Question Virtual Family Tree

Hello
and thank you very much for a great software.

I have this question:
Would it be possible to provide a kind of virtual famliy tree with the following functionality?

Current situation
Imagine you have multiple (large) family trees in your installatiion. Maybe tree1 from your fathers family and tree2 from your mothers family.
All diagrams, lists and reports work well.

Virtual Family Tree
The idea is to create a virtual family tree now, which integrates the tree1, the tree2 and maybe more.
All diagrams, lists, reports and even plugins can work on the virtual family tree.
The _UID can be used to identify links between the trees.

What do you think?

Regards
marwel

Ok, let us assume that all the duplicated persons and all families and all duplicated sources, repositories, notes and shared locations are linked by a unique _UID. But what should happen if the stored information in those records is different? Which one is the valid one? Or should all the information be presented twice (ie two birth dates, two death dates, ...)? Very complicated.

Why not merge those trees? You have to merge the records once and decide which information should be kept or thrown away because it is duplicated. This is the way I recommend.

Thanks for the quick response.

The family trees in the virtual tree could have a priority. If there are different informations in records with the same _UID, the infos from the higher priority tree can be presented (maybe with a hint that there are inconsistencies).
The tree priority corresponds to the order of the trees in the virtual tree list.

By the way: with virtual trees there is no need to have much records doubled in the trees. Only some rare persons which are needed to link the single trees together will have entries in both trees. This makes the data maintenance easier for the editor.

Why not merge? This is easy to answer from my point of view: it's because of data privacy.
There is a perfect user access rights system in webtrees. From the data privacy point of view, a virtual tree could be handled like a real tree. Each user inherits his access rights from the included trees (or gets defined access rights for the virtual tree).

Let me make an example:
FathersTree has the members: child1, child2, user3
MothersTree has the members: child1, child2, user4, user5
VirtualTree has the members: child1, child2

This ensures the data privacy with virtual trees.
On the one hand some users can only see the FathersTree or the MothersTree
and on the other hand some users (e.g. the child1 and child2) have a perfect view on the complete tree
and on the "third" hand the Editor has to maintain less double data.

Sorry for writing so much, but I think it's required to explain some details.

Many years ago we had a discussion about linking together two different trees. I don’t think we will visit that idea again!

I agree at Ken. 

It is not only about a few duplicated persons. All sources, notes, Repositories, locations would be duplicated. All the access rights you mentioned would become very complex. 

Merging the trees is the solution. You can define that a user has only access to for example 3 steps, so a user of the father side can only see private information in father's side of the combined tree.