Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

Dark GDK / DGDK Open Source Project

Author
Message
Faker1089
15
Years of Service
User Offline
Joined: 30th Aug 2009
Location: My own little world....
Posted: 13th May 2010 00:35 Edited at: 13th May 2010 00:37
@matty halewood

I got a new computer so I've been trying to get darkGDK to work with visual c++ 2010 but I haven't been able to get the S3GE project to compile so I'm gonna try a few things before going back to 2008.

I'll see what I can do so if there is a page with what you guys want done I can take a look at, then I can get to work on something.
TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 13th May 2010 02:00 Edited at: 13th May 2010 02:17
Quote: "Tech whats going on with the computer ?"
I've removed all cards except my wireless NIC. Audio/Video sucks, but, it hasn't crashed yet. I'm able to compile and run VC++/DGDK/S3GE.

Quote: "Well, I've been kinda watching how this project has been going, but when I went to boot camp I was afraid of coming back to something and not knowing where it was going and in a sense that's the case now."


Good to see ya back Faker1089. The engine is steadily growing. There is a lot going on. I often jump from one section to another, getting myself confused - lol. As Matty has stated, we're gearing up the Super 3D Game Engine (S3GE) for supporting a intermediate Demo Game: METALOIDZ(working title).

METALOIDZ is a FPS w/ RPG & RTS elements, inspired by Battlezone (1998 Activision) which features Humans/Humanoids in special purpose Cyber-Armor inspired by Japanese Anime.

METALOIDZ pits 3 or more Teams against each other in race to dominate a Cyber Battlezones by building bases and conquering the opposing team's bases. To achieve this, the Teams scavenge/mine for a material called "META" in which you craft more powerful cyber-armors: shields, weapons, and utilities.

The premise and design idea for METALOIDZ mimics that of Battlezone with an additional factor for customizing. The RTS elements is in crafting cyber-armors for scavenging, guarding, blocking, generating, storage, etc. The RPG Elements are in leveling up to produce better cyber-armors: shields, weapons, and utilities. Any way that's whats in my head, and we'll just have to see what we end up with when all of our ideas are combined.

We have isolated DGDK role to a DX9 GFX wrapper. There are doubts that DGDK will be fast enough to handle the demand: dynamic shadows, huge landscapes, networking, collision and game logic of a MMO. This Demo Game will be a testing platform to push DarkGDK to its limits and flush out the systems we develop. It determine DarkGDK's role in project.

*** THE HAVES ***

So here's what the current state of affairs are in the development of the Super 3D Game Engine.

3D Physics

Fulcrum Physics powered by Nvidia Physx dev'd by Matty H. Although, Matty will have to provide the details on what features are available, Fulcrum will power all 3D physics and collision-based interactivity in S3GE. Looking forward to Matty's update.

User Interface

Multi-Application User Interface (MAUI) dev'd by yours truly. MAUI integrates several libraries: Fulcrum Physics (Physx), Box2D Physics, DarkLUA (LUA), XML Importer (irrXML), and HTTP File Transfer (DarkNet). Effectively the `Core` of the Engine, MAUI consolidates all User Interface in S3GE. MAUI has scripted Themes, Schemes, Transitions, and Behaviors for Buttons. I'm adding Behaviors for Textfields this week.

With Buttons and Textfields, we can start developing R & D Tools within S3GE as we buildup the Demo Game. These tools will most likely become full S3GEd Applications. I will add Sliders, List Boxes, and Progress Meter behaviors as along the way, but, these Button & Textfields will be enough to get Apps started.

About networking. I discovered issues with DarkNet with working out the HTTP Transfer. Mike Pryor made an update that fixed the issue.

3D Environment

Automatic Environment Mapping (AEM) dev'd by Haliop. AEM is a procedural Environment Creation System. Content Creation and User Created Content is factored into S3GE's design, AEM is step in that direction.

*** THE HAVE NOTS ***

There are several core subsystems that are missing from S3GE that are required to produce an above average First Person Shooter. Some subsystem can be built on top of other libraries and may require R & D.

Artificial Intelligence

Pathfinding (Navigation Mesh, A*): Recast Nav Mesh Library to be implemented by Marlou. Unfortunately, the state of the lib implementation is unknown.

3D Effects

Particle Engine - TBD
Lighting System - TBD

3D Environment

Terrain (Blitz Terrain) - TBD

Other Issues

Several Contributors have had issues with SVN. It has been decided that the project directories will need to be restructured to minimize conflicts and organize the project more effeciently.

Due to these issues, I've posted several YouTube vids on the GameProgrammerz Channel to show my progress. I'm also finding this media very useful for conveying ideas. In fact, I may start creating more vidz for updating DOSP Project-Trio.

haliop
User Banned
Posted: 13th May 2010 02:32
while walking the dogs at night , seing all the spot lights outside gave me an idea on how to optimize dynamic lighting .
Dynamic Light Triggers. using the volume or zone or region method.

