AFTER
BEFORE
Environment skills improvement.
https://www.instagram.com/pixelappofficial
https://www.youtube.com/@PixelappOfficial
AFTER
BEFORE
Environment skills improvement.
https://www.instagram.com/pixelappofficial
https://www.youtube.com/@PixelappOfficial
not a real game, but more a demo to test p2p networking (the particles should represent the packets being sent from one peer to the other)
Working on a small testing RTS project for a few days now to test Entity Component System and animations :
@RiccardoBlb
Complimenti per lāottimo lavoro di grafica e per la UI. Il risultato tecnico ĆØ davvero notevole!
Hi @TesterY,
are you using a custom pathfinding algorithm, or did you integrate a NavMesh library to handle the movement? Could you give us some more details on how it all comes together?
Hello,
Iām using āRecast4jā library for both pathfinding and NavMesh generation.
very impressive, are you planning to add vehicles (tanks and such) by any chance?
Doing some UI and some improvements. There are some bugs that need to be fix that involves disconnections but are not consistent. The player names use RNG numbers for now.
Thanks you ,iām not really sure. i will probably add some vehicles later but as i said iām just working on this to learn Jmonkeyengine and Entity component system and some other stuffs so whenever i tried to make a real game i donāt need that much code refactors and whatever.
I Was made this models for a project in cryengine ( i was trying to remake Army men rts). here is a gameplay video :
But now iām working with Jmonkeyengine mostly because of shaders which is almost impossible to change or work with in CryEngine for me as itās kinda difficult too while itās much easier to make a small shader or change it in Jmonkeyengine.(And CryEngine just doesnāt get any updates for a while now.they said they are working on shader editor or something but still nothing.and iām a java programmer and jmonkeyengine feels kinda good to work with.)
Wow, everyone is working on awesome projects!
Not directly jMonkeyEngine related, but somewhat since the SDK runs on Netbeans, and one of the goals of the plugin was to more easily develop the SDKā¦
Iāll admit I got a bit carried away by the project I started a while ago. While the intention was to only provide fill in the middle code completion, I went ahead and added instruct type code completion, as well as an AI chat window.
Iāve also added a sort of project wide context through sending the open tab + a summary of imported classes from the same project. This works so much better than copy/pasting to chatgpt.
Now Iāve started using it to develop the plugin. AI all the way downā¦
While I wanted to use local models, I didnāt have the patience to wait for responses, so now Iām using Mistralās API.
I posted about it on reddit a while ago, and the commenters were ecstatic:
Tools/function usage (as in inserting code automatically) is almost working, but I need to refactor it and use automatic json serialization instead of doing it manually).
Itās always a trade-off: Should I work on my projects, or the tools for my projects?
testing recursive metadata rendering
one thing im having a bit of trouble is creating a scrollable list to account for when the expanded metadata panels exceed the parent container size, but it looks doable from my understanding
Scrollable infinite panels within panels is non-trivial to accomplish in Lemur right now. I have a plan to add support for this (after thinking about it for 10 years⦠I finally think I have a way)⦠but itās not there yet.
In your case, you might be able to get away with some clever ViewPort management to show your meta-data⦠the general solution using ViewPorts is hard but in specific cases where a game use-case has something specific in mind, itās possible to make it work.
Panel + scroll bar. An app state that manages a veiwport and keeps it located right over that panelās āinsidesā⦠then render the scrollable UI in that ViewPort.
(Probably this particular part of the topic is worth a separate thread if that explanation wasnāt enough of a head start.)
Progression of average character.
Added Commands and some new features.hereās a test with 700 units and it seems to be working fine :
24 FPS sometimes unplayable. What are your system specs ? GPU ? RAM etc
Looking Good. and Iām impressed with your progress !. Nice to see !!
Iām using a laptop. RTX3050 and 20 GB ram , CPU i5-12500H.
I just tried something new. instead of using bones and Jmonkeyengine animation system you can export all your model animation frames and make your own animation system with those.like that i can have like 10K units with 30 fps.(just keep in mind all this units have components and code that running on all of the each frame all the time and that takes a bit of performance too)
It seems to be working pretty well for an RTS game.because RTS games usually doesnāt have very complex animations.
Here is a demo with 5k units with around 30-40 FPS : (Edited the link)
As i said i can have like 10k units like this but if i use bones and stuff i will probably canāt have more than 1k units in same place.
Yeah, for large numbers of animated units, you have to get creative.
I even imagine you could implement batching if the animation data was compiled into textures⦠then just one index per object would give the animation frame to use. Iāve always wanted to try it but never had a āgood reasonā for my own game. (At least not yetā¦)
Yeah, like this you can have higher poly models too.with a bit of optimization it seems to be very good for an RTS game.(iām not sure if it works that well for games like Mythruna , because that game has dynamic attachments or whatever itās called on the characters which iām not sure this way works that well for those kind of cases).