BUG REPAIR
LegacyUser
✭✭✭✭
Julie Buck said: Please forward this to whoever is in charge of programming. I'm finding a terrible amount of mess in many of the "people" that I'm working on. Example: John Henry Greenwood PID G9K2-WWP. You will note that there have been numerous merges done and this poor many now has 26, yes, that's 26 wives. This is only an example. I've found numerous others like this and have spent HOURS trying to straighten them out. I know how it happens: someone with little knowledge sees the "possible duplicates" message and merges, whether or not things match up. IF the "possible duplicates" vetted more carefully for matching data, it would certainly alleviate a little of the problem
0
Comments
-
Robert Wren said: Welcome back to the users' forum, Julie.
In the current 'open-edit' system in FSTree, anyone can change anything - and the correction responsibility falls on the one who discovers the error. As there is no training directly involved (such as the approved FamilyHistoryGuide https://www.thefhguide.com/) for new users, many users spend MORE TIME correcting than researching (at least I seem to.)
Sorry I'm not able not offer more help, but correct SOURCES help along with helpful messages to those creating (unknowingly?) the problem.0 -
Tom Huber said: One of the areas where we users can help is in communicating what we find with those who have made the changes through the internal message system. Things to keep in mind includes
a) anything much older than a few months is likely going to be past history for whomever made the changes. They likely will not remember what or why they did what they did.
b) many merges were made because the inexperienced users saw the suggestion and because FamilySearch provided it, then the user felt it necessary to merge the possible duplicate.
The new merge process still cannot stop those who do not take the time to make sure they are doing the right thing, but it goes a very long way in slowing down the merge process, which can result in some merges not happening.
As Robert correctly indicates, correct sources are often crucial in slowing down or stopping bad merges, but key to all of this is to communicate, communicate, communicate.
Any suggestions for stopping inexperienced users -- hm... I seem to remember that I was very inexperienced some 50 + years ago and in the same boat, had today's tools been available then -- you may have, short of violating the open-edit nature of today's tree, would certainly be welcome.0 -
Liz said: I do not think that there is any BUG in regard to John Henry Greenwood PID G9K2-WWP - he is an definite ongoing 'Work in Progress'.
Have you taken a look at the Change Log for him?
https://www.familysearch.org/tree/per...
Even today, working on him is going on.
Hopefully, that 'Work in Progress' will be cleaned-up.
Other Users are working on him.
You could communicate directly (Message) with the Users working on him to see what is going on, that may help allay you concerns - you may be able to help. That is what Collaberation in FamilySearch is all about.0 -
Julie Buck said: I don't think HE is a bug. I think that Programming could be a little more selective in the "possible duplicates" that they post on the screen. Too many people think that means they must do a merge, instead of looking first to see if it really is. Yes, I'm aware that other people are working on that one. I only used it as an example. I've seen dozens of others in the same messed up condition. I was hoping that a change could be made to alleviate the issue for the future.0
-
Julie Buck said: I've worked in data base management for years and I know that what we called "bugs' wasn't necessarily a Bug, but something that called for a little better control. For instance, I have looked at numerous "possible duplicate" messages where the principal had the same name, but lived in a completely different country, had a date of birth 20 years different, had a different wife and different children. An inexperienced user (and there are many - as I once was myself) will look at it and do a merge, albeit wrongly. IF Familysearch hadn't ever posted it on the Research ideas, it would not have been done incorrectly because it would have never been there to tempt the inexperienced user.
Should they, for instance, only post ones that at least had the same name and a date of birth within 2 years, and NOT post one that had a different wife or lived in a different location, it would save a lot of headaches.0 -
Colin Cameron said: Sometimes I have been very grateful for the very wide net cast in the 'possible duplicates'. More than once it has found someone who lied about their age, and/or emigrated, and/or remarried, and/or adopted the new wife's children, and/or ... .
I don't know if it's practical, but would it be possible to only present very close matches as suggestions but have a 'show more', or 'search wider' button which would then show the more random results to someone who is actively looking for suggestions?0 -
Julie Buck said: Colin Cameron. Now that is a great idea. Another thing I've thought of is to put them under the "possible duplicates" tab (where they show up anyway) and leave them off of the "research hints". Research hints makes them look like they are a MUST, like fixing the dates and places which are not standardized.0
-
Tom Huber said: On a person's profile, there is a section on the right that is Labeled Tools. In the list is "Find Similar People" which extends a search beyond that of the possible duplicates which, by the way, has been tightened and is fairly effective, except for problems in Europe.
