Discussion: Future of libgdx (backend, functionality)

Anything libgdx related goes here!

Discussion: Future of libgdx (backend, functionality)

Postby captainflyaway » Sun Jul 28, 2019 7:44 am

Hey all!

My Background
I started game developing with a school project a few years ago. As we were learning java, libgdx was the only real option back then - we should code a game not design -> no unity or ue4 was allowed.

I remember that libgdx had a really steep learning curve back then which was not really good.

I tried coding a game but I failed after I realized 1 year into coding that I can never finish the project in any realistic way.

I am studying computer science at the TUM germany now which of course completely changed my view on coding.

Presence
I really like programming rather than dragging stuff in an editor so I tried a few other frameworks (Löve2D, Cocos2dx, CoronoaSDK) and I still think that libgdx feels the best of the options (maybe not in performance or portability but the way you code your game).

Discussion
Things we cannot change easily:
1. The active part of this community knows that there is not a lot of attention anymore.
2. Libgdx is not so much future proof anymore (no Metal or Vulkan background)

Things we can change:
3. Starting a new project is annoying as you have to rewrite (or copy) stuff that you should not really care about to setup every time
4. The framework lacks functionality (usable widgets - apart from buttons, support for vector graphics, alternative pathfinding - e.g. spacial segregation
5. Good and updated tutorials on how to make a game ready for iOS
(6. A modern website and forum)

Things we cannot change:
1. Lack of attention
I do watch a lot of youtube gamedevs in my spare time - non of them use libgdx, the forum has not seen many new games or users lately - if there are new games most of them are a first project (which is nice to see, but nobody will say "wow this amazing tic tac toe game was coded with libgdx".

- maybe you should do a survey to find out why people do not stick with libgdx


2. Libgdx is not futureproof
Apple announced in 2018 that they will deprecate OpenGL and stick to Metal.
- there a not a lot of libgdx games for iOS (probably as first games are not ment to bring in money realistically and maybe because it is a struggle to get a game running on iOS
- also OpenGL will probably stay on iOS devices for a long time

Still the public wants to see a future proof backend (vulkan has the performance and portability for iOS and android) - it is a lot of work and I cannot help with that sadly as my time is very limited (like yours probably)!

Things we can change:
3. Starting a new project is annoying
You should not rewrite the way you render your game each time. The same goes for setting up the a* pathfinder, tiled maps and the UI.
I do not have a lot of time and when I want to start coding a game I want to code it and not worry about all the "annoying parts" over and over again. (Game design!)

I have seen a project of a user who tried to make libgdx more suitable for game jams (probably addressing some of the issues I named).

I think it would be great if you guys published the working functionality of things you have to code again and again every time so other users do not have to care about setting up e.g. the A* (which was rather compilcated for me back then).

4. The frameworks lacks functionality
There are for sure many things that could be included into the framework like an alternative to the A* or a better UI creation tool.
I am sure that people adapted (copied) many useful classes into their project that could be used by many others also! Pls make them publicly available so other people do not have to invest the time like you did.

5. How to publish on iOS
Only some people do publish on iOS anyways but if you do, pls write a good tutorial about it!

6. Modern website and forum
The website and forum work and fulfill their functionality. They do not look high quality or modern.
The design suggests new people that it is not worth to learn libgdx by now.
I know that there are good templates for websites online. Why not use one of them?

I know that most of you will not read the post anyway (it is a lot of text I know). To the ones who actually read it:
Whats your opinion on my points?
Do you have better and or other points I did not mention?

(I think my english is fine, prepositions are not my favourite tho!)
captainflyaway
 
Posts: 237
Joined: Mon Feb 20, 2017 7:52 am

Re: Discussion: Future of libgdx (backend, functionality)

Postby MrStahlfelge » Sun Jul 28, 2019 12:30 pm

1. Come to the discord to have an active community
2. I doubt Apple will remove OpenGL any time, because WebGL relies on it. And no one will buy a device where this is not working.
3. We have some attempts on the discord to improve that.
4. Extensions are a way to improve the situation
5. https://medium.com/@bschulte19e/deployi ... 3796410d82
6. Again, the Discord server is what you are looking for.
MrStahlfelge
 
Posts: 121
Joined: Thu Jun 15, 2017 6:40 am

Re: Discussion: Future of libgdx (backend, functionality)

Postby raeleus » Sun Jul 28, 2019 1:49 pm

I don't know if it was mentioned yet, but you should join our discord server.

https://discord.gg/7c6Wg8H
Join us on Discord for faster and friendlier service.
Create skins visually with Skin Composer.
JOIN THE libGDX Jam of October 2019!
raeleus
 
Posts: 474
Joined: Wed Aug 07, 2013 10:57 pm

Re: Discussion: Future of libgdx (backend, functionality)

Postby captainflyaway » Fri Aug 02, 2019 7:02 pm

I had to practise for an exam but now I can reply.

@raelus I joined the discord server

I know that apple will not remove OpenGL anytime soon. Still to not be or to become a legacy framework libgdx has to adapt (or die if not) at some point.

The discord server does not tackle the issue of going on a website that seems to be written in html by a student as a first project in school. - to exaggerate that a bit

I think one of the bigger competetors of libgdx is cocos2d. I have coded with cocos and I do like libgdx a lot better.
Cocos introduced a metal backend and take a look at their new webiste design - it looks like many people will actually use that.

On the other hand can you really imagine a small studio to come to libgdx and use it for their next project (if it is only for android, sure. But for everything else?)

