In the comments of my rants about how the licensing model of business applications effectively prevents Sharepoint and other applications making best use of the data within the business applications.
Peter suggested taking the data out of CRM, into a separate database and then hooking up sharepoint to that database. Geoff suggests doing a bulk copy of the data into another repository
Unfortunately, my reading of the Microsoft licence explicitly covers that scenario under their term Multiplexing.
In the Microsoft document even if the user accesses the data through “Pooling Hardware or Software” they still need a CAL for the source system if they can “Input, Query or View data”.
They include the diagram below, which even though it refers to Project Server, the principle still applies. (sorry if the image is blurred)
The only way that an end user can see data from the core system without a CAL is if they are removed from the source application by email or hard copy.
Additionally this applies to any number of multiplexing machines that may exist between the end user and the source system and each multiplex system requires an additional CAL.
So my reading of it is that Sharepoint users, Data warehouse users, Scorecard users, Reporting Services Users etc all require CAL licenses for the source application. Is it me?!
The document that explains all this can be found here