We need shared "notes" space in all kind of sources

LegacyUser

Bosch said: When we create a source, there is a space that is common to ALL the attached people that is called "notes". This space is being used by lots of people to add transcriptions of the sources they are linking.

See example:



Sometimes, people is even adding translations of the transcriptions to other languages and that can be really useful when sources are in minor languages. And if there is a mistake, you only have to edit once and the change can be seen in all ancestors using that source.

That doesn't work in this way in indexed sources. In this case, if we change the note in one person that transcription can only be seen in that person and not in the rest of the people indexed from the same record.

It is a shame to think that one day the directly linked sources will be substituted by indexed sources and all those notes will be deleted and lost. We can preserve them easily with the same kind of "shared notes".

LegacyUser

joe martel said: It is a deficiency in the rule that indexed sources from FS can only be attached to one person. Sources today will not be replaced by a record that becomes indexed. Some have done a workaround where they make a copy of that Source, put it in their SourceBox and attach that to multiple people. 

The idea of a indexed Source for a family, that mimics the image was the original design. But I don't expect any solution to the current problem anytime soon.

LegacyUser

Bosch said: "Sources today will not be replaced by a record that becomes indexed"

Not automatically, but if a source is duplicated the previous manually created source will be deleted by people. The is no point on preserving duplicated sources.

"Some have done a workaround where they make a copy of that Source, put it in their SourceBox and attach that to multiple people."

It messes more all. Indexed sources must be attached and duplicates only mess.

"The idea of a indexed Source for a family, that mimics the image was the original design"

I was thinking in a common space for all people in the same record. They have a database connection, so is a change that can be implemented automatically.

LegacyUser

Jordi Kloosterboer said: I copy the notes of a source before I delete the duplicate

LegacyUser

Bosch said: And you have to copy the same in ALL the people in the record. And if one is modified you have to modify the text in ALL the people in the record.

Most people won't do that.

LegacyUser

Tom Huber said: Much of the information in the massive tree is person-centric, which includes sources once they are attached to one or more people in the tree. That is a hindrance (as Joe said) and no solution is on the near horizon. Maybe a year or more down the road, or maybe even as long as it took to get index correction abilities (still not fully functional) -- ten years.

I do not find notes in sources to be particularly useful, particularly if the source has readily available images. For instance why bother putting a note with a census record, for instance. The only time that I felt a note would be necessary is in a census in which the enumerator made an absolute mess of the names. Then it was more a matter of making changes to the title of each person involved in the mess and an applicable note.

Yes, it would have been nice to be able to apply a note to the source, no matter who it is attached to (more than my family's names were a mess), but that wouldn't fix the title.

What can be done is to create a memory with the information and then provide a link in the source(s) to that memory and from the memory to the source(s). Then tag the memory to the various people involved with the source(s).

LegacyUser

Bosch said: "For instance why bother putting a note with a census record, for instance."

I can see you haven't found marriage contracts in latin.

"What can be done is to create a memory with the information and then provide a link in the source(s) to that memory and from the memory to the source(s). Then tag the memory to the various people involved with the source(s)."

What? No! I was talking about avoiding duplicates.

LegacyUser

Juli said: The "linked" source citations are my favorite part of FS, but it seems like the capacity for it is just a side effect or red-headed stepchild: they haven't done anything to address the lopsidedness (if you need to attach it to another person, but don't have it in your source box, then you have to start over at the very beginning), and as Bosch says, there's nothing remotely like it available on indexed sources.

In my ideal world, all sources would be linked, meaning you'd only need to write one citation, regardless of how many people were mentioned. This citation would be event-specific: for example, there would only be one citation for a multi-page (multi-image) probate record, but conversely, on a six-items-per-page register book, there would be separate citations for each of those six register entries. If the image was indexed, then each indexed name would uniquely link to the relevant profile (like it does now), but the "other end" of those links would not be a unique index entry entity, but the single central citation. This would mean that you'd only need to enter a transcription and translation *once*, and then you could go through and attach (link) the citation to all of the people mentioned. This would be very handy on things like marriages, and invaluable on those long-winded goldmines known as funeral notices.

