Hoi, We have always planned the implementation of the GEMET thesaurus as one of the steps in the implementation of the Ultimate Wiktionary. In some discussions we had, we considered that it might be a better idea to implement GEMET in a subset of the Ultimate Wiktionary in stead of in its own datadesign. Once it has been implemented, we will be able to learn many of the lessons we need to learn as reality shatters many false dreams. With this implementation, we implement several of the key functionalities of the UW. It will demonstrate the "eat your own dogfood" idea; the words for a Languages and for a WordRelation need to exist in order to implement their functionality and their localised value. It also demonstrates how a thesaures can be implemented in UW, it will show off how we can have relations between different meanings. The one thing I however like best is, that it demonstrates the core functionality of Ultimate Wiktionary much better. The data design for this is much easier to understand. The absolute core functionality however is without the Collection, the CollectionLanguage and the CollectionMeaning tables. These are needed for the implementation of GEMET. As it is a subset, it does not have many of the features that will exist in the full blown version of the UW. These can however be added one at a time. This allows for more frequent updates and this will propably lead to much more excitement as we will have more often new features to show. More info can be found here: http://meta.wikimedia.org/wiki/Ultimate_Wiktionary_data_design Thanks, GerardM