What is the main thing that you feel would make your experience with this engine better? It can be an engine improvement or something that an engine does not normally provide.
I ask this out of curiosity. There is no actual certainty I’ll do something about it. I would just like to know the experiences of the users and what they feel would make their life easier if “x” was available or “somebody made this”, etc…
In before 100 “asset pipeline” comments. 
see work passion here 
i like nice graphics.
since start, i had a lot problems with shadows(point light shadows mostly). so IMO shadows improvment would be nice.
why? would like to make a nice indoor lighting with shadows, current point/spot shadows have a lot issues. annoying one is for example overlaying of this shadows.
example topic:
that you already seen 
im not sure if this still exist, anyway i had problem with shadows with multiple lights.
there are more problems, like flickering, where need special paramaters to avoid it, and more.
so i always hoped someone will upgrade this shadow system.
Asset store, hint, hint 
Apart from that, decent VFX. All that stock JME provides is mostly from the previous milenium. PBR improved things a bit, but we’re far from good in this department.
I really need to make a decent bloom filter and get into the jme itself one of these days
is our PBR bad? i feel like it look really good, the light probes might be just better. if you could, please explain me what you would upgrade, because im unsure.
edit: back what @louhy said below: some advertisment of engine would be nice to get more developers(where some sooner or later will help). Usually when someone search engine he could get into JME github page and he see screenshots of some not perfect looking games, so he might think “this engine cant make game look better” what is wrong. IMO we should add best youtube videos/screenshots from Monthly WIPs if people agree, to make developers see what nice things JME can do (on this github page and main JME page). Engine advertisment is bad here imo, lets upgrade it.
What i remember myself, when i was looking at JME, i seen nice Terrain with DOF filter video and i thought “its java and it can look like modern games, lets use it!”. So i think its just about some advertisment. (ofc if someone like Java, its one of reasons he will use JME)
Better or more elaborate android support.
For example:
That’s a really good question and I’ve been thinking of starting a similar thread, because I pretty much come up empty when I think about this.
The only thing I would have added would be related to community growth and attracting new users/contributors. Most highly active users here have really old accounts, like 5-10 years. I don’t know what happens to the ones that seem to vanish forever (abducted by aliens or whatever) but it’s a little concerning as old timers seem to “retire” and we don’t get a whole lot of new users to fill in the gaps. This is one reason I’ve put the site back into my regular daily routine to help where I can.
More on the topic, I do agree with the comments above regarding shadows. But I’ve always figured it’s hard to do in a way that looks good in every case.
A mesh manipulation library would be nice too. That’s something I’d have found great use for in the past (and possibly future). But this is hardly something turning people away from using the engine.
Better JavaFX integration maybe, but that’s external to JME. (I don’t think we really have a great solution for keyboard event management yet.)
I have it working but it’s just not live. I have to account for paying for the server that will host it - which is fine - it isn’t even expensive these days - but there’s no point buying the server until everything is in place - and other things like preparing content to be available; There’s no point in an asset store that’s completely empty - and also gives the new posters something to look at when they design their own page. I do keep actively developing it. It is still a “thing”. I’m just building up to it 
I want:
Cleaner engine, i’d like to see sketchy/unfinished parts removed or moved to dedicated branches.
Also i think shaders need some work, currently they look like a battlefield, grouping variables into meaningful structs would already improve them by 100%, even more if we can have consistent naming.
Will it work again in 9.0 and future netbeans versions? Because I am afraid we cannot stick to an old version, we should move forward (as we’ve solved the problems for the newer netbeans versions)
Most of that is related to community size though. Sure we can improve our feedback/review rates, but it’s also a matter of how many active users we have. Not sure what we can do about that though.
Want to see :
Be able to use Effekseer - Particle Effect Creation tool in JME
proper shader node, I guess @nehon and @pspeed had some ideas for it even.
More VFX
and this "Hand drawn" shader 
I cannot remember other features at the moment but will add them if remembered.
i think a lot people also dont know about Shaderblow shaders
its not upgraded for like 5 years, still should work fine anyway.
one of nice example videos:
One thing I’d love to see is more of the awesome “extra” libraries make it into jME core, if their authors are willing to go that route. The argument against that is typically that they’re optional and doing that pushes the maintenance burden onto core devs, but I don’t think that need be the case - (a) once it works and is stable, maintenance is much less time consuming than original development, (b) modern games use a lot of “extras” like ECS that jME just doesn’t address right now, forcing users to look elsewhere or roll their own, © putting these in core may encourage their authors to contribute to core (or at very least maintain their own additions), (d) since jME is modular folks who don’t use them will won’t be impacted, and (e) if maintenance burden becomes an issue little-used libraries can be deprecated if need be.
The issue with incorporating things into core is that you have to wait for glacial JME releases to get updates… and the more stuff you add to core, the more glacial the JME releases will be.
There is also 0 benefit to doing so.
How about keep it simple and rewrite any classes not java 1.8 compliant?
That at least brings the engine to the oracle FU jump off point.
I used to think like that until I worked on wiki. Opened my eyes to just how wise it is not to do that. Core team made the right move to keep engine to what it is now imo. Core functionality.
So many dead projects sent to the contributions graveyard.
Hey it’s not a graveyard if they still work! It’s a “mature project archive”. 
Actually sometimes I wonder if core isn’t already too big. Gradle makes it pretty easy to tack on non-core stuff.
The single thing I need the most is a 2D FX editor that can be handed to a professional artist without hearing in response “With Un***3D I could instead work with a much better tool…”
I would like to have a replacement for the bitmapfont we now (need to) use in our GUI’s.
With Lemur we have an awesome GUI library, but for me, the thing missing to create state of the art GUI’s is the lack of better quality font support. With bitmapfont text isn’t just crispy, good looking and not scaling well in GUI’s.