Proposal for a new system of curators for Historical Profiles

We've got only a week or two before the 'septic tank bursts' (nice analogy Shmuel) so I guess we merge like crazy before that! (without making silly mistakes of course).

Sigh.

Terry, merging like crazy BEFORE hand, won't really help at all. If anything it MIGHT make things worse.

The only thing "gained" by cleaning-up in advance, is that you'll have a cleaner tree covered in mud, rather than the usual mangy sort of tree. This might actually backfire, because for the novice, completing their SPECIFIC pending-merge in a large stack might be harder, than with a single pending-merge.

So really, just carry-on as usual (Tea, dear?) and ride the "wave". We tried getting Geni to see sense. They refused. Ce la vi. I understand not wanting to change development schedules for work already in progress, but I think it would have been better to simply hold on to these features and not release them just yet, until the damage they will cause is minimized or containable!

Two months after the previous wave, caused by the Top-100, I'm about a week or two, from FINALLY cleaning it up.

I predict that this wave will be BIGGER, because while many/most people already have a balanced number of mutual two way collaborations. Not so the managers of the main lines. Most of the power-users are ACTUAL managers in these lines, and they generally collaborate with more people than collaborate with them. This will let a disproportional amount of NEW people to edit these profiles.

Shmuel I don't really understand, what exactly do you think will happens since you fear we will "sink in shit to our necks"

Günther,
what happened LAST time, when Geni introduced the Celebrities / Top-100 feature, was that suddenly it became "too easy" for ALL the users of Geni to find "popular" profiles in the "main" lines. A small fraction of these users, could and did complete several merges all over the tree, between profiles managed by people they were connected to, and the main lines.

They did NOT first ask the owners of the other trees if they had any interest in this merge. They did NOT follow-up and also complete the hundreds of merges that were the result of their first merge.

The direct result of this was that multiple copies of entire trees were now connected at a single point (too late for 'undo') resulting in literally thousands of new duplicates IN the tree, where before there were none.

In this first two DAYS this feature was available, FIVE trees were merged with the main Bible tree that I maintain. In the first two weeks, that number rose to twelve, and in the two months since, about 20! Most of these managers have since joined my Family Group, but not all.

Making Collaboration two-way, will greatly increase the size of the "small fraction" of users who can innocently mess us up this way. There are at least two hundred MORE copies of the Bible Tree on Geni. How many of them will be forced on me this time??

According to Noah, the NEXT change after two-way collaboration, is the "conversion of [historical] profiles to public [shared]", which will make the situation ever WORSE.

This is precisely WHY we need Curators "before" and NOT after this type of change. While the idea of tree-moderators / curators has been talked about before, it was these recent changes that really made this a major issue.

I guess Geni is a firm advocate of MBM ("Management By Miracle") AKA "don't worry, if there's a crisis, we'll have a solution". I never expected that type/level/kind of devoutness in California.

I forgot to add a link to my description of the previous mess, that I wrote for the forum: http://forum.geni.com/topic.php?id=54211

Sounds "interesting" hope w'll survive
When is this implemented

Noah Tutak@Noah-please, please take Shmuel's concerns into consideration as they are shared by many of us who have invested a great deal of ourselves in this project and who are yours and Geni's strongest supporters at the moment. I know that you have a timetable for unveiling the new mutual collaboration and public vs private profiles features, but PLEASE consider the sequencing, and I urge you to implement some system to protect these vulnerable Historical Profiles before opening the gate to allow anyone and everyone to begin merging these.

I know that I, for one, am eager to be a part of the curatorial group, but if this feature is delayed until the problem has multiplied even worse than it is now (and it's definitely at crisis point now), I know that I personally will not be able to afford the time and energy to devote to such a cleanup effort. If you lose your core of dedicated volunteers like myself, I think it will be a major management error. I urge you to reconsider your sequencing.

I strongly support what Shmuel and Pam are saying here - please do things in the right order!

The Historical profiles need to be protected before allowing more people to add to the Big Tree.

For examples of chaos, see one example here, of an area that certainly do NOT need more profiles added/people with edit/merge rights to create more chaos:
http://www.geni.com/discussions/6000000007416019584

@Gunther Kipp, looks as if it's planned for one to two weeks time.
If the mess gets messier I'm not sure I'd have the skill or the time to resolve the issues and yet I don't really want to abandon Geni...
@/Noah, please throw us a lifeline!!!!

Don't know if anyone will be able to access this but would like to share my idea for the proposed Historic Profile and invite comments if youcan see it.
http://assets1.geni.com/photos/p9/1392/8395/53444837a67b069e/Histor...

Terry, It looks like a good list of tools. Doubtless someone could make them operable. However, the question remains, how does one keep out misinformation, sloppy work , disagreements, and so on? How do the curators arbitrate?

As a relative newbie, frankly, I do not know how one looks at one's tree, once you have merged with others. How do you print out your own line?
How can you see the people when the lines are stretched so far?

I have about 800 names, and so many lines coincide with the ninth and tenth generations! I cannot visualize how such a tree would be constructed when all the lines are linked.

It would be nice to have the option of an unlinked version available. And to keep control of my content for a chosen length of time. I prefer to keep the option to edit open to me for as long as I need it. There may not be so many advantages to an outright merge if one cannot be selective about how much is merged and how much can be reworked if necessary.

These are just a few of the issues I am aware of. My time period rarely extends beyond 1550, and hardly includes royalty and titled nobility. My interest is only in history and data.

Just curious and a question from a neophite...

While the new Collaboration policies are being implimented...

Should suggestions be made now about which "HISTORICAL PROFILES' should be protected??

given that I don't expect to be a "curator", but still want to edit my ancestors back to the 1600s (most of which are NOT of historical interest - they were farmers and fishermen), I would make a proposal for defining a "protected" group:

1) Start with Shmuel's "Adam" entry, @Adam / אדם הראשון / آدم . . (or a set of seeds, "Adam" being one)
2) For everyone in the group, add all his/her parents, children and partners, as long as the following is true:
- They are born before <arbitrary cutoff date, such as 1600>
- They have > 10 (or some other number) administrators, indicating that they're the result of a long-drawn-out merge work. (another way would be to look at actual merges in their history tree, but this is not public info)
3) Repeat step 2) until the set stops growing.

Then we should have a connected set that is "protected", but does not include profiles where only 1 or 2 people manage them. It can also be computed in a limited number of passes over the database, something Noah probably finds important.

After the changeover, curators should then have the ability to:
- Move a profile to the "Protected" set, if it is connected to at least one "protected" profile (thus keeping the "protected" set connected)
- Remove a profile from the "Protected" set, thus allowing all collaborators to edit and merge it.

I don't think it makes sense to have automatic addition of entries to the "protected" set once the curators group is established and effective; it violates the Principle of Least Surprise for anyone who hasn't read this discussion.

Noah Tutak@Noa, I have bumped this up on the forum as I really think it's worth serious consideration.... I've yet again (the umpteenth time) come to Empress Mathilda of England to find her in a complete mess. This is getting VERY TIRESOME!!!! We really need some protection for the historic tree and for merges to simply be merged and not duplicated. Another measure that would help considerably and not just with the historic tree is some kind of warning sign for vulnerable profiles. It would be especially helpful if these appeared in the stacking window as well as the tree node.

Pretty please?

Terry Jackson (Switzer) as I've said before, there is merit to the proposal and I believe we'll implement at least some aspects in the future. Right now we are working hard on the new, simpler privacy rules that have discussed in the forum, as well as some other great new features. This is definitely in the discussion though.

Let's also think from a Genealogical way too..what criteria does a curator have to make these calls and do they have to be certified or can anyone do this---also how is the information verified and sourced? Just asking folks!!

i don't know about certified - but certainly sourcing, documentation, and a good history of work would be necessary.

example: i documented a confusion between two pairs of individuals last month, and when I went back yesterday, they had been recombined by an inattentive user, someone who is active on geni, but who just merged similar names and dates without reading the documentation.

I don't think that person should be a curator. maybe I shouldn't either....

(offtopic)
actually this points out something I've been missing on Geni functionality - some place to put information where it WILL show up CLEARLY during the merge process.

At the moment, the only place to put warnings is in the name, and that makes the tree look weird.

Harald, the earlier comments in this discussion do in fact suggest identifying "Historical People", i.e. profiles to protect/curate in very much the same way that you suggest (as does the "formal" proposal that followed this discussion in the forum: http://forum.geni.com/topic.php?id=58161#post-535471 ).

I just have to say that I'm very much in support of a Curator/Moderator system that limits the mess surrounding HPs in the Big Tree.

I must, however, stress the importance of having an easy way to suggest modifications, additions etc. though, and everyone should definitely be able to add comments on these "locked" HP profiles.

Also, it is very important that the vast number of profiles from the 1600s that are only managed by one (or a few) manager(s) are not taken away from these people's control. It would be very annoying to have to go through a Curator to make additions/changes when the profile is not really contested. These profiles should not automatically become locked HPs without the current manager's consent.

The implementation of such a system also needs to provide a solution for people (unaware of the Big Tree) who add profiles for HPs and then later decide to join the Big Tree. How would these mergers be done? Should they just delete their duplicates or merge them into the official profiles in some way?

Maybe one solution for people considering to join the Big Tree would be to give them an option to "overlay" their own tree on top of the "official tree", so they could see how well these trees corresponded. Geni would then warn them where differences occured, and suggest possible duplicates. Then if they wanted to they could replace these sections of their tree with the "official tree" if they wanted to. In this manner one would avoid a forest of merge issues waiting for approvals.

cont. from previous post.

... maybe users should be allowed to keep their own parallell tree on top of the "official tree" with their own profiles, just visible to them?

They would have one version of the profile on which they could remain in complete control, and also have the possibilty of considering the official HP profile. These profiles would be linked, but only you (or possibly a group of people sharing access) would see your own private version of the profile, and it would not disturb the appearance of the Big Tree.

Eivind,
1) the full proposal already enables people to make comments and suggestions about historical profiles. In fact you ALREADY can do that. Just start a PUBLIC discussion and tag the profile you want to comment on. You can tag a profile either by typing a @ followed by their name (a pop-up should appear to allow you to select it), OR the more trustworthy method of putting their Profile-ID within double square brackets, like this [[ 5098603059080070794 ]] without spaces will give you ==> Shmuel-Aharon Kam (Kahn / שמואל-אהרן קם (קאן This will then enable anyone who manages the tagged profile, to see the discussion.

2) "Historical" profiles with only a few managers, are also addressed (no change). But really, how many of these ARE there going to be? Very few people except for VIPs had records kept for them that far back. Even the non-VIPs are going to have hundreds of thousands of descendants in modern times.

3) As to how the actual merge will be done, that too is covered: The manager would have to request it, and a Curator [mostly using the present tools], would do the merges, making the new copy part of the HP.

The idea of the overlay has been suggested, by the counter-Curator proposal, but you would STILL need Curators to manage that as well. Also at what point does something become an overlay rather than the profile itself? The main lines of the historical tree BECAME that way precisely because people kept on merging into these lines. Two years ago, there was nothing unique about this copy (out of hundreds) of Adam of Eden, but now it is "main" with 56 managers (real number of merges much higher).

The BIG advantage of the present Curator proposal, is that it requires very FEW changes from the existing system, making it easier to implement. This lets Geni move ahead faster for everybody, instead of making major changes. What this proposal mostly suggests is ways to remove blocks and prevent damage to the main lines.

As a 'worker' on the historic tree, am I the only one concerned with Geni's proposal to have an automatic accept merges for my(chosen) collaborators feature?

Since merging has been enhanced by being able to select a collaborator to complete the merge that has been great but I have encountered errors by some very trustworthy collaborators that I would choose to trust and also would prefer to think that colleagues would check my requests because like everyone else I am human and make mistakes. It looks like they will go ahead with this whatever the consequences. I am already finding increasingly complicated knots where we once had cleared up and made it quite clear where to put certain profiles when merging so what mess will we end up with when this process is automated without any checking?

The more features like this that are proposed the more I wish they'd already implemented the curator proposal!

Please check out this thread on the forum
http://forum.geni.com/topic.php?id=63081&amp;page=2&amp;replies=27#...

Give it up, the Geni model just won't work.

Curators are here! :-D

http://www.geni.com/discussions/6000000009728208335