Space Combat (madness)

For what's not in 'Top Priority Game Design'. Post your ideas, visions, suggestions for the game, rules, modifications, etc.

Moderator: Oberlus

Post Reply
Message
Author
krum
Creative Contributor
Posts: 244
Joined: Thu Jun 26, 2003 12:58 pm
Location: Bulgaria

#61 Post by krum »

utilae wrote: I just want to clarify whether your understading of phased real time combat is the same as mine.

Phased real time is not really being able to pause in battle. But more you have two phases. The give order phase comes first, where all players give orders for their ships at the same time. When all players are ready, the second phase occurs. In the second phase the battle plays out for X seconds. Orders are executed, ships destroyed, etc. Plays cannot interfare during this phase, only watch, which is why it would be cool to have this part in 3D. Then the order phase occurs again. And so on, etc etc.
Yes, I uderstand phased combat; but you're right that what I described shouldn't be called that.

How often the player can give out orders could be just another factor like ship speed, armor, shield and firepower, etc. Why shouldn't one of the players get to pause for orders a bit more often than the others? Well, besides the issue of balancing it.
Last edited by krum on Sat Sep 25, 2004 8:37 am, edited 1 time in total.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#62 Post by drek »

hmm,

Starting to think system combat just isn't feasible, or rather, would take as much or more time to complete as the rest of the game x10. It would be an epic undertaking, unmatched by other open source games by several factors. We'd be re-writing homeworld.

And planets, suns, etc would have to be beachballs. There's just no way system combat can have the sense of scale that pd's mockup movie featured -and- have an UI that's easy for the player to use in the midst of frantic combat. (course, combat doesn't have to be frantic, but that implies combat that takes an half-hour to an hour to resolve instead of just two minutes.)

Might need to widdle down the dream to something that can be realistically completed. Comes down to this....we can have:

* a completed game,
* a decent UI, with nice eyecandy
* a complex combat system

but only 2 out of the 3. Would an improved moo2-ish combat system with decent graphics really be all that bad?

noelte
Juggernaut
Posts: 872
Joined: Fri Dec 26, 2003 12:42 pm
Location: Germany, Berlin

#63 Post by noelte »

i take
* a completed game,
* a decent UI, with nice eyecandy
and a improved moo2-ish combat system. But that was clear anyway :oops:

As Drek said, the discussed system sounds great, but who's going to implement that. We should keep that issue in mind.
Press any key to continue or any other key to cancel.
Can COWs fly?

krum
Creative Contributor
Posts: 244
Joined: Thu Jun 26, 2003 12:58 pm
Location: Bulgaria

#64 Post by krum »

i'd take complex combat ahead of eyecandy... but above all, good UI.

miu
Graphics Lead Emeritus
Posts: 286
Joined: Mon Aug 25, 2003 2:33 am
Location: Finland/Helsinki

#65 Post by miu »

Problem with systemwide combat is that ship movements and battle happening has to be quite fast to fit everything in few mins - thats problematic if disabling a ship takes a while. Ordering would not have as much detail as in smaller scale and planet-scale model can have some really impressive backgrounds, hmm, can live with both modes, just have to rethink few ideas that I had for planet presentation if we go for planetary-battle instead of system-wide .
Difference between a man and a gentleman is that a man does what he wants, a gentleman does what he should. - Albert Camus

User avatar
pd
Graphics Lead Emeritus
Posts: 1924
Joined: Mon Mar 08, 2004 6:17 pm
Location: 52°16'N 10°31'E

#66 Post by pd »

Starting to think system combat just isn't feasible, or rather, would take as much or more time to complete as the rest of the game x10. It would be an epic undertaking, unmatched by other open source games by several factors. We'd be re-writing homeworld.
well said. this is basicly what i thought of from the beginning. the whole space combat is only a part of freeorion and not the entire game. that's why we should make it fast to play and simple. system wide combat sounds just to comlpicated and would maybe take too long(to play and to implement). i think KISS should aply here too.
Would an improved moo2-ish combat system with decent graphics really be all that bad?
no, it wouldn't in my mind.

