Oni Central Forum

A forum for the Oni community

You are not logged in.

#51 21/10/19 7:29

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

Hmm, sounds interesting.


byproducts are fine, but where's the beef?

Offline

#52 22/10/19 4:41

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

Can you give me your first impression on this ?  ( More about the game system than the obvious errors i have to correct)

https://streamable.com/xcgo3

it's first stage of combat : long range single strikes

Last edited by onibge (22/10/19 4:42)

Offline

#53 22/10/19 8:34

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

It's hard to know what to comment on if you say you have to correct some obvious errors; how do I know which parts are errors and which are by design? ^_^  But I think you won't want to have enemies pushed back so far by attacks, because being pushed back limits the enemy's ability to counterattack, and it even limits the player's ability to chain attacks (Oni has this problem when the "fistsoflegend" cheat is used).

Also, collision and pushback seem inconsistent.  I'm guessing that you have overly complicated physics running, when you really want something very simple.  Oni only looks at whether an attacking bone (a body part that has the "attack" flag on it during an attack animation) collides with any enemy's bone, and then it plays the animation for the enemy getting hit and deducts a preset amount of HP.  There's no calculations for momentum, weight, proximity, center of mass, etc., and that's a good thing for a fighting game.


byproducts are fine, but where's the beef?

Offline

#54 23/10/19 12:34

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

+1, the push back is too strong.   No physics running. I want it arcade style wink But hell, yes, ragdolling is planned.  The pushback runs as long as the collision is running.   The problem on my mind was to know if the game looks playable at a first glance beside all the errors of quick-bad configuration. That's why game devs ask so many people to test the game.  Even if i don"t have to make a decisive choice, the current node is to know if the camera has to be parented (like i did before - just like in Oni)  or the camera has to be on free mode  (here below i made a test) . When the camera is parented, you basically see the back of the player all the time

https://streamable.com/dvs8w

Example of free camera beat them up
https://youtu.be/k4ooiR8mXTM?t=939

The advantage of free camera is that the camera is more stable so more comfortable for eye. The downside is that it's less precise but i can make the character track the enemy when hitting

Last edited by onibge (23/10/19 12:34)

Offline

#55 23/10/19 16:15

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

I think either camera system can work, though I like having control over the camera so I can look quickly in a certain direction if someone is shooting at me or I'm trying to sneak past a certain AI.

On the subject of the combat system, it's honestly hard to say how playable it is without a hands-on test to get a feel for the collision.  Are you planning to release test builds at some point for people to gets their hands on?  I know it's still very early in development.

As for pushback, if I understood you right, I think you want to simply have a fixed amount of knockback that results from each attack rather than applying pushback over a varying period of time.


byproducts are fine, but where's the beef?

Offline

#56 25/10/19 19:48

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

well, i figured out today how to have 2in1 wink    Here below in the video, the camera is set to basic mode but now i know i can do basically all i want

Yes, i think i will go 100% to the arcade method : the npc will play a mirror animation in a 1 to 3 animation system (hitted, blocked, dodge) ( In the video below, for the moment the npc just plays randomly 1 out 3 a damage animation). 

https://streamable.com/e4dgb

For the test releases, i need windows to compile the game into an .exe  I'm planning to share an .exe so people can test.. even if i would prefer contributors hmm   But i can already tell you that the fighting experience is already there

Offline

#57 25/10/19 20:24

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

Looks promising.  Do the fighting controls feel responsive?  It seems a bit floaty at the moment — no crunch or smack behind the successful attacks.

As far as test releases, you don't even have to distribute binaries at this early stage; you could put the development files up somewhere like GitHub and let people make their own builds or just run the game within Blender.  I'm on a Mac, so even if there were Windows builds, it wouldn't be very convenient for me personally.


byproducts are fine, but where's the beef?

Offline

#58 26/10/19 17:23

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

the controls work at 100% but what you see is just non-calibrated animations soften with a 10 frame blending transition. I have to review all the animations ; For example, i should re-bake all punch animations to keep the feet on the the same place.