Volume/Zone/Regoin method.

each dynamic light will have a 3D hidden box around its center and streatched to its farest ray.

if the Player will step inside that region then the dynamic light will get the higest priority of all surronding lights.
but when the Player isnot there, then it Triggers a Static Light effect or Baked lighting.

if a Player is in Region one , and an Enemy is at Region two.
Region one will be the higest priority light
Region two will be the second to higest since there is a dynamic light Actor or Entity inside it the demands further calculations of light, how ever , since Region Two is only Secondary and it has some nice distance from Player or (ppov - player point of view) region two light will be much less accurate and will demand much less GPU power then Region One, with it since its not the Primary Light, Region two light will "calculate" only the dynamic actor shadows and everything else will be Baked or PreRendered.

so the player wont feel like here the light looks good but there it looks ugly, if this is achived properly we will get a super Shadow quality and all that surronds it.

Attached a Hand Drawing Examples.
Faker1089
15
Years of Service
User Offline
Joined: 30th Aug 2009
Location: My own little world....
Posted: 13th May 2010 03:00
I looked around for particle systems and came up with magic particles which has a wrapper for darkGDK already (has an editor for making custom particles more easily). As for a lighing system....I hope someone is good with programming directX stuff because there aren't any external/library independant lighting systems that are going to help us besides dark lights (which I have) but I don't know if TGC would be happy if I started passing it around.
TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 13th May 2010 13:54
It's my understanding that Magic Particles is limited to 2D only and from the looks of it TGC has disconnected most of its links to the product However, I do like the idea of an ready to use Editor.

A Particle Engine was the first Game-related program I wrote. In fact, i've written simple particle engines within a day. I'm highly in favor of developing a custom Particle Solution within S3GE. One that will support Scripted Behaviors, Shaders, Texturizers (AEM & CHIMP), Physics and make use of Animated Sprites and 3D Polygons. Particles can be used to support a variety of visual effects {Fluids, Pyrotechnics, Weather Effects (ie: Rain, Snow, etc), Foilage/Flora, Lensflare, Decals, Meshes (Ammo)}

After dev'ing MAUI I have some new ideas, but being Lazy, I wouldn't start from scratch and would build on top of a decent open source particle engine. Scripted Behaviors, Transitions is feature I would like to add and I think I would need access to the source to do so. I took a gander at Magic Particles, however, its not Open Source. I've been eye-balling several other Particle Engines that are open source and computationally independent from rendering which could be adapted to our S3GE Engine.

http://www.particlesystems.org
http://spark.developpez.com

Anyhow, I'm just rambling on. I have more than enough on my plate and would really like to get into the development of character animation systems for S3GE. Faker1089, if you want to take on integrating Magic Particles into S3GE, go for it.

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 13th May 2010 17:40
Quote: "matty whats up with Fulcrum ?"


haliop - Ive had a small break from it, I unofficially released the latest version in my the Fulcrum thread while I create the documentation and fix a couple of small issues.

Thanks for all the info on static lighting, are you sayng that new objects are created using the lightmap for textures and then placed over the original object? If so, will you have double the poly-count, it seems thats what you are saying but seems a bit slow to me.

Quote: "each dynamic light will have a 3D hidden box around its center and streatched to its farest ray."

Yeh, this kind of optimization is very important although we need to create our different lighting options first.


Quote: "The premise and design idea for METALOIDZ mimics that of Battlezone with an additional factor for customizing. The RTS elements is in crafting cyber-armors for scavenging, guarding, blocking, generating, storage, etc. The RPG Elements are in leveling up to produce better cyber-armors: shields, weapons, and utilities. Any way that's whats in my head, and we'll just have to see what we end up with when all of our ideas are combined."

Techlord - My FPS/RTS is more FPS than RTS.
PART 1.
Three factions, all real players online, the world kinda ended blah blah blah, huge map in antartic or desert somewhere where oil wells can still be tapped, probably 7 locations where oil can be extracted, thats 2 pretty close to each teams spawn and a larger center one, squads of four each have a vehicle, this is there spawn point at all times, if vehicle gets destroyed you spawn back at home base.
To be continued....

Not saying this should be the game, just letting you know what the game I have been wanting to make for a while is, like you said, we will probably take elements from everyones ideas.


Quote: "As for a lighing system....I hope someone is good with programming directX stuff because there aren't any external/library independant lighting systems that are going to help us besides dark lights (which I have) but I don't know if TGC would be happy if I started passing it around. "

Faker - I have been looking at some existing shaders, I should have something solid to show you guys soon.