and so i also take
* a completed game,
* a decent UI, with nice eyecandy

pfew, luckily i made this mockup ;)

User avatar
utilae
Cosmic Dragon
Posts: 2175
Joined: Fri Jun 27, 2003 12:37 am
Location: Auckland, New Zealand

#67 Post by utilae »

drek wrote: We'd be re-writing homeworld.
We don't have to have 3D graphics.
drek wrote: And planets, suns, etc would have to be beachballs. There's just no way system combat can have the sense of scale that pd's mockup movie featured
Sense when do games ever have a scale matching the real world. I'd be happy if things weren't to scale.

To me a system combat would only really have one difference to a planetary combat:
*The map is bigger. And a few planets are dropped at various locations.

Also I don't think we need rediculas things like planets orbiting around the system or even heat from the sun affecting ships, etc.

I don't think we should be going for homeworld style at all. I mean 3D + controlling many ships durign 3D results in a long time to make. You need to make it easy to control in 3D, etc. Of course we don't even need 3D. Top down/Isometric would still be cool.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#68 Post by Geoff the Medio »

drek wrote:Starting to think system combat just isn't feasible, or rather, would take as much or more time to complete as the rest of the game x10. It would be an epic undertaking, unmatched by other open source games by several factors. We'd be re-writing homeworld.
This is unjustified, IMO. There's nothing excessively more difficult to making system combat compared to per-planet combat. The only major difference is that planets are objects in the gamefield, rather than background images. As far as I know, texturing a sphere isn't a hugely difficult task. Issues of comparison with rewriting homeworld are unjustified, and IMO irrelivant to the system-scale vs. planet scale debate.

Are there particular things about system combat that you're thinking of that I'm not? pd, you seem to agree with drek... why?
And planets, suns, etc would have to be beachballs. There's just no way system combat can have the sense of scale that pd's mockup movie featured -and- have an UI that's easy for the player to use in the midst of frantic combat.
This is essentially a realism argument. Scales are going to be completely "unrealistic" no matter what we do. Having planets be the size of my mockup posted earlier would be fine. No, you won't be able to gape at the nifty giant rendered planet image in the background, but this is not a major reason to engage in space combat, for me at least...
miu wrote:Problem with systemwide combat is that ship movements and battle happening has to be quite fast to fit everything in few mins - thats problematic if disabling a ship takes a while.
How is this different form per planet combat?

I'm thinking of combat being a bit slower than a battle between some high-level Starcraft units at the standard "fast" gamespeed. Carriers and BattleCruisers and so forth take a decent pounding before going boom, though can be taken out fairly fast if you have a large group focus firing on a single target. Say, maybe 10 seconds (2 rounds) focus firing to heavily damage / disable a comparatively decked out ship, 15 sec (3 rounds) to put it completely out of comission and 20 sec (4 rounds) to reduce it to space-rubble.

Impaler's talk of ships taking 3 minutes to cross a system is probably too long. Something like a minute at most would be best. The spacing between planets can be unrealistic to make this feasible. (Real planets are futher apart the further away from the sun you go, to a significant degree... this need not be the case in FO)

In FO, the distance between planet radii can be about 1 turn for average speed ships. Since since planets are spaced around the sun angularly, it'd take 2 or 3 turns to get from planet to nearby planet, and about a minute and a half to cross a system of a full 10 panets if the outermost planets are on nearly opposite sides of the system (which needn't ever happen) (and most systems would have less than 10 planets anyway).
planet-scale model can have some really impressive backgrounds
Planet-scale can have planets as background, which do look nice. But a) can't we make nice starfield with nebulae or galaxyscape background at the system level that would look just a good, and b) is a nicer background really a good reaon to decide on a battle system? Yes, graphics are important for the battle system (the whole thing is essentially eye candy), but the background image isn't that important, IMO...

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#69 Post by drek »

This is unjustified, IMO. There's nothing excessively more difficult to making system combat compared to per-planet combat. The only major difference is that planets are objects in the gamefield, rather than background images. As far as I know, texturing a sphere isn't a hugely difficult task. Issues of comparison with rewriting homeworld are unjustified, and IMO irrelivant to the system-scale vs. planet scale debate.
Texturing a sphere would have to be done in either version, and obviously is not an epic task. The beachballs aren't my concern (though obviously the beachballs in scenic combat could look much cooler.)

There's a few extra thingies I think we'd need to do system combat justice:

a: terrian, terrian effects, and pathfinding. includes planets, the various actions that can be performed on planets, stars, good reasons not to run into stars.

b: a much more complex AI. (the computer player for scenic combat would seem much easier to conjure, imho) This is the line item that's really convinced me--going to be hard enough to whip the strategy AI into shape without having to worry about making a non-stuipid tactical AI for a complex combat system.

c: control groups and other rts methods of selecting units (such as band selecting). RTS movement orders, such as waypoints, guard orders, perhaps patrols. In scenic combat, we could do things like the game Spartan--the movement orders could be limited to "advance" "retreat" "flank up/down/left/right" "stand your ground" "attack (target).". Wouldn't even have to issue the orders during combat....as in Spartan all orders (aside from general retreat and general "charge!") could issued before combat begins, during unit placement.

d: a pulled back map of the entire solar system, and perhaps a minimap. The pulled back map isn't as easy as it sounds....have to be able to make the transtion smooth and fast (like in hw) plus represent units, terrian, planets in a clear way (again like in hw)

e: Slightly more advanced camera control: The mildly hard part here is getting UI to these controls to feel right. In scenic combat we could easily get away with just allowing the player to rotate his view around a three or four defined pivot points. (though of course more could be done for extra credit) Also, a method to highlight units hidden by beachballs, assuming the beachballs are as big as they are in your mockup.

Anyway, there'd still plenty of time to decide what sort of combat engine to use. Knowning me, I'll change my mind again next week :P.

Impaler
Creative Contributor
Posts: 1060
Joined: Sun Jun 29, 2003 12:40 am
Location: Tucson, Arizona USA

#70 Post by Impaler »

GRRRRR... Drek this is getting tiring, it seems aparent from your posts that you have linked systemwide combat to "massive Homeworld Graphic display"
when the 2 have nothing to do with each other. One is the scope of the battle the other the display method.

In response to your points.

a - Terrain will be 99% empty space with "beach ball" objects in it, I am no no expert on pathfinding but it would seem to me this would be rather simple as your going shortest path 99% of the time and need only defiate around these BeachBalls. Avoiding enemy ships will be far more complex and that would need to be done in either aproatch.

b - If all we want ships to do is pound on each other then they dont realy need ANY AI now do they. If on the other hand we want to have the player make some desision that might effect the outcome then its nessary to give the player options for risk and reward. Your saying that we should have dumb combat because the best we can make is an AI so dumb it cant move a ship from point A to point B in a few seconds of unsupervised Execution phase.

c - Why couldnt thouse things be done with System wide combat? Were going to likly have some kind of TaskForce organization level in either aproatch.

d - are you saying you want a "scenic combat" area so small and crampt that its all visible at one time and their will thus be no need for a mini map? This sounds like StarWars Rebellions Combat Engine which consisted of a cube of space with thouse oh so important "Scenic Planets" floating in the back ground you seem to be so fond of. It stank by the way.

e - If we chosse to view system wide combat with a 3D viewere we would likly need a better camera control then that needed for a basicaly featureless void of space. Well guess what tacticaly complex and engaging battles over wide areas require more support then do simplistic ones on tiny featureless maps.

We should make a desision based on whats best for game play NOT based on what will let us have 3D graphics. The scope/scale/blance and FUN of combat should be the first priority only THEN should we desided how sophisticated the Graphical display should be based on how much we are capable of. If full system combat requires a top down isometiric view then it is worth it.

Drek if you have not yet done so check out the StarFury Demo I posted earlier its a good example of how system wide combat could work.
Fear is the Mind Killer - Frank Herbert -Dune

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#71 Post by Geoff the Medio »

Impaler wrote:a - Terrain will be 99% empty space with "beach ball" objects in it, I am no no expert on pathfinding but it would seem to me this would be rather simple as your going shortest path 99% of the time and need only defiate around these BeachBalls.
I wouldn't call it "empty space" in that there would be various terrain features on the map other than planets, that would have some strategic value (detection, movement speed reduction, functionaly of ship parts) like gas clouds and asteroid fields and such.

However, with regard to the pathfinding issue, the terrain features are a very minimal consideration. They are likely very limited in size (planets), can be easily flown around (gas clouds, planets) without any difficult decisions to make, or can't be avoided at all (asteroid belt that encircles the star that you want to travel through), meaning there's no decision to be made... (you have to go through it).

There was also some worry about the pathfinding and AI in terms of implimenting player orders between order-giving phases. The differences in this respect are rather minimal between system scale and planet scale... and in either case, if the time between order-giving periods was short (5 seconds), then very little AI would be necessary, beyond following simple orders like "stay in formation", "intercept this ship / fleet", "guard X", "move to this spot" and so forth...

There is also the AI issue of higher level tractics and strategy for the AI players, which is independent of the pathfinding / implimenting player (or AI) orders stuff.

I realize AI discussions are frowned upon, but it's come up, so... One major issue I see with combat AI have to do with how it deals with terrain features, which is mostly independent of system-scale vs. planet-scale. If it really is impossible to write a decent AI that can deal with gas clouds and asteroid fields and the like, then we can say that there is no terrain, and treat the planets as combintation Points-Of-Interest for the AI, and/or immobile weapons platforms. The AI would either plan to get past the defences and drop troops or spies as a suicide run, or would just move around like it would in a system-level map, treating the defence satellites and such the same way it treats ships for combat purposes.

Regarding issues of adjustable map scales and minimaps, I still fail to see why things would be significantly easier or better with a planet-scale map. System-scale gas clouds would be represented as appropriately scaled fuzzy blobs on the map. Asteroid fields would be an appropraitely scaled jumbled of small brown rock-shaped objects. Planets would be appropriately scaled textured spheres.

The transition between various zoom levels does not need to be smooth and pretty. If it's the best that can be done, it would be fine to just have fixed zoom levels that can be switched betwe, just like the current galaxy map. That said, I don't think having a smooth transition would be especially difficult compared to moving the camera around in the plane of the map. An I don't see any reason why zoom levels would be more problematic with a system scale map compared to a planet scale. The maximum out zoom doesn't need to show the whole system, so we can limit the zoom to whatever is useful for the sizes of models compared to scale of the system map that we use.

I don't follow what you (drek) mean by "allowing the player to rotate his view around a three or four defined pivot points." IMO the view should translate when the mouse is moved to the side of the screen, or rotate around the centre of the screen when the player right clicks and moves the mouse. (The default controls in the Rome: Total War demo were quite irritating for me. Rotating the view is of much less use than translating, in general.) Other than those, and perhaps "follow unit" autotranslation, and "rotate about selected unit" with the mouse, what other pivot points are necessary?

Regarding planet beachballs hiding ships behind them, spheres are sufficiently rounded (heh) that we can set the view angle to be high enough above the map plane that there are no problems with planets hiding anything behind them.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#72 Post by drek »

Impaler wrote:GRRRRR... Drek this is getting tiring, it seems aparent from your posts that you have linked systemwide combat to "massive Homeworld Graphic display"
a: my vote, as it were, is only a minor concern to the engine that is eventually chosen. If I were you, I'd worry more about what pd and noelte think, as combined they actually have the skills required to create a combat engine.

b: defining details of the combat engine is still a long ways off. There's no fire here.

