Spring RTS Engine

Most vets have probably looked at the engine at one point or another. Some have even contributed.
That said, you don't need to be highly familiar with the engine to start making games, but it certainly helps, a lot.

Most of the people in https://springrts.com/wiki/Authors#Special_Thanks are gamedevs who have made minor contributions to engine code.

ps. Silentwings, <3 for the wiki praise

d000hgRTS, your feedback as a newcomer to the particularly useful and useless areas of the wiki would be useful, once you get to the point of putting 'pen to paper' in making a game.

FLOZi wrote:d000hgRTS, your feedback as a newcomer to the particularly useful and useless areas of the wiki would be useful, once you get to the point of putting 'pen to paper' in making a game.

Good point. You often find people familiar with an engine/codebase forget all the "well it says X but since version 1.2.6 you have to do Y" things which trip up noobs... I see this from both sides since my work leads me to use an unusually diverse range of technologies. I'm often a noob, and then end up being the experienced one seeing other noobs :)

These are important when working on long running projects

Welcome!

Welcome!

Metal Factions is also under active development, with new units every week and other changes. It's just keeping a low profile.

The Lua scripting can do almost everything. For many mechanics you could want it is probably harder to put in the engine than it is to script it. The TA-ness of most mature projects is due to past limitations and the type of people attracted by Spring. It is easy to not use unit or projectile physics by setting very high speeds or accelerations.

Any Command and Conquer or Age of * game look fairly easy to reproduce in Spring. For Warcraft 1 and 2 it would take a bit of work to make units move in large grid squares. I am not sure whether lua or engine modifications would be better for that. Blizard games past Starcraft have continuous movement and also look easy to reproduce.

You can make a top down camera with lua but you may have to add an option or write a new camera for an orthogonal view.

Populous the Beginning would require a lot of engine work to wrap the map into a torus and then display part of it to give the illusion that it is a sphere.

In the interest of disclosure, SpringRTS isn't the only opensource RTS engine, although I'd say it is the most versatile. Warzone2100 is quite mature as well, and 0.a.d., although still in alpha, is shaping up to be a really impressive work with it's own mod support. 0.a.d very much has it's roots in the the Age of * Games.

darklord42 wrote:In the interest of disclosure, SpringRTS isn't the only opensource RTS engine, although I'd say it is the most versatile. Warzone2100 is quite mature as well, and 0.a.d., although still in alpha, is shaping up to be a really impressive work with it's own mod support. 0.a.d very much has it's roots in the the Age of * Games.

Thanks. I'd heard of Warzone, however my understanding is that's an open-source game where development is driven by the community. Clearly one could fork the project and make their own game, but it's not (I think) designed to separate the engine from the game like Spring?

darklord42 wrote:0.a.d.

0AD looks really good in some aspects -> and if you need mesh deforms they're definitely worth considering. I also find their rendering (eyecandy) to be better than what Spring has to offer. Intel support also seemed slightly better there.

That said, I think Spring has better pathfinding and performance. It's better tested and modding support is better (0AD is really a game that allows modifications, rather than a game-agnostic engine like Spring). Also, we have a better developed infrastructure (lobby servers, clients, download programs/updates, etc.).

0AD community is more centralized. They have one umbrella project for the engine/game/scenario editor & everything else, with specific dev leads/members, and they do do development in that direction.
Spring on the other hand is a lot more decentralized. Engine, game(content), lobby client, lobby server, scenario editor devs are working on separate projects. We still somehow manage to collaborate between each other, but there's a lot more anarchy here

Welcome...

First things fist - you propably guessed it, every piece of code has a history. This project here started out as 3d demoplayer for TA, and while it progressed steadily away, and shed a lot of skin, down below there are still some TA-isms lurking.

Another thing is, that game programming - is ultimately high performance programming.
So while everybody knows about oo-rientated programming concepts and how the code should look like, there are some parts that have to be optimized to run better in the bare-metall world.

We dont do execessive hot-cold splitting to keep stuff readable, but jk is current profiling L2-Cache misses, so if you want to join on that quest, your welcome.

Finally - spring has multithreading support, but the plan is to migrate new features if possible away from the CPU towards the GPUs. OPENCL comes to mind there - but it hasnt been synced yet.

gajop wrote:

darklord42 wrote:0.a.d.

That said, I think Spring has better pathfinding and performance.

Indeed, despite technically having existed before SpringRTS, as early as 2000, they are less mature of a project. I suppose they really only got into gear in the past 7 years. Also, I suppose their goals are more ambitious then Spring in some respects. They aren't just an engine, they are gearing to be a AAA quality game with 12 factions. They actually have hired people in the past to get their game further along, and there is still a lot to do. :)

Their last alpha 18 has been getting a lot better performance wise. Updated to C++11 and Spidermonkey v31 Javascript engine.

did you look at the links I gave you? did you find answers there?