Changes made in FamilySearch

In order to preserve the promise of transparency and open communication, the change log records changes to topics made by both customers and employees. Anyone can dispute a change by clicking the "Dispute" link and entering additional details.


  • October 26, 2018 18:55
    Paul Wrightson, the poster:
    Reason: removed by the poster
  • October 24, 2018 21:57
    Jenn Marie, the poster:
    Reason: removed by the poster
  • October 24, 2018 19:49
    joe martel
    Reason: unkind reply
    Dispute this change
  • October 24, 2018 15:44
    Carol Anne Lethaby, the poster:
    Reason: removed by the poster
  • October 23, 2018 01:33
    joe martel
    Changed reply in topic of Deleting one's family tree from FamilySearch by joe martel to There are probably other approaches that maybe would be better. But there's always a trade-off: keep my tree intact vs. community contribution... In many ways FS FT follows the wikipedia model - any user can edit - open edit, and no one user owns the data. Other approaches include a curated model - where a user can post and another review, or another where a user suggest and another accepts/rejects that. To give some idea of scope and scale earlier this year:<br />- 4 millions persons are added each month<br />- 125 million transactions happen each hour<br />- about 1 million conclusion writes (add a birth, edit a relationship...) happen each day<br />An approval model would need to have users curate probably double that today.<br /><br />I'm always interested in new approaches that would slow bad data coming in and facilitate good changes..
  • October 23, 2018 01:31
    joe martel
    Changed reply in topic of Deleting one's family tree from FamilySearch by joe martel to There are probably other approaches that maybe would be better. But there's always a trade-off: keep my tree intact vs. community contribution... In many ways FS FT follows the wikipedia model - any user can edit - open edit, and no one user owns the data. Other approaches include a curated model - where a user can post and another review, or another where a user suggest and another accepts/rejects that. To give some idea of scope and scale earlier this year:<br />- 4 millions persons are added each month<br />- 125 million transactions happen each hour<br />- about 1 million conclusion writes happen each day<br />An approval model would need to have users curate probably double that today.<br /><br />I'm always interested in new approaches that would slow bad data coming in and facilitate good changes..
  • October 23, 2018 01:27
    joe martel
    Changed reply in topic of Deleting one's family tree from FamilySearch by joe martel to There are probably other approaches that maybe would be better. But there's always a trade-off: keep my tree intact vs. community contribution... In many ways FS FT follows the wikipedia model - any user can edit - open edit, and no one user owns the data. Other approaches include a curated model - where a user can post and another review, or another where a user suggest and another accepts/rejects that. To give some idea of scope and scale earlier this year:<br />- 4 millions persons are added each month<br />- 125 million transactions happen each hour.<br />An approval model would need to have users curate probably double that today.<br /><br />I'm always interested in new approaches that would slow bad data coming in and facilitate good changes..
  • October 21, 2018 16:04
    Paul Wrightson, the poster:
    Reason: removed by the poster
  • October 20, 2018 15:03
    joe martel
    Reason: request of poster
    Dispute this change
  • October 10, 2018 21:11
    Siv Collett, the poster:
    Reason: removed by the poster
  • October 09, 2018 15:51
    joe martel
    Changed reply in topic of How to Report a Problem with FamilySearch in this Forum by Robert Leroy Price to When trying to use Indexing, I get the following message on the screen; "An error has occurred that may affect your ability to use this page." How do I correct this problem?&nbsp;.
    Dispute this change
  • October 09, 2018 02:49
    A van Helsdingen, the poster:
    Reason: removed by the poster
  • October 05, 2018 17:46
    Randy Wilson - Employee, the poster:
    Reason: removed by the poster
  • October 05, 2018 12:27
    Paul Wrightson, the poster:
    Reason: removed by the poster
  • October 04, 2018 02:49
    joe martel
    Changed reply in topic of Can anyone explain why "Find by Name" is failing so badly? by joe martel to A thought experiment, using exact and no equivalent (John not equal to Johnn).&nbsp;<br />Attribute 1 is A1. There are say n attributes.<br /><br />Search for A1="joe"&nbsp; &nbsp;A2="123"&nbsp; &nbsp; A3="blue"<br />My results in scored order might be:<br />R1= Joe&nbsp; &nbsp;123&nbsp; green<br />R2= Joe&nbsp; &nbsp;0 green<br />R3= John 123 blue<br /><br />Why is R1 above R3? Maybe because certain attributes score more than other attributes. I'm guessing surname might be more weighted than say first name?<br /><br />New Search:&nbsp; A2="123"&nbsp; &nbsp;A3="blue"<br />Results:<br />R1= Joe 123 green<br />R2 =John 123 blue<br />R3= &lt;&gt;&nbsp; 123 green<br />The omission of an attribute may lower the score because it didn't result in a match of A1 and some value in the target is preferred. This could be way done the list, say thousands. As you add more attributes it becomes more complex.<br /><br />I know the algorithm is way more complex than this and there are cases tuning has to compromise recall versus precision. I agree its not always intuitive because the algorithm is a black box and to a human unpredictable. That is a human-machine interface problem because the human doesn't know how to tune their query to something that will result in seemingly better results.<br /><a href="https://en.wikipedia.org/wiki/Precision_and_recall" rel="nofollow" target="_blank" title="Link https//enwikipediaorg/wiki/Precision_and_recall">https://en.wikipedia.org/wiki/Precision_and_recall</a>.
  • October 03, 2018 16:36
    Tom Huber, the poster:
    Reason: removed by the poster
  • September 30, 2018 16:46
    Paul Wrightson, the poster:
    Reason: removed by the poster
  • September 29, 2018 22:37
    joe martel
    Changed reply in topic of Interface changes, and tool changes by joe martel to William, totally understand your experience which has been echoed by many users. The interface for FamilySearch has changed many times, (maybe 8 major version times - v8) and each one brought good and not so good. Some users like, some users don't. Most users adapt, or have to adapt. It is the "progress" of technology. I would love to hear specific issues that are troublesome.<br /><br />I'm not familiar with the thin agent. Are you talking about lite:&nbsp;<a href="https://www.familysearch.org/tree/lite" rel="nofollow" target="_blank" title="Link https//wwwfamilysearchorg/tree/lite">https://www.familysearch.org/tree/lite</a>? That product is specifically built for low internet bandwidth..
  • September 29, 2018 21:29
    A van Helsdingen, the poster:
    Reason: removed by the poster
  • September 27, 2018 20:14
    Tyler Peterson
    Reason: Topic expired.
next » « previous