Spring RTS Engine

I will say that I believe there was no consensus on the language to use

I wasn't aware there was much of a choice? How practical is it to build a game engine in languages other than C++?

I thought much the same, but my memories from the period are not in focus. That point did come up a few times, but I don't know what the details were. Perhaps it was about the APIs for Native AI support in various languages, or a specific component they wanted to build separately, I really do not remember which is why I said "I believe..." not "I know..."

Not sure of the entire context here but it's actually becoming more common to build game engines in languages other than C++. Having said that the LIBRARIES the engine uses are nearly always written in C++. A good example of this would be pygame which has a lot of its engine logic in Python classes but lets the SDL and OpenGL libraries do the heavy lifting.

With the popularity of Android I would expect to see a number of new engines targetting the Java platform but Android itself and its core libraries are still written in C++.

There are engines built in C# or delphi too, even java. If its compiled language it does not matter at all. If it's JIT language with runtime framework, it matters less and less now becasue of modern GPU doing most of the hard work in compiled shaders.

But anyway, building a new engine now is silly, if you want to start a game use one of existing engines unless you are uberpro with extreme willpower and years and years of free time

Licho wrote:But anyway, building a new engine now is silly, if you want to start a game use one of existing engines unless you are uberpro with extreme willpower and years and years of free time

What if you can't find an engine to do what you want?

What if you can't find an engine to do what you want?

If you have fuckton of money, or political capital to summon those money via kickstarter: buy a source license on a decent engine and hire a guy to develop that feature.

If you have fuckton of neither: analyse existing open-source engines for least flaws for what you want to do, and then modify them to suit your needs.

Das Bruce wrote:

Licho wrote:But anyway, building a new engine now is silly, if you want to start a game use one of existing engines unless you are uberpro with extreme willpower and years and years of free time

What if you can't find an engine to do what you want?

Like.. what?
Have you see what you can do with say unity?

Have you see what you can do with say unity?

I heard making an RTS with Unity would be a nightmare. Or, more validly, making almost everything from scratch.

Licho wrote:Like.. what?
Have you see what you can do with say unity?

Turn based strategy game.

You can make turn based strategy in any engine - thats the least demanding genre of them all...

Agreed. You can probably make it in Spring with (a lot of) effort.

gajop wrote:Agreed. You can probably make it in Spring with (a lot of) effort.

And you can make a RTS in Source, but that doesn't mean it's a good idea.

The more Unity games I play, the less I feel Unity is suited for anything in particular, although you can do much with it.

Licho wrote:You can make turn based strategy in any engine - thats the least demanding genre of them all...

Well, perhaps not the least demanding genre - logic, non-real-time puzzles - but fairly undemanding.

Das Bruce wrote:

gajop wrote:Agreed. You can probably make it in Spring with (a lot of) effort.

And you can make a RTS in Source, but that doesn't mean it's a good idea.

http://www.nucleardawnthegame.com/

It works, but you have to massively change the source engine.

Forboding and I both played Nuclear Dawn (Sometimes with Sleksa), for those who are curious. It was a good game that did not achieve critical mass in terms of players, I haven't played it for some time.

I heard making an RTS with Unity would be a nightmare. Or, more validly, making almost everything from scratch.

There is a cute little RTS called Castle Story that is done in Unity. And it does seem to be indeed that they did a lot of work from scratch.

Yeah, RTS games generally have some very specific requirements that you don't see in other games - the sheer number of actors in play is one big one, but the huge thing is sync-based multiplayer. You need to design every possible bit of gameplay logic, every microscopic thing that interacts with player actions, right down to programming-language-level, to support sync.

An RTS could leverage a lot of existing graphical and input and audio stuff, but the actual gameplay and netcode would need to be from-scratch, which means the engine needs to support a high-performance language and not a light scripting engine like Lua.