Spring RTS Engine

smoth asked me this in PM but I'm guessing he's not the only one wondering after the 30+ posts on model / engine format changes I've made in the last 2 days. Figured I'd share my answer for the bored and curious.

smoth wrote:So you are taking on all this stuff, I only gave your threads a cursory read(much of it is beyond my understanding) but it seems like you are taking on a whole lot of code or is this more of a fact finding sort of pursuit?

It's a lot of work and I'm not the most qualified to do it either, however it's more like pulling apart a lego model and putting back together differently than building something from scratch. In the process I find out more about how it works and I'm able to make it do things I want.

The desire to do this comes from my experience modding Supreme Commander. I was hell-bent on creating something really new, that hadn't been done in RTS before, and the ENGINE fought me the whole way. Despite anything GPG said about modding it, the reality was, and still is, that beyond the trimmings (models, sounds and basic scripts) you can't force the game to be anything but a typical TA clone (at least, not without some REALLY nasty hacks). It's also poorly documented. Had exactly the same experience with Dawn Of War.

With Supcom me and this other guy, forgotten his handle, we wrote Lua code on top of the engine to do really advanced stuff like MP save and coop play (me) and custom networking (him) but we both found it really slow and frustrating and I decided that was the last time I work on a closed commercial engine.

So I came across to Spring and it's a much better experience. The documentation still sucks but with source code and guys like Auswaschbar, jK, lurker and BrainDamage around I'm generally finding what I need pretty fast.

It'll probably take me about 6 months to get this model / rendering stuff to the point where I'm satisfied, then I'll resume working on my actual mod/game. I only stopped because I have about 100+ models in 3DS format and I got really fucked off with UpSpring. Then, when me and BD got the model loader working (or near enough), I got really fucked off with the renderer (as you can probably tell from my recent threads).

So that's why.

very cool!

win i applaud you for your efforts but now im sort of intrigued about your game . Any news/info and pics of the models? :]

+1 to all your efforts. Don't give up, and remember jK is a supergenius so he is naturally hard to work with. Rumor has it he's just a brain in a jar that's hooked up to the internet.

I've left clues here and there but I really don't want to go into details at this point (kinda ruins both the anticipation and the surprise). Still I can point you to what I was working on BEFORE I left the supcom community and you can speculate from there:

Supreme Commander : Advanced Tactics

The above was written at a time when everyone pretty much assumed that an SDK would be coming out some time after the game was released (GPG were hyping SC as the most moddable RTS ever, etc). Suffice to say it's nearly 4 years since I wrote that and still no sign of an SDK.

Note: 'Ideas people' are NOT welcome. I need people to do *actual work*. We can come up with ideas on our own, thank you.

<3

CarRepairer wrote:+1 to all your efforts. Don't give up, and remember jK is a supergenius so he is naturally hard to work with. Rumor has it he's just a brain in a jar that's hooked up to the internet.

I haven't found him difficult, quite the opposite in fact. Large sections of the manual I've been working on are based on his posts and IRC conversations. Many of his concerns are valid, just not enough for me to change my approach. BrainDamage voiced very similar concerns, that there would be a lot of resistance from mod/game developers. I've yet to see much evidence of that so I still believe I'm taking the right path.

As I said to jK, a lot depends on whether I deliver on my promises and how the new system comes out so it's too soon to for anybody to say what's really for the best.

As long as I don't have to scrap all my work, I don't give a crap. You are not stupid I don't think you are going to fuck anything up. Might seem like blind faith but if something breaks that is what revert is for .

SpliFF wrote:that there would be a lot of resistance from mod/game developers.

Wait till you inadverventently clean up a portion of code with the unintented side-effect of disabling an ugly hack that a mod/game developper came to rely on unknowingly to you.

How much you got it done so far?

The model loader has working (tested) support for OBJ, 3DS, MD2, MD3, MDL, NFF and Collada meshes (many other formats supported but not tested). Only MD5 support is broken that I know of. Also about 90% of the texture/materials code, except that the texturing is busted. I've ceased trying to fix it because I'll be rewriting the relevant parts of the rendering/texture pipeline anyway.

Well it's funny tbh, the day you posted that thread I was dreaming about having a system where OBJ's could be combined via lua, and all of the parameters could be set in a text/lua/tdf sort of deal. Then I read your original post... Was kinda interesting how that worked out

And it's nice to see someone who is actually capable doing something about shitty upspring. Most of us get good a fucking pissed off, cuss out JC, Upspring, spring in general and everyone on the boards for a while... Then turn right around and start dicking with it again. All sound and fury, no teeth because we aren't capable of fixing it even if we wanted to.

SpliFF wrote:Supreme Commander : Advanced Tactics

Sounds a bit like CA (or more accurately my view of what CA is/was aiming to do).

Frankly, I'm insulted

Seriously though I've dropped the whole persistance thing for the moment, it's too much work and yeah the CA team is already doing it. My goals do share some things in common but not as much as you'd think.