I will not bother with Github for the moment, my engine is 5 python scripts for the moment and very simple (which is good). So anyone who wants to contribute could take care to extend/enhance a script.

- the mouse move : view + character rotation
- the button controllers : character location and action launcher  (5 keyboard button + left click) 
- the npc brain :  shortest path, fighting decision, animations
- the car controller
- the library and dictionary of functions

the huge part of the work it's all the things that decorate a game : story, sounds, music, voices, assets, lights .. and so many much animations.   I would like to set up a team but i think that nobody serious can take the project seriously before the game looks serious .. hmm   

Here below, my characters will be made of flesh and not mesh  wink

https://streamable.com/icr38

Offline

#59 26/10/19 17:46

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

Honestly, your game could be using wireframe models and if it has what looks like a fun fighting system then people with skills will want to contribute to the code or assets.  Regular players, as opposed to coders and artists, will of course want to see something more visually appealing before they try out the game.

On a separate note, I was looking into the Blender Game Engine and I'm concerned that it doesn't seem to be supported anymore as of Blender v2.8 which released in July.  Have you considered using Godot, which Blender's developers recommended in place of BGE?  Apparently its scripting language is similar to Python so at this early stage it shouldn't be too much work to port what you have to Godot.


byproducts are fine, but where's the beef?

Offline

#60 27/10/19 6:50

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

The result you see on the video : recorded at 30 screenshots per sec on low quality with low poly models and simple textures. No shaders. My laptop is old but it still can run the game at 60 fps without problem at 1920x1080. But once i screen-capture, the fps drops unless i set to low-settings.  On top of that, the animations are not finished.  Rather than showing low-quality videos, i should show the simplicity of my game system smile   I don't have so much code so i guess it's well designed.

   So why BGE ?  Because, it's simple, fun and can do many things. It would take twice much time for someone to make same on other engine as i have to get familiar with a new gui and lose time with all those little time wasting things. I'm not against the idea to port it on another engine, but i think it's better to have a complete prototype on BGE before migrating to another engine.  Also, there's forks of BGE i can migrate to without lot of problems : upbge and armory.   I've told that Godot is low frame (my game should stick to 60fps on 15 years old laptop).  BGE is not really dead.  A game engine could be considered dead if un-used for 10 years. Same for games, look Oni .. 20 years old and still playable.

For collaboration, let's say it's a delicate question. I'll be happy to share the project if people contribute to it in an useful way. But I'm afraid that some just takes all the work done just to make their own project beside without contributing back. Also, there's so many people who are just "curious" but will end up saying that "they don't have time"  and bullshit excuses, others who love to mess up the code just to show their high knowledge in python without bringing anything useful for the game. Up to now, i never see someone showing up saying " Hey, i can make all the sound " ,  " I can make maps ", "I make all the shooter effects".  I think an indie community game project can work if there's a clear common goal like  ;  "let's make Oni2, who is interested ?" so people can work independently because there's a game model   . But look, you are the only one reacting on this thread hmm 

Did you know a bit Blender, BGE ?

Offline

#61 27/10/19 16:06

Iritscen
Moderator
From: NC, USA
Registered: 22/10/07

Re: my ONI in the Blender Game engine - what i've done so far - check

Ah, I didn't realize you were recording via screenshots and not proper video-recording software.  I do have concerns over performance with both BGE and Godot, but I understand if you feel that it's more productive for you to stick with the engine that you know for now.

I used Blender for a while some years ago, but I'm not practiced with it.  I haven't used BGE at all.


byproducts are fine, but where's the beef?

Offline

#62 31/10/19 11:29

ltemplar
Member
From: Poland
Registered: 23/8/12

Re: my ONI in the Blender Game engine - what i've done so far - check

First project seems interesting, but  community saw so many project that have fallen already , that is a bit sceptical. What You heave showed us is looking promising but allow me to ask first some questions to clarify some things(that for I can't get form the videos that You've showed already):

1.Graphics
"The result you see on the video : recorded at 30 screenshots per sec on low quality with low poly models and simple textures. No shaders."
-Please don't get it wrong , but how should I(or others) know  how well your game will preform with more objects on the screen with full rendering suport(especially real time shadows with are a course since the BGE beginning - are you using some random script to generate these)?
-What types of textures&materials You're plan to include?(all existing in BGE or only few?)
-What style of graphics You will use. Anime/cartoon like or more realistic?

2.AI
How many characters witch complex AI script you can put at the stage without any loss in quality of the game?
-How AI is preforming?
-What commands/fighting capabilities it have?
-What other non combat behaviour You plan to include for it?

3.Game mechanics.
-Weapons stashing will be included?(Can player will be able to have more than one weapon at him/her)
-What types of weapons - only ranged ore melee too?
-Items?
-Inventory: RPG like  or not(similar to oni)

4.Mods
You've mentioned in one of your posts that You plan mod suport for your game. Can you tell us what options will be included?

And once again , don't take absence in your topic as bad sing. The community is quite interested in what You've made(ask Iritscen if you don't believe me) and will try to help (sooner or later), but because there were similar projects in the past(few in BGE too) that failed .. people on the froum/discord are really cautious about anything that sounds like ONI2/ONI remake(me included).

Ltemplar.

Last edited by ltemplar (31/10/19 18:11)

Offline

#63 8/11/19 20:00

onibge
Member
Registered: 29/7/19

Re: my ONI in the Blender Game engine - what i've done so far - check

Hi,

thx for replying. Sorry for replying late, i was afk for 2 weeks.

1°  On my 10 years i5, without recording, my game runs at least at 60fps 1920x1080 without GPU. My models are not that low poly but certainly not optimized (big textures, much poly on hidden zones like mouth, teeth) . So i guess there's clearly some more space left for more nice graphics and assets on screen. I think, this will not be an issue if the aim of the game it's too look like a nice 2005 reboot of Oni.    The style will be the one saw on the videos. Especially the one in the office. But yes, the goal is to have a very pleasant game on screen, but still somehow a good old  game without tons of shaders. Just the necessary as the game must be stick at 60fps in all cases.

2°  For the moment, there's 2 scripts : a navigation one and another one to randomly hit/block.  Actually, both are in the same python file.  Everyone could improve it and even have fun doing it. Scripting AI is the most fun part according to me. For the fighting, it just picks up an action to play in the library of actions. The players does by hitting the keyboard, the AI does according the result of an AI module subject of enhancement.

3°  For the objects, for the moment, can carry one. But i guess i can create inventory for objects you don't have to carry but for objects you hide in pockets like cards and keys.  I classified objects  [ object, weapon ,  car, obstacle, npc ]   according interaction(s) possible :  Take, Use, Throw, Fight.     So an object can be only taken and thrown (or used if aiming a target) , a weapon is exactly like an object but can used too (ammo) . A car can be "taken" and "thrown" ( more like enter and exit the car mode) and "used"  (oil, until chassis broken) . An obstacle, like the car, cannot be taken explicitly (as you can can't carry it) but will interacts with you in a single 1 way action.  An npc, you can only fight it until you grab it and so the npc becomes an object you can take and throw (and even use if you are aiming a target : example, the floor, another npc .. etc)

To summarize :
Take : object, weapon, car, obstacle
Use : weapon, car
Throw : object, weapon, car
Fight : npc (but becomes an object when grabbed )

4°  Well the modding , will be simply modifying/enhance the python scripts and functions. But ofc, you can replace characters with new ones but you just have to ensure the armature system is the same as the animations use these armatures. Just like you can use other animations, but have to ensure the name of the bone channels corresponds to the name of your armature used.  For the objects mentioned in point 3° ,  just need to edit the 3 proprieties and create actions for each of them.  So it's very simple for anyone who can use blender to mod the game. That's why i'm talking about a template sometimes wink


For the helpful help :

- Someone who is very good in python for creating the meta of my game (boot, assets loader, memory optimization, folders)
- Someone excellent in textures, materials, lighting, occulding and shaders (to optimize  beauty/rendering time)
- Anyone who have a passion to create story animations (for cutscenes)
- Actors for voices (probably later smile )

When game reaches a good level, migrate to upbge eevee

Offline

Board footer

Powered by FluxBB