Loading Profile...
At some point today (as far as I can tell), FamilySearch rolled out its long-awaited index editing function.
I kind of want to call it the index editing disfunction.
First of all, it says "Only records with images available on Familysearch currently qualify for name edits", but it says "Unfortunately this record does not yet qualify" when the image is RIGHT THERE UNDER THE EFFING MESSAGE. It comes across as a playground taunt of "psych! Fooled ya!".
Second, on editable entries, it only accepts ASCII characters. There's no documentation for this behavior (or anything else, really, as is usual on FS), but you can try it: try typing (or copying-and-pasting) János or Horváth or Szabó into an index editing field.
Third, it does not allow the given name field to be blank. This means that the index entry for my great-grandmother's youngest sibling, who was stillborn, cannot be corrected (https://www.familysearch.org/ark:/619...). I can't leave the given name blank, like it's supposed to be, but I also can't change the currently-indexed "Halva" to what the register actually says, halvaszületett "stillborn", because it contains a non-ASCII character.
Please, fix the message on uneditable records, and above all, please allow accented characters, so that things can actually be, you know, correctly corrected.
Help get this topic noticed by sharing it on
Twitter,
Facebook, or email.
Twitter,
Facebook, or email.
I don't mean to sound ungrateful, but the new index editing function Needs Help.
Promoted
Responses
-
3The ASCII-only problem is HUGE.
If you edit a field that contains non-ASCII characters, it deletes them. By "edit", I mean "put the cursor anywhere in the field and hit any key." So if you're trying to fix a record that says "Ferencz József" but was misindexed as "Ferencs József", as soon as you delete the 's', the field turns into "Ferenc Jzsef".
You can't even put the correct character(s) in the Notes field, because that's ASCII-only, too -- and it doesn't accept question marks, either. (I didn't check whether it takes asterisks or angle brackets.)
I've been trying to stay polite (or at least not profane) in the notes I add, but I'm afraid I've commited quite a bit of all-caps already. -
4The problem is that this release shows fundamental flaws in the internal SDLC.
Internationalisation should be baked into the process. For code pages that means Unicode support as baseline. ASCII simply does not cut the mustard now. With just ASCII character support you are essentially asking people to do the impossible when correcting many records, as Juli has illustrated.
The Getsatisfaction software can accept à or ù or ç or î as input. Pretty much any web forum software can accept that as input. Heck your own FSFT profile editing software can accept the characters as input! Why can't the transcribed index correction software do the same?
I'm afraid that this release means your software development people have taken a double hit. There is the perception that they are taking inordinately long to release features. On top of that comes the perception that even after taking an inordinately long time to release features, those features are comically incomplete in their execution. Not a good set of perceptions at all.
-
1It only allows ASCII?!!! In this day and age?
Good grief that is dreadful. Anglo-centric to a really, really ridiculous level. Heck even English goes beyond ASCII in some cases! Belovèd and Zöe come to mind as examples.
If this is true then the feature is absolutely not ready to be released. Make it Unicode-compatible before even thinking about releasing it.
BTW the fixed window curse has afflicted things again. The reason statement for the edit Juli has made gets chopped off and the final part is unreadable. Have none of your coders ever heard of scroll bars?Comment -
-
EMPLOYEE
1That is a real catch 22, wait until it's perfect and has all the pieces. I am not sure based on the many complaints patrons have made that they want to wait any longer. I will create a bug report to get the additional characters added. Rather than blank, I would add stillborn or noname . Keep in mind you want it to be searchable and having a blank would cause it to possibly get mixed up with all the other people with that last surname.-
Phil, every indexing project I have ever worked on has instructed one to leave the given name field blank if only a term or description and no name was entered. The previous child in this family was also stillborn, but her birth record was correctly indexed, with no given name: https://www.familysearch.org/ark:/619...
Notice how the entry has only a surname in the "Name" field. Index correction has to allow this; it is 100% correct. -
-
Thanks for making the argument Phil - but no-one expects perfect. Non-ASCII characters and scroll bars are bare minimum, though.
-
-
-
-
-
4The problem is that this release shows fundamental flaws in the internal SDLC.
