I graduated in games concept art but I always got into concept art because I love games and so have been teaching myself all games pipelines from modeling, animation and blueprint.
I’m a little over one month into a game project that I want to take to market for fulfil my dream of becoming an indy dev.
I’m starting to ask myself some tough questions, like how possible is it to make a game and take it to market purly with blueprint.
I did think that down line having someone convert it to C++ for performance. But then it would be out of my reach.
I’m making an RTS - Single player
I’ve done a few searches which point to that it can be done. But I’m more curious about the viability of it all.
Many people don’t realize but several functionalities are available to blueprints via C++ plugins which are part of the engine. All someone need to be successful on a blueprint only game is to have in their arsenal the best plugins for the correct tasks in the game. If you run into a trouble implementing something into blueprint, then you could ask in the forums (if not a project secret) on how to achieve it in blueprint or if it would be better have it as a plugin. Plugins, if done right, will expose everything you need to the blueprints.
I haven’t done too much with blueprints, but if I recall correctly I remember reading somewhere about Epic baking the blueprints to C++ with Paragon to get a little more optimization out. Not sure if that’s a one click thing or something that requires more programming knowledge though. But that aside, you should be able to do a full BP game. I’m sure there are many factors that would contribute to negative performance, not just purely using BP. I’m working on a small project myself that will be done 100% through blueprints, and if necessary a plugin here and there as Nilson mentioned. Good luck! =)
I’m currently developing a 95% blueprint title for consoles. It is 100% possible, and definitely its best to convert anything heavy to C++. As SE_JonF said above as well, there is an option in the project settings to nativize your blueprints to C++ during cook time and you’ll have some performance again by that. But the most important optimization is your own logic and understanding of your game’s needs.
If we want to get into the specifics of what you need to do to work with blueprints optimally is avoid ticking any blueprints as much as well. If you do need to tick them like having some door open, you can just enable tick, let it play, and then disable ticking once you are finished. Ticking or updates are pretty much is your enemy when it comes to game development, so try to make your systems fire once rather than constantly checking for something. Also, having a smaller amount of components in your blueprints will give you some performance too.
There will be instances where you do need to tick all the time such as AI, your player, but there will be ways to optimize those as well. I recently just came up with a ‘ticking LOD’ system to gain performance and allowing more AI to be in the scene.
Just be clever in how your develop your game, and keep those things in mind when creating it.
It is possible, even easy. However while easy it requires huge amounts of self motivation, and persistence. Also biggest trap in indie dev are new ideas in middle of creating game, they usually drift original; concept into new areas, or make me create new project. This way i have multiple games started, while finished only once some kind of vertical slice.
There are multiple great plugins that help, also blender, houdini, quixel substance painter, all those apps are great for small indie teams. Automate as much as you can.
But if you can get some other person into team, do it, and you both need to pressure each other into being active. Do at least one task/thing a day for your project, no matter how small, just add something.
For anyone reading this as of today, it absolutely is possible! I don’t use blueprints aside from creating child actors of my c++ classes, but it is definitely possible. I do agree with Nawrot. Developing a game (blueprints or not) is one thing, because it still takes time and dedication. But the a bigger issue is actually finding the motivation to complete a project. Of coarse c++ is more efficient, as it gives the developers more control and minimal code for whatever purposes.
See this amazing game done only in blueprints by a single person: