Traevaine Entertainment - Panthera Game Project
DEMO OUT
2010 Blender Summer Game Competition Winner (Best Graphics)
_______________________________________________ Basic Information
Project Title : Panthera Game Project
Game Title : Panthera
Type : 1st-/3rd-Person Roleplaying Game (Singleplayer)
Development : begin 2009 | end -
Website: http://www.traevaine.com
Prizes, etc.: 2010 Blender Summer Game Competition “BEST GRAPHICS”
[/B] Development Status [Alpha] Progress [||||||||||||||…][19.6%]
We are the development-team of ‘Traevaine Entertainment’ and we would like to present our game here in the forums.
Our team has a variating amount of members; 8-10 at the moment, all at
the age of 14 to 40.
Our members come from all over the world, but the core development is in Germany.
We are an experienced team of blender-users with a lot of motivation and lots of ideas.
Our team members perfectly match to each other and each one has his own skills.
Currently we are working on the ‘Panthera Game Project’,
a 3rd-person/1st-person singleplayer role playing game
(3rd-prs/1st-prs - RPG).
We aim to get as much as possible out of the BGE.
Team members and several WIP-Shots are listed below.
(Blender 2.49b)
Important note:
For recent updates, please look at the latest entries in this topic.
Looks really good. Do you have an enemy ai yet?
It can be a stumbling block… good to focus on that early.
I use (smart) waypoints rather than A* as it is a good solution for the Blender game engine.
Here is a very good article on the reasons for using waypoints: http://aigamedev.com/open/reviews/alienswarm-node-graph/
Though it doesn’t really describe how to implement them.
The basic idea is that each waypoint collects information about its neighbours and stores that information as properties. AI agents can then access that information and find the entry which is most useful to them (for example the nearest neighbour which is nearst to the agent’s navigation goal).
it works well because there is just a short period of collecting information at runtime, and then the info is available throughout the game. It uses very little processing power.
Anyway, you guys get 5 stars from me. Team development is much more difficult than solo.
I’ve just finished making a script for generating waypoints…
I will be working on it more to add a way of navigating them for the player and enemies, but for now I thought your programmer may find it a useful starting point.
Each way point will store a list of several neighbours as game objects.
You should have your ai players track to the nearest waypoint and when they hit it they can get it’s “near_list” property (by using the hitObjectList fuction in blender) and use the “close_node_search function” to choose the node from the list which is nearest their guide target. Their guide target will vary depending on thier current ai behaviour pattern, but would usially be the player as enemy.
However, you could write other other behaviour patterns for the ai, such as one which would find the nearest “cover” position, or nearest other enemy, if you want them to team up before a battle, or nearest patrol point if you want them on patrol.
This can be a very flexible system, which is 100 times faster than A*… maybe ^^
I will post a template later with some sample ai behaviour patterns. But I thought you may want to take a look at it before then.
I really want to see more quality blender game engine games around on the web, so if I can help other designers I always want to try. ~ though I am kind of busy with my own project too.