It's also not based on TA, more of a WW4 theme with very gritty visuals.

CA has nothing to do with TA - or at least it wont have when its done. Settings and visuals are supposed to be completely different.

I was referring to gameplay. You know, commander builds buildings, building build units, units collect metal, etc... My game is more like S:44 but still different again.

SpliFF wrote:Supreme Commander : Advanced Tactics

"WHY?

Those who have played Supreme Commander or indeed TA will know that SC is basically a base-building game. Although the outcome can be changed by early skirmishing, the game essentially reaches a point where the only viable tactic is to out-produce your opponents. This generally involves reaching T3 (tech-level 3) as fast as possible and building lots of fusion reactors and mass converters. Although taking mass points helps it really has no value after your 4th converter or so. If anything, the effort to take and hold these points detracts from your reactor-building and usually gives an otherwise equal opponent the overall advantage. Once you have a clear economic advantage the rest is trivial; Wether you build t3 bots, t3 gunships, artillery, nukes or t4 monsters your victory is practically assurred as long as you can pump out the most units."

You still feel like this about TA-based games? :)

Some mods are better than others, smaller maps tend to favour attack over porc. Some mods and gadgets have made things better (like mex overdrive, auto mex upgrades). On the whole I still don't believe most TA-based games encourage in-depth strategic or tactical thinking.

Instead I tend to find that trying to micro an assault results in my base production collapsing and engineers falling idle due to lack of orders. Once my economy collapses there's little I can do about it since my enemies will be increasing their production at a logarithmic rate. I believe that's where the main issue lies, ECONOMY RULES TA.

That's not to say this should be "fixed" either. TA has a gameplay style that suits people who like economic management (whether they realise this or not). This style is what most TA players want. You can't make major changes a popular game and expect fans to accept it.

However that doesn't mean you can't create a new game without existing expectations to meet. Lets put it this way; how often have you had to create sophisticated ambushes in a forest? when was the last time you saw a coordinated Normandy-like beach assault? have you ever had to use diplomacy to play a civilian faction off against a human player? Have you even seen an assault team capture a base without damaging structures? When was the last time winning meant capturing your enemies commander and torturing him to reveal troop movements?

I'm not saying my game does any of those things, I'm just saying it to make you think about what "strategy" and "tactics" really mean.

Commercial RTS tries to give players what they expect instead of taking risks on the unknown. That's why it's going to be difficult for a commercial RTS to deliver really original gameplay. I have more flexibility and I intend to use it to create a deeply strategic RTS.

That's about all I'm prepared to say about my game at the moment.

I'm not knocking TA, I just don't get as much enjoyment out of it as many people here. That's why you rarely see me in games. I'm here to write a game, not play TA or its decendants.

Having said that I am ashamed to admit I haven't really tried any alternative mods except KP and Spring 1944. I downloaded P.U.R.E but still haven't actually played it, which is just lazy. I also like the look of SW:IW but again I haven't taken the time to beg whoever it is I need to beg to become a beta tester. From the sounds of it though I think I'd like it.

Spring 1944 is closer to the direction i'm headed but I didn't really enjoy it. I can't point to any specific thing, it's just how I felt at the time. I can't vouch for current versions as this was a while ago.

So yeah, I haven't really told you anything (which is intentional) except that I'm not happy with where things are now. It's not from lack of trying, it's just my tastes. I think I own nearly 60 strategy games ranging over the last 20 years and I really only liked a few of them: TA (for a while), Civ 2/3/4, Alpha Centaurii, Medieval/Rome TW, Populous 3, Combat Mission, and Dungeon Keeper come to mind. Also Epic 40k which is a tabletop game.

The thing that most of those titles (except the Bullfrog ones) have in common is that you have time to think. You rarely feel so rushed you can't do anything but react. You really get a chance to plan and coordinate clever strategies and implement them and your opponent has the same opportunities. When I do play TA I like to play on huge maps and drop the sim speed down to around 50%. Unfortunately it's hard to find a game like this online (Just people thrashing out tried and tested tactics on DSD).

It's a shame really. Hopefully I can do something about it.

SpliFF wrote:The thing that most of those titles (except the Bullfrog ones) have in common is that you have time to think. You rarely feel so rushed you can't do anything but react. You really get a chance to plan and coordinate clever strategies and implement them and your opponent has the same opportunities.

This is relevant to my interests.

I always thought of RTS games as tactical combat, some of them with base building "strategy" part. Which is just management and some general strategy choices like fast rush, large army midway or tech.

In real strategy games, usually turn based 4x, you do plenty of stuff like SpliFF described. Planning ambushes, invasions, diplomacy etc. Designing and building units for a single specific mission, calculating the maximum efficiency. I don't think spring engine (or any real time strategy engine I've seen) is suited for real strategic gameplay. I hope you'll prove me wrong.