All merge information should be selectable and changeable.
All types of merges within the merge SHOULD be optional (i.e., individual children, wives, parents, dates, places, sources). I realize that for some people this may make them more hesitant to merge, but that is NOT a bad thing. Too many bad merges are made already, and making block changes to people's information rather than specific changes is like an ax opposed to a scalpel.
Comments
-
I agree. When you cannot choose what should be kept, then the resulting merge can be just as bad as what others are trying to say are bad merges under the current system because it can include dates, places, and relationships that do not apply to the person but were forces over into the merged profile.
0 -
If they don't apply to the person, then why are the people people being merged in the first place?
0 -
The merges happen because there are so many BAD and duplicative profiles created with false, fantasied dates and relationships with no sources to support anything given in the profile. And yet, that profile will be connected to a well-sourced, well-documented parent or spouse profile with a fully realized family structure and the duplicative profile will already be present in the family, and that profile will have sources to support its dates and relationships.
1 -
Or, to be honest, with well-sourced data, where the name, age/birth year, and parents match, and someone thought it fitted, but where you don't consider it very likely that it is the same person if you have done sufficient family-related research on a person to give a feel for where they were likely to be from.
I mean, because it's in error doesn't have to mean it's carelessness. Attribution is often very difficult.
2 -
On the question, why are the people with different data being merged in the first place — There are thousands of examples that could be given. Here is one: Hannah Smalley LKWB-HMN from Plymouth, MA, daughter of John Smalley, married John Bangs in 1660/61. There are primary sources for that marriage. her husband wrote his will in 1702/3 naming Hannah as executrix, and that will was proved in 1708. Thus, Hannah and John Bangs were married continuously between 1660/1 and 1708.
However, on Find-a-Grave, it is claimed that Samuel Blackford LKR3-KH6 of New Jersey married Hannah Smalley from Plymouth, MA, daughter of John Smalley. Samuel lived in New Jersey which was quite a distance from Plymouth, and he was having children with his wife named Anne between 1678 and 1691. His wife could NOT have been the Hannah Smalley from Plymouth who married John Bangs. Yet, over the years, many people add a new Hannah or a new Samuel and tie the two families together over and over, simply based on this bad Find-a-Grave. And those added profiles NEVER come with sources to support anything within the profile. Those duplicate profiles will have inconsistent dates of birth/death/marriage and various fantasied family members, none of whom have any sources to support their existence. All because the contributor found something on My Heritage or Ancestry or Find-a-Grave or some similar unsupported reference.
Many of us work hard to keep these duplicates and false, mostly empty profiles from clogging up the family profiles and misleading new contributors. So, yes, there are a lot of merges that must be done wherein the duplicate profile has incorrect dates or names or relationships. It is putting an extra – and rather heavy – burden on the volunteers who do this work if we cannot choose the valid information that should be shown after the merge.
2 -
While I can see that it could be easier to be able to have the freedom to pick and choose what to take in a merge, it can also make some huge messes.
It's also not technically feasible to choose some information separately, like children.
Merge is not intended to be a cleanup process.0 -
The statement above from Mary Anna Ebert is one of the core issues with merges. Merging is not part of the cleanup process. Proper methodology requires that any inconsistencies be resolved before you consider a merge. The real issue is that with the old existing system, there was no suggestion that the merge may not be correct and people would automatically do the merge even if it was obvious that the two individuals were not the same. People took the suggestion of a needed merge as a commandment to do the merge.
2 -
I agree that we should have warnings that the merge may not be correct. But those can be added without changing the whole merge process.
1