c: again, I could change my mind next week, or in the next ten minutes. I'm not dead sold on either idea yet.
when the 2 have nothing to do with each other. One is the scope of the battle the other the display method.
Right now, I'm thinking this whole system-based combat thing is something that seems real simple until you actually start implementing it: devil is in the details.
d - are you saying you want a "scenic combat" area so small and crampt that its all visible at one time and their will thus be no need for a mini map? This sounds like StarWars Rebellions Combat Engine which consisted of a cube of space with thouse oh so important "Scenic Planets" floating in the back ground you seem to be so fond of. It stank by the way.
Er, yes. Just a possbiliy, but not a terrible one, esp. if we want combat to be quick and easy to implement.
We should make a desision based on whats best for game play NOT based on what will let us have 3D graphics. The scope/scale/blance and FUN of combat should be the first priority only THEN should we desided how sophisticated the Graphical display should be based on how much we are capable of. If full system combat requires a top down isometiric view then it is worth it.
For me, the fun comes from strategy portion of the game. The combat portion is a test of my empire's high level strategy: can I get the right number of ships of the correct type distributed to the important battles?

If the tactical engine is kept simple, we'll be returning to working on the strategy portion much sooner.
Drek if you have not yet done so check out the StarFury Demo I posted earlier its a good example of how system wide combat could work.
am I missing something? Don't see how it applies. The system map in starcontrol2 would seem a better example.

In any case, if that clunky interface is an example of what we can expect in system engine, then I'm sold on pd's conception :P.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#73 Post by drek »

There was also some worry about the pathfinding and AI in terms of implimenting player orders between order-giving phases.
Pathfinding and AI issues are half the reason why version 1 of Homeworld2 was almost completely scrapped, delaying the game for over a year and a half.

Impaler
Creative Contributor
Posts: 1060
Joined: Sun Jun 29, 2003 12:40 am
Location: Tucson, Arizona USA

#74 Post by Impaler »

Ok everyone listen up, Full system means ONLY that the players fleet is on a Map that has all the planets of the system on it and they can move and attack all of them with their fleets. Nothing else should be asumed beyond that which is what many people started doing this thread started dealing with graphics and visuals and Homeworld type stuff.

StarFury is an example of the scale of a system and planets within it and the kinds of terrain that could be put in it. The movment though is through a 2D plane unlike Homeworlds 3D movment. The games UI is of absolutly NO RELEVANCE WHAT SO EVER and I dont see how you can be "sold" on somone else concept when their IS no plan or mock up for a UI of any sort.

Your coments on Fun coming from Strategy are making it quite clear that you would prefer something like Moo2 ware 2 fleets simply colide in a massive broadside and blow each other up. System based combat is designed to do away with that kind of tacticless fight. Instead you need to manuver for advantage, select targets, hide, use terrain, bluff, feint ect ect. In other words a real tactical engagment. If you dont like tactics then hit the auto resolve putton and watch a replay dont cripple the engine for thouse who might actualy enjoy comanding a meaningfull engagment.

For the 12 time we are not making Homeworld, I would be strongly oposed to ships moving in 3 Dimentions even if we had the capability to do so.
Fear is the Mind Killer - Frank Herbert -Dune

User avatar
utilae
Cosmic Dragon
Posts: 2175
Joined: Fri Jun 27, 2003 12:37 am
Location: Auckland, New Zealand

#75 Post by utilae »

I agree with Impaler. I also think we should settle this earlier, because if we decide on system wide combat, then we could use all the time we can get.

I think we all knew that space combat would take the longest to develop and implement. I am wondering whether all other tasks in making this game are significantly smaller than the task of making space combat. If that is so, then are we overestimating the time it would take to do the other tasks, and therefore have underestimated the time we have to do space combat.

I don't really see a need for 3D, because to think strategically in 3 dimensions is not as easy as thinking strategically in 2 dimensions. An isometric view would be awesome. It would cut down on all the rotating and giving commands in 3D as well as pathfinding in 3D (we should never go there).

Also I don't think we really need to zoom out or in. I mean, if we could scroll around the map fast enough it would be fine, especially if it was phased real time, because a player who could give orders faster wouldn't get any advantage, because all players would have to wait for slower players to give orders.

Post Reply