I know this will likely spark a flame war, but here goes anyway:
I would like propose a new addition to the feature system: Static Meshes. Anyone who has mapped for UT knows what I'm talking about here, and they could be used fairly well in Spring, with a little haxing.
What I'm proposing is a feature that, when loaded, also loads a path file that is also made in a modeler or other program. Basically a series of dots and lines to show where units can and cannot travel, and a set of connector path nodes to connect with the standard Terrain paths. That way the engine won't need to auto-calculate where units can/can't/can-with-difficulty travel, as it's already done. The feature should also have a tag for what terrain type (defined it map file) it is classified as, as well as how much metal it has on it; In metal view, the feature should take the shade of how much metal it has. Such Static Meshes should be inherently indestructable and non-burnable, and should have a decal object for when hit by explosions.
This would enable bridges, multi-level buildings, and would allow AI to be modified for use with such objects relatively easily.
Discuss! *puts on asbestos underwear*
Ideas for model->terrain format: The Technical View
Moderator: Moderators
See discussion of 3D collision meshes for another take on this subject. Long version short: I think that what you're asking for would be less useful than having a new way of determining object -> object -> point -> ray -> map mesh collisions, that would (finally) allow for true 3D collision detection between objects. We could not only do bridges, we could also have, for example, a building (let's use a windmill as an example) that, destroyed, used the Fall event... leaving behind an object that falls, actually blocks LOS, and might impede objects trying to navigate. We could have much more interesting wreckage, buildings that can be destroyed in interactive stages that actually matter past eye-candy, etc.
I really think that Spring would become one of the best RTS game engines ever if it reached a new level of physical interaction. Not pixel-perfect, not perfect period... just more accurate and lifelike, but still low-resolution and fast. What I proposed is a sketch of how this might be done- hopefully y'all will percolate it into something workable, so that we can see some really nice interactive objects in the game engine that just aren't possible right now.
I really think that Spring would become one of the best RTS game engines ever if it reached a new level of physical interaction. Not pixel-perfect, not perfect period... just more accurate and lifelike, but still low-resolution and fast. What I proposed is a sketch of how this might be done- hopefully y'all will percolate it into something workable, so that we can see some really nice interactive objects in the game engine that just aren't possible right now.
Spring game itself...
Jouninkomiko: 1, working on the GUI
SJ: about 1/4 active I think
he is planning on making a better water system.
Me: 1, busy with making the port code fully working on windows, and working on a better terrain system (Buggi and me are discussing sharpmap for this system).
People working on the port, (that is: doing port stuff, no new features so far):
Chris Han: has done a lot, but now seems to have very little time.
Tobi (tvo): ported a bunch of AIs to linux, fixed bugs, makes my code build on linux and is currently busy with a new scons build system.
Quite a few other have contributed to the ported code, but like I said, that didn't include new engine features.
So that's about 2.25 programmers working on actual new features
I have to admit I promised too much, I still have upspring and map features to fix... and I told I was going to support collision meshes sometime.
Apart from my new super awesome terrain renderer I'm not really motivated either...
More info on the terrain renderer when I finished the boring parts of it...
Jouninkomiko: 1, working on the GUI
SJ: about 1/4 active I think
he is planning on making a better water system.Me: 1, busy with making the port code fully working on windows, and working on a better terrain system (Buggi and me are discussing sharpmap for this system).
People working on the port, (that is: doing port stuff, no new features so far):
Chris Han: has done a lot, but now seems to have very little time.
Tobi (tvo): ported a bunch of AIs to linux, fixed bugs, makes my code build on linux and is currently busy with a new scons build system.
Quite a few other have contributed to the ported code, but like I said, that didn't include new engine features.
So that's about 2.25 programmers working on actual new features
I have to admit I promised too much, I still have upspring and map features to fix... and I told I was going to support collision meshes sometime.
Apart from my new super awesome terrain renderer I'm not really motivated either...
More info on the terrain renderer when I finished the boring parts of it...
If I didnt have a great mass of NTAI functionality to restore, 3 rival AI's lynching away at NTAI userbase, 6 big exams in the next 2-3 weeks, a website that needs finishing (darkstars.co.uk is nearly done, I just need to put the content back up and mod the graphics somewhat), a huge backlog of ICT coursework,
then maybe I'd be able to help, especially since there're several areas I have had thoughts for and feel they need expanding or restructuring (such as unit sub-AI dlls and AI interface additions, Options profiles, camera modes, xml support, an array of pretty GroupAI's (a better version of the streaks in the cursor AI but using 3D icons akin to HW2, an information GroupAI that shows data on a 3D map model drawn with the line tools, border lines etc) etc)
But till then I'm busy. After the summer exams though, I'll have 2-3 months of free time to do whatever I want though ^_^
then maybe I'd be able to help, especially since there're several areas I have had thoughts for and feel they need expanding or restructuring (such as unit sub-AI dlls and AI interface additions, Options profiles, camera modes, xml support, an array of pretty GroupAI's (a better version of the streaks in the cursor AI but using 3D icons akin to HW2, an information GroupAI that shows data on a 3D map model drawn with the line tools, border lines etc) etc)
But till then I'm busy. After the summer exams though, I'll have 2-3 months of free time to do whatever I want though ^_^