This is a a medieval RTS game which is similar to an old (1991) game called 'Castles' for MS Dos.
The basic idea is to build a castle and defend it against the enemy during its construction.
This process will take a long time and will require the player to manage resources and money to maximise worker productivity and the size of its army.
There will be politics involved - the user will be approached to make decisions for many different types of scenarios which can affect gameplay.
This is mainly a strategic game rather than a typical RTS type game such as C&C (although it was originally planned to be just like that, but MMF isnt powerful enough for that yet). It is not going to be two teams fighting it out on a map. The enemy will only attack at certain time periods at which time the player will be notified so they can prepare their defences for the attack.
Some of the features;
- Isometric viewpoint with 90 degree rotation
- Grid based pathfinding
- Unit selection box with typical RTS point and click movement system
- Full working economy requiring the user to collect different resources (most likely will be wood, stone, iron, food and gold).
- Political system which requires fine balance between religion, taxes, population morale, etc
- Real-time events which can affect your current situation (such as fires, the plague, bandits, etc)
- Multiple buildings to build including Defence structures and economic buildings
- Multiple unit types including different workers as well as army such as infantry and seige weapons, and maybe boats (if I decide to make water maps)
- Random map generator
- Map editor with event triggers (and maybe scripted events)
Project Forums
Topics
Posts
Last Post
Public Forum Public forum for open project discussion
So I have been playing around with c++ this last week and have read a lot of tutorials for A* pathfinding and have successfully made a working pathfinding example. (I didnt write from scratch, modified an existing source code).
It is still slow when finding a path for many units at the same time, but I have read of a good workaround for this which uses 2 layers of pathfinding to reduce the ammount of lag. it splits the maps up into big nodes (say to represent 4x4 grid squares for example) and small (per tile). It uses the big nodes for the inital path search, and then finds a path using small nodes between each big node. Its a great idea and sounds like it would work great. I am going to try to implement this, and if successful, I will then port it over to an MMF2 extension and see how it goes. If all is successful, this game will be epic!
OK so I think I jumped ahead of myself with my previous post about LUA pathfinding. Although it works perfectly, it is too slow for what I want to do. Instead I have returned back to using the Advanced Pathfinding Object instead, however now I made it so that once a path is found I copy and store the path in LUA. This allows me to do faster lookups of the path information without the overhead of calling on MMF to pass values back into LUA.
No more useless devlogs for a while. I will try to post another video later showing pathfinding in action.
I successfully implemented A* pathfinding using LUA. What does this mean? Well previously I was using the Advanced Pathfinding extension to store all my path information. This worked OK however everytime I needed to retrieve any path information I had to do a call into MMF and return the values which I felt had a bit too much overhead. Using LUA, I can store all of my map data outside of MMF which should result in more speed. So far LUA seems to be able to find the paths extremely fast, however I haven't tested it on big maps yet, so I cannot be sure how effective it will be for what I have planned.
Here is a quick video showing a test engine for isometric rotation.
This only shows some basic terrain and static sprites, however it can easily be applied to moving objects too
I also demonstrate my Z-ordering code which you can see on the trees. As I rotate the map the trees a re-ordered based on their Y position.