in flowgraph interface.

Flowgraph widget is uploaded. Not the cleanest code but it works nicely: https://github.com/UltraEngine/Extras/tree/main/Code/C%2B%2B/CustomWidgets

Also, the /dev branch is more up to date...

I just updated the static lib and tried this successfully with a new c++ project. If it wasn't updated before, it is now.

It might be possible to make glTF files save textures in a subfolder of the model's folder.

I don't quite understand. You want a glTF file to be saved in a different folder than its textures? Normally these all get saved in the same location.

added this functionality in build 193

I do not have an answer to this right now.

Lua is dynamically typed, which can be very convenient in small projects when you want to quickly experiment with things, and can also be problematic in bigger projects when you forget what variable names you used. C# is strongly typed, which gives it better autocompletion support, something that isn't really possible to get perfect in Lua. Coming from a BASIC background, I always saw C# as ugly C syntax combined with a lack of functionality. A language like BlitzMax or PureBasic (or even VB) always seemed much nicer to me. However, it does give you much of the functionality of C++, with fast compile times, and good support for autocompletion in Visual Studio, so I can understand why it would be good for games. For most game developers, it's about the same functionality as C++ with faster compile times and no need for header files. I think C# was Microsoft's reaction to the growing popularity of Java at the time, which kind of explains some of the decisions. I find it strange that a now-dead fad is what killed BASIC, kind of like how the iPhone sent Windows into a decade-long descent into madness before they finally killed off all the cell phone interfaces in Windows 11.

fixed

fixed

I definitely would update your graphics driver.

Fixed.

-4 is VK_ERROR_DEVICE_LOST. Maybe you are running a laptop on battery power?

fixed

Can you post the map file that causes this?

I will probably include it in the pro version, and then have a C#-only .NET version as well. More products in the Steam store = more money.

Once early access is over I think C# support is the logical thing to do. Could be as soon as January.

It's important to know what you are good at and compete along those lines. Providing a lot of game-ready models and videos is not what I am good at. Anything I can create would just be a tiny drop in the bucket of what people demand. It's better to focus on your strengths, which in our case is performance. That's the carrot Ultra has to offer that Leadwerks lacked. As far as game content goes, focusing on the import pipeline has been a lot more beneficial than trying to create a big repository of content. Klepto's HDRI tool completely solves the sky / environment map problem and unlocks infinite content of that type. glTF support has a similar beneficial effect. I am restraining myself from putting up models packs for Ultra because any time I could spend on that would be better spent writing a built-in SketchFab browser for the editor. Similarly, I think our editor extensions will in time be very valuable, but that requires a bigger community before you will start to see the benefits. Other engines are basically selling themselves based on the amount of third-party content they have. That can't be our core value proposition, because it's not possible. The reason people will use Ultra is its performance, and the third-party content will come after that. These steps must occur: Acquire users based on performance advantage. Provide a framework for third-party content to be created (extensions, plugins, components, flowgraph). This is mostly done, but could use some more standard components to demonstrate features. Community builds up third-party content and then you can start promoting that as an advantage. The first step absolutely must happen, or the other two are impossible. It may mean we need to focus harder on VR developers, since they are always having problems with optimization and Ultra can make a big difference in their life.

Static library and Lua bins updated All recent bugs fixed Fixed it so at most only one trigger collision will occur per physics step, per object. Previously a lot of extra collisions were being created, and the door script was calling collectgarbage() so the Lua garbage collector would get run 500 times per second when the player was inside the collision trigger volume. Added Timer:Stop() Added Tags field in entity properties in editor. Separate tags by a comma.

I got it worked out. Wait for a new build...

Fixed in next build that goes up...

I think we should be looking at a December 1 release on Steam, with as much component / game stuff packed in as possible before then. It's clear that is what the immediate value is for the majority of users.

Layers

I've been thinking about this and I think the solution is to do both. I can't imagine forcing people to "build" a character controller out of parts in every single map they create. Just being able to drag a single component to make it work is a hundred times faster. On the other hand, if you can construct a prefab out of parts and customize the behavior a bit, that gives the non-programmer a bit more to dig into. For any programmer that is a horribly tedious way of doing things, but you should have an assumption that some users cannot write code and never will.