There is a lot of tools for that, install4j for example. http://lmgtfy.com/?q=install+builder+java
Those tools are very similar to window’s programs installers.
You should use user’s directory in documents to save application’s data (config, save etc), it not only Java’s rule, my windows applications do it as well.
So you are at the state when you stopped learning new things at all? Without learning the new things, programming would be boring as hell.
I noticed that contributors have a lot of work to do. Maybe it would be useful for them if you write down all the things that are unclear for you, as for someone who is new to Java. Even the ones like ‘classpath’. Kind of a hint for them.
EDIT: IMO you need some GLSL/HLSL guru in your team no matter which engine you choose. Sooner or later there is a need to customize/optimize/rework something graphics related. Or you just need to improve the visual quality of your game, stock solutions are good at the beginning, later you would need to ‘dig into it’.
@Ogli: another option for Windows Application integration is izpack + Lanch4j. Launch4j can pull JVM command line options from a configuration file.
Running an existing gradle build should be simple. JNI - is trying to encapsulate two different programming mindsets in one place. It gets messy.
Many of the issues that you have noted seem to be simply lack of Java experience. Imagine trying to explain how to use visual studio to me, who has never been interested in using a single-platform toolset. Many of the things that are straightforward to you would take me a while to wrap my head around.
I think if game has a soul it don’t need AAA graphics at all.
JME is cool because it allows you to freely involve into open source world. Wealth people could afford some beautiful assets, powerful commercial libraries. But I live in post socialistic country and I am a student with no money and no way to access to systems like PayPal (due to country’s “features”). JMonkey and power of java free software looks super good. Unity and UE4 looks for me like hidden traps.
I’d like to give my 5 cents. I do not make a game but I want to use JME for a professional flight simulator visual system. So I’m one of those guys that experiment a lot with the graphics and I like to test and add my own stuff. I have successfully implemented a few research papers and ideas from all over the web. The result speaks for itself.
So as far as I can answer the initial question: JME goes, graphics wise, as far as you want it to!
That’s sooo 2000-ish to me. Nowadays people just run games from Steam or download an executable. “The uscrt200.dll is maybe unuseful. Delete it? (read: yes-and-pray OR no-and-clutter)”
I think that JNI can mix the disadvantages of using C with the disadvantages of using Java
However, Gradle is a problem for me, too. That’s why I copypasted an example of a working gradle that packs my game and bundles the jre (so that the user don’t even need java). I’ve also made a post about it.
Gradle is really not that problematic to learn at all
Most people started to like it once they learned enough of it, as it allows full java power in a build.
Anyway if you earn money with java at some point, you will meet Gradle (and maven), so learning it now wont hurt.
I don’t understand Gradle at all… and still curse at maven all the time
Especially when you need to do an update and for some reason it corrupt something in the JBoss so you have to do a clean but because the project is so freaking huge that clean take like 10 minutes, but you have a presentation in 30 minutes about something that nobody give a damn about and yet MUST be ready now even if the client change is mind every 3 days.
Coming back to the original question, this is a good point:
I object here. It would be desirable to have the best graphics as “stock solutions” or some variants for different things like realistic versus cel shading. That’s what a rendering engine is made for.
I also think that this is a very sarcastic answer to the original question.
How far can jME graphics go? If you are a genious you can turn it into your own personal UE4!
The thing that the OP might be interested in are some improvements like PBR and better lighting and light culling that have been added recently. And maybe things like shaderblow, some basic decals, CSG solutions and other free graphics-related libraries for jME. I still miss good soft-shadows and sophisticated particle systems and deferred shading. About the VR (virtual reality) I don’t know, but am eager to find out what jME can provide in this area.
I've got questions to the other derailed things too and I object in some points made by others. But since I've been an asshole again, I don't want to continue walking down that shitty road. :chimpanzee_wink:
I think by OP you might mean Ogli. OP was pretty clear when he said Half-life 2 which is a game many years old now. With stock JME and HL-2 assets, you’d come pretty close.
Every game needs a customized rendering pipeline, it is a matter of effects you need to have. Of course, you can put 20 processors and 10 filters but it would kill the performance. For Skullstone I implemented deferred shading, SSAO is calculated INSIDE the main shader, without another pass. I also have items and scene glow merged in one shader, but items’s have other parameters than scene’s glow.
Recently I found some article about Deus Ex rendering - they use UE but they have their own rendering pipeline!
Stock solutions should be only a base for a developer, good game engine should give you strong base and freedom with reimplementing things to meet your own specific needs.
Now I’m sure that when I finish Skullstone I’ll start another project on JME, but to have even better graphics I’ll find some OpenGL guru. Just another person in team, like programmer, concept artist, 3D artist, sound designer…
I dunno about this. I mean. Part of the reason a lot of DIY Unity games look the same is literally what happens with your desirable point. If everyone has access to the same assets, and effects implemented the same way you get games that look the same.
In saying that for a lot of games that’s ok really. Any FPS isn’t really going to benefit too much from wildly different assets gameplay wise.
I guess in the end it really depends on what your focus is when making a game.
Are you the kind of person that has great skills at making unique assets?
Do you crave the kind of control you get from coding things from scratch?
When thinking of a game are you focused on the narrative and who cares about the rest?
For me personally I look forward to the challenge of coding and dread the aspect of making 3 models that are textured and heaven forbid animated… For some people though that is more what their jam is though.
To the original question though… Is it possible to make a HL2 like game? Yes.
Right out of the box without any work to tweak the code etc? Probably not.
Should that make you discount this engine? Enh. Depends on what you want your role in the game creation process to be I guess.
It was not meant to be sarcastic at all. The OP asked a simple question, and while we might want to discuss if C++ is better than Java and what’s the best java build tool (aside from javac) I thought I’d also give a simple answer.
As long as you have a scenegraph and a way of communicating with pure OpenGL/GLSL you can do as good as anyone else (graphics wise). But it takes effort, a lot of time and also some talent. Just like with any other game engine.
Sometimes x)