[Verse-dev] Update and verse related game development.

[eskil at obsession.se]

Hi

Ive been very busy, doing mostly traveling. I want to tank every one  
who made the Siggraph Verse BOF such a great success and also give a  
special thanks to the Crystal space crew for the cool verse demo in  
Lipzig. So for the last year or so I have been working on game  
pipeline related technology using verse, you may have seen the demo I  
did together with Sony at GDC, where we showed verse running on the  
PS3 and the Collada exporter, and later I released the verse export  
tool-kit. So while working in this I have also been working in secret  
on a game using all this technology, and to use as a test platform for  
all kinds of new ideas. Leipzig was the first time it was shown in  
public, and you may have seen the screen shots at quelsolaar.com. The  
entire pipeline for the game is built around verse, and I obviously  
use verse enabled tools to create content for it. The engine can  
connect to a verse server so that you can edit objects WYSIWYG in any  
verse tool and see them in the game engine (much like you can in  
Crystal space).

So a little about the game:

The idea is to create a story based game, but not one where the game  
tells a story but rather on where the game together with the player  
creates a story. Many games today say they are very dynamic and "its  
different every time you play" but this is only in the micro  
perspective, you still have missions, objectives, levels and a story  
and cut scenes. So in a macro perspective it is very structured.

So what we are trying to do is a macro perspective that is much more  
like an RTS game. In a RTS one of your resources may be attacked, and  
then it is up to you to go and defend it or suffer the consequences.  
Imagine building a game and balance it so that the CPU side sometimes  
wins and sometimes looses battles, but neither side ever wins or  
looses the war, then you have an infinite game that ebbs and flows. So  
as a player, the adventure is created by engaging in the events going  
on around the player. You are thrown straight in to the story, and  
what you do will influence the story to come.

The engine has a very novel terrain technology that makes it possible  
to modify the ground simply by pulling it, but rather then just a  
height field you can also build cliffs, dungeons, houses and bridges  
with it. It truly Is a sand box engine, but it is not designed to be a  
sand box game for the players, but rather for the game itself. The  
game can change the environment to constantly give players new things  
to do. The goal is to make players so engaged in the progression of  
events that they wont have time to monkey around. By using procedural  
technology the game can generate dungeons, strong holds, turn cities  
in to ruins, and create endless things for the player to explore.  
Since we can generate as much content as we want, players don't have  
to ever play the same thing twice. Old content will turn in to ruins  
and give the world a sense of history. You may play one day and then  
check in 2 days later only to find that the city you started building  
has grown, a new dam has been built and that the enemy tower is in  
ruins. It is almost like a soap opera that constantly moves forward.

Love is a networked game and you have around 400 players registered to  
each server and maybe around 100 players online at any time. This  
small scale game means that you can get to know most people and make a  
real impact. Say you have a stronghold of the AI enemy that keeps  
sending out enemies and it takes two full days until a player finally  
manages to destroy it (possibly by a series of tasks like finding weak  
spots, gathering right tools, keys and weapons), then the ruins of the  
strong hold will be named after the player and other players who come  
to this place will recognize the significance of the player. In a  
sense we are building characters and story, not threw skill points and  
written story but threw real game play.

The game play will actually be very simple and almost old School  
comprising of first person action and adventuring. All players are on  
the same side and play against AI enemies. Although one idea is lets  
some players act as "Dungeon keepers". One of the key goals is to get  
away form statistics menus and character building. The players should  
rather focus on the environment as that is something that can be  
enjoyed by other players too. While most MMO games are fairly Casual  
in their play style they are very hardcore in the amount of time they  
take to play. Love looks to reverse this and make it more skill based,  
but something that can be enjoyed in short 15 minutes breaks for  
people (like me, I'm sure you have noticed) who don't have a lot of  
time to invest. In the game you don't hut for items, but rather makers  
of items, once a player have gotten the Build-BFG-machine and placed  
in a city, all players in the city can get a BFG. This communal  
sharing is also strengthens the collaborative side of the game and  
means that players who only play occasionally have access to the same  
things as the people who invest more time. Also the player who can  
bring in the cool stuff becomes more significant story wise.

The goal is not just to explore new game play, but also make some  
radical changes to how games are developed. In fact this entire  
project is kind of ridiculous in terms of development: I'm trying to  
develop a AAA game (Who knows what AAA, means but its a FPS with  
impressive graphics, and that usually does it), with no real staff,  
in-house engine, with a full in-house pipeline (Modeling is done in  
Loq Airou) and with a quicker turnaround compared to other games. So  
how is this done? Most things are procedurally generated, but they are  
augmented by hand edited details. These hand edited details are  
automatically added to the the environment. So by modeling a tree and  
giving it to the engine the engine can take that tree, randomize it  
and merge it in to the environment to build entire forests form a  
single tree mesh.

The current state of the development, is that many of the basic  
systems are completed, like networking, user database, terrain  
generation, terrain management, collision detection, animated time of  
day, wind simulation, cloth simulation, Physics simulation, ray  
casting and much more. Currently I'm working on the character  
generation and animation system. It is fully procedural system car can  
create all kinds of characters and creatures and animate them. Some of  
the advantages of using a procedural IK character system is that  
characters can interact better with the environment, like placing a  
foot on terrain, grabbing things and transition between many different  
movements. This system is roughly 60% complete, The goal is to begin a  
closed on-line beta test late this year.

If you have questions about the game that is verse related, you may  
ask them here, but please if you just want to ask something about the  
game, try to keep the non verse related stuff off the list, Thanks

Cheers

E