Spring RTS Engine

Hello fellow springers.

I'll tell you about myself, mostly my relation to Total Annihilation in this first paragraph, if you want to get straight to the point, skip this. I'm fairly new to Total Annihilation. I discovered the game three years ago when someone said that it is the best RTS game he has played. I got the game and all its expansions, and played it quite a lot with a friend. We mostly played against AI, as we're not a competetive bunch and mostly just wanted to have some fun. Later on someone else showed me Spring and its numerous mods, and I was amazed and thrilled. All this in fancy 3d coating and whatnot. Since that day I've regarded Spring and it's TA based mods my favourite RTS. As for my skills, I'm not a great player, most of you would probably kick my ass with ease, but I'm good and experienced enough to know "how shit works". Call me your average Joe player. I've been lurking on the forums now and then, so I have a rough idea on what's going on, but no doubt what I'm going to say next has been suggested a dozen times before and all I'll probably be doing is pour gasoline into fire.

Now, to business. While Spring and its mods (I'll be talking mostly about CA in this post, as I think it is the most advanced of the current mods) is currently aimed for competetive dueling and team matches, there is a great deal of people (me included) who would like to see the single player aspect of the game improved, to learn the ropes of the game and to just have fun on your own. I am talking about campaign(s) here. Let's first go over the problems there are in making a campaign.

From what I gather, the main obstacle between this point and the one where there is a fully working campaign is the AI. The problem with AIs is that while they try to be jacks of all trades, working with whatever mod they're used on and in the end failing more or less in most of them. In a campaign the AI must be modifiable to suit each scenario's needs, so an ordinary "does everything to win using the standard formula" would not do.

Also I recall reading at one point that there was some dispute about mods being suitable for campaigns, as they change all the time, making it hard to balance the campaigns.

I do not now how hard it would be to implement the actual campaign system into Spring, with saving/loading and missions following one another, but if this turns out to be a bottleneck, then my sense of logic is greatly flawed.

And now let's see how we can overcome these things.

I personally can't see a problem with the mod question. So what if mods continuously evolve? Such things happen with all RTS games. What should be done is just take one version of a mod and decide that it will be the version that will be used in the campaign. Of course every mod developer would want their own mod to be the one whose mod is used, but there could be a community vote between BA and CA as I believe those two are the most popular mods out there. I would like to see CA as the campaign mod, as it's graphically more advanced and has all these nice little things that help gameplay, but I digress. This aspect can be discussed more at an other time.

I usually don't talk about stuff I don't understand, and even now I wouldn't want to, but I haven't got any contacts with you people, so I guess I might as well say what I feel about AIs.

I know that this is a sensitive subject on these forums. AI developers are sick and tired of it all and everyone wants AIs to work on their favourite mods. Now if one of these AI gurus were to concentrate on one specific mod and one specific version of that mod, great things could surely be achieved without the hassle of worrying about compatibility with every piece of mod out there. I am 100% certain that a working campaign is a lot better thing to have than to put effort into making an AI work on many mods. Again, I do not know how these things exactly work, but considering how different all the mods can be, making an AI for a specific mod should be slightly easier, at least according to my logic.

In my opinion, the campaign does not need to be anything special. I'm sure most of you have seen the original TA campaigns. Those are certainly enough, at least for starters. Just different scenarios after another. Then someone (my silly accent might not cut it ) can talk through some filters to make the intermission speaks. It doesn't have to be fancy at all, just like it wasn't in the original TA.

Now, before you go all "Why don't you go somewhere else asking for others to do stuff for you!" I can let you know that while I can not help with stuff like AI or implementing the campaign structure into Spring, I can help with the more analog aspects, such as designing and making scenarios in the form of maps and plans. I also know another friend of Spring who will certainly help if we get this thing going.

I have some more ideas in my pocket, but I think that it is proper at this point to ask whether you are with me here. Probably the most important thing would be to get one of the highly skilled AI developers give an approval and help with this task, because otherwise I can't imagine how this could work.

A functional campaign would be a great step in the Spring development. No doubt it would bring in a a lot of new players.

In short, the things we need are:

1. Decide on the mod and the version of this mod to use. EASY to do, but I beg you people, do not start shunning this project even though your favourite mod is not chosen.

