Tree Swarming

The theory is fine, but we need the tools to fix things first.

I have not been into a situation yet where people disagree on the solution, - it is just impossible to correct it because of all the unaccessible zombies where you are not even able to see the relationship or contact the manager, and all those abandoned accounts blocking the lines.

agree. once a deceased node is merged into the big tree all its ancestors should be changed to shared. and should be easier to break situations where they are married to themselves.

Geni have made some fundamental errors in the new privacy rules when not even a mother have access to make her the profiles of her own minor children as private just because they are public profiles and is not collaborating with the inactive manager.

Please bring back the unlimited family member access, - public profiles or not!

Swarming sounds like a really bad idea. Many times while working on a tree, someone has come along and remerged problem profiles I was unmerging. I have found It's best to work on an area alone. There are many hoops one might need to jump through to get it looking alright and another person might not know what you are up to. Also if too many people are working on a tree together it scews up the works and then an error pops up and one can't access the tree at all. I have a few trees I have been working on for a while now trying to get relationship and generations fixed using the best information available. It might not be perfect but after it is looking good, then anyone can go back and tweek it. I think those that get my merging requests know which areas I am working on and I stay away from those I know others are working on . Others can have the ruling families! : )
I have found sometimes unacceptable zombies that are attached to accessable profiles belong to people unaware that they are zombies and since they can't see them, they don't think they exist. Those that I can request to see never get back to me, so when it is feasible, I just send them to outerspace where they belong. Things look a lot better afterwards. Until all are accessable.....looking forward to it.

yes my grandmother shouldnt have to ask me to collaborate when she doesnt even know what that is!

@Jason - Shared/Public profiles have become the problem. Today it is easier to merge profiles if they are marked as private. Earlier we could ask family members to perform the merges, but not today. You have to be a collaborator even if the profiles was added by your inactive brother/sister.

I think I tried something in that direction with this discussion:
http://www.geni.com/discussions/6000000006934472055?msg=60000000085...
But as Bjørn says one keep bumping into Zombies, uncooperative managers and
mysterious errormessage you don't understand the meaning of!!
And if you ask GENI HELP they don't get the message you do!!

Noah,
this is the EXACT opposite of what we need. As it is presently impossible to communicate with users in real-time, multiple people working the same area would only get in each others way. I have had this happen to me with a single other user, and it was CRAZY until I realized what was going on. Luckily I knew them and they agreed to work elsewhere for a while. Never mind that not everyone is quite as careful as others. So more people would be completing more bad merges.

If we can agree on what the tree looks like, we don't need a swarm of people to fix it. Rather one such person would be best for the job...

I have a counter proposal, how about Geni periodically selecting a certain part of the tree and killing off all the zombies in it.

Hopefully this will all get better when they implement this:

http://forum.geni.com/topic.php?id=63711

Can we change it to a Zombie Swarm and get rid of problems that way--a report would be good for that too!

The biggest problem I run into is zombies and other private profiles attracted into a merge stack by the "resolve errors" tool. You can't see that a profile is private in "merge-compare" view. You don't know that an account is inactive in "merge-compare" view.

So I agree: the bigger problem is killing the zombies.

Once that is done we can organize ourselves into zones by historical period / family line to patrol, keep clean, and not step on each other's toes. At the moment I kind of can tell someone is working on an area by the merge requests or their comments in public discussions, so stay away unless we have agreed to specifically work together on it.

I am beginning to worry that until the zombie issue is better resolved on the database and programming side, all user driven initiatives eventually become useless.

Erica,
I agree that the "merge-compare" view IS lacking in these areas you mention. FYI, though, if you click the person's name while in that view, it will take you to the specific to-be-merged profile.

Hi Shmuel,

You are teaching me well but I'm one step ahead on that one. I have learned to study other managers profiles almost Talmudically, trying to ascertain their collaborators (if any), their activity on the site, and their likelihood to respond nicely to a nice note asking them to make profiles public.

Why can't there just be one "flipped switch" for deceased / private profiles? I kind of hate to see Noah doing it one at time.

I'm sure Noah will tire of it himself sooner or later. :-) He will then either automate the process or delegate it...

They could do a one time flip, but unless they shut-down the entire site for a day or three, they would STILL miss quite a few, considering that they are shooting at moving targets. :-D

Would it really be a moving target? The policy going forward is "all deceased profiles above your third great grandparents should be public"

http://forum.geni.com/topic.php?id=69361

The problem with that policy statement is that it says "all DECEASED profiles above...".

The source of the real problem are the [some of the] GEDCOM files being uploaded into the system. Due to limitations in the some of the programs that created them, many of them have thousands of "zombie" profiles - people who ARE dead, but not marked as such in these files, because they lack a year of death. So while they HAVE improved the "filters" to catch these, they don't get them all. The "big-tree" grows at a rate of about 150-200 thousand profiles every day. I presume the overall growth rate of Geni is about double that. So more zombies are being added at any given moment. This is the "moving target".

Many users, certainly the less active or less tech-savvy, don't even know they have zombies. By way of non-critical example. You manage 7000+ profiles. At least 1100 of them are marked as alive. Is your tree really that wide, as to include so many contemporaries??

Do you mean me, Shmuel? I've never imported a GEDCOM.

My profile
Profiles Managed
Focus person: Erica Isabel Howton
Group: Managed By
Filters: Living

(Update list button)

Yup, 1144 living cousins, probably 2/3 with the last name "Howton." They all have a birth date of 1900 or later (I went through that with a fine tooth comb a couple of weeks ago and caught 20 profiles that way). Many may be deceased but I haven't found death dates for them.

They were all entered by hand by me based on census data. I've had no requests to merge anyone named Howton -- and it looks like I have all of them in my tree. :)

Erica,
I wasn't referring to you specifically, just as an example. That's an impressive tree you have! I've seen plenty of people, who manage 10-15-20 thousand profiles, who had many thousands of zombies. Geni actually did do a pretty good job, all considered killing them off, a few months ago, but the problem itself still remains.

Shmuel,

I took it as an opportunity to describe to everyone how they can check for themselves for their own inadvertent private profiles. I've written to collaborators with those instructions (learned from Mr. Lancaster) and they've thanked me, written back saying they had no idea.

Checking (and correcting) your own grouping is a whole lot faster and easier than individual zombie hunting.

BTW those 7000 plus profiles are mostly my American / English side. Sadly on my Jewish side, I have 50 or so living, 15 or so deceased / public. I (and other genealogically inclined relatives, so I don't feel so bad) just can't make the connection to 19th century European records.

Erica - Another way to look for those zombies is to look up ancestors rather than managed by. This will pull a larger list with a variety of managers