Hassan - Still having trouble with exporting the animations? I am looking to do a lighting/collision/model demo, if you've give up on the model let me know and I'll still put it in, but perhaps lying on the floor dead or something

haliop
User Banned
Posted: 13th May 2010 17:49
hassan what program are you using?
pherhaps export it as other formats and use blender to export it as X file with a new exporter shown in the 3D subsection.
Hassan
16
Years of Service
User Offline
Joined: 4th May 2009
Location: <script> alert(1); </script>
Posted: 13th May 2010 18:47
I'm using max 2010, and i gave up, it doesnt support .x exporting by default, and there are only 2 plugins for exporting, both of them sucks, different results, both looks weird

for some reason i feel like the problem is caused by the bone's arrangement, so i will try once again, i hope it goes well, because i want to get this model in my engine's demo too ()

Also, matty, really sorry but, for some personal reasons i can't release the source of the lighting shader

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 13th May 2010 18:54
Quote: "Also, matty, really sorry but, for some personal reasons i can't release the source of the lighting shader"


No problem, I am using one of Evolved's(converted by mista wilson) at the moment, we also have pauli's one and I think he is making more. I may have questions on some of the shaders we use or create if issues arise, it will be cool if you stick around to help us with any possible shader related questions.

haliop
User Banned
Posted: 13th May 2010 21:37
i will animate it using blender, i already have a simple bone setup i will have to adjust it for the long mech's fingers but that shouldnt be too much of a problem and see how it goes.
i'll do Walk and Run just to see if i can export it well.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 13th May 2010 22:14
Quote: " will animate it using blender, i already have a simple bone setup i will have to adjust it for the long mech's fingers but that shouldnt be too much of a problem and see how it goes.
i'll do Walk and Run just to see if i can export it well. "

Cool, Hassan let us know if thats OK.
Haliop, would and idle animation and a run animation be better?

haliop
User Banned
Posted: 13th May 2010 22:53
sure, but before i start.
what kind of animation type would you guys think that can handle this model ?

like i can make it run well , but becouse he is or it is a mech, how should i approach it?
just a normal human figure
or something else..
Tech said something about Anime i'll try to check youtube for something like these types of action figures.
Hassan
16
Years of Service
User Offline
Joined: 4th May 2009
Location: <script> alert(1); </script>
Posted: 13th May 2010 22:59 Edited at: 13th May 2010 23:20
yes do whatever you want with it, just as if it was yours, well, i think it should have idle/run/jump/attack/get hit/die animations

@matty: It's not like im very experienced with shaders, i got alot of help from some of my books, but, alone, i havnt really wrote such "useful" shader, but im going to help if i can



****************************************************************
EDIT
the latest model/texture is at my first post page 25, link:
https://forumfiles.thegamecreators.com/download/2012386
****************************************************************


haliop
User Banned
Posted: 13th May 2010 23:07 Edited at: 13th May 2010 23:09
Quote: "My mecha concepts are Japanese Anime/Manga inspired by Guyver, Gundam, RoboTech, and many others. "


i have some learning to do , when i was in the army the first rule before engaging a task , is to learn what you can on that task, then find the best way (sorth of a path finder) to accomplish that task.

so Youtube will be my best friend for tonight, and hopefully in a few hours i'll get something that represents the Mech idea.
i just dont want to make it walk like a normal joe cause its not...

1st edit:Guyver Episode One.. here i go
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 00:08
Quote: "i just dont want to make it walk like a normal joe cause its not..."

Funny this issue has come up, I had an idea earlier and it goes something like this.
Broken, run down mechs with life-like chracteristics. Do you guys remember the bad droid(general grevious) in 'revenge of the sith', firstly he wore a cloak and secondly he coughed quite alot like he had a breathing problem lol, anyway I thought this was cool. He also walked like a human, crouched over.
You could take this as far as you like, mechs with machine guns because the lasers they shoot out of their eyes don't work anymore, binoculars, smoking cigars(bender) etc.

Anyway I thought this would be a cool concept, give a different look and personality to the game, I know its strange, have a think about it guys .

haliop
User Banned
Posted: 14th May 2010 01:16
well , i did actually thought about a taunt system..
making them dance and stuff but its just time consuming.
what i had in mind is the mecha walk a bit crouched when one leg is higher then the other, one hand is cliping its finger and on the other he has a large old school speaker... lol... ok im tired and youtube sux , i got to see 4 episodes of Guyver and it gave me some ideas but suddenly they stop , flash stoped reacting and i dont have any desire to fix it right now , will continue when i'll wake up.
Hassan
16
Years of Service
User Offline
Joined: 4th May 2009
Location: <script> alert(1); </script>
Posted: 14th May 2010 11:28
So..i re-made the animations and i still get the same results, this plugin is totally dead, good luck haliop

haliop
User Banned
Posted: 14th May 2010 11:39
isnt there anything at the 3D section for exporting animations?
Hassan
16
Years of Service
User Offline
Joined: 4th May 2009
Location: <script> alert(1); </script>
Posted: 14th May 2010 11:55
I asked about my problem there, and got a vague answer, and the guy isn't replying anymore

haliop
User Banned
Posted: 14th May 2010 12:30
i see well there should be a solution not only for this project but for 3D max 2010.. as alot of ppl use it with DarkBasic Pro.


ok , so i have imported the model , there were about 400 duplicate vertices points which were cleared..

and now i'm taking the Biped rig that i have and fully working and combining it with the model. i do not know if it will export but i will try also there is a new exporter for blender which works good with DarkBasic Pro using X files so i think it should be good but i wont know till i'll try.

however , a thought have come to mind.
do we use Assualt Rifles in the game? or is it like the anime mech where it shots lasers from its ass.. i have no clue cause this is a need to know before starting animation , if it uses a weapon of any class that he holds by the hands , then i need to know in order to now just make a run animation but a run animation where he holds a weapon.. altough we can always holster the weapon which is agreed in the game comunity.. well i'll try my best to create a real looking animation but i do need to know if we use weapons or the weapons are inside the mecha body.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 12:57
Quote: "well i'll try my best to create a real looking animation but i do need to know if we use weapons or the weapons are inside the mecha body. "


We just don't know at the moment, I suppose an animation without a gun will be needed in either case so I would go with that. I think finding the right tools for the job is most important for now so any animation will do for that purpose.

