This is a project to create a global family tree using user-provided genealogy details. This would either be a new Wikimedia project, or would be combined with the September 11 memorial wiki into Wikipeople.

A typical entry would have the form 1stPartofName_2ndPartofName_etc_PlaceofBirth_DateofBirth. Every entry would be linked with the rest into a full site-wide tree.


Could also be called WikiTree, a better name than GlobalFamilyTree.



A good idea.


More clarification on this idea: The entry naming format would have to be specified and fixed. The details above were chosen simply to ensure unique entries (how likely is it that two people were born on the same date in the same place with the exact same name?).

More importantly, it is quite likely that there are people in a family tree whose place and/or date of birth are unknown. I'd suggest a more ad-hoc structure, where disambiguation takes place as necessary.

Agreed. Small problem with the above structure is that many people (even famous people) have death documented but birth date unknown. Robin Patterson 02:43, 8 January 2007 (UTC)[reply]

A typical entry would consist of:


Name edit

  • Given name (at birth): Jakub
  • Middle name(s): * Family name (at birth): Dvořák
  • Unique Identifier: I0001

(I in the Unique Identifier is for Individual, following the GEDCOM convention)

Father edit

[[I002|GREG JOHN MAYS]]

Use the UID as the page address to avoid (inevitable) duplication.

Mother edit

First Middle Last - City, State, Nation - YYYY-MM-DD

Sibling 1 edit

First Middle Last - City, State, Nation - YYYY-MM-DD

Spouse edit

First Middle Last - City, State, Nation - YYYY-MM-DD

Child 1 edit

First Middle Last - City, State, Nation - YYYY-MM-DD

Sources edit

  • Personal communication with Child 1
  • Birth records of City, State, Nation
  • LDS database
  • etc ...

Information edit

Details of life, birth date, death date, accomplishments, etc.


There would need to be some sort of overall "site map" that would allow the entire link tree to be viewed. The problem here seems to be ensuring all entries contain the same information in the same format. Maybe mandatory templates?

--Joeljkp 14:34, 24 May 2004 (UTC)[reply]

One of the main advantages of a wiki is its lack of mandatory templates.

It's an advantage only in certain instances. In this case lack of templates would lead to a chaotic mess. Certain facts would be required to make this all work. Example: failure to correctly list the person's parents would make it impossible for the site to take on the "family tree" format. The pages would quickly digress into a scattering of personal bio pages. In certain cases (orphans, abandoned children, unknown father, ect) the tree would be incomplete. But listing only a mother or father would be better than ignoring the fields altogether. Perhaps, for orphans there could be a "non-genetic" parents linking mechanism. Placing a little emphasis on the social form's of the family tree as well. For those people trying to determine actual bloodlines the non-genetic parents options could be toggled. (assuming this evolved into something big, even my lazy non-family tree oriented self tingles at the though of seeing a family tree of all mankind.)

Think of it more as required information.

Also in the terms of the "global view" or planetary family tree form, it would be a good idea to have the visual tree display toggle between patriarchal and matriarchal forms. (Male or Female bloodlines)

One last idea. The family links could have a "truth" percentage. Example: I am 99.999999% sure that I am the offspring of my mother and father (as always there is the tiniest chance, however remote that any one of us could be the milk man's baby, but highly unlikely).

When viewing my family tree I would find that the link between me and my parents at near 100% certainty. And between them and their parents also near 100%. In the case of questionable family ties the percentage would be much lower. Example: My great, great, great, great, great (ect) grandfather was John Doe born somewhere in Unknown Village in Europe. Could John Doe be part of your family line too? Sure but the certainty factor increases as you go back. I could be only 50% sure that I'm related to John Doe so my family tree certainty percentage might decrease significantly after him. Also the percentage could cumulatively decrease (automatic adjustments by the wiki) the farther back I go. I could have a family history spanning several thousand decades but the tops of the family tree would not be as certain.




This'll be so cool! I've got a whole bunch of stuff to put in. One of my great-grandfather's cousins was a genealogist, and that family can be traced back for dozens of generations. (Which brings up the issue of legendary genealogies. Some of my ancestry is described in the Icelandic Eddas, but they're not exactly considered reliable historical documents.) My grandmother on my other side has taken this up as a hobby, and I've got dozens of cousins (and their spouses) to add. So I have both depth and breadth! ^_^ -- Toby Bartels 17:58, 24 May 2004 (UTC)[reply]

Re: template idea - I suppose citing source materials would be quite important to avoid mess, especially when there are conflicting information. Tomos 23:06, 24 May 2004 (UTC)[reply]


I hope all of this will be put on a Wiki of its own, or as an extension of the 911 Wiki as Maveric suggested. If this will be put on the Wikipedia and become really popular it would mean a mess, flooding the recent changes list, etc, and making it even harder to keep the servers perform. If on a separate wiki it can be mirrored separately etc.

It would also make it easier to develop a custom interface for that wiki, e.g. form based for the basic facts. Entering data via a form would make it really easy to store the data as xml or wiki tagged, which would greatly enhance the possibilities of automated processing, e.g. generation of graphical familiy trees. Erik Zachte 00:46, 25 May 2004 (UTC)[reply]


I strongly recommend to do that whole stuff en:gedcom-compatible! And I would appriciate an import-function, so one dont have to save thousands of person each.


What are some good ways to deal with people who change their names over their lifetime, as in the case of marriage? --NeuronExMachina 07:41, 19 Jul 2004 (UTC)


Excellent idea! Very keen on geneology myself. Concerned at how we identify unique individuals. There may be a low chance of two John Smith's being born in the same town and on the same day - but the chance exists nonetheless! In practise, when researching one's family tree - often you know the name of an ancestor but have difficulty finding exact date or place of birth information. A quick scan of http://www.genesconnected.co.uk will demonstrate that the database will soon contain many people with the same name about whom only some details are known - they may or may not be the same person.

To answer the question from NeuronExMachina above, geneology is always indexed by the surname you were born with first.

I would also like to echo the anonymous comment about gedcom. Anyone who has done much family tree work on a computer will almost certainly have all of his research in gedcom format. --Steveguy


I started a sourceforge project called FamilyTreeWiki and created a detailed wiki format for genealogy data before seeing this site. I think many of you will be interested in it. You can find the project page here: http://familytreewiki.sourceforge.net/

The wiki has templates built in for creating new family trees and personal profiles, so users don't have to get too bogged down in the details of the format to enter data.. most people won't want to read a bunch of documentation to just enter data for their immediate family. Soon, I will write open source software that will import GEDCOM files into the wiki and export data from the wiki in GEDCOM format so it can be imported into any genealogy application (to generate charts, etc).

An example of a family tree can be viewed here: http://zwei.homelinux.net/familytreewiki/

Comments and suggestions are welcome. http://zwei.homelinux.net/familytreewiki/UserFeedback

-Jason Dunsmore


You also have to think how you're going to deal with ... how do you say it... bastards. MPS


With MoinMoin, there are a variety of means to block spammers. http://moinmoin.wikiwikiweb.de/SecurityPolicy

-Jason Dunsmore


If you're interested in this, please visit WikiTree.org. Still being improved, but fully functional and in need of contributors/community. --Krubo 16 May 2005



Rodovid.org - MediaWiki based engine for everybody. One global DB. Access to db from any language. Tree for any person in db can be viewed immediately. --195.137.203.26 17:22, 2 February 2006 (UTC)[reply]


CoNcE Proyect - a nice graphic engine for show genealogic trees, and a simple interface to introduce your familiar history.


Mention should also be made of the Genealogy Wikia, "GFDL" and supported by well-known wiki experts Angela and Jimbo: less structured than WikiTree or Rodovid and therefore of interest to people who like to use their own formats. Robin Patterson 02:43, 8 January 2007 (UTC)[reply]

Update: Wikia sites migrated to CC-BY-SA soon after Wikipedia did. The Genealogy Wikia is now named "Familypedia". It incorporates Semantic MediaWiki, which allows for semi-automatic linking and a variety of displays for individuals and for the places where people had documented events. It is free to use and supported by advertisements. Data entry is partly by forms but contributors may add anything in the free text areas. The resulting articles have a Wikipedia-style appearance, with an infobox for the individual's details and commonly tables in the text for siblings and children. Subpages detail ancestry and descendants.
Robin Patterson (talk) 02:48, 7 February 2017 (UTC)[reply]

I really like this idea. I love Geneaology and would certanly contribute to this project.


The best implementation I've seen of this is WeRelate.org.


Merge proposed: See Wikipeople.


This would be a project that would cause sleepless nights for the Ombudsman's office.


I believe gedcoms are the wrong way forward. Commercial websites already exist which use gedcoms formats and anyone interesting in this should try them out. The problem is that this would allow mass uploading by people who will never look again. Who will clean all that up? Secondly, surely one aim of a genealogy wiki has to be to provide a way that people can compare speculations and probable connections, whereas a gendcom format is going to encourage people to always pick between "black and white". In other words, I believe the wiki has to be based upon entries concerning individuals or even individual units of evidence, while a gedcom demands that a person be defined by a network of relationships which are by implication proven.


It appears that the specific term "Global Family Tree" is already under copyrighted use here: Global Family Tree.


I'm the founder of WikiTree.com. It's a highly customized implementation of Mediawiki, but it's Mediawiki and you'll recognize it. I'm also the creator of WikiAnswers so the customization style is similar. If anyone has questions, comments, suggestions, or insults I'd enthusiastically welcome them. E-mail me (chris@). WikiTree 22:07, 11 August 2010 (UTC)[reply]