This is the official research blog for the Distributed Family Tree, an open network of genealogical data and metadata. In a nutshell, the big idea is that we can combine all available genealogical information on the Internet into a single distributed network. The foundation for this network is the substance of the Master's Thesis that I am currently working on.
Having documented how search will work, implementing it was a breeze:
This screenshot shows results for a search of PhpGedView sites for “Jared”. There is no similarity test yet, so results are shown in the order found. Subsequent searches bring up cached results, but database queries are running far too slow for my taste. The theme of this project seems to be “premature optimization” ;).
Filed in Genesis
Comment feed for this post
[…] If you’ll look at the most recent screenshot of Genesis, you’ll notice the search box is disabled and the “Go” button is changed to a “Stop” button. Contrary to the original spec, I made it so that when you search, you have to hit the “Stop” button and wait for all the threads to die before the search box is enabled and you can search again. This really lengthened the cycle when I was testing searches because the threads can take a long time to properly die. Finally getting fed up with, I decided to return to the original spec. […]
[…] months ago Yakov Shafranovich made an excellent recommendation: why not just use Google!? I didn’t quite realize at the time how insightful he […]
Have you looked at Google’s Custom Search Engine. For example, the following query searchs PhpGedView pages across the web:
http://www.google.com/search?q=inurl:individual.php inurl:pid=
And the following quiery pulls up GEDCOM files:
http://www.google.com/search?q=adam filetype:ged
Thanks for the tip! Search agents could definitely be written which leverage Google search to find data. I’m currently working with PhpGedView sites directly because I anticipate some sort of feedback cycle: one where my work impacts the software that runs PGV sites so that they can automatically (or semi-automatically) link up together, creating a distributed network of genealogy in the process. Google search results would certainly bring in more data, but that data won’t be as actionable in this respect. I hope to go down that avenue the future, however.