I am putting a demo together, showing off dynamic lights/shadows and collision, shall I wait to include an animated mech?

haliop
User Banned
Posted: 14th May 2010 14:56
you should do the demo with a spehere or something and implent the model when its done , you can also take the model presented here , and use it just move it without animation , then when animation is done you already have everything else, how does the lighting coming along?
haliop
User Banned
Posted: 14th May 2010 17:44
remodeling Mecha's
Hands
-----

i am still at the Bone configuration setup. but i have notice something that i as an induvidual who will play this game and as a gamer.. causing me hard time understanding about mecha.

where is the hand?
if we will use any weapons that is not stationary on the Mecha's body and we want it to hold it , can anyone of you see it holding it with its huge finger nails? i cant .. i try to imagine but it means that the weapons will have to be so big..

so 2 thought have come to mind.
one : build a normal thumb and hand behind the large finger nails and use the finger nails as "Wolverine swords" for a decent meele attack , with the option to costumize these swords.

two : tech has mentioned something about a Blue Plasma Glue or something for the joints, so we can actually make a Blue Plasma hand , that is a constant FX on the screen that could be really awesome thumb and fingers and the hand are FX animation rather then a Limb however constant FX will probably cost alot of gpu therefor returning to solution 1.

anyone disagree before i'll start to model?
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 19:26
Do whatever you think is best, there is no guarantee that this particular model will fit into the final game, depending on what art direction we take.

The main thing is that we know we have all the tools and all the tools are compatable with each-other, I have just come to the conclusion that 3DWS is not suitable for our needs, I'm looking at mapscape at the moment as it seems to be a good option as lighting and physics seems to work well whether exported as .x or .dbo.
I have done a video, I'll upload it to you-tube and post back later.

haliop
User Banned
Posted: 14th May 2010 20:29
i dont reccomend using MapScape.
actually i dont reccomend using any application rather then the 3D Render and Modeling Application like 3D Studio Max , Maya , Bryce , Blender etc..

and i'll explain why:

since this is a demo for the Engine and not a demo for ourselves.
we need to approach level creation as something anyone can do wether expriance or not with game development.

keeping this in mide , if we use Exterior Editors like MapScape or any other World Builder we actually demand the less exprianced users to watch the demo , but not only to understand S3GE code but also to understand how to work with MapScape or anyother World Builder.

now , if we keep it to the well known 3D Render or Modeling application as mentioned above , anyone can use them and who ever is coming to the Game Development area will have to use one of them , therefor he will expriance the 3D Modeling and 3D World Building in the Same Software.

with that said , currently there is no way to run from Particle Systems as they are more advanced in nature (LOD wise) , so we can use an exterior Particle System / Editor.

i still think we should find a solution to lighting using DarkGDK or DirectX or both , but not go too far.
plus we must understand that there is no rush.. if we want a decent demo we shouldnot find a quick solution , we can talk and think about new solution all the time it will benefith us all once we will know how's one think we can adjust ourself to that and create at the end a much better demo where all the parts may look simple but are actually very well thought and have a good balance.

cant wait for the video im really excited looking at it .
for now i've put a very simple altough high poly about 300 for a hand , one thumb and combined fingers, the Bone Structure im using is very detailed and very easy to manipulate and scale and everything Blender is really good when it comes to this area, one just needs to practice to get very high results.

just no need to rush , lets make this demo the best way we can.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 21:03
Sorry haliop, you have lost me in quite a few places there.
Why not add mapscape to that list, max, maya etc?
After all, it was made for DBPro and DGDK, it could be our recommended CSG editor. For someone like me, who has never modelled anything, CSG editors allow me to show ideas quick, obviously more complex stuff will need max/blender etc.

Quote: "keeping this in mind , if we use Exterior Editors like MapScape or any other World Builder we actually demand the less exprianced users to watch the demo , but not only to understand S3GE code but also to understand how to work with MapScape or anyother World Builder."


I don't demand anything from anyone, coders code, modellers model, as long as everything is compatable we'll be fine.


Quote: "i still think we should find a solution to lighting using DarkGDK or DirectX or both , but not go too far.
plus we must understand that there is no rush.. if we want a decent demo we shouldnot find a quick solution ,"


I think you may be mis-understanding the purpose of this demo/video, I am wrapping one of evolved's shaders so S3GE users can just do this:

Now they have a dynamic light that casts shadows in a few lines of code because S3GE does all the work.
The demo is just to show the dosp team the light effect and then see how fast it is on various computers.