No matter where the possible duplicates list shows up, there are those who move ahead and merge the profile with the existing one.
Short of removing the possible duplicates altogether and relying on two "tools" -- one to replace the removed possible duplicates and one to find Similar people -- it gets back to working with the inexperienced (which we all were at one time) and communicate with them why what they've done was a problem.
The inexperienced go beyond the merges and apply source hints because they show up on the list. They can cause problems with perfectly good dates and places by overwriting a details place with a "standard" place, and so on.
The key gets back to communication, recognizing that the user who made the changes is likely related to the person whose profile they've applied merges, sources, and overwritten user-entered dates because of the "pin" that appears next to a matching user-entry and standard.
In general, I've found using several of the following to be effective in communicating with users who have made inappropriate changes-- Thanks for their interest in making the person's record as accurate as possible.
Do I want to send them a clue-by-four through the internet? Of course, but that is not effective. This is one area where the adage "You attract more flies with honey" really applies.
-- The person or family involved and my relationship.
-- My thoughts and sources with respect to the changes they made.
-- The corrections I made to their incorrect changes and why I did it.
-- Request that before they make changes that they study the record, including the sources that are attached, any notes and stories that may be included in memories.
-- Remind them (if they have not provided a source or a reason) that sources are crucial to establishing conclusions and facts, and that a person's reasoning is needed to let others know what research and thinking was done to reach those conclusions.
-- What I did to correct what I perceived to be incorrect material.
-- Thank them in closing for their interest in making the record as complete as possible.0 -
Paul said: This problem causes me so much exasperation that I often wish the "possible duplicate" feature could be completely dropped.
True, I am not a member of the LDS church, so from a purely secular standpoint I see having lots of duplicates in Family Tree to be far less a worry than having "composite" individuals - i.e. several separate persons all rolled into one.
At the very least, the program must not be so "loose" as to allow impossible matches. Being offered a possible match relating to a person with a different surname and who had a family on the other side of the world during the exact same time period DOES happen. Periodically, the FamilySearch engineers request examples to help them work on improving the algorithm applied. Sadly, nothing ever seems to change.0 -
Julie Buck said: Amen Paul. Right now it's too loose. I've had one instance where the name was the same but the people were from England and had birth, marriage and death dates in England (at the old family home) and up comes a "Hint" . We found your relative in New Jersey. Different wife, different children, and all being born at the same time as the children in England. This was early 1700's. Yes they did travel from continent to continent, but not a lot. Someone merged the two and was quite irate when I "unmerged" them.0
-
Julie Buck said: Communicating is great, but it takes hours to clean up the mess. If it weren't quite so easy, or if it was more clearly identified? Is there a place to "replace the removed person"? I'm not aware of that function.0
-
Tom Huber said: If a merge has recently occurred, I'll send a message to the user who merged the two, explaining my actions (unmerging the merge) and why I did it. Because the message starts with my thanks for their interest, I've never encountered an irate user over the action I've taken.0
-
Julie Buck said: Darn, I was hoping to find a way to UNMERGE after the fact. I've done, like you, an UNMERGE, and explained to the user who did it, and never had an irate user. But the one I'm working on now was merged (26 times - and I'm not exagerating). Some of them a long time ago.0
-
Paul said: Unfortunately, politeness is only a secondary factor when these situations are encountered. I get short, polite replies like "Thank you or "Sorry!" But that hardly compensates for the amount of hours it has taken me to fix the incorrect merges - which almost certainly have been made following a rather far-fetched "possible duplicate" suggestion from "FamilySearch".
The other annoying thing is how they are immediately offered again as a possible duplicate, and have be recorded as "Not a match". If I move on from the person page too quickly another user will probably repeat the incorrect merge, because "FamilySearch" is "STILL" suggesting it is!0 -
Tom Huber said: The best that can be done (if changes are made to the surviving record after the merge) is to restore the merged record and clean up the existing record so that the merged results are replaced with the original entries.
If no changes have been made after the merge, then the records can be "unmerged" but there still may remain some artifacts from the merged records.0
This discussion has been closed.