NEW 'TreeSync' with My Heritage Announced
LegacyUser
✭✭✭✭
Robert Wren said: ICYMI, (I recently discovered what that means) My Heritage & FamilySearch announced a new synchronised TreeSync pilot program between the two websites.
Those eligible users may download their 'branch' of FS Tree into MyH (individual) tree to be able to take advantage of resources in both. THEN some sort of 'semi-auto' synchronized sharing of changes will occur at the users initiation.
Trying not to sound too skeptical, i refer to this topic from last year.
https://getsatisfaction.com/familysea...
This proposal DOES seem to be much improved, but . . .
A code is needed to join the pilot testing of this new feature.
Those eligible users may download their 'branch' of FS Tree into MyH (individual) tree to be able to take advantage of resources in both. THEN some sort of 'semi-auto' synchronized sharing of changes will occur at the users initiation.
Trying not to sound too skeptical, i refer to this topic from last year.
https://getsatisfaction.com/familysea...
This proposal DOES seem to be much improved, but . . .
A code is needed to join the pilot testing of this new feature.
Tagged:
0
Comments
-
Robert Wren said: Here's a blog post from LAST year about the 2018 proposal, which appears quite similar:
https://blog.myheritage.com/2018/03/n... /
This may get thou to the MyH presentation Friday at 1:30
https://www.rootstech.org/salt-lake/l...0 -
Gordon Collett said: A more direct link to the presentation announcing this is:
https://www.rootstech.org/video/getti...
He starts taking about it at around 15 minutes into his presentation.
Last year's beta rolled out caused a few disaster's with the source page in certain situations. I would assume they have spent the last year reworking the entire system to prevent the same kind of problems.
Potentially the most exciting part of this is that by exporting to My Heritage, one can take advantage of their consistency checker.0 -
Jeff Wiseman said: Any synchronizing should NEVER be "automatic" or "Semi-automatic". It should ALWAYS be manually, initiated and limited in scope to a single PID at a time (and within that, a single vital at a time). It is one thing to have automatic mechanisms that can report when differences occur (very, very useful), and quite another to have some pre-programmed response to compensate via synchronization for those changes based on what a single individual has in his own "private tree" (aka, their account at My Heritage at the moment).
The ability for people with private trees outside of FamilyTree to correlate their information easily with FamilyTree can be really valuable (I do it all the time with Ancestral Quest). But it can be very EASY to go a step too far...
It's probably pre-mature on my part, but it seems to me that this type of thing, if not extremely well managed and controlled, has a great potential to become a "My Tree War-Mongering" feature.
I do have some concerns over FS's ability to exercise "extremely well managed and controlled" power over these new capabilities. This is in light of FS's historically apparent priority of allowing large blocks of data to be easily loaded into the FamilyTree via GEDCOM files, over the damage and extreme frustrations that it has continued to cause over the last few years.
I also have reservations about both FS's and MH's judgement and approach to some of these things based on the last attempt a few months back when MH accounts could totally overwrite EVERYTHING in a PID and glutting the change histories by replacing all vitals with vitals of the same exact values. This trashes the synchronizing histories of ALL OTHER third party databases such as Rootsmagic, Ancestral Quest, Ancestry.com, etc. And for some reason it was considered acceptable that the testing on this feature wasn't don't in a beta protected area, but rather in the actual production database!
So please forgive my anxiety but it's hard not to be really concerned over this...
Remember:To err is human. But to really screw things up, you need a computer.
and:The computer has the incredible capability of destroying data faster and more efficiently than any human being is capable of.
0 -
A van Helsdingen said: Will this opportunity be available only to Latter Day Saints only, or to non-LDS as well?0
-
Jeff Wiseman said: By the way, if anyone wants to watch the presentation by My Heritage at Rootstech on Friday at 1:30, you can find it at the 6th hour 29th minute mark on the live stream at:
https://www.youtube.com/watch?v=UVIEp...
By the way, one thing that I WAS really impressed with is that this sync feature that is being rolled out is has been specifically designed for FamilySearch members who are looking to expand their research capabilities into MyHeritage. Although the Sync *IS* semi automatic (i.e., you have to push a button to start it running through your entire tree), if there is a conflict between your MyHertige tree and the FamilyTree, the FamilySearch data will override that in the MyHeritage tree.
Although they didn't get into a lot of detail, it sounds like the mechanism works sort of like this:
1) Initially you create your MH tree by importing a FS tree based on you as the root. This means that everything in your MH Tree initially matches everything in the FS tree.
2) If you change something in your MH tree and the FS tree for that data hasn't been updated since your last sync, when you run your sync, it will copy your new data into the tree.
3) If something has been changed in the FS tree, when you sync your MH tree with it, the data in your MH tree will be updated with that from the FS tree.
Here is where it is interesting! If a "collision" occurs (I assume by collision, they mean both the FS data and MH data have had changes since the last sync), the recent changes in FS data will always take precedence over the recent changes in the MH data thus replacing the MH data with the FS data.
It's nice that in the case of collisions, the FS data remains stable. However, I'm not sure how MH will deal with the fact that changes to data in the MH tree can be blown away by data in the FS tree during a collision.
I'm not sure if my assumptions are correct here, but if they are, it is encouraging.
By the way, it also appears that MH has added "Reason for Change" boxes on everything modeled after the FS structure. Assuming those transfer as well during a sync is also a good step forward.0 -
Jeff Wiseman said: In the presentation (after illustrating that in a data collision during sync, the FSFT data changes will always get priority), the presenter emphasized that this is specifically intended for FamilySearch "members" who want to expand their research capabilities into MH.
I do not know what he mean by "members" (i.e., church members or just members of the FamilyTree website), although it is conceivable that it might have the same restrictions as the current Ancestry.com relationship (e.g., if you want to sync between Ancestry.com and FS you must do it from an Ancestry LDS partner account. MH does have LDS specific accounts if I remember correctly.0 -
Jeff Wiseman said: That's probably a better reference than the 11 hour one I provided. I tried to find one on the Rootsteck site but I must have missed it.0
-
David Newton said: Automatic and semi-automatic sync should NEVER happen with FSFT when things are posted to FSFT. Automatic gathering of what the changes are? Certainly. Automatic or semi-automatic application of those changes? Absolutely not. Each change should need to be done manually. Will that take longer? Yes. Do I care if people moan about it taking longer? No, except that they should be told to go away if they moan. Doing each change manually can still be fast and efficient if the user interface is correct. Doing each change semi-automatically would have far too much potential for breaking parts of FSFT or introducing vast amounts of junk into FSFT. Same reason for stopping GEDCOM imports.0
-
Robert Wren said: Thanks for the additional analysis, Jeff. I'll admit I was a bit 'stunned' on hearing the presentation on Friday. I think you have presented it somewhat as I remember. I had hoped there was more a one person change at a time but wasn't sure and I hadn't reviewed the video.
I think it proposed 8 generation (with descendants??) and I don't think the gave an option for less. 8 Gen is about 1,000 people (parents) plus a large quantity of descendants. - So we are talking a fairly large group here -and consider how many might take advantage of this offer now - that could be a huge amount of potential changes coming in a short period of time in the future. https://www.tamurajones.net/AvgNumOfN...
A better PILOT program might be done with a user selected 2 generation sample from FS rather than a '1,000's of people' sample.
When discussing this with a couple others, their first comments were 'It looks like MyH wants more data!'.
I agree with David Newton that it should be one at a time, not bulk.
BUT, it's all being done by a computer, so what could possibly go wrong, go wrong, go wrong, . . . . . ...
The Pilot code to do this is available, I'm reluctant to post it0 -
Jeff Wiseman said: They didn't feel it appropriate to have it live streamed or recorded yet so I wouldn't post it. They will make it available when ready.0
-
Jeff Wiseman said: Again, if this really works as they say with changes to FSFT taking priority during syncs with MH the main problem that will occur will be on the MH side (and as I said I have no idea how they will handle it).
Say a person imports his or her part of the FSFT into MH creating a new tree for themselves. If they were then to immediately go through and put inappropriate changes into their tree and then immediately do a sync, all of those changes would be updated in the FSFT (assuming nobody was working on those PIDs at the same time).
Now, since everybody is using watches on people they've done significant work on :-) They spot the incorrect changes right away and go in and reverse them with reasons nicely documented. That might be a bit of work, but the NEXT time that person does a sync, all of that CORRECTED data will now be back-hauled directly into that person's private tree in MH regardless of whether or not they had done further changes in those same areas since their last sync!
Although it doesn't require examining each PID during the sync (that would have happened during the time the person was updating their tree in MH), the fact that ultimately updates in FSFT will work themselves back into everyone's personal trees at MH seems very promising!
This kind of synchronization can be complicated with a myriad of potential gotchas and other side effects (apparently they've been working on it for a year now), so I hope they can get it stable before a full release of the feature. I can think of a couple ways that this could be abused (e.g., loading GEDCOMs into MH prior to a sync) so I also hope those things are being considered.0 -
joe martel said: The only way to sync MH to FS is via the user's private tree being loaded from FS to start with. No other way to upload is allowed into that MH private tree, except by the user doing it one at a time, so no Gedcom import is possible. Let us know if anyone has seen any behaviors that are concerning.0
-
Robert Wren said: Joe, would you happen to know if there is any limit to the number of trees that contain an specific PID could be transferred into MH?
For example, say my 6gg father is a Mayflower ancestor; could a thousand descendants transfer 'their' tree branch to MH. As 'everyone' seeks 'famous' ancestors, that might create chaos in potential FSTree changes.0 -
joe martel said: Really wouldn't be much different using a record manager.0
-
Robert Wren said: So there are no limits? I assume a record manager is e.g. Ancestral File or Roots Magic? If so, I don't quite understand the correlation.0
-
Jeff Wiseman said: Joe, that looks like a great way to "keep the holes plugged"
There was an odd occurrence reported just 6 days ago. Whether it's related to this or not is hard to tell, but it is really odd enough that it should be investigated if you haven't already:
https://getsatisfaction.com/familysea...0 -
Jeff Wiseman said: Robert, just for reference here is the "perimeter" diagram that was given. As you can see, it is about the 18 minute mark into the rootstech video at the link that Gordon Collett gave above in his reply to the topic.
I don't know if the sync is totally limited to those in the "perimeter" as shown or if the scope of the sync can be extended by adding to what you have imported from FS
0 -
joe martel said: Sorry, I was referring to the last sentence and that if a bunch of users pull down the same famous ancestor tree they can each push their changes in a record manager in a similar way to MH.0
-
Stewart Millar said: Not related to the discussion content - just a glitch on the chart above - the child of a "Great-Uncle" is not a "2nd cousin once removed" but rather a "1st cousin once removed"
(children of all uncle/aunts at any level are always "1st cousins" removed by the respective generation gap)0 -
Jeff Wiseman said: Nice catch :-) I wonder if anybody pointed that out after the presentation?
Of course it also brings into question where the real "Perimeter" limit that they import into MH actually is since the GRAPHIC itself shows that only 1st cousins and 1st cousins once removed are included. Per the graphic 2nd cousins are never included, but per the text 2nd cousins once removed would be included.
Well, we should know in a few weeks :-)0 -
randystebbing said: The tree sync feature (still in beta), when released, is for LDS church members who have a MyHeritage LDSAccess linked account.0
-
Michael W. McCormick, AG® said: Does anyone know where the help file is?
Is there a way to tell MyHeritage after I import from FamilySearch to import another generation or person on a line?
If I add a 9th generation on the MH sync tree will that be ignored in the sync as out of scope or override what is already in FamilySearch?
Really I would like to be able to bring in a more distant collateral line into my MH sync tree even if I have to import each additional name one at a time. That way the people I add from MH can be easily brought into FS on that collateral line I’m working on. Is is possible?0 -
Tom Huber said: This is an old discussion and dealt with a feature that was being introduced at the time. Any requests for help with the current situation can be posted in one of the discussion threads dealing with the latest problems of the MH sync feature or better yet, directly to my heritage.0
This discussion has been closed.