Internationalisation should be baked into the process. For code pages that means Unicode support as baseline. ASCII simply does not cut the mustard now. With just ASCII character support you are essentially asking people to do the impossible when correcting many records, as Juli has illustrated.
The Getsatisfaction software can accept à or ù or ç or î as input. Pretty much any web forum software can accept that as input. Heck your own FSFT profile editing software can accept the characters as input! Why can't the transcribed index correction software do the same?
I'm afraid that this release means your software development people have taken a double hit. There is the perception that they are taking inordinately long to release features. On top of that comes the perception that even after taking an inordinately long time to release features, those features are comically incomplete in their execution. Not a good set of perceptions at all. -
-
3The ASCII-only problem is HUGE.
If you edit a field that contains non-ASCII characters, it deletes them. By "edit", I mean "put the cursor anywhere in the field and hit any key." So if you're trying to fix a record that says "Ferencz József" but was misindexed as "Ferencs József", as soon as you delete the 's', the field turns into "Ferenc Jzsef".
You can't even put the correct character(s) in the Notes field, because that's ASCII-only, too -- and it doesn't accept question marks, either. (I didn't check whether it takes asterisks or angle brackets.)
I've been trying to stay polite (or at least not profane) in the notes I add, but I'm afraid I've commited quite a bit of all-caps already. -
-
I checked on this and it was working so an engineer is checking as to why it's now failing.
-
What was working? Non-ASCII characters?
-
-
Everything was working as far as character sets (International and English) as we had a number of International users submit changes that were accepted. Something must have broke or got backed out and why engineering is looking at what it was.
-
-
-
-
-
I have discovered that it's not just the given name field that it doesn't allow to be blank: it doesn't allow _any_ name field to be blank.
This makes the instruction to "highlight the full name" rather difficult to follow in many cases, because you end up filling in a parent's surname from elsewhere on the image. (It's also a rather nonsensical instruction when fixing misindexed illegitimate children: the whole point is that "ismeretlen atya" is NOT a name...)
And what these changes do to search results is rather hard to parse: a single child's misindexed baptism somehow multiplies into "children". -
-
Juli - can you give me the link or name you are trying to change? As I understand it ismeretlen atya means unknown father. How would you change it if the record says no father or unknown ? I assume you are saying blank or make it blank. Sorry to ask the questions but I am trying to understand it and use it as a bug justification.
-
I searched for given name = ismeretlen ("unknown") in the Hungary Catholic Church Records collection. There are seven of them, all indexed as parents. I changed them to "- -", for lack of better ideas, but they really all should be blanks.
https://www.familysearch.org/ark:/619... ("unknown father")
https://www.familysearch.org/ark:/619... ("his/her father unknown")
https://www.familysearch.org/ark:/619... (unknown mother, and no, I don't know how that's possible)
https://www.familysearch.org/ark:/619... ("his/her father unknown")
https://www.familysearch.org/ark:/619... ("his/her father unknown")
https://www.familysearch.org/ark:/619... ("his/her father unknown")
https://www.familysearch.org/ark:/619... (foundling, nothing but a given name recorded) -
-
-
-
-
1This is great! We can edit the indexes?
Would it be too much to ask, How do we do this? Seeing this thread, I've hunted all over FamilySearch and Indexing and I see no door marked "Edit". Is this the Review function that was rolled out in 2017?- view 2 more comments
-
-
-
Thanks for that Phil - just a shame that the comms team didn't tell us that to start with rather than sending you on a chase...
-
-
Not really a chase. A blog should come this week hopefully.
-
-
-
-
1The other side of the "name can't be blank" problem is when the index missed something, so there's no entry to correct.
Are there any plans to add an "add new" ability to the index correction function? It'd need two parts:
- Add a new name to an existing record (for example if the indexers missed the spouse's name), and
- Add an entirely new record (for cases where the indexers skipped a line, for example).
The second part would have to start from the image, perhaps at the bottom of the Index Information tab: "Add New Record".
The first part would only need to be offered if the indexed record was missing any name fields from that project: if the indexing project included six name fields, but the indexed record only contains five, then there should be an "add [missing name field]" button/link. Alternately, and perhaps more simply, the system could always give the full complement of name fields in the edit screens: even if no spouse was entered, if the indexing project included a spouse field, there could be a (blank) spouse field, with an Edit link/button. -
-
Yes those are on the docket.
-
-
2What is the situation with collections where church members can see images, but non church members (including me) must view the images at a Family Search Centre, or in some cases at a FamilySearch Affiliate Library
I looked at India Births and Baptisms, 1786-1947
https://www.familysearch.org/search/c...
and used Smith as a Search term
Once I started clicking through the results, I could click through to a statement, example on
https://familysearch.org/ark:/61903/1...
"Name editing is now available. Only records with images available on Familysearch currently qualify for name edits. Unfortunately this record does not yet qualify. As we continue to make progress more records may qualify for editing."
Also on the same page "Images Available
To view these images you must do one of the following:
Sign in to FamilySearch as a member of The Church of Jesus Christ of Latter-day Saints
Access the site at a family history center
Do church members get this same statement as they can in fact see images? If I was at a family history center would I still get the same message or would I be able to edit?
If editing has not yet been rolled out to such collections, what will be the final situation?-
It says the same thing on collections where anyone can see the images but nobody's gotten around to flipping the switch yet. (Try Slovakia Church Books, for example: https://www.familysearch.org/ark:/619...) So no, I don't think you would be able to edit the index on India Births and Baptisms even if you showed up at your local Mormons.
-
-
The way it's supposed to work is if you can see the images home, FHC, affiliate public, LDS and they aren't hosted outside FS then you should be able to edit them, so no FMP, Ancestry.com etc. The collection you show should be editable once the falg is flipped to turn them on.
-
-
-
-
-
Then of course you get collections where people CAN see the images, for example on Ancestry or Findmypast, but can't see them on Familysearch if they aren't at an FHC.
Tricky little conundrum isn't it. -
-
I’m
disappointed
I got all excited because I finally found an indexed record where I can correct the name of the mother, BUT it doesn't work because the zoom into image function is dead.
If I simply click the image to get a closer look it goes to the proper image viewer window, but the image disappears, and the only thing visible is a gray background. It DOES display the proper film and image numbers, but not the image.
If I click the edit button to correct the name, the same thing happens in that the image disappears. The instructions are to highlight the name, but that is not possible when the image is blank.
Sigh.... And I was so looking forward to using this new feature.- view 16 more comments
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Thank you!!!!
-
-
The MAc should work now.
-
-
-
-
An acquaintance pointed out another problem with the index correction function: on Ellis Island records, it only allows editing of other names on the record (if any), not the name of the principal. This means that on records with only the principal's name indexed, it's yet another instance of the "what effing edit link??" problem.
For example: https://www.familysearch.org/ark:/619...
(The guy's name is correctly indexed. It's the "residence place" that needs correction, but of course that's not part of the function yet.)- view 6 more comments
-
-
-
-
-
-
-
Juli - Thanks those are bugs and I opened a bug report. What I was told is only primary name right now so my guess is on the 2nd case is we think the other name is a primary name. Gald we have you folks pointing these things out.
-
-
And we're glad we have you to fix them!!
-
-
-
-
I’m
frustrated
I will report on a number of things I've seen on this edit function which does NOT work at all! for me on two browsers (Chrome and Firefox on a Windows 7 64-bit system - both updated with the most recent version)
The problem is found on this page for sources on Christopher Columbus Hall:
https://www.familysearch.org/tree/per...
The 1913 entry for the death of Sidney Morriss Hall (Christopher's son) contains (in the indexed Information portion) a Spouse's Name that says ", [1913]" with the Edit function next to it.
There is NO spouse listed here on this document, and I'm unsure why the 1913 in brackets is in this location. But my intention is to try to remove the 1913 part, or put in an empty or blank data field, as there is NO spouse listed on the source.
