I’ve talked before about how I think there needs to be more semantic standards adoption in Genealogy. I’ve been thinking a lot about what that would look like. Here are the key factors I think the perfect “Family Tree” would include:
- Each individual has their own, unique URL. In a sense, this would give the dead identity.
- Individual pages should be wikis, ie, anyone can edit them, and they can add *anything* to the page (not a pre-defined list of fields). A predefined list of fields can be defined through semantic markup though, which should aid search. There could be discussion pages for each individual similar to wikipedia for any disagreements.
- Each individual page lists parents and siblings and other relationships through simple, standardized markup such as FOAF or XFN
- Each “owner” should own an additional “family tree” index. This is where all the relationships between individuals are mapped out. A simple site map syntax ought to solve this (which could be organized in a pretty tree like format through CSS if you chose). Individuals in that tree ought to be able to identify their place in the tree.
- It should be 100% open (perhaps with privacy controls just for living individuals – still need to work that one out, but we already have many examples to learn from)
- The system hosting the “family tree” should be 100% open source – this ensures that anyone else who wants to host a compatible system can host it with no worries. I foresee many of these, all compatible, federated, and individuals and family trees linking across multiple domains and hosting environments. I’d like to host one of these for my extended family, for instance, where we can store all the stories and family histories we’ve archived over the years – others could easily contribute.
- Content within each individual page (for dead individuals, at least) should be licensable according to specified, open licenses, similar to Flickr.
- The system hosting should detect when there are other duplicates on the web – perhaps Google could provide an API off their index to make this easy (or maybe FamilySearch.org could do something like this). It should automatically flag records that are potential duplicates and share where the duplicate records are.
- If I want to host my own family tree, I should have the option to download my family tree from any system, and be given all the files necessary to host my own system on my own servers for others to access. It would work exactly the same as all the other systems, be 100% compatible with all the other systems, and ideally, if I agree another record is correct, ought to link to records on other systems when I don’t need to host the data myself (to avoid duplicate records if I don’t want them).
I think with all these factors in place, a fully distributable, 100% owned by the community, system can be created to host genealogy data. No one organization ever has to own the data, and anyone who wants to own their own piece of the puzzle can take it, free of charge. Any search engine on the planet can help index this data. It’s the essence of open.
So who will be the first to create it? Or maybe someone wants to fund it/donate and I’ll hire the developers?
These opinions are obviously my own and not necessarily those of my employer.
I'm liking it too… now where did I put that seed money I had laying around… 😉
I like where this thinking is headed. It's not clear to me how the federation will work without getting too messy though. Maybe you could do just a post that outlines that point?
Yeah, maybe I'll go into that at some point. Some study into XFN might help
you understand. That will at least handle the mapping and identify the types
of relationships, and then each “owner” of the data could own a sitemap of
sorts that mapped out all the relationships, except they would likely link
across multiple domains (this would, in essence, be each individual's family
tree). They could submit these to search services for easy indexing of the
data. (and through this people could build APIs off this stuff similar to
what Google has done with the Social Graph API)
People like to be able to browse their family tree in a single interface. Would this type of structure where each node of data could be hosted in separate locations, work in a consistent interface or would it change depending on the source? I'm having a hard time visualizing this as well.
Yes – ideally you should be able to view your entire family tree if the
hosted source wants to categorize all your data. All you need is a tree – it
doesn't matter where that tree links to if all the data is standardized.
While it doesn't offer hosting on multiple domains, Dallan Quass at WeRelate.org has worked on a similar wiki-style solution. It might be worth collaborating with him since he has worked on it for several years and has come up with some good solutions for collaboration. For example, here is the link to George Washington: http://www.werelate.org/wiki/Person:George_Washington_%286%29
Good point. I know Dallan well. I had forgotten about that. I'll talk to
him.
Jesse,
Great ideas… I wonder if I could get the web over-lords to let us have the address:
Bruce Christensen.(dot)Christensen and you could have: Jesse Stay.Stay
I built something very basic like this for my own purposes (http://prather.org/family). I wanted to add wiki-style content addition, as well as more linked data (there is FOAF for individuals but not for families for example, and nothing using XFN). The code that i have (such that it is) is available on Github.
The problem I ran into is the funding issue. Hobby projects have taken a back seat recently. If you do find funding, I would love to have the opportunity to work on this 😉
I wish you would open source this Dallan – I'd really love to install a
version on my own servers for my personal family tree.
This perfect family tree is good article for me,i learned this blog to root is domain and leafs are web sits,thanks for sharing this blog.web hosting review