Spring RTS Engine

oh hai!

I've been thinking about realising my own game ideas for some years now and having recently discovered spring (hai, I'm new here etc.) almost simultaneously (!!1) to playing with the thought of making a very Simulation-Like Naval game I thought I'd take a shot and ask if anyone would be intrested developing it with me or even to play it for that matter before my squirrel sized attentionspan inflates the ball before I can get it rolling.

The game I have in mind would be somewhat like the Harpoon series - just without the tremendous amount of fail when it comes to usabillity, bugs, gameplay and the like. For those of you who don't know the Harpoon series, which for your mental sake I sincerely hope, here a summary:

Basically you see everything from a tactical scale of quite epic proportions (e.g. all of north Atlantic) and you command carriergroups and their aircraft, subs, destroyers, sometimes airfields and you're given realistic scenarios where you need to attack positions/ships, search for subs etc. Everything takes place in realtime (fast forward yes please); everything trimmed to realism which ranges from setting depths for subs, speed for ships/subs/aircraft, selecting armaments, selecting munitions, having a fuel system, having realistic combat behavior to getting messages from your Officers (you're the Admiral... I suppose). The latter makes up alot of the games athmosphere which is probably the only reason I like it. Theres just a little something about hearing "Vampire! Vampire! Vampire!" whenever a enemy missle is detected.

My idea for a Naval Sim for Spring was to take all the good bits from Harpoon, add Silent Hunter 3 - like action cameras and actually beeing able to zoom in from the epic scale to your units beeing able to see things like launches, hits and take-offs.

Now however to the actual point of this post; a few questions I'd like opinions on:

- Does the concept exceed the mod scale, requiring major alterations the the actual engine (I'm not THAT afraid of this... maybe I should be)

- are epic map scales even realisable in Spring (see above)

- are people intrested in creating such a game

- is it worth putting up with the warfare nerds (I don't use this term lightly as I personally see it as an insult) flaming and crying about every inaccuary based on their totally not biased views ("A F15 could shoot down 10 russian jets easily1!!1!")

some clarifications:
- Yes I realise this post probably sounds like a standard "omg my idea is soooo awesome someone develop with me but I cant code model skin or think for myself, kthx!", but while I admit sucking at both modelling and skinning the only thing I exceed in is studying software engineering.

- Yes I realise the concept could hardly be farther from beeing a actual Multiplayer game or beeing a RTS in the sense of TA

- Of course it's a rough concept, requiring and anticipating construcive ideas (edit)

I may be wrong but it seems that the spring may be somewhat lacking in what you need for such a game.

*edit* also I should say that I found your post to be a good one, there is nothing wrong with asking for help in your fact finding. It is good to see someone first ask what is possible before assuming spring can handle it. Such a feasibility check is important.

On map scales: you could just make your ships small...

Guessmyname wrote:On map scales: you could just make your ships small...

But most maps with sea have too big and tall masses of land that would look unrealistic if we counted the sea as "sea sized" :) Making maps with only tiny pieces of land and everthing else covered in water wouldnt be very hard tough.

I think that from the features he mentioned most arent inviable depth of subs, and maybe speed... the fancy customization could be made with lighting quick morphing I hope :)

Couple of key issues with making ships in spring you ought to know..

Ships behave exactly the same movement wise as tanks, they just only go over water. Hence simulating any sort of even remotely realistic looking ship movement is just right out of the window, ships will turn on their center axis as if they were a simple point mass object. This doesnt look to horrible on small ships, but for any sort of long battleship/ship of the line/dreadnought thingy that you might want a ponderous naval feel to it will look silly and stupid.

Another aspect is that since most ships are much longer than they are wide, you will currently get clipping issues since spring only handles spherical hitboxes(spheres) at the moment. This means that making the hitbox as long as the ship will also make it that wide, which will be awkward looking and function awkwardly. There is/was a patch in the works to add rectangular boxes, or at least elongated spheres.. so hopefully that will fix this, but the naval movement will still exist.

Almost anything else from a surface standpoint is pretty much possible. Though subs are somewhat limited in being able to shift depths on a functional level..

Do you know lua?

If not, start studying- this would take a LOT of it.

As for the game concept- is this RTS or RTT? Sounds like RTT (lets not get caught up in the semantics of the what is tactics/what is strategy and just say 'It doesnt have an economy' is what im getting at). The number one thing i always ask people who come up with new mod concepts is 'Whats your economic structure?' because its usually the cornerstone of the whole game (and the least considered, they'd rather think about what cool units to have).

Are you planning on making scenarios and having this singleplayer, or making a serious, balanced, multiplayer? Spring is mostly about the multiplayer...

A simple 'You get points, spend points on force, now FIGHT' would do i suppose.

Either way, you might want to start with something less ambitious, an easy mini-mod based on a neat little concept thats easy to execute in spring, just to learn how its done.

Ever taken a look at the War on Sea mod for the original Total Annihilation? It seems just what you're looking for.

