Also I know you're "not a flutter dev channel" but for as long as you're using it for gamedev please do keep sharing your experiences because Flutter gamedev is a very interesting topic!
Man, this is incredible! You deserve a million stars for your contributions and dedication. I've been working with Flutter for 2 years and your work has always been inspiring to me, A big Thank you for being sincerly!
Why is this game in Flutter? Why not? *pulls out well-creased copy of 'Advanced Programming Techniques for the Acorn Electron', Jim McGregor & Alan Watt, 1984*
I came directly from Flutter start up video, and I liked so much your explanation I wanted to see more things from you, leading me here. Thanks for the video, really interesting and informative!! Keep it up!
Very nice video. It's great that you didn't go for the easy way of just using a dependency, and recreated the dependency to match your art style. That's dedication, can't wait to see the game in the stores 🙌
This is really cool! And sounds familiar. I already made 2 simple games in Flutter and now working on a bigger one. Seems I need to create a dev log for it and Steam page too! Keep up the good work!
Thanks for talking about the topic of gamedev in Flutter. I think with the new Flutter engine and shader support it will be more attractive to expand beyond pure app dev. I would be interested in knowing more about the tech behind drawing the curved landscape as well :)
great content as always! I also like to push flutter in non-standard way and like how flexible it is. As for gamedev, I'd wonder if it would be good fit for an RTS game like red alert I am also looking forward to 3d capabilities of impeller presented at one of the conferences. Maybe I should compile engine again and check it out :)
I'm taking up an AI course for my masters because of how much "knights of san francisco" inspired me on NLP. Now knowing that This kind of video games are also possible with flutter, I'm more than confident to fixate on my project idea. I know to develop mobile apps, but not the core principles of programming to make the right decisions. I hope I learn them and AI to eventually fulfil my dreams that you have inspired 🤩
I love your videos. I love that they reach beyond the "normal mobile app" development, which is where large majority of Flutter use cases are. Flutter seems as a really versatile, powerful and modern framework. I have not yet learned much Flutter except for the very basics. I am still struggling with the concept and paradigm of widgets as a blueprint for some real objects instead of widgets as "live" object with properties which I can access and change (which is the paradigm which I am too used to from other frameworks)... But once I manage to switch my mindset to the new paradigm, I believe I will be able to quickly progress. I am now professionally stuck and overwhelmed with work in Qt framework (both fortunately and unfortunately :) ) but I hope some day in the future I will be able to save some time and really dive deeper into Flutter... And I hope that some day Flutter will expand to mainstream also in other areas than just mobile development. Like games for example.
With Rive, I'm pretty sure it would be harder to scale (as the 3D widgets are literally just regular Waveform .obj files, rotating in 3D and flickering). But making something cool-looking as a one-off would definitely be easier in Rive.
Absolutely happy that you answered my question ❤️, Flutter is said to have better control over each pixel on the screen. My next question is in which platform have you currently developed this game for? Mobile, Web, Desktop. Any limitations you see in extending the game cross-platform?
I'm going for desktop (incl. possibly web-desktop). The only (but important) reason I'm not going with mobile or SteamDeck is that I don't know how to effectively translate mouse+keyboard to a touch/controller environment. I'm sure it's possible, but it will require a long and hard look at the UI.
Super prace. Diky za to, co pro Flutter a jeho herni ekosystem delas. Chci zacit s migraci Casual Games Toolkitu do ClojureDartu. Rekl bys, ze to je stale nejlepsi zpusob jak zacit game development ve flutteru (pokud vynecham Flame)? Nebo tu jsou i jina vykopnuti, ktera bys doporucil?
Web (on the desktop) should be fine. Performance should also be fine, even on mobile devices. The big issue is the control scheme (currently mouse+keyboard). Thinking about also supporting touch is almost like creating a separate game in parallel. I want to focus on the desktop first, and maybe do mobile/touch later.
@@filiphracek Interesting! I have been working with a hobby project of mine, which is procedurally generated isometric map game. I bet it has a lot of similarities with this project. I just used joystick for the movement. Not the best solution but it is usable in all paltforms. Only problems I have had with the different platforms were with isolates/webworkers on web and the game also does some heavy drawing (500 000 vertices, 60 fps) which seems to be the limit for my Android phone.
