The International Computer Game Collection (ICS) is a project of the Computer Games Museum Berlin, the Entertainment Software Self-Regulation, the Computer Games Collections of the Centre for Computer Games Research at the University of Potsdam (DIGAREC), the game-Association and the Stiftung Digitale Spielekultur. The goal is to establish an institution for collecting, documenting and preserving the phenomenon of digital Gaming. The total amount of the united collections is about 58,000 games and still growing. Recently the website of the ICS was published (www.interntionale-computerspielesammlung.de). The visitor could search through the data from different collections, which are already ingested. Currently approx. 40,000 are games recorded.
This first phase of the project was financed by the German Federal Ministry of Transport and Digital Infrastructure. It includes the amalgamation of the recorded data in one database system. The open source framework Collective Access with the Pawtucket frontend was used for this purpose. In the following phase, the data will be consolidated and enriched to improve the general quality of data. This necessary but time-consuming process must be carried out in the second phase of the project, which also includes the physical union of the collections themselves in one storage location.
The ICS intends to serve research and documentation purposes. The stablished information system is also used as a collection management. It includes several different modules for internal processes such as cataloguing, loans management, conservation and long-term archiving needs. For legal reasons, however, these modules are not publicly accessible. Both the development of the system and the input and control of data will be completed once the collections have physically brought together in one location.
The basis of the management system was developed at the Computer Games Museum about 10 years ago. At that time no suitable database system was available, which could be used for managing a collection of computer games, particularly regarding authority files and aspects of long-term preservation. It stands to reason to develop an appropriate concept and to establish Collective Access as database framework. This concept was refined, changed and expanded in the following years.
For practical reasons, the metadata describing Games are divided into three aspects: Work, Version and Object. Since there are numerous compilations of games in the collection (e. g. “Gold Games”, “Stronghold Compilation” etc.), the system is structured in such a way that the game theme, its mechanics and other content related aspects could be recorded in the “Work” division, the platform with the system requirements and developer information can be found in the “Version” division. The physical information (sales packaging, guides, manuals, data carrier), the origin and the publisher are recorded in the “Object” division. This approach offers enough flexibility to capture objects with several games without repeating any data.
The similarities of this concept with FRBR (Functional Requirements for Bibliographic Records) is evident. The most obvious difference to FBFR is, that in our concept “manifestation” and “item” are brought together in the “Object” division. Without going into further similarities and differences at this point, it should be noted that this structure is convenient for further amendment of information on the game, its themes, tropes, settings and game mechanics. New metadata fields could be integrated in the “Work” division (i. e. locations) and doesn’t have to be recorded for every item manually.
Due to technical limitations references to web resources (Linked Open Data, Authority files) have been a problem at the beginning. But the new version of Collective Access offers a very flexible way to use these sources such as Wikidata, Wikipedia, Gemeinsame Normdatei (GND), Art & Architecture Thesaurus and Thesaurus of Geographic Names (both Getty Research Institute), Virtual International Authority File (VIAF) and ICONCLASS (still in experimental stage). Involving these web services should improve clarity and enrich the information structure. Titles in different language are imported on-the-fly from Wikidata, while for additional information on companies and individuals are imported from Wikipedia, if available.
At least, I would like to address a problem that has not yet been finally resolved. When defining the “Works” division, it was assumed that all relevant information on game content and game mechanics should be registered here. It turned out, however, that this split-up not always a constructive way to record game characteristics. Let’s take “Mario Golf” as an example. It is available for various platforms (Nintendo 64, Game Boy Colour, Wii etc.). In this case the one Work entry approach is not very useful, because the gameplay varies in the different systems. This resulted in the necessity to create different Work entries (e. g. Mario Golf Version GBC, Mario Golf Wii). Being aware of this, it turns out that these works are linked to each other via a defined relationship. It would make more sense to define a “Series” division, which is not yet intended according to the current conception. Wikidata (https://www.wikidata.org/wiki/Q1898865) and mobygames (https://www.mobygames.com/game-group/mario-golf-series) use a similar kind of concept.
The publication of the database on the website of the ICS at this time was intended primarily to inform the public about the scope of the collection, even if the mentioned deficiencies are present in the documentation. This information system doesn´t intend to replace existing game databases but it intends to offer additional kinds of information for game research. We will go on, improve the data and expand the use of web services to make this collection more efficient and enriched.
Stiftung Digitale Spielekultur