2. The campaign system implemented into Spring, saving and loading, campaign progress storing and such. Shouldn't be a big deal.

3. An AI that could be tailored with minor scenario specific modificiations that the scenario author could implement with relative ease. I do not know whether it would be easier to modify an existing AI or to create a completely new one. This is probably the most important thing.

4. Ideas for scenarios and mapping to make those reality. There should be a mix of basic base vs. base, but also imaginative and/or weird scenarios, where the player (and especially the scenario author!) must think of new things. I for one can certainly help with these things.

5. Finishing up work. intermission speeches and such. The story for the campaign could be figured out later once all the scenarios and such have been made. These are all minor stuff, with which I'm sure everyone can help.

Let's fucking do this.

my opinion is just that ba and ca shouldnt get a campaign(well if they want to they can make it:P) i think there should be a mod completely focussed on a campaign, ta based or not

we will not have stuff like

wtf my mod should get a campaign, it isnt fair

it would be much easyer to manege and thats why it would be better!, well no more time to write now, i will post again later

nota has missions.

Hoi wrote:my opinion is just that ba and ca shouldnt get a campaign(well if they want to they can make it:P) i think there should be a mod completely focussed on a campaign, ta based or not

That kind attitude, however, can be highly problematic, as it has that "they should do it" factor. CA and BA are already there, and they are both very suitable for a campaign, so I say that in order to actually get somewhere, an existing mod should be used. Also the campaign would act as an introduction to the game. If a completely new mod would be used, it wouldn't introduce the player to anything.

Fatal wrote:

Hoi wrote:my opinion is just that ba and ca shouldnt get a campaign(well if they want to they can make it:P) i think there should be a mod completely focussed on a campaign, ta based or not

That kind attitude, however, can be highly problematic, as it has that "they should do it" factor. CA and BA are already there, and they are both very suitable for a campaign, so I say that in order to actually get somewhere, an existing mod should be used. Also the campaign would act as an introduction to the game. If a completely new mod would be used, it wouldn't introduce the player to anything.

i think you get me wrong, what i mean is that we take ba or ca as a base (or a mix) and from that just make it the regular campaign mod, with like new campaigns and (some) changes from the base mod every month orso (as a diffrent download, and with a diffrent name than the base mod, like tacampaign or something)

i tried a few, but nothing really "felt" like a real mission/campaign. i think fatal is right saying that we need a fixed mod-version and a dedicated ai to improve mission development. perhaps we could merge ca's high poly models and nice effects with ba-gameplay and create a "completely-balanced-annihilation-sp-campaign-mod"

Hoi wrote: i think you get me wrong, what i mean is that we take ba or ca as a base (or a mix) and from that just make it the regular campaign mod, with like new campaigns and (some) changes from the base mod every month orso (as a diffrent download, and with a diffrent name than the base mod, like tacampaign or something)

Ok, I think I see what you mean there. Yes, that would be a pretty good compromise, but I say we save that until shit actually hits the fan and we really need to compromise. Until then, let's just say that we use an existing mod for this. After all, this is nothing but a concept yet. Before we go into details, such as choosing the mod, we need to get some people to support this. I just asked this person I mentioned in the first post, and he was behind me on this. So at this point we have two fellows willing to make maps.

Honestly, if this thing does start coming off the ground, I'll back up Fatal's invitation and help out with stuff. I have a basic idea of how to make maps, and I'll help with scenario planning ect ect.

That's about as good as I am tbh.

Sorry, you are stepping into a minefield of community angst here... I think you have probably just pissed off 90% of the active community in one post... but I digress.

First of all, I don't think I've run into anyone who felt the currently available AIs were unsuited to being mission opponents so that's an interesting take on the matter. Effectively, the mission making abilities of spring do things like spawning units manually and/or then giving them orders while also supporting triggers by giving access to useful game variables. As such, a simple stable AI could work quite well within these parameters without a new MOD/Campaign specific one being drafted up. There are a number of AIs that take a mod specific configuration file to understand how mods work which should be good enough for what you are looking for.

However, if this is not the case and a new AI needs to be created... that's an awful lot to ask of someone when the engine and the community barely support missions as it is.

