a prototype of the seminars Game Design and Game Programming in the third semester of Master’s studies at the University of Applied Sciences Erfurt (FHE)
Motivation and Overview
In the winter semester 2004/2005, the seminars Gamedesign were held by Mrs. Professor Ulrike Spierling and Gameprogrammierung by Dr. Paul Grimm at the FHE. The goals were to develop a Gamedesign with following paper prototypes and a prototypic implementation of this Design in an almost freely choosable Game engine. The only given constraint was a reference to Erfurt, optionally an Edutainment character.
For “Der Waidjunker”, the chosen genre would be an economy-simulation. More information about this can be found in the section “Gameidea”. As technical platform was the Multiplatform Game engine Crystalspace selected, as seen in the section “Game-Engine”. For the Contentproduction, considerably Gimp and Blender were used. The project was developed on a Linuxsystem first and finally ported to Windows, whereby the chosen Game engine was of large advantage. The Download-Section refers to used software and the Game itself.
In the Setting of Erfurt in the 16th Century, the player tries to gain wealth and prestige by trading with Erfurt-typical goods such as waid, to attain finally the status of the Waidjunker, the probably most outstanding and richest townsman of Erfurt. Therefor, the player has the following activities at the disposal:
- Buy raw materials of the city or driving dealers
- Trade goods with higher order with driving dealers
- Rent or establich manufacturing plants for improving of raw materials
- Invest in the city to gain reputation
- Establish trading-routes to secure income
- arrange supplies by order of other traders in order to win much profit and reputation
The Edutainment character is generated by a realistic representation of Erfurt in 16th Century and by a pronounced informations system. The player himself can at any time get informations about buildings in Erfurt and inform him about the background to the merchandise and their production. The system should be completely integrated into the play, whereby it is open the user at any time whether he wants information or not.
The game is divided into temporary limited rounds, whereby 4 rounds result in one year – thus equating with the seasons. The game is limited to 160 rounds at maximum, thus 40 years.
Computer and video games attain increasingly meaning, which reflects itself also at the offer of Gameengines at the market. Despite the selection from some commercial Engines, the decision was made for Crystalspace (CS) – a freely available, initiated by Jorrit Tyberghein, open source engine, which is licensed throught the LGPL. The license permitted also commercial shipping of games developed through CS without having to reveal the sources of it (which was required by the GPL).
CS is developed very modular, it provides Plugins for almost each opportunity: Particle system, animated models, support for pixel and Vertex Shader, a multiplicity of image-formats, virtual file system (for better platform-independance), even a complete window-based user interface.
However, CS is missing a tool that helps with the production of contents and logic itself – the probably largest disadvantage of the system. It has tather to be seen as a Framework, which however reached enormous extent and is surprisingly well documented. Nevertheless, by the Framework character one is dependent on the use of other tools for contentcreation, while the game-logic has to be implemented completely in C++ or a Scripting Language.
Despite the mentioned, good documentation, the concept of CS requires a relatively long training period. There is both an API documentation of the CS interfaces as well as manuals with tutorials of the use of the engine, which however leaves open questions at important topics. The Mailinglist of the project was often by large assistance, where one can interact with the developers and users as well as taking its extensive archives for reference – with approximately 20-50 mails per day, a substantial library was already formed there.
Due to its roots in the *nix environment, CS sets on a close co-operation with established, already existing applications, instead of trying to recreate a complex somewhat own, whose development time would not correspond to its increase in value. Therefore CS possesses e.g. interfaces to Blender, Worldcraft or id’s Radiant in order to generate level data, the free Toolkit cal3d, md2/3 or Maya for the production of Models as well as a multiplicity of available picture and audio formats.
For “The Waidjunker”, only Blender was used for either Level- and Model-Data. It possesses export possibilities for the free cal3d-Toolkit, a skeletal animation system, with which animated models in the prototype are represented. Blender is a powerful, free tool to provide and manipulate 3D-models, from simple polygone over complex buildings up to their exact textureing. It requires a longer training period due to its somewhat hardheaded usability, even if the user established paradigms of use for other modellers such as 3DMax or Maya. For the conversion of the level to CS, the Tool “blend2cs” was used, whereas blender-versions higher than 2.35 can be used with built-in cal3d-Export. Otherwise the necessary Python script must be downloaded.
In order to become the requirement at authenticity fairly, a map Erfurt around 1500 and about 50 photos of material still existing building fronts were taken.
On December 14th 2004, the prototype was pr während deesented with the opening of the Virtual Reality laboratory to the University of Applied Sciences Erfurt (FHE) before public. The avowed goals of the prototype were:
- Test of the play concept on playability
- Mediate the game-feeling
- Representation of fundamental play concepts and how these interlink
- Training into a 3D/Game-Engine
- Tuning of the gamedesign
The development period of the first idea up to the prototype was limited, particularly with approx. 8 weeks parallel to further studies, quite scarce in view of the necessary and something underestimated training period. Not all play concepts could be converted, yet the product for the authors is quite satisfying. With the experiences by the prototype development it has to be assumed that another 6 months of two-head-development is needed until the game would reach production-state.
For an executable Version, please contact the authors Tobias Tost or Clemens Sutor.
Idea: Tobias Tost
Game-Design: Tobias Tost, Clemens Sutor
GUI-Design: Clemens Sutor
Modelling, Logics: Tobias Tost