Follow Us
Be a Fan
Geni Forum » General Discussion
Killing Zombies
(68 posts)-
Member
Login to Send PM
@noah: "... Profiles outside the importers max family (4th cousins and closer and inlaws out to first cousins) are marked as public profiles
"One way we could deal with most of the zombies is to retroactively apply these rules to all GEDCOM imports."
I hope you are no longer contemplating applying this retroactively. Are you?
Posted 7 months ago # -
.
Posted 7 months ago # -
Mark, maybe Geni isn't planning on doing it anymore, I just found this post:
http://forum.geni.com/topic.php?id=59131&page=2#post-548001
So I guess if your profiles are marked as living, you are probably okay.
Posted 7 months ago # -
Member
Login to Send PM
@Aditya: Thanks. It certainly seems like they abandoned the idea.
It would have been proper to state it here, where the idea was floated explicitly, as well as that other thread, where the idea was first proposed. Geni can not expect us to read each thread, especially when the name of the thread has little to do with what Geni says in the thread.
Posted 7 months ago # -
Geni Team
Login to Send PM
@markm that is correct. Based on feedback, we have decided to only convert deceased profiles that don't have a manager in their max family group to public profiles. We are not going to convert living profiles. We will need to find another way to deal with 'zombies'.
Posted 7 months ago # -
Member
Login to Send PM
@noah: I'm very pleased with that decision of yours. Some people - not me - were also concerned about keeping private the deceased parents of living profiles. Have you decided what to do about that?
Posted 7 months ago # -
Geni Team
Login to Send PM
@markm In most cases I imagine that these would be in the max family of a manager, so they would remain private as well.
Posted 7 months ago # -
Member
Login to Send PM
@markm that is correct. Based on feedback, we have decided to only convert deceased profiles that don't have a manager in their max family group to public profiles. We are not going to convert living profiles. We will need to find another way to deal with 'zombies'.
@Noah: Is finding another solution to deal with zombies a high priority for Geni? Do you have anything at all in the works?
I'm looking for an optimistic "stay tuned" here.
Dave K.
Posted 7 months ago # -
Geni Team
Login to Send PM
@dkaleita we're actively discussing it. It will move up in priority after we release our privacy fixes. The more I think about it though, the more I wonder whether just changing them from living to deceased will really fix the issue. Isn't the real issue that they are managed by users who are not interested in merging them?
Posted 7 months ago # -
Member
Login to Send PM
Isn't the real issue that they are managed by users who are not interested in merging them?
Not typically. Of my 300+ collaborators and Family Group members, every one I've ever contacted who created zombie profiles had no idea that one or more of their profiles were incorrectly marked as living until I pointed it out to them.
I could understand if Geni didn't want to change the living/deceased status on profiles that have only the original manager and no other co-managers or collaborators. But if someone accepts merges and collaborators, then they should have no objection to having the deceased status of their profiles auto-corrected.
In any case, forcing people to omit dates of death or deceased status from profiles solely to ensure that the profile doesn't get merged is just crazy. If Geni wants to provide a separate field to prevent a profile from being merged, that's fine I suppose. But using the living/deceased field for this unrelated purpose is just wrong.
Dave K.
Posted 7 months ago # -
Member
Login to Send PM
@dkaleita: I never heard anyone in any thread say that they force dead profiles to be living just to avoid merging. Where have you heard this?
My main concern is protecting the privacy of my profiles, even the far distant (I wish there was a better name to call these profiles) profiles. If they are marked as living, they ought not to be forced public.
@Noah: I agree with you, the real issue is figuring out better ways to merge. Part of the problem is the users who don't wish, or don't bother, or don't know how to merge. Another part of the problem, I feel, is that merging on Geni is a cumbersone, time-consuming activity. I have stated frequently on other threads, that Geni ought to be shifting its focus to merging entire trees, rather than merging profiles.
Merging trees ought to be some sort of batch operation, where entire trees are compared for matches, and a manager can select N of them for merging in a single stroke. There are many ways of actually implementing this, I am sure we can come up with creative ways.
Noah, Mike: Any thoughts?
Posted 7 months ago # -
Member
Login to Send PM
@dkaleita: I never heard anyone in any thread say that they force dead profiles to be living just to avoid merging. Where have you heard this?
I thought that's what Noah was implying in his post just above mine- the one that I quoted.
Posted 7 months ago # -
Geni Team
Login to Send PM
@markm we're working on some bulk merging tools, probably much along the same lines you are thinking. We need to make some back end changes first though, and we can't do those until we make our upcoming privacy changes. The good news is that merging should be faster as soon as we make our privacy changes. We have a lot of great things coming up this year, I'm excited to roll them out.
Posted 7 months ago # -
Member
Login to Send PM
@dkaleita: You may be right, but I doubt it. I have the feeling that what Noah meant was that the same folks who don't take the time to kill their zombies, are also the same folks who don't take the time to respond to merge requests. So even if the zombies are killed, those same folks will still be slow to respond to merge requests.
I guess we ought to let Noah explain what he meant, though.
@noah: After all of this wrangling (and bickering), I hope Geni will open up a new thread to discuss its latest privacy change proposals before any implementation.
Posted 7 months ago # -
Member
Login to Send PM
I have the feeling that what Noah meant was that the same folks who don't take the time to kill their zombies, are also the same folks who don't take the time to respond to merge requests. So even if the zombies are killed, those same folks will still be slow to respond to merge requests.
@markm: I'm sure that you're right about that. But my point to Noah is that doesn't necessarily mean that those people don't WANT to merge. I think that, if they just weren't so busy having a life outside of Geni, they probably WOULD want to merge. In that [likely] case, killing their zombies for them would be doing those people a favor.
Posted 7 months ago # -
Member
Login to Send PM
@dkaleita: I'm not so sure I agree with you. My guess is it is a mixture. There are certainly people who do not wish to merge, I know because I have some who feel that way in my ownfamily.
The point that I, and others, have tried to make over and over again is that all opinions and all needs ought to be respected.
Posted 7 months ago # -
Member
Login to Send PM
<p>I would like to make a proposal to Geni, to kill off a great deal of the zombies, and enhance the historical portions of the trees, as an alternative to what Noah proposed above:</p>
<p>I recommend that Geni automatically (and iteratively) do the following:</p>
<p>1. All parents of people born before 1900 be converted to "deceased".</p>
<p>2. All children of parents born before 1800 be converted to "deceased".</p>
<p>3. All siblings of people born before 1850 be converted to "deceased".</p>
<p>4. All spouses of people born before 1850 be converted to "deceased".</p>
<p>If the above five steps are repeated iteratively, I believe that will successfully kill off a great many zombies.</p>
<p>Thoughts?
</p>There's no need to do it iteratively. Your sole condition is birthdate, and your algorithm never changes a birthdate. Your first pass would kill a good number, but your second pass wouldn't kill a single zombie.
We need additional conditions taking generations of offspring and generations of offspring of close relatives into consideration.
Posted 7 months ago # -
Member
Login to Send PM
@erinspice: You are absolutely right. Even my later seven step suggestion is not iterative.
"We need additional conditions taking generations of offspring and generations of offspring of close relatives into consideration." - Any ideas along these lines?
Posted 7 months ago # -
Member
Login to Send PM
I have even more ideas: a) Not just parents, but ALL Ancestors of those born before before 1900, can be converted to zombies, and b) more consideration given to spouses and their family.
So, Here is my latest offering:
I recommend that Geni automatically do the following:
1. ALL ANCESTORS of people born before 1900 be converted to "deceased".
2. All siblings (and the spouses of their siblings, and the ancestors of their spouses) of people born before 1850 be converted to "deceased".
3. All spouses of people (and the ancestors of their spouses) born before 1850 be converted to "deceased".
4. All children of parents born before 1800 be converted to "deceased".
5. All children and grandchildren of parents born before 1700 be converted to "deceased".
6. All children, grandchildren, and great-grandchildren of parents born before 1600 be converted to "deceased".
7. Etc. for each century going back as far as needed (i.e., add one more descendant generation for each century we move back).If the above seven steps are performed, I believe that will successfully kill off a great many zombies.
Thoughts?
Posted 7 months ago # -
Member
Login to Send PM
I like your list of rules...some comments:
#1 - For this rule ANCESTOR needs to be clearly defined. Let's call the person born before 1900 person X. I think ANCESTOR for X should include:
- all grandparents of X
- all great aunts, great uncles of X and their spouses as well as the ANCESTORS for these spouses
- all aunts and uncles of X that you know are older than the X's parent. This would need to be determined based on the "birth order" data. There is a chance that person X has an aunt or uncle that is younger than X is.#4 - #7
- I would expand these rules to include the spouses of the children, grandchildren, etc as well as their siblings. Once you've flagged these sons-in-law and daughters-in-law as deceased you could then apply rule #1 to them.
- I think a distinction needs to be made for the birth year of the father vs. the mother. What is the oldest that a mother can have a baby...let's say 50 as a rough guess. According to wikipedia the oldest person to ever live was 122 years old
http://en.wikipedia.org/wiki/Oldest_people#Ten_oldest_people_everSo say a 50 year old mother gave birth to a child 122 years ago, that would put the mom at being born in 1838. So for rule #4 say "a children of a mother born before 1838 be converted to deceased". For rules #5, #6, #7, etc step back in 50 year blocks (or whatever # is agreed on as the oldest age a mother can give birth) instead of 100 year blocks.
Then repeat rules #4 - #7 but based on the birth year of the father and use 100 year blocks since males can become parents at much older ages.
Posted 7 months ago #
Reply »
You must log in to post.