Edit: Basically what I am trying to say: I think libgdx should be more attractive to the outside.
Last edited by captainflyaway on Fri Aug 02, 2019 8:36 pm, edited 1 time in total.
captainflyaway
 
Posts: 237
Joined: Mon Feb 20, 2017 7:52 am

Re: Discussion: Future of libgdx (backend, functionality)

Postby tomski » Fri Aug 02, 2019 8:31 pm

You are basing these stats on people that frequent the forum? They are unfounded. There are existing studios and game developers that actively use libgdx for all platforms, iOS included. There are many users of the iOS backend, and many games and non games that are deployed using libgdx.

This goes for all games deployed using LibGDX, you would only know about them if the developers for some reason decide to tell everyone, which is pretty unlikely. Why would they do that? They aren't gaining anything from it, and aren't required to give attribution in any way either. Most people that come to this forum are generally beginners, hobbyists, or people with the odd technical question. They don't come to market their game (to other game developers)


In regards to other things, everything is completely open source, anyone is free to contribute and push things along. There is only so much time available, and as everyone is working on this pro bono, most of the time goes towards maintaining the codebase.
tomski
 
Posts: 909
Joined: Wed Jun 26, 2013 6:06 pm

Re: Discussion: Future of libgdx (backend, functionality)

Postby shatterblast » Fri Aug 02, 2019 9:58 pm

GWT completely is awesome. I think League of Legends uses something similar for a few of its web site backgrounds with an overlay for actual web page interaction. Granted, they feed videos into different pages as well, so different pages carry different things. HTML5 seems crazy in what it allows while staying safe. GWT just takes advantage of that.

If this post gets too old, then the following link probably won't work right, but if you look, you can easily pull off that background with a mixture of GWT and Spine. (Not accounting for artistic talent, though.) You can get a basic transparent overlay with a simple GUI to work with some simple JavaScript maybe, which of course does not employ GWT. I am sure that the blue button in the screen's middle is JavaScript-based.
https://signup.na.leagueoflegends.com/e ... rrer=index

In this link, you can pull similar results on the left. It can switch between one of two backgrounds. The same general concept of a mix with GWT and Spine still applies.
https://www.dndbeyond.com/sign-in?returnUrl=

I have seen a few others, and League of Legends also posseses several I have not.
shatterblast
 
Posts: 414
Joined: Sun Jul 06, 2014 1:14 pm

Re: Discussion: Future of libgdx (backend, functionality)

Postby captainflyaway » Sat Aug 03, 2019 7:16 am

@tomski yes and no - yes, there are of course developers and teams using libgdx. I am not talling about existing users but anout new users. The forum mainly is used by beginners, you are completely right. And how many new users are there? I do not know, but I think it is not too many.

Or the other way around: I think there can be a lot more. Most pupils learn java, many people now java. Libgdx is the only framework that use java (I think).

Of course libgdx is open source and everyone can contribute. Nobody will (and there I am sure) adapt libgdx to Metal or Vulkan tho. I am not capable of that and I do not have the time to work on a new backend anyway. Still a new backend is the only option for this framework in the future. Not in the next year or two, but at some point.

@shatterblast I do not know what youre talking about exactly. Do you talk about a new option for creating a website with GWT? -I am not saying that html is bad and cannot achieve a good website. Badlogicgames.com just is not the eyecandy it could be.
captainflyaway
 
Posts: 237
Joined: Mon Feb 20, 2017 7:52 am

Re: Discussion: Future of libgdx (backend, functionality)

Postby tomski » Sat Aug 03, 2019 3:00 pm

Yes, and just like when robovm was bought and closed by micrsoft, we had another option running within the month. There is a deprecation of opengl on ios, but that is not an indiciation that they will be stripping it immediately any time soon. There are plenty of options going forward if and when that does happen, and making one now is generally not a huge priority.

See Apple doing this with openssl, quicktime, etc. Deprecated for many many years, and still widely used.

They have receieved a huge backlash, and the move will just stop the few people targetting these platforms for their production games. iirc there have been large studios that have said if they will actually remove support, they will just stop developing for the platform. Apple do have a history of making some 'unique' decisions, so if they do actually completely forceably remove opengl and gles support on their platforms, (which i actually doubt they will personally for quite some time), we can make the backend support this. For most peoples use cases of libgdx, it will be pretty trivial to make a shim using meta/molten or whatever else.
tomski
 
Posts: 909
Joined: Wed Jun 26, 2013 6:06 pm

Re: Discussion: Future of libgdx (backend, functionality)

Postby captainflyaway » Sat Aug 03, 2019 3:33 pm

@tomski Sometimes it is not about what is the practice or what is actually happening. I am your opinion ("opinion" is the wrong word as it is a "fact" that apple will not quit openGL any time (soon)).

I think the typical indie dev reads about openGL deprecation and if he is looking for a new framework/engine to use he will pick one that supports the newest standards (e.g. Metal).

I mean people literally change their opinion on who they will elect as the next president because the candidate coughes in an interview. It does not make sense but it is still happening.
captainflyaway
 
Posts: 237
Joined: Mon Feb 20, 2017 7:52 am

Re: Discussion: Future of libgdx (backend, functionality)

Postby yobowargames » Mon Aug 05, 2019 11:38 pm

You are basing these stats on people that frequent the forum? They are unfounded. There are existing studios and game developers that actively use libgdx for all platforms, iOS included. There are many users of the iOS backend, and many games and non games that are deployed using libgdx.


To reinforce the above comment by Tomski:

I am currently working on my 4th & 5th commercial games with LibGDX.
I release on Desktop, Android, iOS, Mac and Linux
I make most of my money from Desktop with iOS a distant second place
I read the forums and occasionally answer if I have specific experience with the problem
I follow the discord channel as I can see what other people work on and find useful tools they have added - My two favorites are an in-game console and wrapper for gameanalytics.com.
yobowargames
 
Posts: 11
Joined: Sat Oct 27, 2018 4:25 pm

Next

Return to Libgdx

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 1 guest