The Yggdrasil Project

Transcription

1 The Yggdrasil Project Project Charter Copyright 2013 Christoffer Owe

2 Version Date Changes Initial version

3 1 Project Scope and Deliverables The idea behind Yggdrasil is to create a Wikipedia of genealogical information a database where anyone can enter or modify information, but with a strong focus on sources and reliability. As Wikipedia has become the world s leading encyclopedia, the ultimate goal for Yggdrasil is to become the leading collection of genealogical information. 1.1 Challenges with Genealogy Today Until the 70s all genealogy were done using pen and paper. In the 80s the first applications arrived, and the genealogist could finally store his data in a database with search capabilities. After the World Wide Web was created in the 90s, more and more genealogical sources have become available online. And it s also much easier to find other genealogists working on the same topics as you, even in other parts of the world, using different kinds of social media. Although an impressive increase in efficiency has been achieved in genealogy during the last 30 years, there are still several challenges: New researchers are redoing a lot of work that has already been done before as the works of previous (serious) researchers are not available. Even if all previous work was easily available, there would be more than enough genealogical tasks to work on. Most newer bygdebøker (Norwegian local history books focusing on genealogy) are written by experts who know the available sources better than most amateur genealogists. The result would therefore not necessarily be better if all genealogists go straight to the sources themselves instead of reusing what has been done. On the other side, the bygdebøker doesn t contain source references, and they do contain errors. Middle age genealogy is not for amateurs, as a deep knowledge of the sources is required. This should be left to the experts and those with a special interest in the topic. But it s hard for everyone else to keep up with the latest research. The same questions are being asked again and again in forums with the risk that the experts get tired of answering. And errors disproved years ago are still being spread across the Internet. Even if the research has been published in a journal, it is hard for amateurs without special knowledge to keep themselves up to date. Many people have relatives and ancestors abroad. Genealogy abroad is usually both more expensive and more time consuming, which may effectively prevent research in these areas for many genealogists. In addition, language may be a show stopper, as well as proper knowledge of the relevant sources. Local genealogists will probably be in a much better position to do a good job. And perhaps someone has already done the job? Both the Norwegian local history books (bygdebøker) and other publications contain errors and deficiencies. Some are corrected in newer publications, but most corrections are too small to be published in the traditional way. Probably, many such corrections and additions are being stored in each genealogist s private database while other genealogists will have to make the same discoveries on their own. There are several large databases (as for example Geni and myheritage) with the goal of collecting and combining as much genealogical data as possible. But, as many have pointed out, these are almost worthless because the quality is unknown and source references are, at least in most cases, completely lacking.

4 1.2 The IT of Genealogy The market consists of Type traditional desktop applications web-based applications web-based services online solutions with the goal of combining uploaded data into one big tree Examples TMG, Legacy, Family Tree Maker PhpGedView, GeneWeb MyHeritage, MyFamily.com (Ancestry.com), Family Tree (FamilySearch.org) Geni, Werelate.org, WorldConnect (Ancestry.com), Family Tree (FamilySearch.org), wikitree.com All of these solutions/applications focus on the conclusion part of the genealogy. In addition there are several databases with online source information available but without any lineage linking. Examples are Ancestry.com, FamilySearch.org 1 and the Norwegian Digitalarkivet 2. All the existing web-based services, perhaps except wikitree.org, support Gedcom-import. The idea is that everyone may upload their database, and then the databases can be linked together. Consequently there is a massive job of combining the uploaded trees and merging the persons. And uploaded Gedcoms always contain a lot of errors and rarely source references. The value of those large databases are thus very limited. Some databases with no Gedcom import, good source references and lineage linking has been created and some are planned. They all have a limited scope though. Examples are: Íslendingabók is a database of all inhabitants of Iceland. It is however only available for citizens and inhabitants of Iceland and it is not available in English. Nasjonalt historisk befolkningsregister is a Norwegian effort of creating a lineage linked database based on the online church records and censuses available through Digitalarkivet. 3 It doesn t seem to support other sources or older data. The linking will partly be done automatically and partly by volunteers. kinpedia.net (formerly wikitree.org) also has the goal of creating one unified tree and seems to be the project most similar to Yggdrasil. It is currently closed down for reconstruction. There seems to be no existing solution so far which addresses all the issues discussed in section Requirements Note: The list is not comprehensive and will be expanded during the project. All information must always be free First, all information entered by users must always be public domain [check available licenses to prevent misuse of the information]. Yggdrasil is not the place to store your own private tree. Secondly, all information must be freely accessible, i.e. without any required membership with or without payments. No 1 Note that both Ancestry.com and FamilySearch.org also offer other services, as mentioned in the application table

5 Everyone may add information Event-based data model All changes to genealogical information must be logged Both person centric and source centric registration of data must be supported subscription will be needed to access the information. This introduces a risk which must be mitigated. However, this works for Wikipedia, and Wikipedia would not have gained its position if contributors needed some kind of approval. The data model should not have a family entity as in Gedcom. Instead children are linked to their parents using parent-child relationships and spouses are linked together using marriage events. 1.4 Scope Although the solution must be continuously improved, the following concepts must be implemented before go-live: User management, including functionality for anyone to register a new user Basic functionality to register all data Support for multiple concurrent users Functionality for an administrator to block users and, if necessary, reverse all changes by this user. Logging of changes Functionality to view all genealogical data in a simple and easily understandable way Functionality to view all changes and who did the change and when Interactive ancestor chart The list is not comprehensive. 1.5 Implementing and Maintaining the Solution The project consists of the following phases: 1. Pilot 2. Implementation 3. Testing 4. Go-live Pilot Phase The pilot will be a fully working minimalistic solution for entering data. The pilot will be published on web, but write access will be limited to selected users. Data entered in the pilot will be used for testing purposes in the implementation phase. The data entered should be real and only real genealogical data, and it should be comprehensive enough to cover most combinations of genealogical data.

6 1.5.2 Implementation Phase This phase will expand the pilot with a focus on the following areas: User interface Change log Multiuser support Proper search functions for persons and sources Support for entering source data Functions to merge persons and sources Other functionality discovered Test Phase For the test phase several volunteer testers are required. The testers should be familiar with the general objectives of the project Go-live Go-live in this context means general availability, also for entering data. Anyone may register themselves as a user and enter data. The implementation project is finished by the go-live, but the solution must be continuously improved and updated Application Management An organization is required for supporting the users and handling of any technical issues. Additional functionality should be prioritized based on feedback from the users and the long-term strategy. See section 3 for how the application management should be organized. 1.6 Data Registration After go-live, anyone may start a data entry project with the goal of for example entering as much data as possible for a family or a geographical area. If possible the Yggdrasil project itself should include some data entry subprojects. This will of course depend on available resources. 2 Risks There are two groups of risks: Risks that may prevent or delay the implementation project, and risks that may prevent people from using it. The main risk of the first type is the initial dependency of one developer. It must be a goal to create a team of developers, architects an web designers, but it is unknown when this can be achieved. Either in has to be achieved with volunteers, or there has to be a sufficient income. For the success of Yggdrasil it s very important that people find it useful enough to enter information and that even more people find it useful for reading information. Risks that may prevent this are: Risk Bad user interface Too much garbage in the data Actions to mitigate the risk The test phase 1. Functions to report vandalism and very bad quality 2. Function to block users spamming or vandalizing

7 3. Function to automatically reduce the quality of data for a selected sources (moderators or administrators only) 4. Function to delete data for a selected source (moderators or administrators only) 5. Functions to modify or delete existing data for general users 3 Project Organization The project will initially consist of only the project founder, Christoffer Owe. During the implementation phase it will be considered to include people with different skills. In addition to technical skills, proper genealogical experience is also required. The following technical skills will be needed: PHP programming SQL Web design and user interface design For the test phase it is necessary to include several testers. The testers should be genealogists with a technical understanding and who agrees with the goals of Yggdrasil. At some point, most likely after the go-live, the solution will be too large to be maintained by a team of volunteers only. Then funding must be raised and a formal organization must be created. This should be a non-profit organization with the sole purpose of ensuring the continued support and maintenance of the solution and the intentions behind the solution. 4 Project Funding The project will initially be 100% based on voluntary work. If the project succeeds, funding will be needed, both for the operating the solution and for application maintenance and development. As one of the premises of the solution is to be always free both to retrieve and add information, any subscription type of income is out of the question. That leaves two alternatives: donations or advertising. Donations would be the preferred funding option, but it s probably unlikely that donations alone would give enough income to maintain and expand the solution. We don t expect rich companies to donate moneys to Yggdrasil. The other option is advertising. It s likely that advertising will be necessary. If so, it must not be too dominating, and it must be clearly distinguished from the genealogical information.