LegacyUser

Jeff Wiseman said:

I was thinking in a common space for all people in the same record

So it sounds like you are suggesting that all citations to a census image be the common citation that is attached to all records in the tree taken from that census (all 500,000+ of them)?

That really is in direct opposition to the whole concept of a quality citation. That is, a citation which uniquely identifies not only the source, but the specific location and identification of the data in that source, as well as any notes specific to that data that is pertinent to the record which the citation is being attached to.

I definitely do NOT want shared notes in all citations. I mean, can you just imagine if a citation to a census image shared notes with all other citations to that census, every single note that is entered for every single person in that census will exist in that same note space (e.g., "George's last name was not indexed correctly--the correct spelling is..."). It could be HUGE!

No, that is why in the FSFT when you use the automatic linking of "sources" (i.e., Index file entries) to person records, you are NOT creating a citation to the digital image, you are creating a citation to a specific entry in the INDEX FILE derived that digital image. The Index file is the source here and NOT the digital image.

It all depends on what you are citing in the citation that you attach to a record in FSFT (note that FS uses the term "source" rather ambiguously here. The things that exit in a person's "Source List" are not really sources, they are citations to sources).

You can be directly citing a digital image where you must identify in the citation where that image is and what part of the image applies to the citation.

On the other hand, you can cite a specific entry in an Index file that was derived from a digital image. In this case, the source being cited is the INDEX itself and NOT the digital image that the index was derived from. This is the way that the source linker in FSFT always sets things up.

I really don't want to have the system set up so that somebody adding a note to a citation attached to a single person record, changes the notes in thousands of other citations where those notes are in attachments to other people where the note does not even apply.

LegacyUser

Bosch said: Stop talking about census, please. Nobody talked about census. It a completely different situation of a marriage source.

LegacyUser

Tom Huber said: Correct, I do not have marriage contracts in latin. Almost all of my ancestral lines came to North America and, lacking available ships' lists, cannot track them back across the Atlantic. So virtually all of the records I find are in some form of English.

I understand about duplicating notes, but until FS gets away from person-centric sourcing (which is unlikely given the current state of the massive tree and sources), there isn't much hope for the near future.

LegacyUser

Jeff Wiseman said: Actually, it is not different at all. It only differs in size. If you have a marriage record that has been indexed, it is exactly the same situation as when dealing with a census that has been indexed. You can cite the marriage paper image directly, or you can cite specific people in the index that was derived from that image.

LegacyUser

Tom Huber said: I use that "ideal world" in my local database.

I have one citation for the record and use that same citation (and image) for each person in that source. When I print out my "Family History" book -- Descendants of (name of ancestor) -- I have only one copy of the image with links to it from each of the people who appear in the book who are mentioned in that source.

It took me several tries with my local program before I got things down to where I could do that.

To me, the person-centric sourcing was a bad move in designing FamilySearch and sourcing. It means that a new title has to be generated for each person to which the same source is attached. Talk about adding unnecessary complexity to what is basically a very easy method of sourcing.

LegacyUser

Bosch said: This proposal is not against person-centric sourcing. It just needs certain programing. Notes can be added to the original record that is linked to various ancestors and this information be extracted to all sources attached to all ancestors where they are attached, in the same way that all them have a link to the same image.

LegacyUser

Bosch said: Not, it is completely different. If you want to understand a record to be sure indexation is right you need to read all the record and a transcription and a translation can help. A typical marriage record has 6 persons and all them are very closed related. To show the same text to all them can help.

LegacyUser

Tom Huber said: Many of us have problems with the person-centric sourcing -- actually not the idea of sourcing, but the way the source is set up in the person's record. If the source was source-centric, then a note in one copy would appear wherever the source was attached.

Your request would be automatically handled with source-centric sourcing. As it is, the means to get there will require a massive redesign of the sourcing routines so that notes can be carried over to each copy of the source attach to persons in the tree.

FamilySearch would be better off starting with a new sourcing model -- one that is source-centric, rather than person centric. I think it could be done without having to reattach every source, but I do not know if FS is of the mind to get away from the very much flawed person-centric model.

LegacyUser

Jeff Wiseman said: I agree with regard to the transcription and translation. However, the original suggestion was to enforce "shared "notes" space in all kinds of sources. The most common type of sources being used right now are index files and not the image files they were derived from. In many cases we don't even have the image files. We only have the indexed files that were derived from them.

In FS they are used in doing searches, supporting the hints engine and the source linker, as well as automatic titling of citations. In general all citations to index files are for one specific person, meaning that any one of those citations will normally, at most, attach to a single person in the shared tree. So they should not have their notes set up and shared as though the citation were pointing directly to the original image source and shared across multiple reason records.

Of course, when you DO create citations directly to the image sources (thus bypassing the index file sources) then it might be beneficial to include shared notes that apply to the entire source, and this is currently already supported if you go through the source box and cloning it to multiple people. But the minute you have a person specific type note in a citation to a general image source (instead of a person specific index file source), you have no idea who else is going to get a copy of that note in their own citation to that source.

Index files are not going away, and they are deeply entrenched in the structures used at FSFT. Citations to index files are, by definition, person-centric. And as such will never need to have shared notes. You can get away with shared notes between citations directly to image sources that don't get attached to too many other people, but any information notes unique to just one of the people is forced upon everyone else's citation. For small records such as marriage, you might get away with it. For large ones like censuses, it would be completely non-sustainable.

And suggesting that a different type of method be used depending on how big the original source is would be kinda ridiculous because citations should always behave in the same predictable manner.

LegacyUser

Bosch said: Speaking about all kinds I mean indexed and created from image that are the two kinds I know and use. I didn't imagine that someone would talk about census when it obviously doesn't make sense.

"And suggesting that a different type of method be used depending on how big the original source is would be kinda ridiculous because citations should always behave in the same predictable manner."

What? Depending of the kind. It's just programming. Marriage and census sources are already different.

LegacyUser

Juli said: Jeff, your "500,000+" concept is a total red herring. That's not at all what anybody meant. A census enumeration citation would be specific to a family group, which may be one or two images.

But I agree that for citation purposes, a marriage is exactly the same as a census. The only difference is that the "scope" is more intuitive on a marriage.

LegacyUser

Jeff Wiseman said: Tom, I'm not sure how successful such a venture would be. The reason the model is person-centric is because of all the sources being cited are INDEX FILES and not image files. When you get the URL for an indexed source in FSFT, it is pointing to a location in the Index file. So each URL does exist for a separate person in the index. But one of the advantages of it working the way it does, is that the creation of citations when attaching Index entry sources can be automated to a large degree and less experienced people don't have to figure out how to create effective citations by hand. Also the presence of more than one citation to a specific location in an index file is always either a mistake or an indication of a possible duplicate.

In the AQ model (which I really do like), you really can't reference a source without going through/using a citation (other than kludging a blank citation). But unlike referencing Index files in FSFT, a citation in AQ can be SHARED between different person's records. But that is because in AQ, you as the user can directly manipulate not only the citation assignments, but also their creation and the creation of the source references that they cite. In FS, the support of the sources database is mainly taken out of your hands.

That limitation is very unlikely to change, so getting away from the influence of structures promoted by the presence of index files seems really unlikely.

LegacyUser

Jeff Wiseman said: Yea, I realized some of that after I hit "Comment". However the original digital image source is not just a few pages, it can be an entire film. That film, of course, would not necessarily be the entire census, but it could still be a whole lot of people.

I suspect that the "Family" or "Residence" groups is totally a function of the indexing and that information is contained within the index file in a way that the source linker tool can use it when attaching to multiple people. However, each attachment is still a citation unique to each person. Everyone does not get a common, shared citation attached to them. However, the citing the family or residence entity is unnecessary since the citation titling generation usually handles that (e.g., Jane Smith in the household of Peter Smith)

To make a citation shared you need to do it manually using the source box, and by not citing the index file but by citing the image source directly

LegacyUser

Juli said: As I have pointed out before, the current source citation structure on FamilySearch, with its emphasis not just on indexed (derivative) records, but on individual names within those derivative records, is, how shall I put this, bass-ackwards. My analogy is that we're being forced to cite the card from the card catalog instead of the actual book in the library.

Linked/grouped citations *can* still be person-centric. You just need to revise the "primary anchor point" -- instead of millions of individual points, one for each and every indexed name, you'd have a somewhat smaller set (though still millions), one for each event, household, or similar entity. Individual indexed names would each be linked to such an anchor citation entity on one end, and (hopefully, eventually) to a profile at the other end. Establishing those connections for an indexed source would be unchanged from the current Source Linker functionality; the only difference would be that the citation generated by the process would be for the whole group (event/household) instead of separately for every name. It would still be automatically generated based on the metadata for the collection and the index entries.

LegacyUser

joe martel said: My approach to historical records and evidence is decidedly old-school. But the records themselves are old-school. They almost always written down on some substrate like paper. We have the luxury of capturing those as images on film and now digitally, making it easier to reproduce and reference.

But in the old day (and still do) make a paper copy of the record and annotate it, transcribe it right on the paper and put it in the folder for that family, that has a page for each person.

Today, I still believe the image is king. So I try to create one source that has that image in it. That image is typically just of the critical piece, just the household in a census, not the whole page. In that Source I transcribe the image into the Notes, create a citation so I can find it again (typically a URL and the records' real citation) and Attach that one Source to all the people in the record.

But then I get the FS hint to the persona in the record. I Attach those as well, each persona hint to each Person. That makes FS happy - it accounts for that persona being Attached to a FSFT Person.

LegacyUser

Tom Huber said: I have nothing against person-centric entries on a person's source page, but when it comes to the record of the source itself, that needs to remain source-centric so that a note entered for one record's source appears on all instances of that source.

I agree that the citation needs to be source- and not person-centric. The title on a person's source page certainly can be person-centric, but the source and citation itself needs to remain as a source-centric source and citation.

LegacyUser

ATP said: Tom Huber,

Tried to post this under your name, but, wouldn't allow. I'm not sure what you mean as "person centric", and, if you don't mind, would you define the terms in which you use the phrase. It is my thinking it is 'person centric", also, but, I'm wondering if we might using the phrase in different terms.

Thanks...

LegacyUser

Tom Huber said: I define person-centric as a source record that is oriented to a person, rather than a source. For instance, the typical FamilySearch citation is person centric:

"United States Census, 1870", database with images, FamilySearch (https://familysearch.org/ark:/61903/1... : 19 March 2020), William Newman in entry for J T Philippi, 1870.

and

"Idaho Death Certificates, 1938-1961," database, FamilySearch (https://familysearch.org/ark:/61903/1... : 1 February 2016), William A Newman in entry for Homer Ray Newman, 04 Aug 1950; citing Death, Lewiston, Nez Perce, Idaho, United States, Department of Health and Welfare, Boise..

In the above examples, I've highlighted the person that is person-centric in the citation.

Note that the 1870 census citation is absolutely useless if I am searching on a different site or in the NARA records.

I don't have a problem with titles in a person's source list being person-centric, but the citation should always be source-centric. For instance, a decent citation for the 1870 census would be

"U.S. Federal Census, 1870, for for Washington County, Oregon, Post Office: Forest Grove, NARA Series M593, Roll 1288, Pg 12

Note that the Post Office is not the place location, but the serving post office for that page.