Saturday, March 29, 2014

FamilySearch is Corrupting Sourced Entries; Needs to Stop Immediately (updated)

Note, March 31: I just saw a response from FamilySearch that says this was all a misunderstanding of what was actually going on. That may be the case, but FamilySearch should not disregard some underlying issues here:

  • Why did the sources disappear? Why did it show that George Jarvis only had Legacy NFS sources? Why weren't we able to see any other sources there or in his change history? What if this had been a permanent change and his old sources never came back? How could we tell?
  • FamilySearch is currently and systematically adding thousands or hundreds of thousands of mostly useless sources to Family Tree entries. This creates countless hours of useless busy work for people as they go through and delete the sources and glean a bit of information here and there that they can usually easily reproduce. (I don't think I've seen more than a small handful of Notes or Legacy Sources that contained useful information.)
  • It is very difficult to get authoritative information from FamilySearch. I just typed "legacy nfs source" into the FamilySearch Learning Center and did not get an answer that could tell me what is going on. Neither did the FamilySearch blog. (I've been following one months-old discussion on the blog. People keep asking questions. Usually other blog readers answer them. No one from FamilySearch is answering their questions.)
  • The workers in my local family history center are always running on a high level of stress due to the regular changes in Family Tree, most of which they only find out about when sitting down with a patron. I have had to reassure them numerous times that things will eventually work out and to just be patient, but it's an uphill battle. They feel very out of the loop and unable to provide support to members. (And no, I have not asked for this feedback; it's offered freely and regularly to any and all patrons.)
  • Yes, I understand the theory behind the program. I am a loyal user of Family Tree as well as a regular editor at Wikipedia. I believe in open access and user-controlled data. I know a few devoted Church members and highly trained and reputable genealogists who refuse to use the program out of principle because of the ability of anyone to change the data, and I've tried to talk them through their concerns, but in the end sometimes these few people have to be left to use their own systems. In other cases, my explanations have helped users understand how to make sourced changes and respect other users and their data and deal with conflict with others working on shared lines and understand what a wonderful tool Family Tree can be.
  • But the theory behind the program, and the actual experience of using it, are two different things. I regularly see a great deal of frustration among users. I have sat down many times to train people to use it and every now and then they have a seamless experience, but more often they get repeated technical glitches, something as simple as not being able to upload or label a picture, and so I have to bring out the constant refrain of, "Keep trying." "Come back another time." "Use a different browser." "Refresh the page." "Call or email FamilySearch for help." The interconnection of data and figuring out which buttons to push are not always intuitive even for a long-time user, let alone a new one.
  • Family Tree is much improved from the old NewFamilySearch system, but FamilySearch certainly can't claim that there are no underlying problems with Family Tree, either of its technical operation or of communication with the users of the program and the volunteers at the local FamilySearch centers, and they certainly can't claim to be surprised if and when these technical issues blow up in public.

* * *

As far as we can tell, FamilySearch is employing a battalion of volunteers to go in and corrupt sourced entries on Family Tree.

Named people are migrating data from the old NewFamilySearch by hand. Instead of a carefully and thoroughly sourced entry, these sources now look like this:

Basically, most or all of the sources have been stripped from an entry and replaced with nonsense.
For example, if I had added a census source and created a citation and copied the information out of the census about family members and the pertinent data contained in the census, now it looks something like this:

No data. Just nonsense.

In the case of George Jarvis (LWYL-M7G), his entry had been carefully and thoroughly sourced by myself, noted genealogy lecturer James Tanner, and many devoted members of the Jarvis family including Sharon Simnitt, Danelle Curtis, and family website manager Mark Jarvis.

A few of us have been trying to correct the mess, but there does not seem to be any way to restore the previous contents of the Sources and the entry is so corrupted that it could take months to get it back to where it was before FamilySearch started making corrupting the data.

This is outrageous. It is compromising any reputation FamilySearch had left with the serious genealogical community. It is compromising its integrity. It is compromising the trust I had that my work will be preserved, and if changes were made by other family members, we could negotiate and come to a reasonable conclusion.

This is not a case of what Ron Tanner at FamilySearch calls "my-tree-itis." This is clear cutting of the virgin forest.

Fix this problem immediately, FamilySearch.

Ed.—As noted in the comments, George's entry has been fixed (thanks, FamilySearch); I will add a picture showing what we saw before the fix happened since the first graphic above doesn't show that the sources ended.

Part of the systematic and ongoing problem for users of Family Tree is that the development team needs to work on it as people are using it, and when changes are happening, we're still using it and tend to freak out when years' worth of Sources disappear or other glitches happen. But, as said below, the theory behind the system is great, and, if working correctly, will be a great benefit to the Church and the field of genealogy.

Ed. (March 31)—Someone asked me yesterday what all this meant. She could tell I sounded irked, but didn't understand the technical details. I explained that many problems that show up while using Family Search are pertinent only to the individual line, such as family disagreements over genealogical details, problems with using merge, or the need to reclaim ordinance reservations made under a previous system. Those usually can be solved by a simple note to FamilySearch, where a friendly volunteer will help resolve the issue.

Other problems, like the one described in this post, seem to involve system-wide changes or institutional attempts to change a problem. This change seemed to be creating the potential for far-reaching corruption of data in the system, and needed to be brought to the attention of FamilySearch immediately. Perhaps they already knew about the problem. Perhaps they didn't. Perhaps it was actually limited to this one entry. Perhaps it was widespread. How could we tell? None of this was mentioned on the FamilySearch blog. (Which I do read consistently.)

So, there are a few ways of getting immediate attention for an issue like this, and writing a blog post like this (not that anyone at FamilySearch reads my blog) followed by my dad reprinting it on his blog (which they do) is one of them.

Ed. (March 31)—This post has provoked a passionate discussion over at Genealogy's Star. Sigh. One commenter notes that the migrated sources are being given the identity of the people who added the source to NewFamilySearch. That makes it look like the person actually came in and made the change on Family Tree, although they seem not to have done so.


  1. I just took a deep breath and went to look at John Tanner's entry (KWJ1-K2F) but that has not been touched. (Great sigh of relief.) Luckily FamilySearch has added the capability to print out the family records, so I will go through and create PDF files of everyone I have on my watch list. (126 people. This is going to take awhile.)

    I'd seen some of these Legacy NFS changes happen on other sourced entries, but hadn't noticed until my dad sent a note last night that they replaced real sources with the corrupt entries.

  2. I just checked again and George Jarvis's sources are now showing up. I don't have time to check any of the other entries that had migrated NFS sources. I'll get back to this later.

  3. ...and thank you for fixing that in George's entry, FamilySearch. Hopefully the root of the problem will be identified and fixed. I do very much agree with the theory behind the open sourced family tree and have found the opportunities for collaboration extremely useful, but problems like these tend to erode trust in the integrity of the system, and that is a shame.

  4. Yikes. I've mostly just been adding photos and documents for now. Hopefully they will be safe. Thanks for the notification, Amy. I'll keep my eye out on things!

  5. This is a good reminder to have things backed up!

  6. These things are really a pain. I usually just delete them because, as the writer said, you can't see any information in them anyway.

  7. Oh, man. It's been awhile since this all blew up. The major concern wasn't the legacy sources, although those were strange; the problem was that all the other sources were disappearing. It was evidently just a computer glitch, and I still see it sometimes on entries with a lot of sources.

    When you delete Legacy sources, make sure you check for content and note why you're making the change. I usually say something like, "This is a data fragment migrated over from newFamilySearch and contains no useful information."