I'm also thinking of building a 2D game, but I can't decide whether Flutter is ideal. It's gonna be a top down space sim that involves newtonian physics based flight model with WASD/joystick controls and trading/strategy management, so a UI toolkit like Flutter would be great, but I'm not sure if it's powerful enough to handle detailed physics simulation on a small scale with ship traffic/decision making on a grand scale, say 1000 NPC ships.
This is really hard to answer. I'm assuming the worst case scenario, when the ~1000 ships are at the same place at the same time. That should still be fine as long as you're not doing something inefficient. But then imagine the 1000 ships all start firing cannons, so now there's 10,000 individual physics entities flying all around? That could very well be a problem. Not that I ever tested it, but I feel that's something for a "big boy" game SDK, something with C++ probably. My game does have physical simulation and tens to hundreds of objects on the map at any time. It seems to run fine. FWIW, I think that grand scale traffic and decision making should be fine. This is not something you need to run updates for on every frame for every entity in the universe.
Hi Filip, amazing video as everytime Btw, I can't find the article where you explain how you make this style of videos can you please help me find it ? thank you
I want to know how the 3D mobile game produced by Flutter is? I will connect the Bluetooth sensor to transmit the data control model posture in real time. Can flutter withstand such rendering pressure? I have tried react native before, but unfortunately the rendering process of js can’t hold on. I don‘t know if you have any understanding of this.
@@filiphracek hi filip, thanks for replying, I really appreciate it😍. sometimes, applying fxxa or aa on top of the shader gives the SKSL error "Program is too large" . you have made a video about vertices, these vertices will be drawn without aa, right? can you enlighten us and make a video how the flutter team implemented anti alias to their canvas? i read somewhere shaders can be rendered in several stages. but can we do it in flutter?
Wouldn't it be beneficial for flutter to have an official 3D toolkit as well? I mean embedding on align Class an optional Z coord system or something like it on other already existing flutter widgets, or on variations like AnimatedContainer could be ElevatedContainer or something like it then you could order widgets and choose overlay clip behaviour etc? I can think of some very cool use cases for games and even software / VR UI's (compatility with apple vision?).
That's a great thought experiment. In a way, widgets already have a sense of depth (via `Stack`). I worry that trying to make Flutter too universal (2D _and_ 3D) would make it unnecessarily complicated. But I like the idea, and there's probably a future in which "Flutter3D" exists.
@@filiphracek you are right, the goals that I had in mind was actually for creating something like a camera and 3D coordinate system, for laying out 3D models and “original” 2D widgets from flutter on this system, enabling the possibility of game development and vision OS for the future, but getting the camera and coordinate system embedded on the official framework could and would probably be not a good idea.
Aren't you afraid that Flutter is limiting you? You've explained very well why you think it is a good fit for Giant Robot Game, but do you think you will always keep creating these type of GUI-games? I mean, I love Flutter, but when me and my son mess around in Roblox for example, I really feel like it's made for games in the same way I feel that the Flutter is build for UI.
For sure it's limiting. Flame makes some game-y things easier, but it's not the same as a dedicated game engine. I myself use Pico-8 to make retro 2d platformers with my kid, and that's obviously way better and more straightforward for that task than Flutter. So I guess it's - Flutter's a good choice for me now. It might not be in the future. Then again, if I decide to create a fully 3D game at some point, it'll probably be because I have enough money and clout to hire 3D programmers and artists. I don't trust myself to do 3D game programming by myself, now or in the foreseeable future.
Thank you for your answer. I can relate a lot. The games Ive made and am currently making are fine to do with a UI framework like Flutter. And since I am experienced with Flutter already, it makes it easy to start without having to learn a whole new thing.
I believe there will never be a native or official support for 3D rendering in Flutter because it is not included in the company’s plans (2D UI only). Everyone only talks about the efficiency of the rendering engine, but no one wants to make full-fledged 3D support, for which they would not be ashamed. Therefore, you will have to wait another 5-7 years to see any changes or just use Godot or webview as usual.