This demo/video I am putting together is no way representing the game we will make, don't think I am rushing ahead, I am just trying to find us options for our game and for S3GE.

haliop
User Banned
Posted: 14th May 2010 21:19
Ok i think i just went with my crazy head places.. nvm that.

i still think we should consider building the level with Blender / Maya and not mapscape
what you say is true about mapscape... but i dont think its a way to take, altough , it can be added to the list of programs attached with s3ge until we'll create a level editor of our own that is.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 21:26
Yeh sure, build levels with anything you like, if you are familiar with blender the build a simple test level, I could test it to make sure Fulcrum can make a tri-mesh from it and that various lighting effects also work on it. If they don't, we then need to look for solutions or workarounds or change the lighting/collision to suit. In an ideal world everything would work but thats not always the case, we need to know what our options are.

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 14th May 2010 21:54 Edited at: 14th May 2010 21:55
VIDEO




This lighting option is quite cpu intense although I get 90-100 fps on my laptop, thats with collision, I will post an exe for everyone to try and see how it performs.

I see this lighting as an option for small enclosed spaces, at the moment the light has to be placed where the camera is but I'll look into changing that to give more options.

I had trouble lighting x and dbo files exported from 3DWS but mapscape does both perfectly(and its free).

Isocadia
16
Years of Service
User Offline
Joined: 24th Jul 2009
Location:
Posted: 14th May 2010 22:08
NICE,

Isocadia

Hassan
16
Years of Service
User Offline
Joined: 4th May 2009
Location: <script> alert(1); </script>
Posted: 14th May 2010 23:12 Edited at: 14th May 2010 23:12
very nice indeed!!! also, you should make the character stop when it hits a wall or an object, makes it more more realistic

i think the fps is somewhat good, because shadows already consume much cpu, and maybe physics too?

oh before i forget



haliop
User Banned
Posted: 14th May 2010 23:22 Edited at: 14th May 2010 23:25
hi matty here is a simple x file from blender , nothing much in it , just a huge plane and some tall boxes making it look like a non-textured city, if you want to check it pls do .

what do you mean when you say tri-mesh , i havent tried fulcrum yet.

ohh forget the city map , a country side with the same idea.
you may need to scale it im not sure.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 15th May 2010 01:20 Edited at: 15th May 2010 01:49
Quote: "what do you mean when you say tri-mesh , i havent tried fulcrum yet."


You can give any static complex object collision in one command, either FulcrumPhy::makeTriangleMesh() or FulcrumPhy::makeTriangleMeshFromDBO().
This is a great way to set up collision without any effort although for a real game you would have two maps, one for visuals and then another simpler version for collision, or you could create your collision world out of primitive objects, I think making it as you make your map is better.

So, anyone making official maps/buildings/structures for our game, keep another simpler version as you go along, we can use this for collision.

EDIT:
Quote: "hi matty here is a simple x file from blender , nothing much in it , just a huge plane and some tall boxes making it look like a non-textured city, if you want to check it pls do ."


It works, but you need to do the following:

If you dont, everything still gets created but the positions/rotations are wrong for some reason. Also, you may want to rotate your maps so they are the correct way up in DarkGDK, +90 or -90 obout the x-axis should do it, I think.

haliop
User Banned
Posted: 15th May 2010 10:25 Edited at: 15th May 2010 11:49
yeah what i do is



btw i use the same object id for the mesh id then turning the mesh id to the same object id that was before , therefor excuting much less calculations.

here is a better example for what is written above:


this gives us a better control over objects and less calculations
for getting a free object or a free mesh.
so if the object is 100 mesh is 100 also and so on and so..
it might not work if its in the loop as other meshes and objects may interfer, but for loading an object like a map or any other in the beggining of the project before entering the main loop , it should work , and save us some coding time same as saving us to retrive from the resource manager a number of the current editable ID number.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 15th May 2010 15:26
haliop - The dynamic lights are not working properly with that map you give me, it could be because of the way its textured, could you make me a small room, fully textured with maybe a box in the middle or something, thanks.

What program was used to make the mech warrior? That seems to light quite nicely.

haliop
User Banned
Posted: 15th May 2010 19:51 Edited at: 15th May 2010 19:57
mech was built with 3D studio Max 2010.

yeah , now after using blender to build a map , i would like to do it in mapscape.. much more easier..

i think S3GE should have a built in Map Editor.. we must have one.

ive done some CSG oparations in blender , and when you look on the a box mesh after cuting a cylinder inside it , man thats not a pretty look.

also blender dont give a damn about preformance , it just throwing all kind of vertice inside the model that shouldnot really be there , thats weird man.. i dont think blender is a good program for this .

puting that aside.

--------------------
UDK - has a fully featured CSG builder inside it with lightmaps and everything, and it is extremlly easy to use , we need something like that.

the question that i have is:

we want an open world expriance
therefor we need one big Terrain and Model for Buildings and such.
can we use MapScape as the Buildings Modeler and the GDK Terrain System (or Blitz) .

btw , does blitz work for you guys?
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 15th May 2010 21:10 Edited at: 15th May 2010 21:11
The last GDK version of blitz terrain was built wrong and then kaedroho started having trouble with his computer, I'm hoping we will have blitz terrain soon.


Techlord - You are very quiet, do you agree with the direction we are going in? If not, let us know what you think.

I have been thinking about character animation/movement, I wrote a plug-in for myself once that was to take care of all this for me but it was really ugly, I never realised before I started, how complex it can get.

So, I think I have come up with a good design to take care of all this for us in a general way that can be re-used for any character(non-character) model.
Before I thrash out the design, I need to know that you(tech) have not already implemented something similar? So, do you have anything so far that will handle animations for us?
I can't really go into too much detail without diagrams as it gets a little abstract, if you do have anything, let me know what it is and how it is used, thanks.

haliop
User Banned
Posted: 15th May 2010 22:58
i hope blitzz will be aviable soon also as we must use or something similar i dont know anything about how to implent LOD and im not so good in c++.. nvm that.

animation will be finished tommarow. im putting my self a deadline cause im just playing with it with no results..

so , you can expect it tommarow i just hope i'll be able to export it and that the shader will work on it.\
one more thing , i just played Battlefield bad company 2 , and when i looked at the shadows , i just saw damn.. thats some nasty shadows. what they did is , only player models have high soft shadows everything else is very low , you can actually see the Pixels like the only games .
and its still an awesome game, how ever to the ones having DirectX 11 they have a Soft Shadows that supports these kind of cards and they (the ones who own DirectX 11 cards) see a much smother results , everyone else suffer from a pixel not perfected shadows, so its something to think about.
Faker1089
15
Years of Service
User Offline
Joined: 30th Aug 2009
Location: My own little world....
Posted: 16th May 2010 00:34
So you are suggesting we use somewhat nasty shadows for most of the objects except for the players' and those that will add to the environment? I can see how that will be beneficial for the program speed especially on older computers.
Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 16th May 2010 01:46
I think static lighting/shadows is the way to go most of the time, you will notice on my lighting video that everything is lit dynamically, I don't see this as the lighting we will use most of the time.
In my opinion, we need to light the map/level with static lightmapping inside the map/level creating tool, then find a way to just add shadows for the players/vehicles and any other dynamic objects, but on top of the lightmap. Hopefully there will be a shader out there that does this, it needs investigating.

Also, I think any game we make should cater to the high-end gaming PC's but also check for specs and disable certain things for lesser computers. You don't want to show off your engine on a mediocre PC, we want to push it as far as it can go but still make our games playable to most people, if not everyone.

Networking:
Has anyone ever done any, preferbably with DarkNet?
I would say we could do with getting a small demo of two players meeting on a local network. Then look into general techniques for smooth online gaming over the internet. I'm not sure what the game is yet but we have no AI at the moment so online may be the way to go, what do you think?

haliop
User Banned
Posted: 16th May 2010 07:41
yeah man , in my opinion Ai is useless these days unless you are building a Single Player or a Co - Op which is single player with a friend .. lol.

however , if we are building Ai and pathFinding it could take a while to a acomplish a good or decent feel, we dont want the old style Ai , shoot , run , take cover .. we need something to blow your mind, i do suggest Evolutanry Ai but thats really not for now.

ok ,im starting with the Mecha Animation, its 7:39 am here in israel, i'll post back in the evening and im going to rush it in any way so it will be done until about 8 pm, i got 12 hours and im free today , lets do this! yeah baby yeah! /"do i make you horny? "
(Austin Powers , why all movies have to be Trilogy and not Quadlogy .. ).
TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 16th May 2010 12:47 Edited at: 16th May 2010 12:59
Wow, you guys are busy! Sorry for the disappearing act. My PC has been out of commission, installed a new GeForce 7600 GS yesterday. Yes its old, but, I rather spend 100 then a 1000 at this point and time. I'm back in action and coding MAUI Textbox Behaviors.

Quote: "
1st edit:Guyver Episode One.. here i go"
The Guyver series is one of my favorites. I'm also a fan of the Armored Core Franchise - too bad the First Person view sucked.

Quote: "Techlord - My FPS/RTS is more FPS than RTS.
PART 1.
Three factions, all real players online, the world kinda ended blah blah blah, huge map in antartic or desert somewhere where oil wells can still be tapped, probably 7 locations where oil can be extracted, thats 2 pretty close to each teams spawn and a larger center one, squads of four each have a vehicle, this is there spawn point at all times, if vehicle gets destroyed you spawn back at home base.
To be continued...."


Have you ever played Tribes? The its the game in which the Garage Games Torque Engine powered. It had similar objectives, and other variants of capture-the-flag.

No reason we cannot blend FPS/RTS/RPG elements into a single tight game play and blend the themes. For example, take the same idea in the quotes above, adlib with a different theme:

Quote: "Three factions, all real players online, on a alien world, with huge map in frozen iceland to desert wastelands. There are Seven 7 locations where META can be extracted, thats 2 pretty close to each teams spawn and a larger center one, squads of four each have a Cargo Ship, this is there spawn point at all times, if Cargo Ship gets destroyed you spawn back at home base."


I've played many games. I can blend themes in my head simultaneously while I'm playing. Swapping out Zombies for Aliens for Demons for Enemy Soldiers. Swapping out Cars for Marine Craft for Flying Saucers for Horses w/ Rocket Packs. You get the picture. The purpose defines the Game Entity, not the appearance and I adopt this philosphy in my software design.

