Hacker News new | past | comments | ask | show | jobs | submit login
Videos of Godotcon 2023 (ccc.de)
250 points by mdtrooper on Nov 7, 2023 | hide | past | favorite | 45 comments



I gave a lightning talk about Godot as a wallpaper engine replacement via ScreenPlay[1]. I hacked this together the week before the convention and I hope to release it by the end of the month.

[1] https://screen-play.app/


I want to love an app like this, but the burning question for me is, how CPU intensive is it to run? Is it comparable to running a video in QuickTime or more like keeping a website open?


Good question! There are several ways I want to combat this:

- Query all windows and check if a full screen or maximized window exist, and then pause the wallpaper. I have a prototype for handling this on Windows.

- Limit frame rate via Godots Engine.max_fps. Also in general it does 100% depend on the wallpaper of resource intensive it will be.


I haven't done any Windows GUI programming in a long time, but can't you use GetDesktopWindow and IsWindowVisible instead of querying all other Windows? Apologies if this is obvious and doesn't work.


I haven't tried this method, thanks I will check it out!


Ohh. I had been looking for something like this. I own Wallpaper Engine but I needed a replacement for it on Linux. Have you studied porting this to Linux as well?


Looks really cool. Is there a way to download this for MacOS without having to go through Steam?


I'm absolutely thrilled using Godot and Bevy for hobby projects.

And with Blender you have a competitve 3D modelling solution.

Good times for open source gamedev right now.


And midjourney/dalle can generate passable dev art for textures, sprites and mostly anything 2d.

And proton makes you not really have to worry about cross compiling if you don't want to.

'Good times' is underselling it.


GenAI for art is indeed godsend for this humble software developer without any skills in that field.


Do you have any good resources on how to make consistent characters or art styles? I have been meaning to dig into it for a while. I'm sure it's a combination of prompts and image2image training....been thinking about building a workflow tool for this for a while but haven't had time to dedicate to it.

Even my naive prompts are getting impressive results with midjourney and dalle3...I imagine waiting a few months to a year this problem will be solved in a much better fashion than I would be able to hack together


For that kind of consistency you'd create LoRAs


> Godot and Bevy

I'm guessing you're using these separately right? Not together?


Yes. I'm rapidly switching between Techs in my free time.

A nice contrast to regular work ;)


The great thing about working with Bevy is it forces you to finish a project within a few months before the APIs all change again.

(I kid! It's great that Bevy is developing so quickly.)


Heh, yeah, my work is the opposite, so I've been sticking with Bevy for quite some time at this point. Tried Godot, but I really enjoyed ECS so hard to change now...


You can use ECS in Godot with the use flecs.dev ECS library if you compile as Gdextension, or use Bevy ECS with the rust bindings


Really interesting talk at the top of the list, https://media.ccc.de/v/godotcon2023-57866-swift-godot-fixing...

… of all the strange places to find a good summary of nice swift features.


It's Miguel from Ximian!


the man has a whole career out of pushing corporate-controlled languages into wider and more open use


Thank you for pointing this out. It's not a problem for a lot of people but the fundamental ethics driven FOSS user should at least be made aware that he sided with Microsoft during the worst Linux is cancer era of MS PR.


I found the talk about SwiftGodot amazing. Do we really need to be reminded in this context about what the speaker did 20+ years ago? Like, any relation to the talk or to Godot AT ALL?


What year was this?


Cancer quote is Balmer in 2001. That same year Miguel started Mono. But he did other things that held back the FOSS movement like support Microsoft for OOXML.

"Advocacy of Microsoft open technologies edit

De Icaza endorsed Microsoft's Office Open XML (OOXML) document standard,[15][16][17] disagreeing with a lot of the widespread criticism in the open source and free-software community.

He also developed Mono – a free and open-source alternative to Microsoft's .NET Framework – for GNOME.[18] This has raised much disagreement due to the patents that Microsoft holds on the .NET Framework.

De Icaza was criticized by Richard Stallman on the Software Freedom Day 2009, who labeled him as "Traitor to the Free Software Community".[19] Icaza responded on his blog to Stallman with the remark that he believes in a "world of possibility" and that he is open for discussions on ways to improve the pool of open source and free software.[20]"

Complicated man. I wouldn't want his legacy.


Yep. He's been hacking away at this for a while - it's been fun to keep tabs.


I wonder if Godot can be analogous to Blender in the sense of their impact on the respective industries. I got my feet wet in the 3D design world and my impression is that Blender already won and the established commercial products began adopting Blender conventions and ideas to stay relevant. I'm not professional in any means and I'm sure there's much more than I've been exposed to but the power of Blender was mind boggling, is there a chance of Godot becoming the game engine?


