I’m sorry if this is not the right section but I just wanted to know if there is any future for blueprints after nativization got removed in UE5. Should we except a total removal of BPs or at least a performance boost?
You’re going to launch 5000 traces every frame, but synthetic tests with absurd examples surprisingly win C++.
But in a normal game the possibility of having an optimization problem because you use blueprints is very remote,
What are we talking about 1.6 extra, if you have a buget of 8 milliseconds for code? A buget reserved for something intense in real time.
you end up removing the shadows before
As a newbie, I’d be lying if I say that I’ve understood a word lol, sorry but thank you anyway.
Why do think there is no future?
I don’t know but after they removed nativization which, as far as I know, improves BP performance, I’m worried that BPs would get removed as whole and replaced with that Lua-like scripting language called Verse.
I didn’t even notice it getting removed, because frankly I never used it.
I don’t think you should stress too much about it tbh. A lot of the skills from scripting with blueprints are going to carry other to other environments even if blueprints were to go away. Nativization going away isn’t a strong indicator for blueprints going away anytime soon, either. The whole UE workflow is built around using C++ and BPs in tandem.
Blueprint isn’t going anywhere. Based on the release notes for Gameplay framework they are buffing BP.
-
Blueprint Header View
Similar to the Blueprint Nativization tool, the Blueprint Header View is a new method you can use to quickly convert Unreal Engine Blueprint Classes and Blueprint Structs to C++ code.
Epics flagship game “Fortnite” heavily uses BP. The Lyra game project does as well.
Blueprints are intended to solve a common problem:
- I have a C++ object (a dog, a door, a vehicle, a lamp, …)
- I need to re-use this object with small changes in various parts of the game
- I may want to change the mesh or animations for the dog
- I may want to change how injured a dog needs to be to attack, or run away
- I may need to change how fast a door opens
- I may need to change what triggers the door to open
- I may need to change the color of the lamp
- I may need to wire up an on/off switch for the lamp
- and so forth
You can make blueprints without any execution code in them; just override the base object with new meshes, new colors, new tuning values. This is a common and legitimate thing to do.
But, it turns out, some customizations like this really do want a bit of code. “Is the player strong enough to open this door?” “Does the lamp ramp up and flicker when turned on when power isn’t strong in the building?” “Does this car play occasional backfire particle and sound systems?”
For these simple behavioral configurations, we have Blueprint Scripting. As long as you don’t do scary math or iterate over many sub-objects, perhaps every tick, then Blueprint Scripting won’t show up in your performance profile, but it still provides a lot of value to AAA, C+±based developers.
When you get started with Unreal Engine, you may write a lot of “code” in Blueprint. That’s fine – it’s a great way to learn the engine. As long as you don’t try to push the limits, or ship on restricted systems (like the Nintendo Switch,) then your game will probably be just fine, even though Blueprints aren’t CPU speed demons. Only if your game starts bogging down, would you need to move the heavy bits to C++. And, for most indie development, that point never comes. Or, if it comes, it would bog down in C++, too, because the problem is algorithmic, not CPU cycle performance.
Experienced developers know what parts will need the performance, and go to C++ early. For the cases where you chose wrong, the good news is that converting a blueprint function or even full class to C++ is much easier than writing that C++ class from scratch – you already did all the hard work (figure out how it should work,) and now you just need to write out some properties and functions / methods. That’s almost mechanical, and shouldn’t take very much time.
So, no, Blueprints aren’t dead, and blueprints don’t require nativization to be quite valuable in game development.
Verse is almost nothing like Lua, at all.
The closest I’ve seen is Prolog, and maybe a bit of Erlang.
If you treat Verse like Lua, you will commit really grave performance sins, and will keep banging your head against systems that work totally different than you expect.
Sorry, i just wanted to create some flames
But actually I’m scared too
I have always liked programming logic but I have never felt comfortable programming in any language, for me it is horrible.
I have a strong dyslexia and I am also very clueless, #includes are my nemisis, Syntax rules steamrollers of illusion, a complete nightmare and I always ended up very tired .
With bluepritns everything changed for me and I have a lot of fun, you could almost say I’m addicted.
Thank you all for the very helpful replies