Java Community Project: Space Real Time Strategy Specification
Feel free to ask question, or correct any mistakes.
Hi bobjob,
Obviously you’re not going to share your grand ideas here when interest and dedication haven’t been shown or proven, but what’s the unique selling point?
Design/story are inter-dependent, under unique selling points… what will this RTS have over existing RTS’s already out there or under production by a bigger studio?
One recent RTS I found better than the average collect minerals/build units is Supreme Commander. It went for macro over micro full of destruction on a grand scale. I’ve never tried to play multiplayer, but the campaigns became really easy once you hole up. I’m sure vs a human player it won’t be the case which offers skilled players so many options on how to beat the other player(s).
Of the RTS’s I’ve played over the years. Bold titles are the ones I would choose to study from.
Spellforce Series - Story driven, game play is okay, heroes are way over powered but once again not the all in one solution. Could of used better scripting for in game cut scenes and voice acting.
Supreme Commander Series - LOTS! of units, options, variety of strategy. Simple resource management, but hard to master the perfect balance between energy vs mass.
World at War - Nothing special about this RTS, I don’t even remember the story and plot and the resources needed… it had issues with camera.
Warcraft 2 & 3 - cream of the crop in terms of replay value (yay bnet). Blizzard never disappoints. Hero units are great but not invincible, micro management over massing units. Great level editor.
Starcraft Universe - once again, Blizzard delivers another great RTS, but this time micro or macro is possible depending on the map. Same as WC3 almost but a different theme and game play style.
Age of Empire series - historical / mythical driven, too many resources imo, but well done and well made.
LOTR on the Xbox - I played it all the once, but the control were good for a console, and I was surprised. I remain neutral on this once since I haven’t finished it.
MOO3 - Terrible implementation of the MOO universe… too complex of planetary needs when the best points about MOO was building kick ass ships to take over the galaxy.
Pax Imperia (I forget the spelling) - MOO wannabe… terrible balance in terms of weapons… all you needed was to tech to smart bombs. Although I liked their simple planetary/ship design concept.
That’s all comes to mind now, I know I’ve played other’s that I liked but can’t remember their names or developers.
Jack
gamlessx25
for cutscenes I was thinking maybe just use video files. We could always just take some in game models and subsurf them with some quick action and explosions, at least for the early stages of development (would be extremely easy for an air strike cut scene).
I was wondering if you had a prefrence when it came to modelling? Do you mind rigging skeletons?
Another Idea for the theme would be construct an RTS from the same universe as the space trade game. Mostly spaceship combat.
- simple enough pathing.
- easy to implement effects
- easy effective skybox backdrops
- static models, no need for rigging or many keyframes.
dude I was looking at some of your models, there so cool. It also answered my question about the rigging.
Another question though, Do you know how to Tangent map models(Bake normal maps) from high poly to a dot3 texture map?
I am asking as I plan to do alot of work on the 3D side of things I want to make sure the game supports basic shaders.
Hey bobjob,
Yeah, I can do normal/displacement maps no problem. I practiced them a lot during school, but never produced a highly detailed to showcase since all the projects I modeled and rigged for had 1 programmer and that was impossible to implement in a 4 month period.
I agree with space sci-fi theme, boxes are much easier to model/texture/animate verse characters. Although Mudbox makes character modeling tons easier now. I’m working on a demonized black sheep model for my Conquerclub avatar, will put that up as soon as it’s done.
Jack
Great start bobjob!
Making a really configurable RTS is definitely an interesting project. Is the intention to fix the theme initially and then migrate to something modable, or start with building a configurable system? My opinion is fixing the theme initially and working on getting something working while keeping the modding system in mind but not a priority is going to work better.
Other questions:
- Will the 3D version support both lightweight software mode and opengl/directx hardware?
- Multiplayer - I definitely think it’s something that makes the project interesting, just because there are so many ways you could approach RTS networking
- If you’re going to use 2D and 3D, is that going to rule out using 3D terrains? Or is there something inbetween like some set of fix elevations?
Kev
There would be no priority on making the development tools stable, or open to the end user at the start, only to speed up development of the game, instead of spending to much time scripting.
As for the theme Im personally swaying alot to a space/scifi (If anyone has other ideas please tell me).
Another bonus in this would be (other than what was said before):
Basic RTS layouts seems to be a baseGird[][] and airGrid[][]. I was thinking if there isnt a base grid defined in the mapeditor, then it could turn into a “heavy weight unit” grid[][], smaller in size maybe baseGrid.length / 3; It means for the space scenes, there could be large carrier ship, or space stations in the lower part, and “leight weight” ship on the heigher grid. Also means faster pathing calculation for big ships instead of taking into account multiple small blocks when pathing. Simple and effective.
3D version will be openGL only, optimised also for slow computers (eg basic lighting, display lists, and single texture mapping).
2D version I highly recommend doing it using openGL, I think there are alot of good slick programmers would be nice if someone could use it to render the 2D version, If someone is confident that a 2D version could be done using software and still rendered fast enough through and applet or webstart, feel free to do so.
When the actual development(of the core game) begins. I will highly recommend starting it out as a multiplayer, Ill problably end up programming the networking stuff. I was thinking I may be focusing on 3D/Networking side of things as I already have completed version of code suitable for the project, ill probably end up porting some of that over.
Yeah I was thinking about this. I really dont know a nice way to approach the 2D side of things.
In a space game elevations wont matter to much (but Id really like support a variety height) for impressive tile maps when there used.
A “space scene” 2D would work great, draw the lower scene, then higher scene, turn Depth testing on, and re render the base scene with GREATER, set alpha channels to maybe 0.25f or something like that for an effective visual.
Yet for a version with the tile map on. I could propose the visuals of startcraft 1, But thats a bit demanding for a “simple version”. Another option would be to break up the tile map into its height vectors, then at runtime render the high points with lighter colours, and lower points with a darker colour.
But I honestly dont think I would be working on the 2D side of things, so it really comes down to the capabilities of the person who takes the banner for 2D. A simple problem to overcome would be the use of bridges, and again it would just be a second call to render the base layer with GREATER depth testing on and alpha blending on.
Added:
Some form of direct communication is needed be it video or just voice communication. Just a forum thread or just email will lead to miscommunication and people being frustrated.
As for multiplayer, it would be great to play a fraction/base as team.
To start, I made an mistake about Pax Imperia, it was an RTS I actually liked!!
The one with the over powered smart bomb was called Ascension.
Pax Imperia 2 is something we should all check out to get an idea of a good space rts. It’s Low poly 3D units moving in a 2.5D space environment. Plants are always at the bottom right while ship to ship combat revolves around fleet formations and ship design.
Point defenses are key since missiles and fighter/bombers are available on CVN classed vessels. But to counter, you load up a BB class ship with long/mid range particle weapons to take slow moving targets out from far away. I loved this game! gonna try to find it and play it for the next bit on my one free night on Mondays.
Ill check out some shots in youtube. Gameless I was thinking a might just throw together a quick mock version, to give a basic idea of the game machanics. How long till you could get a few models ready? maybe even some quick UV mapped arrow heads, for basic fighter crafts.
So far the participants are few, so there is no need to make the communication more complex.
I’ll take my sketch book along when I do laundry tomorrow. Simple models with normal maps shouldn’t take too long.
The international high distributed project that I was in for the last year has clearly showed that ‘just’ using mailing lists, email and some issue tracking system made things pretty dawn helpless. Even using IM here and there did help did’ly squat. Only thing that ‘kinda’ worked was Skype meetings. And that was for what I considered pretty trivial things.
Since I’m only helping on the art side, I just need a deadline date on art assets. Communication is more vital between the programmers.
As for the game’s theme and style, Star Trek Armada came to mind as a great space RTS in the classic RTS format. Lets see if I can find that and play it for insights on what’s good and bad. Been so long since I played it.
Will be at the computer when MSN is marked busy, chat then bobjob.
Jack
good to know. Same as me.
A friend started working on some basic ships. will be nice once we got a few basics models to show off.
Im not sure about if they will be the models used in the end (probably will be), becaue in order to keep in it sync with the other game, may have to wait for the decisions.
Just a few things to note for models(at least for the mock version until more programmers come on board):
- Triangulate mesh (convert polygons to triangles)
- export keyframes to Wavefront “.obj”
I am interested in helping with this. I am a Java programmer for my job. I haven’t completed any games except 4k ones. I hope I can stick with this one to the end. My time will vary, but I will put as much as I can. I have already been added to the Google code site. Are you going to take advantage of it?
I just sent Kev a PM in order to regester to the site.
Glad to here more interest in the project.
I wouldnt mind checking out some of your 4k games, do you have a site?
Once I work out how the site works :persecutioncomplex:, Ill upload a basic version for starters.
No I don’t have one. You can check them all out at the Java 4K site.
This is my most recent one. http://marber8.freehostia.com/planet.jnlp. I think it is a personal best for me.
Not a problem, just have to switch from Edit Poly to Edit Mesh before exporting.