Not sure it will, maybe, but with Unreal and Unity, Godot is still far behind in 3d. As where with Blender there is no other alternative.

Godot is like years behind modern engine for 3d. And there is not a single chance it can keep up with Unreal which is used by everyone.

It can be an industry standard for indie / mobile game.


I think Godot catching up to Unity is pretty likely within the next 2-3 years. Unity is a mess of tech debt, mismanaged, and has just burned all their good will. Godot should be able to improve much faster due to the lack of tech debt, and a lot of indie devs are motivated to help.

I agree that catching up to Unreal is not going to happen any time soon; they're way further ahead on rendering features, and constantly improving.


Depends how well they manage to provide the same .NET development experience for game consoles, including AOT toolchains, debugging and profilers.


For Blender it’s more the other direction: they’re moving to the way of working of other tool. Before this change (2.8 ?), it was a pain for someone used to navigate any of the big player to drop into Blender.

The fact that Autodesk has since bought nearly every big package (notable exception: SideFX Houdini, Maxon Cinema4D and the Foundry Modo) helped a lot. They’re not exactly well loved, be it in special effects, architecture or CAD (though for CAD they’re all bad at the end of the day)

I like that Blender is a credible option. I would have loved to have this kind of option when I was younger - way simpler than finding Maya, 3Ds or other big package… with, erhmm, free licence let’s say.


Oh yes, I was first exposed to version 3.0 and all the tutorials that were a bit old were hard to follow and even the infamous Doughnut tutorial was re-filmed for the new UI but I also recall about some commercial pro software taking the Blender route as its popularity among the certain creatives exploded.


It’s certainly possible, I am an early adapter of Godot. It’s one of the talks we used to have on the social platforms back then. They certainly seem to manage things in the same style.

In time Godot could be quite formidable, and I argue in many ways 4.x already is. One thing is that rapid prototyping in Godot is actually fun


The GIS in Godot vision talk a bit unlucky, but amazing food for thought nevertheless. The combinatorial repurposing of open source platforms, apps, libraries etc for use cases beyond what they designed for is one of the most fascinating possibilities opening up with FOSS. It is helped alot if people adopt open standards for data exchange.


At least we didn't have to wait too long for them!


Expected to see videos of a couple of guys sitting on a bench.


I really want Godot to succeed. Good luck all!


It's still for a ways to go though. Much of the documentation I've read (2D physics) is incorrect or outdated. It also kind of assumes you already know everything, which is somewhat annoying.

Still, it's making good progress.


For those like me familiar only with the play, "Godot is a cross-platform, free and open-source game engine released under the permissive MIT license. It was initially developed by Argentine software developers Juan Linietsky and Ariel Manzur for several companies in Latin America prior to its public release in 2014."

https://en.wikipedia.org/wiki/Godot_(game_engine)


Hurray I love getting Godot news. Lots of conference content for me to listen to over the next few days :)


Any comments about how the team is going to spend the Unity-incident windfall?


Not sure they need to do much in terms of marketing at least. When I started using Godot in earnest last year, only a few nerds seemed to know it. In the past few months I haven't come across a single indie developer not aware of it.


Gave a look into Godot, find out the engine is deeply integrated with some custom scripting language that leads to performance issues [0] (Unreal learnt from this and got rid UnrealScript), support object-oriented design over ECS [1] and are still trying to support OpenGL when the rest of the industry is dropping it in favour of focusing on modern APIs like Vulkan [2] but progress is at least being made with Godot 4.

[0] https://sampruden.github.io/posts/godot-is-not-the-new-unity...

[1] https://godotengine.org/article/why-isnt-godot-ecs-based-gam...

[2] https://forum.unity.com/threads/opengl-deprecation-and-remov...


>Gave a look into Godot, find out the engine is deeply integrated with some custom scripting language that leads to performance issues

>Godot API is designed around GDScript: This is also not true. In fact, until Godot 4.1, typed GDScript did calls via "ptrcall" syntax, and the argument encoding was a bottleneck. As a result, we created a special path for GDScript to call more efficiently. [0]

[0] https://gist.github.com/reduz/cb05fe96079e46785f08a79ec3b0ef...

I don't have to write anything new, since you didn't bother to write anything new either.


A great reply to your [0] is this: https://news.ycombinator.com/item?id=37598985


OpenGL support is currently necessary for targeting web and certain other platforms. The industry might be trending away from it, but it is still very much used.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: