STOP 🛑 allowing Gedcoms to be uploaded to the main tree
Comments
-
I do not want to sound like a conspiracy supporter, but it seems that now that a lot more documents are transcribed and available there is a bigger problem with GEDCOM uploads. Some people have noted that some of these seem to be aimed at deliberately causing problems.
A possibility that this is done to get people to move to the paid platforms. Just my pessimistic view on it. I had people use some of my details of living relatives without permission on a paid platform. Some seem to want to gather as much info as possible without any thought on the correctness of the data.
I support a request to disable the upload of GEDCOM files without a review. If a review process is enabled this will at least assist with managing the mass addition of duplicates and incorrect links.
4 -
As someone who also spent hours merging and repairing work done by multiple GEDCOM uploads, I also 100% support not allowing GEDCOM imports to enter the collaborative tree.
If this continues to be allowed, please develop a way to restore the tree before the GEDCOM import took place. We are able to UNMERGE; we should be able to UNGEDCOM. We need a way to restore the collaborative family tree before the GEDCOM upload happened.
My heart is with those working on the 3000+ mess. It's so sad thinking of the efforts of so many buried under a GEDCOM import!!
5 -
I add my support to this request to stop GEDCOM uploads, AND practices that are just as bad or worse...the same problem happens with 3rd party synchronization and the census projects run by the RLL (where hundreds of thousands of names from census records are programmatically added to Family Tree).
There really isn't a good reason to do mass additions of any kind to Family Tree. People think they're saving time, but they aren't. Mass additions are virtually always fraught with errors and duplicates. Any time saved is more than cancelled out by the many hours required to clean up the messes that result.
There's also a tremendous opportunity cost. Those of us who spend hundreds of hours cleaning up these messes are painfully aware that we would rather have spent our time more productively--for example, finding and adding valid names to Family Tree.
It's one thing if people want to upload a GEDCOM to a private tree on a site like Ancestry. But these types of additions absolutely should not be permitted in a shared tree because the duplicates and errors cause serious problems for so many other users.
5 -
This issue has been going on for at least 7-10 years.
I am cleaning up a huge mess - from one gedcom uploaded by a single person, containing almost every one a duplicate. This upload was last year. Not the first one I encountered since there were several ones but last gedcom mess I came across before this one was in 2018. This one is a nightmare, hundreds and it's German one, a big change from the past nightmares which were mostly New England.
1 -
Yes stop it.
4 -
Hi, I posted a comment here on YouTube. >>>
How to Upload Your Family Tree to FamilySearch.org
I watched this video about the time it came out, but was new to FamilySearch service. I was interested to see what she had to say. I went back to look when I saw this thread and although she is a very experienced and knowledgeable genealogist many viewers (myself included) are not. So I posted the polite comment about being very careful and how it can cause serious problems. It was deleted almost immediately. I noticed that the comment about 100 names had also been deleted.
1 -
Hi I uploaded 1 gedcom tree due to wanting to move from Ancestry.com to FamilySearch.org
1) no notice that it goes into a live tree - expected it to be separated with option to import new data (no warning signs to read up on pass chats about Gedcom uploads) - it shows as an easy steps - 1) upload 2) compare 3) view 4) add people..
2) the GEDCOM file was a download from Ancestry.com It show no matches (after i did the compare tree) - and I expected most to be matches and would have taken the matches from FamilySearch as correct - I was stuck between a half in tree and half out - so i added names (thinking i am adding 1 person - not a whole family) - did not understand how it works - and with no access to help (ps no chat available for South Africa - all other countries have a chat)
3) the gedcom area does not give any support (have been trying to get hold of a DBA/Engineer from Familysearch.org to have a look at the data (was told there may be a problem with Ancestry.com gedcom files as it did not show any matches, but instead created duplicates
4) to add a person - it does not only add the person, but the rest of the family - no warning sign or notice that it not just the person that you want to add, but the whole family
5) the facility does not even have a search function - with the tree disjointed - creating more problems - with a page that give 10 names at a time - finding any new person to be added - next to impossible
6) it does not move from the first page of the list - if you close the page and reopen - it jumps back to the first person - (so you cant see what was done and what still needs to be done) - or have a function to say completed/verify
7) no notice of root magic on the upload of gedcom files as a better option (first time i heard of it was above and will looking to it)
8) made a suggestion on community page - with a request for FamilySearch to come back to me - nobody so far..
anybody that have access to FamilySearch staff - Engineer - if you can please get them to come back to me and look at what can be done
2 -
Great example - I am glad someone is showing examples of the process.
1) no notice that it goes into live tree.
Um, sure it does ... https://www.familysearch.org/search/genealogies
It seems obvious if you are using the Compare feature - you are comparing it with the 'live' Family Tree.
Furthermore - if you select from Already in Family Tree you are presented with comparison of GEDCOM (on left) to that person in Family Tree (on right):
this is quite explicit - though I do agree perhaps more attention/direction could be shown.
what is not clear:
if you do select to update/save the individual from the GECOM (on left) exactly what it will replace - NOTE: it appears only those items with -> REPLACE will be replaced.
This does seem to indicate non-destructive replacement of items mentioned by FamilySearch's Jimmy Zimmerman:
2) 31882 of 37793 is way more than half
3) Support is always available. The ? icon> Contact Us
4) It shows you what you are going to -> REPLACE
I did not try nor see an option for -> REPLACE of the Family - so I would only expect the family to be replaced if you went through each OR if there were a ->REPLACE where it lists SPOUSE and CHILDREN.
5) No search function - this is as designed. If you are going to add persons through this upload process - it wants you to do so one profile at a time.
6) can't help you there - I had no one to add. Only if you add someone would I expect the indication to change. Obviously you shouldn't be replacing persons Already in Tree without knowing what it is going to replace.
7) Refer to the video link above in my response.
8) here is a response - I am a Community member (not a FamilySearch representative).
0 -
NOTE: Above image for Invalid and Living - it looks like upload of these IS prevented.
FamilySearch Image moderation has blocked part of my above response. In the video (below) Mr. Zimmerman mentions how the Genealogies> GEDCOM COMPARE process does so [paraphrasing: safely and non-destructively]. So I for one am a little befuddled by the thread in this conversation ... IF the process is safe and non-destructive - what are we talking about here? It would seem that IF the problem is not Genealogies> GEDCOM COMPARE then it probably is the process through one of the 3rd party tree management solutions - which Jimmy mentions DOES have bulk/batch capacity... Either way it does require the user to process through the profiles and explicitly indicate they are 'replacing' the existing profiles. To do so maliciously/destructively should earn a stern talking to from FamilySearch representatives. They have promised in the past (sorry no quote or reference) to throttle/limit accounts which make bulk edits/changes - especially malicious ones. So reports of bulk edits should be referred to FamilySearch and they should handle that...
The blocked image in question is a link to FamilySearch's Jimmy Zimmerman's Rootstech 2022 (or earlier) video (obviously this image moderation makes no sense?? No referral to internal resources??):
https://www.familysearch.org/rootstech/session/add-an-existing-family-tree-to-familysearch
0 -
FamilySearch's Jimmy Zimmerman - in the above referenced video - indicates the preferred use case for Genealogies - [paraphrasing again: ... new to FamilySearch that has created a GEDCOM file ...]. Why would anyone want to prevent upload of new/good information into Family Tree?
So in reference/contrast - this thread may be talking about someone with a GEDCOM file that already has contributed to Family Tree OR is possibly REPLACING/Updating (yes potentially maliciously) profiles that already exist in Family Tree (contributed by someone already). The implication from Mr. Zimmerman's statement - is that once someone contributes a GEDCOM file initially - they SHOULD continue working in Family Tree OR with more experience know what they are doing with 3rd party tree management solutions.
Either way - whether a new user or someone more experienced with a GEDCOM file - the Genealogies Upload/Compare process with Family Tree appears to be much 'safer and non-destructive' than potential bulk edits using (referenced above) 3rd party tree management software. My comments are meant to help this thread focus the thread/argument appropriately - and indicate there are current solutions to 'malicious edits' ... report them to FamilySearch and let FamilySearch take care of the problem (as they have indicated they would in the past).
That such possible 'malicious edits' can cause heartburn (or worse) - I completely (well at least very well if not completely ... I am in sympathy) understand. That is why I suggest Ideas to FamilySearch to restrict open-edit ... especially for the most recent 4/5 generations (the people in your tree whom you are most likely to personally know something about) and which you most likely would not like to see 'malicious edits' made to their profiles.
I'm sorry to take this thread afield - but my view of Ideas is that such are opportunities for expression from differing points of view ... what may start as an Idea in one place may profitably be taken in another direction to another more profitable end. ...
It is clear from Relatives at Rootstech (and other such similar campaigns/features and applications) that common ancestors and their descendants or near relations can relatively (punny) easily be identified. Since this is the case I would propose (yes this is controversial) that open-edit of the most recent 4/5 generations be restricted to those identifiable accounts/profiles with such identifiable relation to a common ancestor. There is no need for open-edit privilege to the entire world for which there is no near relation. There is enough disagreement sometimes between near relations as to edit of Family Tree profiles - why broadcast such disagreements worldwide - such that a worldwide audience can comment? IF such 'non-related persons' believe they have edits to contribute to a Tree profile they could be given that opportunity to 'submit for edit' through Discover Pages or some yet to be proposed/developed path. Open-edit could remain beyond whichever limiting generation would be selected (4/5) - as collaboration typically is not as much of an issue for further generations - although I can see the need to expand restricted open-edit further as some lines may be more well documented. All I mean by these proposals - is that I would not like to see my nearest relations' profiles destroyed by persons unfamiliar with the line or unknowingly merging with similarly named persons. I find that 'bad merges' are potentially much more destructive in Family Tree than most other actions (excepting outright wholesale replacement).
0 -
There is no need for open-edit privilege to the entire world for which there is no near relation.
Here I strongly disagree.
- Over 99.9% of the world tree has no direct descendants contributing to Family Tree. Said another way, less than 1 in 1000 persons living today has contributed to Family Tree.
- When/if those direct descendants come to Family Tree and start putting in their ancestors, I want the tree to be ready to connect them with their ancestors as quickly as possible, before they spend much time creating duplicate profiles and re-creating fantasy family trees.
- Much of the world tree has no descendants: children who died in infancy, young men killed in wars, adults in religious orders and other walks of life who had no children.
- Many ancestry trees stop at brick walls. One of the best ways to deconstruct brick walls is to work from the far side: do descendancy trees of all families that could be the missing ancestors.
- Going even bigger, and achieving the greatest efficiency and common good, worldwide surname studies generate many trees (ancestries and descendancies) and put them together.
The only reason I work on Family Tree is because it is open-edit. Open-edit is what makes a world tree possible.
5 -
Very good points for open-edit (you are correct some of my argument is probably short-sighted on these points. But my argument is quite focused - whether that focus was communicated or not.) - my points are meant for those living family that do have 'near relations' in the Tree. For obvious reasons those decendants don't want their relations mis-represented in the Tree. Also you might be making my point somewhat ... if you are working more than 4/5 generations back on the Tree - it does little to affect the 4/5 generations I mention ( I don't know maybe about the last century to several previous centuries?) - depending on a generation in a tree. yes I understand you are descendancy researching as far as you can). Whether ancestry or descendancy researching - the main point that I believe you have agreed with me is that a correct representation (not fantasy) be produced.
A question: if ancestry research reaches a brick-wall - because there are no more records - and descendancy research starts further back but in a different location (obviously because records might exist there) - won't that descendancy research run into the same brickwall or gap in records from the other side?? I'm not seeing how the descendancy research finds a way through the brick wall - if the records don't exist. Certainly one can hypothesize a connection (in either direction) - but until there is a record one cannot prove such connection (and even then the records may be so 'thin' that such an assumption may be misdirected). Certainly there are very old collections which may include well-documented persons but connecting 'living person(s)' to them may be limited and not so common?
What I am 100% certain of: I have ancestors that lived and I am a descendant. But I may not be able to find them all - there aren't records for ?50-75%? of them (or at least not 'commonly available. That FamilySearch has lines extending back further than I can substantiate - hopefully means they have sources that I cannot seem to find. What are those sources?) Is it going to be profitable for me to pick a person - I don't know around 1600-1700 and start researching them - hoping they join up to my ancestry? Or should I start by finding the records of the ancestors that I do know and build a base from there to start 'jumping off points'? I don't know - I feel like ancestry research is what I mainly should do first - and then once I get back a ways - do descendancy of my brick-walled ones ... then if I get done with that (nope not done here) maybe think about trying more ancestry work... that's just my preference.
Say 99.9% of people living joined FamilySearch and started contributing - who would they start by contributing? I think they would start with 'near relations' and then try to link into Tree somewhere. That there are those very few (such as dontiknowyou) trying to connect the Tree for everyone before they have a chance to enter their 'near relations' - that would be an infinitely smaller group than the ~70-90% who are focusing on 'near relations'? Yes, I typically do not have the focus of creating the Tree for others. I don't randomly pick a name of interest in a 'fruitful record collection' and start descendancy/ancestry research. There are some few extended/non-connected trees I may have contributed to - but by-and-large they are connected relations and mostly near. I think those that do have that extended/non-connected focus have legitimate reasons to contribute - but perhaps their Tree contribution may also contain 'fantasy' - that only when a 'near relation' gets involved will be able to point out. Fantasy should certainly not be the mode of operation for anyone contributing to Family Tree (nor am I saying that is dontiknowyou's intent) - but certainly ancestry research has no more of a monopoly on 'fantasy trees' than descendancy research.
Perhaps dontiknowyou - and myself are on 'opposite ends of the Tree' - and both have legitimate reasons for our point of view and contributions to Tree. I appreciate dontiknowyou's perspective about Tree contributions that are not the typical way in which I contribute. From my interactions with dontiknowyou - I would rank his/her (i don't recall) Tree contribution abilities above/beyond my own - but that doesn't negate my perspective also.
0 -
@genthusiast oh golly your response is off topic do you agree or disagree that Gedcoms Shouldn't be allowed. ? I stubbled on to another one that was uploaded in December 2021 and only now just recently has been merged by someone else not the person that uploaded the Gedcom and please note no documents attached to the profile either Profile 😔 Ishe Yacov Katz ha Cohen GD89-4KX
1 -
@Roz14 yep, that's a big cluster of duplicates. However, it appears only some of the mess was added using the GEDCOM upload tool. Which is the point some of us are trying to make here: GEDCOM is not the problem.
The problem is users with not enough experience making mistakes. I have seen and merged far worse tangles of duplicates. As Family Tree contributors most of our experience comes from making mistakes, and for us the solution is easy: avert our gaze and wait. Stop watching that part of the tree and let those who care enough to upload a GEDCOM realize they made a mess and fix it. There is no rush.
1 -
@genthusiast actually in most cases I am coming in after most of the near relations trees have already been contributed to Family Tree. Most contributors very soon hit brick walls, and stop. So for one surname there are, say, 10,000 profiles and 5,000 trees, a tree being any profile with the surname and no parent.
A surname study tree is like a severely pruned nursery seedling: very tall, very skinny, with lots of lateral branches cut short. It is specially prepared to be grafted into other trees. As a rule this grafting involves finding duplicates and merging them.
We all tend to think we are working in isolation, and our family is no one else's business. That is true but only to a very limited extent. The farther back you go, the more other people have, individually, an equal or greater interest than you do. So for example here is the ancestry tree of someone I'll call Bob:
Bob is the guy at the base, in the small grey circle. Circled in Bob's ancestry tree is just one brick wall in the 6th generation that has frustrated genealogists for decades. Let's call this guy with no parents Jeff. Now look at Jeff's descendancy tree:
Jeff is in the center. His descendant Bob is on the rim, at the bottom. From this tree we can see that Bob's own living descendants are not the only family historians stopped by the brick wall at Jeff. Jeff has many descendants. (In general, any ancestor has far more descendants than most descendants have ancestors. That is because everyone has 2 parents but historically most people had far more than 2 children.)
In addition to Bob's family there are easily 100 other families also stuck at the same brick wall. Much of Jeff's descendancy is incomplete; most of those men and women who appear as dead ends in Jeff's decendancy tree probably do have descendants; we just haven't made the connections yet.
All of this is no tangent. On the contrary, this is why I like that FamilySearch allows uploading of GEDCOMs into Family Tree.
0 -
@dontiknowyou OK so your point of view is correct and so is mine I just worded it differently. And no, I can't avert my gaze when I am using Family Search Daily and see these terrible twists in the Tree. Do you Agree that Gedcom shouldn't be allowed to be uploaded so freely by inexperienced uses?
0 -
Do you Agree that Gedcom shouldn't be allowed to be uploaded so freely by inexperienced uses?
No.
0 -
@dontiknowyou so do you have a solution at all? Or should we avert our attention when more people go to private trees and just use family search as a database for documents.
0 -
@Roz14 yes in fact I have a solution, actually several.
My personal favorite solution is very simple: work on Family Tree, to make Family Tree where it is happening.
Yes, I know many people just come here to take away historical records and pieces of Family Tree to use elsewhere. I used to be annoyed to find chunks of my own trees uploaded to Genealogies as someone else's work. Now I just ignore Genealogies. I am so far beyond that. Once profiles have sources attached and are polished, to include attaching all direct relations and attaching their sources to them and polishing them too, rarely does a GEDCOM upload make any dent on them.
Years ago a genealogist friend told me what a nightmare his progenitor's tree was on Family Tree. I checked it out, and yes it was a mess. It was a huge tree but almost no sources had been attached and no profiles had been polished. Dates and places were not standardized. Duplicates were everywhere, breeding like mice. Edit warring was constant.
I stepped in and began working from the top down, following only the progenitor's surname. Of course descendants with the progenitor's surname are a tiny minority of all the progenitor's descendants (see Galton-Watson process), but the good results from attaching sources and polishing soon became evident to other contributors and now, years later, that progenitor's descendancy tree is is looking very fine. And it is enormous.
Here are just the first 4 generations:
1 -
I've only peripherally encountered the results of GEDCOM imports. The vast majority of my merging efforts are due to legacy profiles: all those little floating branchlets resulting from old extraction projects, that FamilySearch for some unfathomable-to-me reason "seeded" the tree with, basically a decade ago now. So in that respect, I agree that GEDCOM import isn't by itself the problem.
The problem is that the compare-and-add step of the import process is severely flawed and full of miscommunication. It doesn't show you what's already in the tree, floods you with false matches so that you have no chance of noticing the real matches, and fails to find matches that are clearly there. And all of this is after a series of steps that never once mention anything like "collaborative" or "open-edit" -- they talk about "the Family Tree community", but that doesn't convey anything about the singularity of the tree.
It's like putting someone in a room full of what look like wicks, handing him a match, and telling him to go ahead and light the lamps and candles -- but failing to tell him that some of the wicks are actually fuses leading to sticks of dynamite somewhere in the distance. Now, if he crouches down to the floor and peers through the cracks, he can discern which ones are fuses, but since nobody's told him of the possibility, what are the chances of him doing that? Naturally, therefore, he sets off some distant explosions, but they're undetectable in the room he's in, so the cleanup is left to other people who stumble across the mess, sometimes years later.
3 -
oh golly your response is off topic do you agree or disagree that Gedcoms Shouldn't be allowed. ? do you have a solution at all?
Sorry you feel I'm off topic - I believe it is on-topic - but you are certainly welcome to request a fork from moderation if you so wish. I disagree. solution: GEDCOM is a tool - tools should be used responsibly - and the system should allow 'safe use'. There possibly could be changes to further compare/restrict merges into Family Tree - mainly dependent upon 'already in Tree matching'. I dislike merges of non-sourced profiles which 'overwrite' sourced profiles as much as anyone (it seems like Compare/merge should safely preserve sources) - but without knowing more specifics I couldn't comment as to where the fault lies. From what I saw when using the Compare process - the process appears safe. I would need to investigate further to have more of an opinion - but I certainly sympathize with whatever happened to replace sourced profiles with non-sourced. I am not looking forward to more such open-edits of my near relations - that is why my Ideas ...
I understand that Jeff's descendancy (cousins) could be much more numerous ... I agree they will have an interest in finding his parents. I also agree that nearby trees may possibly have some 'association' with him/his family (one never knows where a source may turn up).
0 -
@Wanda Botes, I am a FamilySearch employee but do not work directly on the Tree Import/Merge feature. I can't promise that I will be active on this community page, but I'm happy to forward your suggestions and issues to the team that works on the Gedcom upload process. Please feel free to message me directly.
In reading through the responses it seems like @genthusiast answered most of you concerns pretty well, I'm also not sure about #6. That seems particularly bad because it might encourage a user to merge more than one person at a time to avoid losing their place.
Below are some of my opinions about the FamilySearch Tree import tool:
The Tree Import tool is good for importing a personal Gedcom (e.g. from Ancestry, My Heritage, etc). This seems useful on its own as a place to store and reference at any time.
The part of the Import tool that helps identify individuals to merge or add into FamilySearch is also useful.
The actual Merge process works well for users who know the Gedcom data well and are experienced with merging persons in FamilySearch. However, the merge process isn't a great fit for someone new to FamilySearch or not expert with the data being merged. The problem is that merging two trees of data can be very complicated, and merging into a system that you are not familiar with magnifies that complexity.
The Merge part of Import should be used with great care. Here are some Safe-Guards and things to consider.
- FamilyTree Import does not allow a Mass import (Other tools may do that.) The FamilySearch FamilyTree Import Product only allows users to add people to the FamilyTree "One at a Time".
- Persons from GEDCOM that match records in the tree are considered duplicates and can't be added.
- Living records and records with too little information also can't be added.
- High confidence matches are not allowed to be added.
- Possible (medium confidence matches) Duplicates require the user to state that they aren't a match and add them anyway. (Not a good idea in my opinion, if your not an expert with the data or are new to FamilySearch)
- Low-confidence matches won't shown as duplicates in FamilyTree during the merge process and should not show as duplicates after the merge. It's still a very good idea to manually check these for duplicates. Sometimes, persons can't be matched well because there isn't enough data provided. If you are an expert with the data then you should be able to pick that up before adding the person into FamilyTree.
- Merging can take several steps especially when relationships are involved. If you look at the result in the middle of a merge sequence the data can appear very broken, especially to others observing the process. (This is one reason why it is much better to just merge one individual at a time)
I'm a huge Fan of the "Open Edit" FamilyTree. I think it is one of the big things that makes FamilySearch great! I think it's really important that we all work together on the public FamilyTree to build and improve it. I've watched FamilyTree data get sooo much better over time. I've had the experience of having to fix other people's data and I've also had the experience where "my data" was not correct. There are periods of time when my own FamilySearch FamilyTree has been in a very "broken state". However, it has always recovered and become more accurate in the end.
I believe in this shared environment we need to be more patient than we might otherwise be with each other and let things settle. If you are the one merging new data in, then it would be good to work with others as much as possible. If you are viewing the changes that appear in a broken state then message and reach out to the person who is making the changes. If you find they are in the middle of a merge process , then just be patient and let the person complete the process.
1 -
At the risk of continuing this tiresome 'war of GEDCOM' I have to disagree - hopefully the following will demonstrate why...
I have experienced what I would consider a few 'malicious edits' - atleast non-collaborative ... but that is beside the point ... Whether experienced to a greater extent as this thread deals with or not - that such edits occur is because of the 'open-edit' nature of the FS Family Tree platform. And yes I advocate for 'restricting open-edit' - precisely for these 'potential malicious edits' to remove well-sourced ancestral profiles with completely differing profiles - or worse non-sourced profiles (I almost exclusively focus on ancestral lines).
To advocate for 🛑 STOPping all GEDCOM uploads can be logically shown to be a non sequitir (hopefully I am using that correctly. Disclaimer: I do not claim to be a GEDCOM expert.) as follows:
- GEDCOM is both a file specification and data structure which has changed over time. There are different GEDCOM standards/implementations between various applications.
- FamilySearch Family Tree implements a propriety brand of GEDCOM X ->> Family Tree IS GEDCOM.
- Therefore to request "STOP 🛑 allowing Gedcoms to be uploaded to the main tree" explicitly means - you wish Family Tree to cease. Such would stop any further change or development toward being a 'one-world Tree'.
Now I recognize that is not what y'all are requesting ... You folks seem to be more in alignment with my request to 'restrict open-edit' than you realize? You would prefer that Family Tree not have 'bad GEDCOM' merges or removal of 'well-sourced' profiles ( I would prefer that as well). @dontiknowyou and @Joel N Thornton on the otherhand (with much experience I might add - and sorry if I misrepresent anything here...) - says we are misguided in these Ideas/requests precisely because Family Tree is 'open-edit' ... [paraphrasing] these problems are not a big concern ... someone will come along and clean them up. My point on the otherhand is - 'why allow the messes to begin with?' And also why allow a good representation of my 'near relations' to be 'messed' with? If the resulting conclusion in a couple decades or after I die is worse or no better than the current representation - why allow it to change?
Can't all us smart people come together and come up with some idea(s) which allow open-edit/collaborative aspects of Family Tree but also protects the 'good/well-sourced' profile contributions above 'non-sourced' junk we should all be concerned about? My basic proposal is (as I have advocated elsewhere - a hybrid curated/open-edit approach) ... atleast something in the range of the most recent 4/5 generations (those whom from both family lore and research I am likely to know something about) allow to be 'locked/read only' to most of the world (but still allow for edit submissions - but to whom?) but allow priority edit for 'near relations' (descendants/cousins - those most interested) - and yes even allow that group to expand/contract. Beyond whichever generation - the Tree could remain 'open-edit'.
There are difficulties with trying to change the model. And to be sure what is represented in this thread is that the initial GEDCOM came from Ancestry ... and supposedly the upload (via Genealogies) 'maliciously edited' Family Tree. If that is the case then to understand what has occurred I suggest we all try downloading our Ancestry GEDCOM, uploading to FamilySearch Genealogies - and experiment (obviously don't break profiles in Family Tree) - but try to figure out what is going wrong with the process. Only in becoming more knowledgeable about what is going wrong can I further comment on a recommended solution (and likely even then I won't have enough expertise to suggest much other than better COMPARING/matching...). It seems an obvious solution would be similar to what occurs in Merges - not allow overwriting of 'well-sourced' profiles (have all GEDCOM merge with 'Already existing in Tree' profiles to go one way - not allow overwriting - since as I understand Genealogies strips GEDCOM of Sources, Media, etc.). Also reporting 'malicious edits' to FamilySearch - which can with expertise examine said GEDCOM and the changed profiles - should be able to pinpoint the 'problem' and know what the specific solution would be for each specific case... Additionally I believe FamilySearch has promised in the past to throttle/limit mass edits to Tree - there used to be a threshold - and I believe they mentioned banning of user profiles causing 'destruction' or something similar.
GEDCOM is not the problem and it is the problem (when used badly... by a user) ... but which versions/applications ... If you want a solution - become a part of it ... If I were a developer this is one of the 'solutions' I would be working on (to be sure there are a few others I would want to render differently than current Family Tree). I would go out to the FamilySearch GEDCOM Github and develop away... I would be trying to implement a solution that provided these positive changes to FamilySearch Family Tree. We should all want Family Tree to represent true profiles that can maintain their integrity in preference to 'bad merges' - but additionally we should all want ease of upload of new profiles (think internationally - new users for example ... not banning all because there is some bad).
Anyway ... I am sorry if some think my posts irrelevant to the thread - I do not view them that way ... Ideas are ideas ... FamilySearch is content to let these ideas be discussed (within Community Code of Conduct - which by the way is still not functional through the ?).
Hopefully some solutions will be presented by someone if not FamilySearch ... isn't that the point of Community?
0 -
@genthusiast, the underlying code of Family Tree being a flavor of GEDCOM is a red herring.
What people would like to stop is batch processes that have the potential of adding a great deal of junk to the Tree in a short amount of time. There are multiple such processes, but the one that starts with an upload to Genealogies is the one that ends up labeled in the Tree with the "GEDCOM data" text in every reason box, so that's the one everyone knows about and hates.
The question of batch uploads to a collaborative effort is completely independent of how you feel about open-edit. WikiTree struggles with it, and it has fairly significant edit restrictions.
Those of us with experience working in a collaborative tree know with full certainty that in the long run, it is always less work to add profiles one at a time. The messes created by one wrong click in a batch process take longer to clean up than it would have taken to add each person individually, and the fight with getting a batch process to cough up all of the information one wants to see to make good decisions is also time-consuming and frustrating.
The problem is that less-experienced people don't see it that way. What they see is "I've done all this work on another platform and I don't want to do it again. Why can't I just copy all of it over to your platform?" This is why the various import processes exist, despite their inherent flaws.
Because of the trail it leaves in reason boxes, the GEDCOM upload becomes the poster child for the whole problematic concept, and that's why people want to stop it in particular. What we really want is for FamilySearch to answer "Can't I just copy all of my work over all at once?" with a clear and emphatic No.
3 -
@Julia Szent-Györgyi (seems to again want to lecture me against my contributions to Ideas. Julia - until FamilySearch wants to limit my contributions to Ideas - please stop denigrating mine. I have the free right of self-expression - as do you - and I will defend my right to do so. You are not a Community moderator.)
the underlying code of Family Tree being a flavor of GEDCOM is a red herring.
hmmm, red-herring? GEDCOM X, GEDCOM 5... GEDCOM is GEDCOM and has flavors ... yes? This thread says STOP GEDCOM uploads ... which? That's the point I am making - no red-herring about that - and I mention that quite specifically... Yes Genealogies is the main focus but why leave the Idea there to what appears to be a safer process than 3rd party applications (which have more batch processes)? This brings in the idea of 'open-edit' immediately because that is the model of Family Tree. Different GEDCOM versions communicating with Family Tree do so because of this model.
What people would like to stop is batch processes ... Genealogies is the one that ends up labeled ...
Yes, and I am for restricting the process too! Since you appear to be claiming superior expertise then ... what caused the 'problem' in this thread? Can you specifically comment? Was it a batch process? (not so from the discussion - Genealogies doesn't appear to have a 'batch process' and FamilySearch mentions profiles must be edited one at a time...) ... Anyway this goes to my point about FamilySearch is in the position to pinpoint exact nature of GEDCOM COMPARE processes ...
The question of batch uploads to a collaborative effort is completely independent of how you feel about open-edit.
No - open-edit is not extraneous (my feelings are but the model isn't at all. and my feelings comment on the model...which is relevant to the 'batch processes'.) - the FamilySearch FAMILY TREE IS BUILT UPON GEDCOM. Since that is already established (read above) it is very relevant. If 3rd party apps could not successfully batch to Family Tree GEDCOM format - or Genealogies convert GEDCOM upload to a COMPARE process which edits based upon open-edit - what are we talking about??
The problem is that less-experienced people ...
I understand why people want to upload/COMPARE GEDCOM.
What we really want is for FamilySearch to answer "Can't I just copy all of my work over all at once?" with a clear and emphatic No.
FamilySearch has already done so in Genealogies - look at the process - that is what I am pointing out ...
aratyat
0 -
Why allow the messes to begin with?
This is the heart of the issue, isn't it?
The answer is, in a collaborative, open-edit system where everyone is welcome, messes must be allowed and tolerated. We all start out as beginners and making a mess now and then is part of the process of gaining experience and becoming an expert.
Now, I am aware of a few individuals here who pretend to be beginners but actually are vandals. They need to be dealt with individually, and there are processes available for that. If it is true that, as claimed here on Community in the last few days, an individual used the FamilySearch GEDCOM import tool to load 3,000 or 37,000 profiles in a single day, that strongly suggests the individual used a 'bot operating on the FamilySearch web user interface.
Taking away GEDCOM imports or any other tools vandals sometimes use, or restricting use of tools to only sanctioned contributors, would create a social hierarchy and alienate many. And alienation leads quickly to stagnation. Just look at WikiTree.
3 -
Was it a batch process? (not so from the discussion - Genealogies doesn't appear to have a 'batch process' and FamilySearch mentions profiles must be edited one at a time...)
It is entirely possible to write a bot to act directly on the user interface. That is a common method of denial of service attack.
The question I have been asking for days now is: Is that what happened? How many profiles were added via GEDCOM upload in how much time?
0 -
 They need to be dealt with individually, and there are processes available for that.Â
I agree and have stated such. But I don't know the exact nature of what occurred and don't presume to (which is somewhat at the heart of this thread) ...
But if you look the supposed 'perpetrator' appears to have commented ... claiming ignorance ... DDOS is not ignorance. Distinguishing what actually occurred is a first priority.
...messes must be allowed and tolerated. ...And alienation leads quickly to stagnation ...
I think you will agree - messes and destruction are two different animals? All my comments have as a main thrust - prevention of 'malicious edits' to be considered 'destruction'. As such perhaps edits per time threshold will need to be enforced (probably a unfortunate 'alienating' side effect of these situations). There might be a determinable difference between a user 'mindlessly clicking' and a DDOS/bot attack (a computer/bot would probably have more transactions in a shorter timeframe)? What is better alienation of criminals/destruction or allowing them the same access? I think fighting against destructive persons is the only good option.
The idea that messes can be cleaned up is fine and I am glad of that. But it does not persuade me that such a course is preferrable to letting good data take presidence - and not allow it to be changed to a mess to begin with. For this reason I continue to be persuaded that some process in addition to those currently in place - should protect good data. Why should I have to allow well-sourced profiles to be destroyed (non-sourced), reconstituted - hopefully as the same person but possibly as a different person - different families linked in incorrectly - making it so much more difficult for someone to come along and figure it out and re-enter something hopefully close to what was there a decade or more previously? Or if you take the other point of view - am I just worrying without need - the millenium will just take care of it - in which case why continue adding data into Family Tree currently anyway? ... works, effort ... (effort for naught?) ...
0 -
I apologize if my tone came across wrong. I didn't intend to lecture anyone.
The file format underlying the Tree, and any discussion deriving from it, is very much a red herring to the topic that I believe this thread is talking about, which is the process that results in all those messes labeled with "GEDCOM data" in the Tree.
Yes, the GEDCOM compare-and-add process is masquerading as "one at a time" -- but all it means in practice is sequential mindless clicks. FS is constantly adding extraneous clicks to all of its processes (see Search and now Find for prime examples), so having lots of them is really par for the course.
0 -
Apology accepted. It just seems that my 'feelings/themes' on open-edit - get so criticized ... look I'm just contributing to Ideas - no one else has to like my contributions (I do). And I wish Community (and myself for that matter) could better organize to be more persuasive/helpful ... especially on these issues of critical importance.
The file format underlying the Tree ... red herring to ... this thread ...
I don't believe it is - because as mentioned - if the GEDCOM COMPARE processes (from whichever source) could not communicate with Family Tree GEDCOM structure - then we wouldn't be having this conversation. Yes - I don't know all the sources that could results in 'GEDCOM data' label (perhaps someone else does).
... masquerading as "one at a time" -- but all it means in practice is sequential mindless clicks.Â
But the Genealogies COMPARE process only allows -> REPLACE of certain artifacts - no Sources, etc. - from what I saw. So I don't understand why the 'destruction' referred to in this thread - unless as mentioned something more nefarious was going on? What replaced the sourced profiles is my first question - which yes - refers back to GEDCOM version and account log (which only FamilySearch will be able to forensically investigate).
If there are organized DDOS/bot attacks against Family Tree (as there seems to be such increasing destructive network attacks, ransomware encryption, etc. these days - these people just make me sick and tired ... ) I hope FamilySearch can not only block/prosecute them to full extent - but hopefully prevent such from occurring (yes even restrict open-edit) or else restore data loss from backups. Zero trust does not mesh well to open-edit? I wish I could provide better Ideas ...
0
