One or two dimensions?

This post is about a typical question in dimensional modeling. The question is about how to cluster the attributes for analysis into dimension tables in your data ware house. Usually this is done within the ETL process of your BI system, so this might not be a typical question for the SSAS design. However, the question might come up in the context of your data source views and also in the context of usability (which makes it exactly a question of the SSAS design).

In the elementary school of BI we learn that our master data tables do not necessarily be the dimensions, that our transactional data tables do not necessarily be our fact tables but that the dimensional model is somewhat re-grouping all the information that is stored in columns of those tables to a flexible, yet easy to understand model of fact and dimension tables.