Pages

Sunday, June 19, 2016

The Blind Men and the FamilySearch Family Tree

File:Blind men and elephant2.jpg



My reference to the Blind Men and the Elephant is transparently intended. In the case of the FamilySearch.org Family Tree the real question is who are the Blind Men? The users? The developers? A little of both? I have been reading comments from various people who profess to "know" what is going on. There are several problems with these opinions. None of them account for all of the data. Put another way, none of the supposed analyses account for what is observed as happening in the Family Tree.

So what is happening?

It is apparent from my direct observation and the observation of many others that a huge number of "newly discoverable" entries are being either added to the Family Tree or are now discoverable by the search engine. One explanation by the Ancestry Insider in a post entitled, "Barrage of Records" Causing Problems" dated June 15, 2016 addresses some of the observed issues. I could ask a number of questions that are suggested by the AI's commentary, but I see no reason to raise the issues those questions would suggest.

The bottom line of the situation that exists in the Family Tree presently is that there is a defect in the ability of users to merge obvious duplicates. Further, the search engines involved with the program are unable to find and display all of the duplicates. I have shown how this is discoverable several times in past posts, but I will do it again here.

MyHeritage.com has access to all of the entries in the FamilySearch.org Family Tree. One of my ancestors, Nathaniel Potter shows 284 Record Matches. All of these, except for the first 6 are records found in the FamilySearch.org Family Tree. This means there are 278 duplicate entries in the Family Tree discoverable by MyHeritage.com. In paging through these duplicates it is evident that they relate to the same person. Here is a screenshot from MyHeritage.com.



Here is one of the records on the FamilySearch.org Family Tree.



Here is what I find when I look for duplicates for this randomly selected record:


So again, obviously the search engines are not finding all the duplicates. This situation has existed for months (if not years) and is not a "new" development. What is a new development is that many of these "duplicates" are now discoverable and can be merged. Some still cannot be merged. But in some cases, we have been able to eliminate all of the duplicates for a specific person. 

Now, I cannot believe that these hundreds of duplicates came from the recent move by the Membership Department. The duplicates came from multiple submissions of the same people for over a hundred years of accumulated submissions. They have been in the data all along. The imperfections in the program cannot yet deal with either the number of duplicates or their specific origin. It is obvious that some of the issues do come from the "Membership Records" whether or not those records actually represent people who were or are members of the Church. 

The net effect on the user is an inability to actually work with the program and make any progress on some of the lines usually referred to as containing IOUSs. Another issue is the fact that many of the duplicates are converted into duplicate ordinance requests. 

Now, the bogey man here is the new.FamilySearch.org involvement with the Family Tree. I am sure that the explanation of what is happening is much more technically involved than my generalized speculations. But what I can do is "reconstruct" the cause from the effect. In my favor, I may be one of the Blind Men, but I have a very powerful computer attached to the Internet and many, many years of experience in deciphering the internal workings of computer programs and detecting bugs. In this case the Family Tree elephant is my "black box." I try to work out how the program works or does not work from the observable product.  

My best guess today is that intentionally or not, the fact that we can now find more of the huge pile of duplicate entries in the database indicates that FamilySearch is making progress in "fixing" the program. We are steadily seeing the functionality of the program increasing and the area closed off by the "construction" grow smaller and smaller. I do not claim to have any "inside" knowledge of insight into the time table involved. My guess is that there is still a significant amount of programming to be done, but I can also foresee that many of the issues now preventing merges are being eliminated. 

As always, the Family Tree program is the solution, not the problem. If you think the program is a problem, then you are the problem. Get to work and do what you can do with the program and stop worrying about the future changes. Let me worry about them. I will do this anyway. 

No comments:

Post a Comment