Monday, September 13, 2010

Mailbox Monday: Community Reference Links

Dear Ancestry Insider,

How does this differ from Ancestry.com's "Web Links" that are available (though not prominently) to subscribers on Family Trees? There, too, one links to other websites referencing the individual.

Nolichucky Roots

Dear Nolichucky,

I'm giving you my understanding of what I’ve been told. But this is an emerging technology. I may have misunderstood, or the technology may change before you, the public, ever see it.

I should clarify that Ancestry.com Family Trees are also available to non-subscribers, for free. But I digress…

Let me illustrate how Community Reference Links would differ from web links. Let's say that you have a Nolichucky Smith in your Ancestry.com Family Tree.

Imagine someone finds a record about him on FamilySearch and attaches it to the new FamilySearch Tree (once that is possible). You would automatically get a shaky leaf in your tree. You could attach the record from FamilySearch to Nolichucky in your tree. Something similar happens to all the Nolichucky Smiths on all participating websites.

You find the 1880 census record for Nolichucky Smith on Ancestry.com and attach it. The link shows up on Nolichucky Smith in the new FamilySearch Tree and so forth.

Community Reference Links are worthwhile for pay websites—and this is important—because non-subscribers must still pay to access the linked records. Pay websites get a source of highly targeted potential subscribers. Click-through rates will outperform any other source.

Community Reference Links are good for you, an end user, because you find out whenever anyone finds a record about your ancestors. You find out what websites have those records. You can make more informed decisions about website subscriptions. You find free records and photos.

So you see, Community Reference Links operate more like shaky leaves and member connections than website links.

-- The Insider

Broken Links

Dear Ancestry Insider,

The Community Reference Links you discussed sounds really great except for one big problem. Web links don't always remain constant. Webmasters seem to have a propensity for redesigning things periodically and when they do, sometimes the URL for a particular object changes. This results in the infamous broken links often seen on web sites that have a lot of off-site connections. Unless and until this problem is solved, I will continue to download everything I can to my own storage facility where I know I can find it.

John

Dear John,

I have found someone else whom I think the world of. Oops. Wrong dear John letter. I digress…

John, you are so right. CRLs is not an attempt to solve the issue of broken links. I would continue creating your own copy of anything you want to keep.

While solving broken links is not the focus of CRLs, breaks would undermine the value of the link community. FamilySearch is giving the problem some thought. One line of thinking is to place vendors under contract to honor forever any links the vendor submits to the community. Another approach would be to require URLs with some sort of embedded identifier. The identifier is used to automatically mend broken links.

Thanks for the feedback.
-- The Insider

 

Apologies to you who already read these comments online. Comments and my replies are edited for length, style, or clarification.

3 comments:

  1. One way FamilySearch has solved the problem in part regarding changing URLs is something they are doing with FamilySearch Wiki.

    How it works: Go to a name in either Record Search Pilot, especially one that has an image that was indexed through FamilySearch Indexing. You will see a link with the name when you go to the page for that name where you can click for more info. Click on that link, it will go to a page in the wiki.

    The page will have a box at the top created via template with the info on the direct URL for the collection. Additionally, and this is the key, at the bottom there will be a 'collection ID' number. However, it's not visible as it was created in white I think using another template. To see it, mouse over the blank area at the end of the article as if you were selecting text, eventually it will appear.

    This was one of FamilySearch's first attempts to solve the CRL problem, while it is very basic, it does show what might be done once the technology is more developed.

    ReplyDelete
  2. Here's another way to prevent broken links: http://www.doi.org/

    If the advantages to the publisher sites are as great as AI says, they'll happily use something like this to ensure that they keep getting those click-throughs.

    ReplyDelete
  3. AI, you said, "Community Reference Links are worthwhile for pay websites—and this is important—because non-subscribers must still pay to access the linked records. Pay websites get a source of highly targeted potential subscribers. Click-through rates will outperform any other source."

    Am I misinterpreting, or do you mean that newFamilySearch, once released to public access, will operate as a marketing device for Ancestry.com and other pay database sites?

    ReplyDelete