I was wondering, wouldnt it be awesome if there was a visual programming framework that can be adopted by other open source engines. If that framework was coded in such a way that it makes it easier to other developers to make it work with
and so on
There are so many awesome open source engines - some of which with amazing features that gdevelop doesnt have yet.
libgdx for example can do 3d games and export not only to html5 and native, but also native ios and android!
You may ask, but what about the level editor?
For 3d games:
Recently blender (open source 3d package) got an addon that allows people to code a gdx game inside blender:
blenderartists.org/forum/showthr … ML5-export
For 2d games:
Overlap2d is an amazing open source level editor that supports mainly libgdx
Here is the website of libgdx:
Now imagine if somebody made a visual programming framework that can create libgdx code and collaborated with BDX and Overlap2d projects - both of which have communities who have been requesting such system for years. Suddenly you would have something that a lot of people are using and you would have a game development suite that is very mature, completely open source and has both a huge community and a lot of developers.
I am not sure how that would be doable, but its just an idea.
Abstracting the event system from GDevelop to make it standalone and usable with other framworks would indeed be interesting - I’ve already thought about it from time to time
Still to make something really usable the event system would need to have a strong integration with one of these frameworks/library.
The risk of making something too abstract is at the end to build something that nobody use: people loves GDevelop because they can use it to create games without programming and without having to setup complicated things.
Creating an event system that can be plugged into Pixi.js or Three.js would be interesting, but to get something really polished and used you must take care of the workflow and integrating things well so that you do not end up with a mess of tools nobody use.
But surely if I have to build a GDevelop 4 from scratch, I would take the path of having an event system that can be easily plugged into a library like Pixi.js or Three.js!
So all in all: yes that would be cool, but it’s something really abstract that need to be backed by a product/platform offering a nice workflow. But GDevelop could slowly migrate into this kind of thing by allowing more and more external tools to be integrated with the software - offering you a development environment that is simple to use yet open with plugin for powerful tools.
Since you already thought about it I would like to take the opportunity and mention AGK2 here.
AGK 2 got really useful features and more to come, check out it documentation.
It is also support all existing platforms (except HTML5) even Linux support is coming slowly both 32 and 64bit.
You can also code AGK2 in C++
So, if next time you think about it, think AGK 2 as well.
Personally I would be interested to load GDevelop scenes in AGK 2 also would be nice to use the advantages of AGK 2 through an event system.
I would even pay for it if it matters. You could sell it through TGC store, or team up with TGC and sell it on steam as an extension/DLC to AGK2 (extension support is coming), or you can also try Steam greenlight and sell it as a stand alone app…
AGK2 will be developed for a long time in my opinion and just getting better and better with more and better features with each release…
I agree with you that an engine should offer a complete solution - meaning it should include an asset system and level editor.
There are many open source engines that do not have a visual programming framework at the moment - but do offer all the rest. There is a niche here that is not being exploited!
My proposal is to turn gdevelop’s event sheet system into a visual scripting tool for other engines that do not have visual scripting. Make it capable of generating code and integrating with other open source engines. Injecting it into the workflow of people who use other engines, but prefer to do visual programming.
Three.js is very tempting as it supports 3d game development, as well as 2d! You also get webgl acceleration.
I think we should really aim to bring gdevelop to the next level with version 4 and take advantage of it’s open source nature to lure more developers and users to get involved. That is one way to do it.
You should partner with an engine that is:
- Open source
- Mature and already has a big userbase and a number of projects
- Can compile to html5 so games can be ported to other platforms
- Can compile to native windows,linux
- Features in native work just as fine as features in html5
Supports 3d games as well as 2d games (construct2 users have been asking for that for ages - construct3 wont have it) - and they run in html5
Other Open source engines that cover those points:
- Godot- uses a very simple scripting language of its own - similar to python
-Superpowers (soon to be open sourced, source code available to people who funded it) - it uses truetype for coding! - can do 2d and 3d games - has its own level editor. Its not very mature yet, but their development pace and community growth is staggering!
I think Ideally gdevelop visual programming framework should be so adoptable that you would be able to integrate it in other open source engines that already have everything, but a visual programming framework. And since they share it, their developers could collaborate on it and that way commit code to gdevelop as well.