Web search sourcing

[Warren White]

First off, I am not a programmer, so I don't know the details of implementation of this idea. I do know that some web-based tree applications implement what seems to me to be a logical step in sourcing. Can it be done in Reunion?

What I envision is that following a successful web search against a person in my file, shouldn't it be possible for Reunion to 'grab' the source information from the web site and bring it into Reunion? It seems a logical extension of the 'search the web' feature.

I would be most interested in Leister's thoughts on this, as well as what other users might think.

[ljirwin42]

Re: Web search sourcing

What I envision is that following a successful web search against a person in my file, shouldn't it be possible for Reunion to 'grab' the source information from the web site and bring it into Reunion? It seems a logical extension of the 'search the web' feature.

Warren, I appreciate your thinking of new ways to use Reunion. But this particular idea leaves out the most important function of data-gathering. What we need to do is not to search-grab-import. It's to search-EVALATE-import. Reunion's existing web-search feature is useful, but I find myself narrowing it down to just a couple of "faves" because the rest produces too much un-substantiated, un-documented "genie-junk." It's not the fault of Reunion, it's just that so much that's out there is unreliable. So to automate and amplify the importation of raw data would, for me anyway, be a step backwards. We need better filters, not finders.
Lee James Irwin
Fairfax, Virginia

[Warren White]

Re: Web search sourcing

Lee,
You make a good point, but one that need not preclude making this a function of Reunion.

I would not ever use this function without evaluating the strength and accuracy of the source. But, for example, if you find a family on an Ancestry census, if the people, places and times lead you to the conclusion that it is 'your' family, why can't Reunion take the source information from Ancestry (whatever they provide) and put it into my Reunion file rather than having me retype all of that data?

Frankly, I am surprised that nearly 200 people have read this, and it looks like only you and I are talking. Perhaps I am talking about something only I think is a good idea.

Bummer!

[theKiwi]

Re: Web search sourcing

This is a great idea - some organisations have worked on and support such a scheme - I think it's currently at least more prevalent in the world of libraries.

But I agree - just because someone might abuse it, doesn't make it a bad idea for those of us who wouldn't abuse it.

Roger

[Michael Talibard]

Re: Web search sourcing

Not often I don't agree with Roger

[theKiwi]

Re: Web search sourcing

Wouldn't it have to make assumptions about how you lay out sources? Users have their own rather different preferences.

There are already some de facto standards for this - Evidence Explained for example - and at least some of the places that currently do provide easy sourcing of online resources do follow some standard or another - as I noted typically in the world of Libraries currently - and people who want to use "easy" online sourcing are probably more than happy to conform their systems to these methods of online sourcing I imagine.

Roger

[Mark]

Re: Web search sourcing

What I envision is that following a successful web search against a person in my file, shouldn't it be possible for Reunion to 'grab' the source information from the web site and bring it into Reunion?

This isn't possible with the current web search.

All the web search does is build a URL (web address) and tell your preferred web browser to open that URL.

The URL contains all of the search criteria (names, dates, etc.), Reunion isn't directly interacting with these websites and doesn't see the response that the website returns. Those responses are all in HTML, anyways, which would be very problematic to try to parse through to find the pertinent source information. In many cases there may be no way to identify what bits of text in the HTML are the source information, depending on how the website is designed.

There are, to my knowledge, very few websites that would allow us to pass queries directly to their databases (instead of our current method of using the web browser as an intermediary) and receive a response that we could parse into usable data. Doing that would require a very different method than what we're currently using, and it would only work with sites that provide an API we could use to interface with their databases.

I hope that helps explain the situation