Source handling during (distribution) aggregation
There are different possibilities to handle sources during aggregation of factual (distribution and structured descriptive) data. This was already discussed for E+M in #4366 (but unclear what the final result was) and in attachement Discussion-Transmission_Referenzen.txt of #2625 (no aggregation to higher taxon wanted, but explaining text required, especially if primary data exists also on higher level). The problem is that the source list becomes long on higher taxon level.
Explicit sources for computed descriptions: this solution has been selected in additivity for structured descriptive data aggregation. A taxon level computed description has been created for each taxon. This "computed" description got sources from all specimen level descriptions (and also literature taxon level descriptions) attached. These sources are NOT displayed in the dataportal taxon page but a link is provided to the detail page for this description. On this detail page each single description element of the description has a link to its original source (which can be primary sources or again links to CDM base descriptions). This could probably be done for distributions, too. It is a bit similar to the detail distribution page in E+M Berlin Model, where each single attributing primary distribution data is shown. The solution also scales because it links from higher to lower level information, so it avoids having long source lists on higher level information while it is still possible to find the primary source for each information easily.
Problem: model change needed to allow OriginalSourceBase to link to CDM descriptions (and maybe other CDM objects): #8673
Simply add all sources to aggregation
Add all sources to aggregation but try to deduplicate sources (reference matching, microreference concatenation)
Like before but exclude sources for distributions of status with lower priority as the aggregated distribution (Berlin Model solution, but result not used in BM data portal)