New Viewer Issue, Records Disappear
Answers
-
There is a current known problem, especially with the 1950 census, but also with many other records. The new Image Viewer/Editor tool has a flaw. If someone edits the record, to correct the index, the record disappears. For more info, see this thread. There are links to other threads on the topic in that thread.
0 -
Thank you, I reviewed the thread. When I work with visitors and guests, I'll just have to tell them that FS rushed something out that wasn't ready and now we all are paying the price. With all the obvious frustration, why doesn't FS simply remove the new search viewer, until they get it fixed? I tried the new version, didn't like it, and am using the current version.
Until FS starts involving actual users for testing before rolling our developer "fixes," these problems will continue. Of course, FS could document user requirements before "fixing" what works!
3 -
As Áine implied, the answer to your title question -- why doesn't FS give a reason -- is that it wasn't FS doing the removal. It was the gremlins in the seriously-underbaked index editor that they rushed out the door for the 1950 census and still haven't managed to fix. In fact, if anything, it seems to be getting worse -- although this might be an artifact of it being rolled out in more and more indexed collections.
1 -
The good news (it's a stretch) - the records that disappeared do sometimes reappear, several days/weeks after the edit.
0 -
I was made aware of a new wrinkle in the index-editor tool's propensity for deleting entries instead of fixing them: editing one entry deleted the next one.
In the baptisms of István and his twin sister Katalin, the indexers misread their father's surname of Konderák as Kondesak.
A descendant of one of their siblings came along and corrected István's entry, as can be seen in the index panel on the right. His index detail page now shows just the new name, as if it were original.
It's only on the new editor's details pane that experienced users of FS can recognize the "this name has been edited" pattern.
The individual fields have different entries than the combined field, which is what happens when someone edits the name. This is ...not ideal, but hardly worth noticing, in the grand scheme of things on FS.
What is worth noticing is what happens if you look at Katalin. Her entry has not been touched; it is exactly as originally indexed, however many decades ago. (These Catholic baptisms were done back in the IGI days.)
However, clicking the box-and-arrow after her name, which should go to her individual index detail page, instead results in the "removed" error message.
In other words, editing one index entry has resulted in the removal of a different index entry.
FamilySearch, PLEASE turn off the edit-anything tool until its gremlins have all been chased down and eliminated. They're seriously corrupting your data and compromising your entire website's utility.
1 -
Appreciate the hopeful news. The solution still seems to be to take it off line, and fix it. Many of us would be happy to continue working on reviewing the 1950 census, to get it right. Of if FS doesn't want to do that, simply posting another line to the effect of this is a temporary software problem that we are working to fix.
However, it is FS - unless some other external entity has control over the software development and implementation. I don't know what part of FS handles software development, but a basic principle is that requirements should drive development, and before a product or upgrade is rolled out, there are processes for verification and validation that it meets requirements and does what it is intended to do. Absent V&V, you are guaranteed to have problems. Without involving actual users for both requirements and testing before release, FS is doomed to continue frustrating users around the world.
6 -
@Sam Sulser This is the sort of thing we are seeing daily. Please, we need to know that someone somewhere is working on a fix.
Thank you.
3 -
Multiple times daily, now, and affecting all parts of the system. Such as hints.
I don't know how to even recover from this situation. How can that hint be dealt with? It can't be attached, it can't be dismissed.
3 -
Just a quick addendum. I can still see the Record direct, as before, but not via Search. One point I did not make previously is that the displayed Record has this on it:
I assume this is a standard symptom of the current widespread index corruption issue.
0 -
This still isn't fixed. I guess I will use this post to report every time family groups and full pages of indexed people disappear during my attempts to group households and/or designate "Principals."
I grouped a household on a Ramsey County, Minnesota 1895 Census (Henke's and Hitz), and all of the names have now disappeared from the index. Here is the link to the Fredrick Henke entry: https://www.familysearch.org/ark:/61903/1:1:MQ64-X2V
Here is another one where all of the names disappeared from the index when I grouped a household and designated Principals and relationships. Fred Henke: https://www.familysearch.org/ark:/61903/1:1:MQ64-Q6K
1 -
Yes, this is being looked at. The information you guys give us is put in front of the team. In addition they are collecting feedback on that page (which they actually look at - a lot). Keeping submitting feedback!
Sam🤓
0 -
@Sam Sulser, how can we submit feedback on the nonfunctional hint? There is no Feedback tab to be seen anywhere for that.
And how do the engineers expect to be told about indexed records that have disappeared from searches? The whole problem/point is that they're gone. Users can't tell if that's because the record doesn't exist, or because the index entry no longer does.
As I've said before, this bug/gremlin should have been an absolute and immediate deal-breaker, way back in June when it was first reported. The index editor should have been taken offline, all of its edits should have been reverted, and it should not have been put anywhere in reach of the production database until this problem was fully and completely eradicated. The current situation, where people are allowed to blithely continue destroying the database with this badly-broken tool, is inexplicable and unfathomable.
6 -
We absolutely need a group where we can work with the developers. It was very useful for the New Person Page and other changes. The problem is not going to fix itself, and the developers/engineers can't fix what we cannot tell them about.
3 -
Reading these posts makes me grateful that nearly all of the sources I attach to my relatives' profiles can't be edited! I completely agree with the suggestion that this tool should be withdrawn from FamilySearch until a fix is found, as its use currently appears to be doing more harm than good.
Unfortunately, this is not the only feature that is completely broken at the moment, so we can only hope these problems (dating back several months) will soon be resolved.
2 -
Hey guys. I have merged several discussions about the new editing tool causing records to disappear into the discussion that is being tracked. I have also changed the title.
1 -
I hope FamilySearch will take from this the need to communicate better and be transparent. When a bug or problem first comes to light, post a message to alert all about it, how it is manifested, that it is being worked on, estimated resolution date is whatever, etc. This especially needs to be communicated internally among the user facing staff, the developers, the managers, etc. Your credibility and trust are at stake. One other problem that became apparent from different responses is that FS is too much in individual silos. To the using "customer" FS means the software, the database, the records, etc., without differentiation.
4 -
All of the names on Page 6 of 11 of this Harrison County, Iowa 1950 Census were all in a single group, so I tried to separate them into separate groups. It looked like it had worked, but when I reloaded the page, everything seemed messed up: https://www.familysearch.org/ark:/61903/3:1:3QHN-GQHW-TKLC?cc=4464515
The groups all disappeared, and now when I click on ATTACH TO TREE on any of the names, it goes to the "Something Went Wrong: Error finding source" error for all of them.
It is very troubling how many of these census pages I seem to be destroying just by trying to group the households, designate only the Head as Principal, and assigning relationships. Sometimes it works, so I will keep trying. But often it doesn't work, so the indexing on many many pages is seemingly getting destroyed, although I am sure there must be a way to recover that.
UPDATE
In the above example, I grouped the households on the 1950 Census page that included the following family of three. Before I made the groups in the editor, I was able to click on the source documents on their individual pages to go to the census page. However, after making the groups in the editor, clicking on the sources on their individual pages now gives the "Something Went Wrong: Error finding source" warning:
Henry George Johnson (Head: LV9F-2PQ) --> Error finding source: /ark:/61903/1:1:6FQF-J1YY
Anna Mabel Waples (Wife: KLB4-DDV) --> Error finding source: /ark:/61903/1:1:6FQF-J1YT
George William Johnson (Son: GKRK-JMY) --> Error finding source: /ark:/61903/1:1:6FQF-J1YB
This makes me wonder how many other IDs have had their sources unlinked due to the failed editing of all of these census pages. Can this ever be recovered? How many more will be destroyed before this critical bug is fixed?
I think the bug tends to pop up more often when household groups are made that span between pages, but then it tends to mess up a lot of stuff on both pages, if not the whole set of pages. This seems VERY DESTRUCTIVE, unless I am just not understanding what is going on.
1 -
@SteveLinke In case you haven't seen this comment which @Stephanie V. made elsewhere https://community.familysearch.org/en/discussion/comment/543743/#Comment_543743
Due to an issue with the FamilySearch Rights Management system, records in part of certain Historical Records Collections were mistakenly restricted following user record edits. This problem was fixed and validated on February 13, 2024. Record editing should no longer cause any kind of restriction or cause records to “disappear”. For the records that were impacted due to the defect, the data is not lost, just restricted from view. All of the records that were impacted will be completely restored in the coming days and weeks as we work through and restore the data. Thank you for your patience and also for reporting problems as you see them.
1 -
Thanks for your reply, Áine. That notice says that the problem was fixed on 2/13, but my editing of the census pages on 2/14 caused the sources to disappear. Thus, I would have to conclude that the issue I am experiencing is separate from the one that was fixed.
Would you (or anybody else out there) be willing to see what happens if you try to group households? The below link should bring you to another page in the Harrison County, Iowa 1950 Census that is not properly separated into households (and has not been messed up by me yet). Create household groups by clicking on a Head name, Relationships, DRAG NAMES, drag name to NEW RECORD GROUP. Then, back out and drag the rest of the names over the new group, and then assign the relevant relationships. Wait a few minutes and reload the page. Now, also try to group the name at the top (Charlotte M Herman) with her family on the previous page and see what happens.
The links to the various iterations of the census editors are ver confusing, but, try this link, then go to Page 8 (for example): https://www.familysearch.org/ark:/61903/3:1:3QHJ-5QHW-TK2P
Note that, in this particular set of census pages, it looks like the heads of the households were properly left as the Principals, and the Principals "flag" was turned off for the rest of the members of the household. However, the households were not put into separate groups, such that when the Source Linker comes up to attach people to the Family Tree, the entire page of people show up. In many other cases, not only are there no separate household groups, but also every person shows up as a Principal.
1 -
No, I would not try to group a household, and most folks I know are shying away from editing until we know the problem is fully corrected. The issue was "validated." I take that to mean someone in charge finally agreed with the months of user reports, not that it was repaired so it won't happen anymore.
1 -
Ayup, @Stephanie V., whatever they think they've fixed, it's not fixed.
I only edited the first two families, but the entire page's index has vanished in the old viewer (https://www.familysearch.org/ark:/61903/3:1:3QHJ-5QHW-TK25), and all of the details page links in the new viewer (https://www.familysearch.org/ark:/61903/3:1:3QHJ-5QHW-TK25?view=index&action=view&personArk=%2Fark%3A%2F61903%2F1%3A1%3A6FQF-GCKS) come up Not Found for edited people (such as https://www.familysearch.org/ark:/61903/1:1:6FQF-GCKS) and Removed for unedited people (such as https://www.familysearch.org/ark:/61903/1:1:6FQF-PBFW).
I'm afraid we need to continue refraining from any kind of index editing on FamilySearch. The index editor is currently nonfunctional, and any attempts at using it are liable to be highly damaging to the data.
I have submitted Feedback via the tab on the page. I may have overdone the exclamation points a little, but the continuing availability of this fatally broken tool is terrifyingly awful.
3 -
Thanks for trying, Julia Szent-Györgyi. At least I know that I am not crazy to think the editor is really messing things up.
The quote from @Stephanie V. that "This problem was fixed and validated on February 13, 2024. Record editing should no longer cause any kind of restriction or cause records to 'disappear'" clearly does not apply to this census record editor issue, which is ongoing.
It makes me wonder how many pages I have inadvertently ruined, how many pages other users have inadvertently ruined, and--most importantly--whether FamilySearch will be able to identify and repair all of those ruined pages sprinkled among the hundreds of thousands of pages that are indexed. I am still learning how all of this works, and probably only FamilySearch knows for sure, but it is very troubling.
I have provided comments through the Feedback button and through posts here. If Stephanie V or anybody else at FamilySearch is reading any of this, please do something to correct this.
1 -
There were many of us working on the 1950 census when FS appeared to decide to stop human work and rush it out. One thought is to separate the verified 1950 work, then pull the rest back and allow users to work on it. In the interim, I've found Ancestry has the 1950 census available, as well as other records currently unavailable on FS, so suggest FSCs and those with Ancestry access stop trying to work with a broken program and just use Ancestry. I've not searched MyHeritage or FindMyPast, but they also may have working programs for these disappearing records.
1 -
Example: familysearch.org/ark:/61903/3:1:9392-MK9R-GZ?i=255
Click on any of the names and you get the error message 'This Record has been removed'
1 -
A known error with several existing threads. The problem stems from the New Image Viewer/Editor.
3 -
I am researching this new occurrence. Our engineers are diligently working on this but it is quite complex. Thanks for your patience. Nothing is being lost or deleted, it is just hidden.
2 -
If you change any of the existing data of records belong to the very image, disapperared records will turn to visible and available again.
For example if you have 6 record groups linked to an image of a birth register, change (or set) every principal's sex and all the data "comes back" though not all the data of the strucure will be visible. (After editing data of page Nr. "n", pages "n-1" and "n+1", image index on them will appear as before, on page "n" there will be less...)
To edit a permanently unavailable record, open the previous or the next image of the film, enter the editor and turn to the image that contains the records you need. The changes need to be done in one go to have effect on the data.
This might help until bug will be fixed...
0 -
@Maile L is there a page that lets the community see the specific live data and/or application issues that have been identified (whether or not as a result of reported incidents), plus the progress made against them, and any advice on what users should be avoiding doing until they are resolved?
1 -
@MandyShaw1 I believe having a place for "Known Issues" has been discussed, but that is not currently available.
1