Here are the problems in trying to edit this data or use the edit function:
1. I have to click Edit TWICE to get it to work. [seen in Firefox and Chrome]
2. After I have clicked the Edit button twice, a page pops up that says "Improve the Name" followed by "Loading..." and nothing more. The image is displayed on the right side, with a green "Feedback" button shown. [seen in Firefox and Chrome]
3. Clicking on the Feedback button doesn't work [seen in Firefox]
4. [Chrome only, can't get to this point in Firefox] When I click Feedback, I get a popup screen asking me to make a choice between:
"Specific Feedback I'd like to give feedback on a specific part of this page" or
"Generic Feedback I'd like to give general feedback on the entire website"
When I select Specific Feedback I'm asked to select a part of the page, and I pick the non-functional/stuck "Loading..." portion.
A new pop-up appears with a series of emoji's listed. (Whaaaaa?)
I pick the Sad face one that says "Dislike" (It didn't seem appropriate to pick "Hate" (which leads to darkness - Yoda) and "Neutral", so "Dislike" it is...
The pop-up expands with a text entry box labeled "What would you like to share with us?", and a selector bar rating from 0-10 that says "How likely are you to recommend us to your friends and colleagues?"
Wait... WHAT? I'm trying to comment on a specific part of this page, and you're asking a question more appropriate to an evaluation of the entire website?
It turns out this field is MANDATORY, as I tried to enter my email address in the optional text box labeled "Email (optional)" and hit Submit without putting a rating there (it wasn't appropriate contextually), so I gave a low score and hit submit.
So, the feedback function doesn't work to varying degrees on Chrome and Firefox on a Windows 7 64 bit system. The source code seems to have some references to Apple Itunes. I don't get that... but it's clear that it's NOT working.
And in submitting this entry here, I'm going to click the frowny face, and fill in "frustrated". (this whole emotional/mood thing is ridiculous, honestly - how did we get to this point????) -
-
Justin - Thanks for the feedback. Edit not loading I was able to duplicate on Windows 10 as well and I opened a bug report. The feedback function is a new option and really is generic because it used on multiple feature pages. When it's submitted we capture the link to the page so we know where you were when you submitted the feedback. I am not sure why it's not working well on Windows 7 but I am going to check it out. It's my understanding it will be turned off soon as it was intended to gather information on search and error corrections as they rolled out and now that both are at 100% it will be removed soon.
-
-
Did Phil or any other employee address the issue that I do not see an option to edit my edits? Is that coming? I would like to be able to edit my own edits in case I make a mistake or a poor "note". I made some today, but noticed no option to edit them. I see I can add more, but that is not the right solution for what I am asking. -
-
I saw the banner that editing is not yet available on this record on everything I checked until today, but today I found a record I could edit the name on. So I thought that meant it is all fixed now, but apparently not.
The very next record I went to said it is not possible to edit yet (even though clearly the image is there).
Does FamilySearch need a few more weeks to process the collections to make them compatible or what is going on?
-
-
1You should be able to edit your edit and make corrections. I haven't tried one yet but I'll check it out. I did see it done so if it doesn't work we will need to know. The records need to flip a bit so they can be editable and we are about 80% right now.
- view 1 more comment
-
-
Hope I have understood your request, Michael, and the following will be of help:
https://www.familysearch.org/ask/sale... -
-
I skimmed what you shared, but do still not find any indication in the website itself or shared articles about any ability to modify previously made edits. ??
-
-
-
-
1Good news: index corrections are no longer ASCII only! Thank you, Phil, for whatever part you played in getting this fixed.
Names still can't be blank, and the way corrections display in search results is disconcerting at best: a single child turns into two children. It's especially confusing when a single baptism with three given names was indexed as three people; in the post-correction search results, it suddenly looks like the parents had sextuplets.
I once again came up against a need for _adding_ name fields. I found a child's baptism by browsing, because it was totally misindexed: they skipped the father and invented a name for the mother. (No such name appears anywhere in that entry.) I corrected the mother's name, and added the father's in the note, but the notes are not searchable.
