Reporting a glitch: Disappearing index
I sent this issue to FamilySearch support's email, they don't seem to understand and told me to post it here. Hopefully this gets attention from the tech team.
I do a lot of work on immigrants, and when I see an immigration record where it was digitally transcribed wrong or has a birthplace that hasn't been read, I try making edits to improve their searchability and the information the index provides. This is a very useful feature and helps everyone and I'm glad to use it. I've been using it for a couple years to good effect (sometimes you end up back at the same record for someone else and it saves time trying to read it).
But when working on recent transcriptions, specifically the Galveston immigration records and the St Louis marriage license records, I've had problems at least 3 times where when making a single edit to an indexed person, after I save the edit, an issue will prompt above the person's name saying I need to sign in or "we are unable to show if this person is attached." When I refresh, the person's individual digital index disappears completely.
If I try to find the person in the main search again, they will show up with all the edits I made, but when clicking on that record, it will say "We're unable to show this record." After a day, the person will disappear from the search altogether.
As an example, for this page I updated the date and location of immigration for every person. Now, everyone is missing and the system claims it was never indexed.
https://www.familysearch.org/ark:/61903/3:1:3Q9M-C9YK-X9D6-C
Regarding the man named Jaroslav, I navigated on the main search and found him with my updated date and location, but when clicking on the record it says "We're unable to show this record."
https://www.familysearch.org/ark:/61903/1:1:CKY4-DRN2?lang=en
He now today no longer even shows up in the main search.
Regarding a man on this list, Valentin Honc, another user successfully attached this source to his profile. I believe this preserved his digital index which you can see here. It also shows my updated location and date as well as the original record.
https://www.familysearch.org/ark:/61903/1:1:CKY4-D5N2?lang=en
But when you click on the record, it says "This image has not been indexed." It seems the index became detached from the page.
This incident also happened here, when I made edits to the men Jan and Arnost Dybal. They disappeared but the rest of the passengers remain.
https://www.familysearch.org/ark:/61903/3:1:3Q9M-C9YV-NS4T-M?lang=en&i=193&cc=3348478
If I try to recall, it also happened with the St Louis marriage licenses, where I updated a woman's name to be Mrs. (to specify it was her previously married name). Once I made that single edit, both the man and the woman disappeared from the record. Unfortunately I don't remember what couple, but luckily I noticed and manually added the source to their profiles afterwards.
Please review this glitch. I now know to avoid making edits to the St Louis marriage licenses and the Galveston records, but others might not. And the work of volunteers going through the whole book will be for nothing if this isn't fixed.
I have a theory that the reason this happens is the record was recently transcribed. Because of that, the system has not finished classifying it as a completed index (so the system is still in "transcribe mode"). So once anyone makes an edit, the system will reclassify the index as incomplete and it will disappear from the search, prompted back to the drawing board to be reviewed. Is that right? In that case, hopefully all these people are still in limbo somewhere and can be retrieved by approval. I have seen similar issues prompted when working on Ellis Island records, but when I refresh, the person is still there. This makes me believe it's because of how long ago the person was indexed.
Answers
-
@SerraNola could you have a look at this please?
0 -
A similar problem is happening with WWI and WWII draft cards:
0 -
@Jonny Stipek your post was very detailed. I have looked at each link you provided. I will make the engineers aware of this. Thank you for your diligence in helping us to improve the usability of FamilySearch.
0 -
First, your efforts to index and edit records are truly valuable, and I apologize that it has not been a reliable and satisfying experience. Many in the Community know we’ve had an ongoing issue where some edited records don’t stay consistently visible after they’re saved and it’s not always the same fix to take care of the problem.
Behind the scenes, every time a record is indexed or edited, the system automatically re-checks and reapplies access rules. Sometimes that new check applies different permissions than the record originally had. In some cases, this is correct—for example, when an incorrect date is fixed and the record now clearly does belong in a restricted time period. However, often the new permission is wrong, and the challenge for engineering is determining what triggered it.
I appreciate your clear description of how the error occurred, which made it easy to replicate. I have a hunch the issue with passenger lists may be related to the addition of a new special permissions tag, as I see it is attached to this film. As they are still implementing this, some permissions may not be correctly assigned. I will ensure that a work item is created for this passenger list collection and will also examine the St Louis marriage collection.
@Nyx773 I can confirm that the missing indexes in World War I and World War II collections are part of this same problem, and specific work items have been created to address these draft registration–type record collections. This is an active issue for our Rights Management team, who are fully engaged in finding and fixing the root cause.
Thank you for your report. When there are updates or if more information is needed, I will post here.
4 -
@SerraNola
I have another case of disappearing index. This time for Immigration Index Cards.
This morning I was working on 3 profiles for people named Jacob/Jakub/Jacobus Warzecha ( L114-6X9, GLVX-TGH and GLVX-XF4), trying to figure out if any are duplicates and which, if any, is the Jakub Warzecha who immigrated on 8 March 1900 and other mysteries.
At some previous date (probably June and/or July 2025), I added indexes for the person in the upper-right-hand corner of these cards, the person named in the "Accompanied" box:
Adam for Jakub Warzecha
Jakub for Adam Warzecha
I attached the cards to Adam(us) Warzecha G9M2-F3H back in July, but not to anyone named Jakub (since I don't know who is who).
While looking at these particular index cards again this morning, I created links between the names. Since I don't know if they are brothers or cousins or if they are related and since "Other" doesn't work, I chose "Associate".
I then stepped away for a few hours and now the indexes have disappeared.
HOWEVER, UNLIKE with the previously reported problem records above the indexing is still visible and the image is still visible when looking at 1) a hint on the details tab or 2) the files on the Sources tab.
1) Click on the hint available for Jacobus Warzecha GLVX-TGH on the detail tab, and you'll see the indexing and the image of the index card. But when you click on "View Original Document", the page opens and states "This image has not been indexed." All the indexing that was there this morning is now gone.
[note: I'm not ready to declare this is the correct person to attach this record. and note: the July 1900 immigration record for L114-6X9 may be wrongly attached, maybe he immigrated in March 1900]
2) Click on the two March 1900 immigration records for Adam(us) Warzecha G9M2-F3H on the Sources tab, again it shows all the indexing and the image, but again when you click on "View Original Document" all the indexing that was there this morning is now gone.
[note: this might not be the correct Adam Warzecha who immigrated in March 1900 since there is another immigration record for 1901 also attached]0 -
Thank you for your efforts in documenting this problem. This collection includes the newly introduced permissions tag added to passenger lists, which may be related to the issue. Although I was unable to reproduce the indexing loss with the previous report, your documentation will enable me to present further evidence. I appreciate your help; it has been very valuable.
2