I'm all for a theme other then modern, because we can define the rules and content and not have to compete with the high expectations for realism. Whats in my head as far as media goes, is based on my personal media development skill. Also, a major feature I desire to achieve within the demo is some limited form of entity customization that can be developed into a more comprehensive content creation system later on. Players want flexibility, so do Developers. S3GE needs to showcase this power from the start. Otherwise its just another FPS Creator demo

Quote: "So, I think I have come up with a good design to take care of all this for us in a general way that can be re-used for any character(non-character) model.
Before I thrash out the design, I need to know that you(tech) have not already implemented something similar? So, do you have anything so far that will handle animations for us?
I can't really go into too much detail without diagrams as it gets a little abstract, if you do have anything, let me know what it is and how it is used, thanks."
At the moment, I only have ideas for integrating pre-processed and on-the-fly animation. 3D Animation is the final frontier for me as a game developer. I'm very interested in your design. Perhaps, we can co-develop on it.

Quote: "Networking:
Has anyone ever done any, preferbably with DarkNet?
I would say we could do with getting a small demo of two players meeting on a local network. Then look into general techniques for smooth online gaming over the internet. I'm not sure what the game is yet but we have no AI at the moment so online may be the way to go, what do you think?"
We do have AI. AI is implemented in various layers and S3GE has scripting capability. Event-Triggers are a part of the AI system. Auto-Steering can be implemented easily in many ways. A simple procedural method: establish target point, turn entity (if necessary), move entity forward. However, advance techniques suchs A* pathfinding Algo and Navigational Meshes are better suited for complex geometry navigation.

I've integrated HTTP with DarkNet and reviewed the `Cube World` Server/Client demo that comes with the DarkNet Tuts. I have also taken a gander at the Recast Nav Mesh Lib. Both may require the equivalent amount time to implement properly. Although, I'm a big advocate for networking and would pursue its implementation first, it is not a substitute for pathfinding. Pathfinding can be used for various purposes to include Editing Tools. There is no reason why it cannot be implemented simultaneously. The question ultimately becomes, whos going to implement it?

haliop
User Banned
Posted: 16th May 2010 14:02
wait wait wait , why do we need a Path finder and why do we actually need Ai ? we are building a Multi-Player Demo..
Multi - Player Player Player Player.. i dont see why we need Ai at this moment.
TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 16th May 2010 15:11 Edited at: 16th May 2010 15:20
Quote: "wait wait wait , why do we need a Path finder and why do we actually need Ai ? we are building a Multi-Player Demo..
Multi - Player Player Player Player.. i dont see why we need Ai at this moment. "
AI is inherent to any computer program. A conditional construct such as IF, is AI in its simplest form. Auto-Steering aka Pathfinding is a part of AI. Someone could ask why we need networking? For me, its not a question of why, but more so a question of who?

I have already worked with DarkNet, I have elected to work on the Server/Client code firstly. This does NOT mean someone else cannot implement a Lib such RecastNavigation simultaneously. in fact, I was under the impression marlou was working on it. Unfortunately, we do not know marlou's progress or intentions within the project

Auto-steering can be as simple as taking a `bullet` and moving it in a straight line or as complex as navigating a bot in a 3D maze. There are techniques to do it. We just need someone who will implement them. Thus, the questions remains: whos going to implement pathfinding?

PS: IMO, both multiplayer only and single player only games have short life spans. This is not my intention for games developed with S3GE.

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 16th May 2010 15:40
Quote: "I've played many games. I can blend themes in my head simultaneously while I'm playing. Swapping out Zombies for Aliens for Demons for Enemy Soldiers. Swapping out Cars for Marine Craft for Flying Saucers for Horses w/ Rocket Packs. You get the picture. The purpose defines the Game Entity, not the appearance and I adopt this philosphy in my software design."


Yeh, I agree completely, I make up the story/setting to fit the gameplay I desire.

Quote: "At the moment, I only have ideas for integrating pre-processed and on-the-fly animation. 3D Animation is the final frontier for me as a game developer. I'm very interested in your design. Perhaps, we can co-develop on it."


Sorry, I may have been a little vague, I am actually just talking about a system to handle the pre-processed animations.
Without any thought or planning you can end up with messy non re-usable code like this:

I am looking at ways to split the problem up into several self contained classes which when working together will simplify this problem and lead to faster code without 'if-else' statements, and be re-usable with minimum effort.
When making a game, I see this as one of the first problems you come across, I was just wondering if you have anything already?

Quote: " There is no reason why it cannot be implemented simultaneously. The question ultimately becomes, whos going to implement it? "

I don't really see the link between pathfinding and networking, why do they need to be implemented simultaneously?
It sound like you may be the best person to implement the networking, but I know you are already doing alot, we need someone to step up to the challenge

TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 16th May 2010 16:18 Edited at: 16th May 2010 16:24
Quote: "I don't really see the link between pathfinding and networking, why do they need to be implemented simultaneously?"
Did you know that networking (more particularly network routing) uses the Dijkstra's pathfinding algorithm to find the shortest path?

