A part of my family tree is maintained by another person, who delivers record updates through a Gedcom that covers that entire subtree (he uses another genealogy program). How do I avoid that updated and unchanged records of that subtree appear twice in my assembled FTB tree when I import that Gedcom? It seems Gedcom import just adds records and does not replace records for existing people. Should I first manually remove such people from FTB before the import?
If I then publicise the new data, will that restart the smart matches mechanism for these persons over again? I can imagine the administrators from other sites may be upset when they are asked to review the information for such persons repeatedly.
The method you propose is impracticle in my case: the other person maintains his family tree with another tool (good old PAF). Regularly I receive a Gedcom file which I merge as a subtree in my tree which I then publicise. As an example, last time I received a Gedcom file with 3 additional persons, 14 changed records (small additions or changes) and 350 unchanged records. You see, if I merge everything and then manually remove the duplicates, this becomes a nightmare.
Would the following work: I maintain 2 FTB projects: one in which I maintain my own dataonly (FTB1). The other one(FTB2) is completely emptied each time I want to publicise. I the import in FTB2 the content of FTB1 and the newly received Gedcom file. I then publicise the merged project.
I suppose this should work, but I fear that the Smart Matches tool on your server will consider each record as a new one (it was never flagged as a "duplicate") and would recompare my data with the other family trees and signal all matches again. And this every time I do a publish. No doubt I would get angry reactions.
This all becomes very technical. How does your server identify each person? If it is e.g. by name birth date/place, it should not be a problem. If it is by some internal reference number,it (thinks it) sees a new person.
However, the point is that the subtree (including changes, additons and unchanged records) will still regularly be delivered in gedcom format. I cannot convince the other editor to switch to yet another editing environment. So I would use the two-FTB approach, described earlier, resulting in a re-publish of the entire tree every one or two months.
As explained before, I am very worried that the Smart Matches mechanism might swamp other family tree publishers with "new" (but in reality identical) matches every time I re-publish the tree. I suppose every person would possibly be considered as a new one each time. Could you elaborate on that, please?
You can have different Smart Matches settings for each of your online trees - if you follow the instructions I have sent you, you can prevent receiving the same matches for your second tree by disabling Smart Matches for the second tree alone.
If you enable Smart Matching for both trees, other site managers will indeed receive matches for both of the trees, but you will not receive matches to other trees you manage.
Seems we don't understand each other. As described earlier, I want to merge two trees (my own subtree as FTB1 import of Gedcom into a prevously cleanedFTB2) on my local machine and publish online the merged result, hence only one tree is published at any time.
My question is: will Smart Matches send around new match reports to other family tree administrators for all persons in the published tree (the published version of FTB2) or only for the newly added persons (compared to the previously published version - the changes may originate from either FTB1 or from the Gedcom input). This all depends on how person lists are maintained within your server and is a technical design matter.
Obviously in a family tree there are many people with the same name. A person added to the tree should trigger Smart Matches, even if he/she has an already used name. How does the tool differentiate? If done by e.g. birth date, that would solve the issue. I hope it is not by RIN number, since that one is reallocated each time for every person by FTB in my scenario.
I hope this thread was also useful for other administrators, since many work in mixed environments.
Please note, that the Smart Matching engine comdines all the entered inforamtion about each individual in your tree project (name, birth dates, relatives) in order to search for suitable matches for that individual.