If anyone could show us a working tutorial MISSION that goes through the basics of a mod's gameplay (with the empty AI that just sits there and takes it) and I think you would go a long way towards people getting interested in campaigns.

One other beef I have is that there is a LUA parser built into the engine to handle missions but it doesn't have direct access to the more advanced features of the engine so you can't play sound, display images or draw on the map, or even have instructional text show up somewhere other then the chat-box... so you'll either need to wait for some engine dev to care enough to add this or get someone to LUA up a good gadget that can interface with the mission scripts to get anything more then very basic functionality, which is the main reason I gave up on the whole thing (too lazy to do it myself).

So anyways... in summary, make ONE GOOD MISSION that shows everyone how it can be done and I think campaigns will follow nearly behind.

As I recall, if someone could make a proper mission that could be started by the command line luascript interface and then drop the results to a file then the campaign could be managed by a simple program(like a one week job using high-school level programming techniques).

Good words there, Sinbad. As I said, I am certainly no expert on AI matters. If the case is as you say, that an AI could be used in a campaign and made to do specific things on a specific scenario, then that's one obstacle out of the way. However, it MUST be such that a mapper can easily determine the things an AI does, even if they're simple and consist of mainly "bomb this zone", "make land attacks from these directions in this strenght" and "start firing with the bertha at the 20 minute mark", because that would be great for this purpose.

I'll think about the other things you said for a while, and post my thoughts on them later.

Well, we (Fatal and I) should probably make the tutorial mission at least, to show that we are dedicated to this idea.

Yes you should but first make sure you know about all of springs quirks and buttons and abilities.

Fatal wrote:However, it MUST be such that a mapper can easily determine the things an AI does

are you saying that you will need to make a map to make a mission? you should be ble to make a mission with an existing map
@ the 2 people above: LOL 2 cats!

Hoi, I get what you're saying, but practically all Spring maps are just made for more or less balanced gameplay, which certainly is not what campaign scenarios are about. There needs to be interesting terrain, and this tutorial map/mission will give good practice to me and Osiris.

Fatal wrote:Hoi, I get what you're saying, but practically all Spring maps are just made for more or less balanced gameplay, which certainly is not what campaign scenarios are about. There needs to be interesting terrain, and this tutorial map/mission will give good practice to me and Osiris.

i still think im right, you can fit a map in your campaign but you can also fit your campaign in a map

Put simply, any AI developer able of implementing what you suggest would not have the time to read your post properly. If they did then theyre either vastly talented or they skim read like I did.

AI dvelopment at the moment may have seen a small resurgence but AI devs are almost entirely focused inwards at the moment trying to engineer the enw AI interface.

May I also point out that there is and shouldn't be a 'spring campaign'. That doesn't stop the <insert gamename> team building a <insert game name> campaign.

SinbadEV wrote:Sorry, you are stepping into a minefield of community angst here... I think you have probably just pissed off 90% of the active community in one post...

truth.



again, I want to mention... NOTA HAS MISSIONS

So anyways... in summary, make ONE GOOD MISSION that shows everyone how it can be done and I think campaigns will follow nearly behind.

Chickens! In the end it's just a mission with special units. There is also a 90% done CA tutorial.

One other beef I have is that there is a LUA parser built into the engine to handle missions but it doesn't have direct access to the more advanced features of the engine

That system is obsolete, it should not be used anymore. Use LuaRules. You don't need an interface for a gadget, the gadget is itself is the mission script.

AI dlls are not necessary for missions, probably not even very useful. You have all the tools you need to control units.

We have everything we need, except time. Traditional missions are not very replayable and so, in my opinion, there are better investments in modding time.

I must stress out that the missions do not need to be greatly script dependant. What I have in mind is similar to the campaign in original TA. That campaign was simple, it worked, and is what we should be aiming at first. Once we have a campaign of SOME KIND, we can start considering fancy maps with many scripted events. Right now what is what Spring needs is something simple and relatively easy to make, so that it gets actually done.

we have had these threads, this is not a new idea, spring is not able to complete support it. Campaigns are not possible only missions.

THIS WE crap also needs to get fixed..

LESS TALKING AND MORE DOING, you want missions, get started, you want campaigns, get started on the code to make it happen. If you can do neither shut up.