My point is, retrofitting is more time consuming. It seems we are ruling out pathfinding because no one has signed up to implement it. Well, with the exception of myself and Faker1089, no one has signed up to implement the server/client. Both systems requires implementation. Pathfinding requires the management of one state, networking requires two. Do not rule out pathfinding as its implemented in one form or another.

Quote: "It sound like you may be the best person to implement the networking, but I know you are already doing alot, we need someone to step up to the challenge"
We are at the mercy of volunteers. I intend to work on server/client, but, MAUI has priority at the moment cause its in half-ass usable state. Textboxes are the toughest GUI behaviors to implement.



Looks like scripted (AI) Animation to me.

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 16th May 2010 17:11
Quote: "My point is, retrofitting is more time consuming"

I'm sure I am probably missing something as I just don't see it, I do know however that you have been in this game alot longer than me so I will trust what you are saying, which I think is that we need pathfinding and networking to be implemented together?

Quote: "Looks like scripted (AI) Animation to me."

Yeh, once implemented, it would be best controlled using scripts, or better, a tool that allows you to visualise what you are doing and how the model looks when transitioning from one animation to another(via interpolation). The easiest way to show you what I have in mind is to show you what the tool might look like that will make setting up model animations easier, I may attempt a little drawing later.
Although I don't want to rush ahead and do something that you already have plans for, I suppose I just need the go-ahead.

Matty H
16
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 16th May 2010 18:51 Edited at: 16th May 2010 18:55


This is what I'm talking about at its most basic level.
A button or event can be mapped to any of the three types here, state, animation or abstract animation.
You could have it so the space key changes 'state' from 'upright' to 'crouched', this would automatically point the abstract animations to their 'crouched' animations.
'Animation' is just a class/structure, at its basic level it just has two integers,a start frame and an end frame.
'Abstract animation' is just a pointer that holds the address of one of its possible(linked) animations, depending on what 'state' your character is in.

Anyway, it gets a little more complicated, but this diagram is also what I would see a tool looking like, and you just create animations and link them together as I have shown.

Anyway, I don't want to go too far with this if Techlord already has a system in place inside maui or something else, let me know what you guys think, thanks.

P.S, I have not put alot of work into this and will drop it instantly if needed, so don't be afraid to tell me straight lol.

TechLord
22
Years of Service
User Offline
Joined: 19th Dec 2002
Location: TheGameDevStore.com
Posted: 17th May 2010 08:46 Edited at: 17th May 2010 09:13
Matty, that diagram looks great. I only have some vague ideas for dealing with some procedural techniques. Have been gazing at libraries like Cal3D. We can build on your system.

JTK
15
Years of Service
User Offline
Joined: 10th Feb 2010
Location:
Posted: 17th May 2010 09:33
Ok, I've been watching this thread for awhile now, so I feel I should speak up before you all progress any further. At the risk of being flamed here, just know that I say these things to help out.

First, I think it's great that you all are in this project together. You're all trying to make a very good (suite of) product(s) and I look forward to seeing the results, but;

And maybe I need further understanding here: I think some areas are better left alone (for the time being)...

MAUI - Not quite sure of the concept, but to me it seems like a data-driven UI (data being stored in XML-script files). This may be well and good - you would have to have a UI editor (not unlike a resource editor) to pull it off. From what I've seen so far, I think it's rather convoluted and difficult to follow. I like the idea of data-driven UI (for any application - including games) but for S3GE, is should be easy. That's not what I'm seeing so far... I know you've spent alot of time thinking about this tech, but make it easier - we already have WINPROC, dbKeyState() and similar functions to work with. Give us a UI editor and your golden...

ANIMATION - With Matty's diagram, what I see are various ANIMATION STATES; STATES being the keyword here. According to matty's diagram, animation states should be handled by AI... not an easy task. Similarly, by design (again according to the diagram), states need to be fluid - again not a bad thing... You should focus on the AI aspect!

From the "outside looking in", I think you guys need to focus on other areas for now:

AEM - sounds promising. I'm liking the sounds of it and the results of what's been done. I'd like to see this progress to a usable tool...

FULCRUM - Works great, so far, with GDK. I haven't used the "advanced" features yet but what I've seen so far is wonderful. I love the idea that I can attach a GDK object to physics and (more or less) forget about it - Fulcrum handles the details...

LIGHTING - Again, from the "outside lookin in", I don't think you should worry about this aspect just yet...

From what I have gathered from my readings of this thread, you all are looking to make a great "game-engine". Go for it! But I feel I should say that you already *have* a great game engine: GDK... Make the tools that will make your already great engine even better!

You all have the great ideas. Focus on those, build the suite of tools and then, if it's not good enough, add the other stuff...

Sorry, I had to say something. It seems to me like everyone is "meandering" along their own path - which sounds good, but doesn't move the project forward...

I apologize if I've offended anyone, I'm just trying to help you all get some focus here...

JTK

Login to post a reply

Server time is: 2025-08-08 12:05:10
Your offset time is: 2025-08-08 12:05:10