The above-mentioned "sextuplets?" case also has me musing on ways to merge duplicate index entries. It's complicated: sometimes, an event was indexed twice because the register page was filmed twice. This means that duplicate index entries can correspond to different images, but they're duplicates none the less. Other times, an event was indexed twice because it appears in two different registers, such as an original and a bishop's copy. In such cases, I do not consider the index entries to be duplicates, because they're based on different pieces of paper that may have subtly different information on them. (Bishop's copies are often missing the death notations in the margins, for example.) I cannot think of a way for the computer to tell the two scenarios apart. However, even an index-merge function restricted to entries originating from the same image would be useful, for correcting the cases where a single person's multiple names were misindexed as multiple people.-
Well, I just did my first corrections and they worked OK - I think - thanks to reading this thread and others. Even more impressive, I was correcting a Danish 1880 census
Danish? Don't ask... It's me being pedantic, refusing to allow some garbage on a relative's profile re an event in the USA to pass and ending up in Danish records partly to correct the USA stuff and partly as an intellectual challenge. They're rather good actually - if you can read Danish, which I can't. However (getting back to indexing) even I now know that "fødte" means "born", it's not a name so shouldn't have been indexed as one. No problem removing that one though, because it was in the middle of other text in the name field. -
-
-
-
-
Here is a link to the blog https://www.familysearch.org/blog/en/...
-
-
The most recent edit of the name is the one that appears in source citations, in the family list on the record index page, and on the index data tab of the image viewer.
I'd like to be able to reorder them or set a default.
For example, I put two edits. One was to do what the image actually says, and the other was to add the actual spelling. There is an option allowing for both types of edits. However now the actual name appears in all those places, but best practice calls for the display name to be what the image actually shows.- view 1 more comment
-
-
Whenever the original record is something like an abbreviation -- Wm for William, Chas for Charles, and so on, the "record is incorrect" needs to reflect the full name as an alternative, not as a correction.
Hopefully, that will be something that can be done because right now, who are we to correct how the person used or signed their name, especially if they were literate?
The record needs to reflect the name as it was written, so that a search will produce the written version as well as a spelled out version if the record was an abbreviation.
O'Leary is really "of Leary" and that is something we don't want to see. -
-
"best practice calls for the display name to be what the image actually shows". I'd phrase it more like "best practice calls for the name displayed in the citation to be what the image actually shows". That may be what you meant.
Yes - it's an interesting point - when working on my home system, I will always cite something using the name from the source. If it's wrong in the source, I will still cite the incorrect text from the source but add "[sic]" after it.
Indexing needs to cover all the bases - citing needs to use what's in the source. Yeah, never thought of that up until now. -
-
-
-
I have a record that I'd like to edit the gender. The indexer incorrectly made it female, but Clarence as a female name? https://www.familysearch.org/tree/per...
Unfortunately, I didn't catch this a long time ago in my haste to provide names for baptisms for the youth, and as I'm reviewing my reserved list, I'm finding work no longer needed (due to merge's, mostly), and adding new sources to people which gives additional details or family members.
But this one.. needs the gender fixed. Is this an area that can be considered for editing? (I'm thinking not, as a change like that can invalidate a LOT of ordinance work very fast!)
But perhaps a law of witnesses here? (More than 2 validate it needs to be changed?)-
Gender can certainly be changed, but it can have an adverse impact on the ordinance page.
I had a situation with an end-of-line ancestor where the ordinances were "not available" even though they had been performed.
After merging all the records that I could find, the "not available" no longer existed. I think what might have happened is that the records with the earliest valid ordinances match the sex of the current (combined record). That is the only situation where such a change might have allowed what previously had to be handled by support, to be "okay" after the merges.
I still have to wonder, despite the assurance that all ordinances would remain for the person, and that the earliest valid ordinance dates would be used. For some oddball reason, it really looks like some dates are ignored if a lot of merges take place.
Because there is no change log for the ordinance page (severe pushback from Ron Tanner), we have no way to check the final disposition of any ordinance dates and where it came from. I still want to see a change log for the ordinance page, even though it could not be viewed by those who did not have Church membership. -
-
-
-
-
Huh... this might be a bug!
I changed the gender, and the ordinance work done til now still stands! is that because the endowment wasn't done yet?- view 2 more comments
-
-
-
All people who are baptized are confirmed. Not sure where the male only thing comes from. Take a look at ordinances for females.
I'm thinking that because there really isn't a difference until endowment or sealing to spouse stage, a change in gender would not affect the previous ordinances. (I don't think there are any differences for initiatiory, but I haven't asked anybody in the proper environment. There hasn't been a need to, honestly.) -
-
Ulp, you are correct. My mistake, I was thinking of the ordination, but that takes part during the initiatory ordinance.
And yes, the proper environment to discuss the difference in wording is in the temple. -
-
-
-
I think an easy model to follow is what Ancestry does.
You have the indexed name, and then additional names listed (and searchable) whether that is a correction to the indexing or image.
Since I haven't explored the editing option much here on familysearch (my 1st try failed due to improper characters in the field - being debugged as I type), it's sounding like only 1 "correction" is being shown here on familysearch?
Sounds odd... but perhaps there's some purpose in that. Wouldn't be as robust as I think we'd like to see.-
No, all of the corrections show (and are searchable), but it has to choose one for the citation's title, and it appears to go with the most recent one for that purpose. (Unless you're editing an index entry that you attached ages ago, in which case it has the misindexed name in the citation title.)
-
-
The way it will work if it's not already, The original and up to 10 additional aliases will be added to the search. The primary will show but the others will be in the search.
-
-
-
-
-
I just found another one related to race. Adding sources, many family members match in one census, and it says they are negro. But a previous census says they're white.
So I look at the source, and I see what appears to be sheet after sheet of N, even when they or their parents came from Maggyar, Hungary, Slovakia, etc. I saw a sporadic W (too few, statistically speaking), and then after several pages I finally come across B for black. When I go back to compare the W and the N, they look fairly similar, almost like someone filled in that column later, and just rushed it, because the B for black is in very heavy ink, as if they were overwriting something else that was written there.
Can't do anything about race there... but there are some interesting indexing errors!- view 1 more comment
-
-
I have a white ancestor who was the father of black or mulatto children. His black or mulatto daughter married a black man and had 8 children. One of the eight children became a doctor in Chicago, Illinois for the mob and always showed white on all of his censuses.
-
-
That sounds like a story to add to the memories section! :-)
-
-
-
-
Was there ever an explanation as to why an indexed image cannot be corrected?
I'm trying to correct an entry on this page: https://www.familysearch.org/ark:/619... but I get a notice saying the index to this image can't be edited.
Just wondering what the criteria for an indexed image being eligible for editing.- view 8 more comments
-
-
-
-
-
-
-
-
-
The above is from https://en.wikipedia.org/wiki/Fair_us...
I'm not sure that someone could collect and publish for profit, all the obituaries appearing in any given newspaper. -
-
You wouldn't think somebody would do that but you would be surprised how many try that. Some countries are worse than others but it does happen.
-
-
-
-
All
I finally got one of those "Indexed" Records that I am supposed to be able to "Edit"; but, CANNOT, there is NO "Edit" 'Link'!
And I have read both,
the 'Blog'
Editing Names on Indexed Records—FamilySearch Update
https://www.familysearch.org/blog/en/...
and,
the 'Knowledge Article'
How do I fix indexing or transcription errors in historical records?
https://www.familysearch.org/ask/sale...
... Oh, hum!
What a 'Nightmare'!
Brett
ps: Record was from the "British Newspaper Archive, Family Notices".-
My guess: The FindMyPast / British Library version of the "British Newspaper Archive, Family Notices" collection is, I believe, the original for this FS collection and is continually (every week) being updated with new images and indexes to match.
I have no idea how often the FS version will be updated (not least because I don't know the exact relationship between the FS and FMP/BL stuff) but presumably it will be, in which case it would be a nightmare to merge a revised FMP/BL index with FS-only index corrections. So my guess will be that you'll never be able to correct this collection (or the similar "British Newspaper Archives, Obituaries"). But it is just a guess...
For me, the annoying thing is not that I can't do the corrections but that, yet again, the FS Communications team gave us the high-level vision while not warning us that (a) we'd need to wait while collections were switched over and (b) for possibly several, probably perfectly sensible reasons, some collections would never be correctable. High-level vision - yes. Requisite detail - no. Communication before the event - no. -
-
-
-
-
Somebody, I can't find it now indicated that the panel for editing brings up his full name. I opened a bug on it and the contact name is what is used and you can change that. Many people took the default which used your full name for a contact name and why it showed.
-