It all depends on what sacrifices you are willing to make. I don't think you could make the whole ocean the map and still have a carrier group consist of individual ships, the smallest footprint (ground collision size) of a unit is 1x1 and map sizes can't realistically exceed 24x24 "screens" (with 64 footprints to a screen or something on that scale). You can't have different scales of simulation (i.e. you can't simulate a battlegroup as a point and only use units when zoomed in or a battle happens) so your units would have to be visible on the scale of the map.

Additionally you cannot detect the location of an impact so if you wanted e.g. rear hits to destroy the propulsion that won't work. Your description doesn't seem to involve that though.

Fast forwarding isn't really a good thing in multiplayer, one player might want to FF while the other wants to remain at realtime.

The combat is feasible I think, different sub depths can probably be done with MoveCtrl too.

lets summarize. (I like to do that)

Major issues:
- Map scales not appropriate for the engine (engine alteration required)
- spherical hitboxes (engine alteration required)
- movement behavior inappropriate for Naval units (engine alteration required)
- no different scales of simulation (engine alteration required? I'm pretty sure it's possible in concept)
- RTS or RTT? SP or MP?

Spring might not be a very appropriate engine for the concept so I might keep looking for a engine that suits better.
As for the gameplay: I'm still thinking and as I said the concept is kinda rough. I'm playing with the thought of not making reallife weaponry and vessels but using imaginary hardware which I can bend to suit better (and reduce vet's whining) additionally to making it somekind of global scale tactics game where you earn points for capturing enemy positions, which you in turn spend on improving your own positions and building/researching new ships and weaponry.

The idea of a arcade-like "Total Dominiation" comes into mind.

As for lua: I assume you mean the script language. Looks just like every other C based language so it shouldn't be that much of a deal.

You're thinking of BOS, which is individual unit scripting/FK animation. Spring's lua interface gives the lua language (which isn't very syntactically similar to C*) mod-side control in most areas of the engine, and pretty much opens up essentially limitless possibilities in terms of what is implementable.

$Icy wrote:lets summarize. (I like to do that)

Major issues:
- Map scales not appropriate for the engine (engine alteration required)
- spherical hitboxes (engine alteration required)
- movement behavior inappropriate for Naval units (engine alteration required)
- no different scales of simulation (engine alteration required? I'm pretty sure it's possible in concept)
- RTS or RTT? SP or MP?

Spring might not be a very appropriate engine for the concept so I might keep looking for a engine that suits better.
As for the gameplay: I'm still thinking and as I said the concept is kinda rough. I'm playing with the thought of not making reallife weaponry and vessels but using imaginary hardware which I can bend to suit better (and reduce vet's whining) additionally to making it somekind of global scale tactics game where you earn points for capturing enemy positions, which you in turn spend on improving your own positions and building/researching new ships and weaponry.

The idea of a arcade-like "Total Dominiation" comes into mind.

As for lua: I assume you mean the script language. Looks just like every other C based language so it shouldn't be that much of a deal.

Map scale I think won't be a big issue for you, but you will need specially designed maps like you said.
A patch is being made so you can use something else than spherical hitboxes IIRC.
Different simulation scales is indeed impossible, and requires you to basically recode the complete engine.
You can make certain RTT principes, like deployment, and reinforcements, with Lua. SP would require a Lua to AI link, MP shouldn't be a problem.

Spring is a fine engine to use, I think you won't find any engine better suited for this. The 'global scale' game is indeed impossible, but the capturing system and the points for reinforcements and research is doable with Lua (although it will be hard).

Anyway, if you want to actually execute this, and need some tips/help, feel free to ask me or the other forum members. Many people here (including me) are working on their own mods, so we know the situation you are in. :) But please don't count on other people to do lots of work for you.

I just remembered that I once read in some thread that it was possible to have units with multiple damageable parts on spring, and I suppose this allows multiple hit-spheres (cause there would be no point in having more than one part if they couldnt be hit individually); Wouldnt this solve the problem of ship length and damageable parts?

Additionally you cannot detect the location of an impact so if you wanted e.g. rear hits to destroy the propulsion that won't work.

<- you can do this with simple hitbyweapon script

zwzsg wrote:

KDR_11k wrote:Additionally you cannot detect the location of an impact so if you wanted e.g. rear hits to destroy the propulsion that won't work.

I beg to differ.

fixed

I wouldn't call such HitByWeapon scripts simple, or else they'd have been used earlier, and I wouldn't have to regularly remind people about them up to this day. We have to keep in mind that most people who write unit scripts have so little idea of what they're doing they'd argue fiercly for hours when I tell them they can drop out the "if(true)" of their walkscript. So to count on them thinking about using an arc tangent in an HitByWeapon function is quite optimistic.

Also, last time I tried, Spring had the peculiarity of passing null arguments to HitByWeapon for weapon without impulsefactor.

Yeah, it's the impulse direction in Spring. Also it has a tendency of pointing in the wrong direction if the shot is fast enough.

