jMonkeyEngine 3 has a good performance for heavy Android games?

I am creating a startup Minecraft clone for Android, which after the success, cast a PC version. But I saw many topics people reporting performance problems in jME Android which made me very apprehensive because the last thing I want is to make a 2-year game that does not work on plataform planned. So I decided to take the question asking this question to you: If Mythruna were converted to Android, which fps wait? If the answer is less than 60 fps, what do to improve? Thanks for listening.

Myrhruna would not run on Android… just way too much geometry. Way too many textures. It has nothing to do with JME.

If a similar game to Minecraft - Pocket Edition was created in jME, its performance would be good?

Things are much improved in jME3.1 in terms of performance. Note that lighting is still somewhat slow on Android despite all the improvements, this is not specific to jME, many Android games use unshaded or vertex lit graphics to avoid this.

If Mythruna was properly ported to Android, it would run just fine, but the graphics will look much worse than the desktop version because normal mapping has been disabled and the view distance reduced significantly.

Yes… assuming that the developer isn’t afraid of writing shaders if needed, jme can rival virtually every game engine.

Okay, so it looks good! :grin:

@Dimalenus, this actually reads “if the developer is sufficently experienced to make it work”.
Many people start with a minecraft clone even on desktop and have very poor performance because of game design problems.
A minecraft type game is already a huge challenge on desktop, it’s a lot more complicated than it looks.
Now… on android the challenge is even harder, you have to know the ins and outs of about everything in the rendering pipeline, memory consumption, partionning of the scene, etc…
And this, is not JME specific.
So the question here is : Are you experienced enough to make a minecraft clone in 2 years on android?
If you feel so then go for it. JME is a wise choice, because you’ll be able to fine tune every aspect of your pipeline.
If you don’t, then I’m not sure there is an engine out there that will help you to make this out of the box.

1 Like

Performance on phones is really hard. All phones and tablets are a bit different for a start. Secondly just because it works doesn’t mean it won’t drain battery life or even overheat! We had that happen! So you really have to work to get things to play nice across phones. And then there is weird performance patterns on that one phone while it works well on others.

Yea you really want to keep things so far down in the low end, you are just unlikely to have too many problems. You can’t really push phones/tablets.

And for that reason even testing is hard. In the past we have used a testing service. You pay a fee and they test across many tablets and phones and give you a report at the end of the week.

Thank you all for answers, you convinced me it’s a good idea to use jME for my Android project, maybe with some Greedy Meshing and code optimization, I can perform well. Why Android? Post an application in the Play Store provides an excellent advertising and facilitates the sales process, and that Play Store has an efficient system of automatic updates. :smirk:

This actually actually reads “if the developer has lots of time to invest in learning new things”, which I think is the case of the OP. I assume he don’t have business deadlines, just passion and will to learn, so JME is the perfect toolbox for that :wink:

What’s wrong with using the default emoticons? :wink: