Dear Ancestry Insider,
Could you report on the recent dumping of duplicate records into the FamilySearch Family Tree? Where did they come from? Why can't we merge them? Why does FamilySearch continue to dump duplicate data into a tree while at the same time begging us to stop doing the same?
In case you're wondering what I'm talking about, see https://getsatisfaction.com/familysearch/topics/barrage-of-records-attributed-to-lds-membership-creating-problematic-situations [titled “Barrage of Records Attributed to LDS Membership Creating Problematic Situations”] and other issues linked in the comments.
Some say the data came from Ancestral File or other data sets FamilySearch has (and that they might dump even more duplicate data). I don't understand why they would keep dumping that into nFS. I thought nFS was pre-populated with that data from the beginning?
Please enlighten us.
Signed,
Justin YorkPS. It would also be nice to know when nFS is really going to die. It's been slated to be disconnected "in the next 6-12 months" for years now.
Dear Justin,
I’ve searched around and can’t find any public comment by FamilySearch about the events you’re referring to. But I think I can answer your question with information FamilySearch has previously made public.
First, it helps to understand some history.
The New FamilySearch tree (NFS) had a unique, advanced architecture. The architecture allowed you to extract information about a person from an historical record. That extract was called an inner person. The architecture allowed these historical extracts to be combined into what was called an outer person. Brilliant architecture: one inner person per historical record extract, one outer person per historical person. Inner persons could be compared and contrasted as required in the genealogical proof standard before reaching a conclusion. It was both a research tree and a conclusion tree.
Special inner persons were created and linked to records in a database kept by the Membership Department of The Church of Jesus Christ of Latter-Day Saints. (These database records are not membership records, although they are commonly called that. But I digress…) When changes are made to that database, the changes are reflected in the inner person. It is as though you are looking at a church record, glance away, and the writing has changed when you look back. The effects of this unusual source behavior are mitigated by the separation of inner and outer persons.
While the NFS architecture was brilliant, the implementation was not. FamilySearch didn’t distinguish between inner and outer persons in the user interface. They created inner persons from low-quality records like Ancestral File and Pedigree Resource Files. They didn’t require sources in an architecture that required sources (how do you create an inner person when there is no source to extract?), so they had to blur the distinction between inner and outer persons. Everything went downhill from there. The implementation of NFS proved unstable.
FamilySearch decided to abandon it in favor of a more traditional conclusion tree, FamilySearch Family Tree. There are no inner persons. Sources are handled in the usual way as attachments.
Ron Tanner talks about replacing an airplane’s engine while inflight. The analogy should be extended. FamilySearch has rebuilt the entire aircraft inflight, with the final step being the replacement of the engine. The problem is that the new aircraft is designed for the new engine. An elaborate set of springs, pulleys, cables, gears, bailing wire, and a lot of duct tape are required to synchronize the old engine with the new aircraft. The synchronization has unfortunate consequences. Luggage is sporadically soiled, crushed, torn open, mashed together, destroyed, or sucked out of holes in the fuselage.
In most of the cases I’m reading about, persons created by “LDS Church Membership” cannot be merged into other duplicates. That would delete the ones created by LDS Church Membership. However, those duplicates can be merged into the LDS Church Membership persons. If they can’t be merged either way, see “Cannot Merge Duplicate Records in Family Tree” in the FamilySearch help system.
I think even the FamilySearch software engineers would like to know when NFS will be shut down. Software engineering time estimates are famously wrong. The last official pronouncement I remember seeing said the shutdown will occur “in 2016.” New people keep appearing on the plane and the old engine is barely keeping it in the air. Each Sunday the plane is clipping the trees. Family Tree will either switch over to the new engine soon, or it will crash and burn.
Airplane image copyright 2011, Joe Loong. Used under license.
I too am frustrated with all my names that are showing up as duplicates. I talked to family search and they have no answer either, just membership records. It has taken me hours to merge them, several of them, multiple times.
ReplyDeletehttps://billsfamilyhistorycenter.blogspot.com/2016/06/an-easier-way-to-find-duplicates.html Best explanation I've seen (and I'm a support missionary also)
ReplyDeleteRon Tanner made the following comment on Get Satisfaction a couple of weeks ago:
ReplyDeleteRon Tanner (Family Tree Guy) 12 days ago
We have responded to several other posts. This is an artifact of us still synchronizing with NewFamilySearch, a predecessor of Family Tree. The membership department has a copy of Ancestral File records and for their own purposes they decided to make these visible in their systems. Unlike Family Tree, NewFamilySearch is significantly sensitive to membership changes. When these AF records were exposed they were created in NFS and then synchronized over to Family Tree. When we are complete in our transition from NFS than synchronization will stop and this will not happen as we expect no contributions other than a potential source from membership databases.
https://getsatisfaction.com/familysearch/topics/familysearch-managers-are-treating-patrons-with-disdain
The best explanation I've seen was on Bill Buchanan's blog:
An Easier Way to Find Duplicates?
You may be aware that in May, Membership Department reconnected to the Family Tree, after being disconnected since August 2014. As a result 3 different types of records from membership were brought into FT.
1. Records of members, that were previously hidden because no death was recorded. Yea!!!! We have been waiting for this to happen!
2. Records of non-member family members who are deceased. As I understand it, these are chiefly created to allow non-member family members to be shown in ward and stake directories. Okay, but confusing! These incomplete records need to be the surviving records in the case of a merge.
3. Ancestral File records, of which Membership Department had a copy. This happened by accident. Ouch!!! These records can mess up family lines for generations.
http://billsfamilyhistorycenter.blogspot.com/2016/06/an-easier-way-to-find-duplicates.html
Ron Tanner posted this on Get Satisfaction today:
ReplyDeleteRon Tanner (Family Tree Guy) 12 hours ago
There was a bug that did not show the life sketch refreshed but no data is being lost. No ordinances are being lost. We are looking to see if we can remove them without adverse issues. I would wait until June 27th to see if they are gone, if not, then merge.
https://getsatisfaction.com/familysearch/topics/officially-how-should-i-tell-my-stake-to-handle-the-new-family-tree-duplicates
It seems to me this is a pretty significant issue and that there was definitely a failure to communicate on FamilySearch's part. Just for the names I had a watch on there were dozens of duplicate membership records dumped in starting in early May. In some cases the duplicate records could be merged, but only if the membership record was the surviving record. In other cases, the duplicates could not be merged at all. I opened a support case and was simply told to "be patient" and when the interface with nFS was taken down I would be able to merge the duplicates.
There was absolutely no response from FS to this issue until a month or so after the problem occurred and then only after after dozens of posts appeared on Get Satisfaction. There was no guidance from FS on what users were supposed to do about the duplicates until today!
It really seems like FS needs to do a much better job of communicating with it's end users when issues like this arise.
At the beginning of June I started to see all of these duplicates, especially in my four generation ancestors. Most of these recent additions needed temple ordinances -- even though the same people on my current tree had their ordinances recorded and sources already there. With no direction from FS, I thought I needed to merge the duplicates. The comment that the duplicates newly introduced could not be merged into my current tree but the positions could be switched is accurate. So far most of the duplications have resulted in no problems.
ReplyDeleteHowever in one case the ancestor suddenly had a note stating he was alive or the details were "confidential" have frustrated me. Now his temple ordinance dates are hidden from view. This ancestor was born in 1864 in Sweden. He died in 1941. He was just a poor farmer in Minnesota, definitely no one whose information needed to be confidential. (I spent over 25 years trying to find his birth/christening record and determine his other family members so the disappearing ordinances mean a lot to me.)
I wrote FS immediately about the situation but no response as of yet. Why aren't these issues discussed by FS on their blog, Facebook page, mass emails, or some other way?
Communication is encouraged between researchers so why doesn't FS follow their own guidelines?
Over the past couple of years, I have tried to keep up with all the "changes". It is amazing how many hours can be spent in updating records and sources (sources already listed in my tree but the records were newly linked to the individuals details page as suggested through FS and indexing transcriptions, etc.); moving ordinances for individuals and couples to temple file (to keep personal family files smaller in accordance with FS requests); dealing with the release and resubmission of pre-2010 temple files which couldn't be done due to FS glitches; re-labeling photos for the Gallery so they sort properly by surnames (also a FS request), etc.
Users of Family Tree need to be kept informed when such matters as these duplications are dumped into the system. The time and efforts of the users should be respected. The more we have to "clean up" the less research we can do. This also affects indexing. I have basically stopped indexing in order to straighten out all of these other challenges.
I am certain that other researchers are in the same boat as I am. Time for the captains to pay attention to the rest of us and not assume everything is alright when FS users are seeing water leaking into our boats!