Also it has a tendency of pointing in the wrong direction if the shot is fast enough.

That's because the shot's actually on the other side of the object before hit detection occurs. Adjust the CollisionSize of the weapon to mitigate this issue, but be aware that it is circumstantial, and, in the case of the new collision systems, I have not tested whether or not it reports accurately or not at this time, since I don't use this feature in P.U.R.E.

In a real sim, at least at the scale proposed by the OP, this is a non-issue, anyhow. Things would either intersect with the target's space, or not- after that, whether they "really hit" or not is in the realm of simulation mechanics, and has nothing to do with Spring's usual handling of ballistics, etc- in fact, all weapons in his game would probably be slow-moving Cannons without gravity, or Missiles with guidance. At naval-war scales, they'd be very, very slow- we're talking about weapons firing over kilometers of distance, even for the short-ranged stuff, and while they're fast, they ain't lasers.

Hit detection would never be a real problem. All units could be one square, and the only big issue would be stacking vs. scale, and even that is something that could be dealt with, via some Lua.

Simulating facings is, imo, out of scope of the game design proposed- Harpoon is an icon war, for those of you who've never played it, attempting to emulate very complex RL naval warfare at a huge scale- it's entirely un-necessary to use Spring's handling of such things that far in depth, and would just butterfly-effect the game design into oblivion very rapidly, as well as massively raising the complexity.



As for the OP's game design... meh... I got confused reading what you wanted.

Basically, Spring's just fine, for a true copy of Harpoon. It'd require one heck of a lot of Lua to achieve, but it's entirely possible to realize, if you have the programming skill to hack it.

However, as I got further down this thread, I saw people bringing up all sorts of crap you don't even want to actually deal with, in a strategic-level simulation of modern naval warfare, like the turning-rates of ships, and showing them graphically, etc., which is completely irrelevant- a ship would have to be so tiny that it'd be practically invisible, and it'd be incredibly pointless to try and do rendering of stuff like hitboxes at "true scale"- a real simulation wouldn't bother anyhow. Seriously, have none of you people ever played a game like this? It's icon-wars for a reason, and that reason is that anything further than that is completely pointless. IRL, naval officers don't train with anything more graphically sophisticated, when learning about strategic-level stuff- it's all about where you put your icons, and what their capabilities are. That, and the skill / luck of the captains and crews involved. But as an admiral or whatever, you don't control that. Which is why IRL, large-scale warfare sims use a lot of probability to simulate very complex events.

The big problems with your proposed design have nothing to do with any of that stuff, once you get past the idea that there's no point in being able to "see the action" when the "action" consists of mainly robots (guided missiles) and aircraft, with the occasional ASW action amidst convoys traveling veeeeery sloooooowly across a huuuuuge expanse of ocean. Forget the fantasy of being able to zoom in and see a sailor on the deck of a carrier, then zoom out and control the entire war.

It pales when you think of these things you'd need to program:

1. You'd have to write an all-Lua description and display of sensor warfare, for multiple levels / types of threat. I.E., passive vs. active sonar, radar, etc., plus ECM / ECCM, etc.

Without that element, alone... this ain't a sim of anything real. It's probably the most important element, tbh, and Spring has nothing like a realistic depiction of same, natively. You can write one- you can write practically anything, with Lua. But it's not going to be easy.

2. You'd have to deal with issuing orders in terms of Rules of Engagement, etc., to various fleet units, and take control over combat away from Spring's CAI. Not too hard, but not fun, either.

3. You'd have to write a custom damage model, using either COB or Lua. Not hard at all. Seriously, the weapon sim's the easiest part to do, including removing capabilities from a unit based on status.

4. You'd have to write a custom way to define and delineate game goals and scenarios. In Spring, generally, we blow each other's stuff up, one guy wins. The complexity is entirely in how we reach said goals.

In a real war sim, you need to have detailed objectives (by this day, deliver X percent cargo ships to Y location, or you lose). Can be done, but it'd probably require being tied to specific maps.




Meh, this is already too long, so here's my final word.

Overall, your idea's about as do-able with Spring as with any engine that exists, that I know of. I don't think there are very many other engines, if any, with the sheer number of programming tools available in Spring to realize complex projects of this kind, that are nice and stable, so if you don't want to write an engine... this is probably your best hope.

TBH, though, it sounds like you just want to mod Harpoon itself, and fix certain things. I know they have scenario editors, etc., for that game engine... so what are your major beefs about doing so? Maybe explaining why Harpoon is not all you wanted it to be would help us point you in a better direction here.

Bah, NO ESCORT MISSIONS! Those fail.

Your description doesn't sound like an issue to me, with low unit counts you can feasibly implement your own sensor system (and with Spring SVN you can make the units actually visible instead of just lua-drawn icons). I'd ignore the movement issues for now.

Escort missions are fine, as long as you DIRECTLY CONTROL the units you are escorting.