Monday, August 30, 2010

FamilySearch In a Corner

As I explain in my charter, FamilySearch releases a lot of information to the public, but only in out of the way corners. That’s where I come in. I try to be in those corners. This past weekend one of those corners happened to be the Salt Lake Family History Expo.

(I told a coworker that attending Ron Tanner’s presentations is the only way a FamilySearch employee can find out what FamilySearch is working on. But I digress…)

Tanner was off his best game—but only a bit. “Two weddings in two weeks,” he explained. Between tales of Jelly Bellies, birthdays, weddings, and the DMV, Tanner was able to slip us some juicy news nuggets that you won’t hear anywhere else.

Here’s the new news:

  • Records from Record Search are being transferred to the Historical Records area of the FamilySearch Beta. The data transfer, containing 447 collections and 600 million names, should be completed by the end of the month. (We all assumed he meant September.)
  • Historical records from the current www.familysearch.org will be transferred after that.
  • All new collections will go directly to the FamilySearch Beta.
  • Ancestor discussions were recently added to the new FamilySearch tree. The length of discussions will soon increase from the current limitation of 500 characters.
  • The filtering of Historical Records search results will be improved soon, along the lines of Record Search.
  • The Family History Library Catalog search on FamilySearch Beta will be dramatically improved by the end of September.
  • Disputes can no longer be added to the FamilySearch Tree. Soon, all the disputes on an ancestor will be moved into a discussion called “Legacy Disputes” and deleted from details.
  • The new FamilySearch Tree rollout map. Credit MormonTimes.com. The new FamilySearch Asia rollout is set for a couple of months from now. Tanner showed us a photograph of the rollout map and told us we could see the original by visiting FamilySearch headquarters in the Joseph Smith Memorial Building in Salt Lake City. Ride the elevator to the 4th floor. The map is in the lobby.
  • beta.familysearch.org will replace the current www.familysearch.org website by the end of this year. The new website brings together FamilySearch Beta, RecordSearch Pilot, the Wiki, FamilySearch Indexing, Forums, the Family History Library Catalog, and all other collections and content from the current www.familysearch.org. Specialized websites such as consultant.familysearch.org will continue to be separate.
  • The general public can begin preparing for access to the new FamilySearch Tree by registering now for an account on the FamilySearch Beta. When the Tree is ready, they will be given access using that same account.
  • Some time after this year the new FamilySearch Tree will become “Family Tree” on the new website.

The Ultimate Future

Tanner reiterated his vision for the future of the new FamilySearch Tree. However, changing product managers—generally considered a healthy practice—can change the course of a product. Just three weeks ago product assignments were adjusted, giving Tanner complete responsibility for the Tree.

Tanner admitted that the current design didn’t go far enough from changing “my tree” and “your tree” into “our tree” since it maintains “my conclusion” and “your conclusion.” The many alternatives that exist today under the Details page will be pulled out. In their place, the system will have links to the original sources of the information. These include Ancestral File, Pedigree Resource File, and the other sources used to populate the Tree.

The tree will allow only one conclusion. This will solve the IOUS problem that slows down the system and makes it impossible to merge all copies of a person. “Combine and Separate” will be replaced with “Merge.” I’ve stated before that I don’t think FamilySearch has any corporate memory in product management. It remains to be seen if Tanner can avoid Merge pitfalls.

Tanner’s future direction is to change the Tree so that anyone can correct it. Anyone can contribute to it. Anyone can contribute artifacts to it. Anyone can change the information that you contributed to the Tree—maybe for the better. When someone changes your contribution, you’ll be notified and given the ability to undo the change. Ancestor Discussions will allow contact with others.

Tanner said his pet feature was Community Requests. You can post a request for a volunteer to lookup a record, take a gravestone picture, or translate a language you don’t read.

Most importantly, his stated goal for the Family Tree is that it be so genealogically correct that even the best of genealogists will want to use it.

Tanner was asked what the timeline was for these features. “Before we die or slightly thereafter,” he quipped, artfully avoiding being painted into a corner. Stay tuned…

8 comments:

  1. Occasionally I do a general search on the old FamilySearch site for certain ancestors, just to see if any correct additions have been made there.

    For one search I did today there are 7 IGI entries, 1 Ancestral File entry and 2 Pedigree Resource File items.

    Of the 7 IGI entries one has correct spouse and birthplace; one has right birth year, right State and death date; one has the same as the previous entry plus wrong spouse; one has right birth year, wrong County, wrong parents; one has right birth year, right Township and County.

    The Ancestral File entry has 'right' birth date estimated from age at death on gravestone, right death date, wrong place, wrong parents, wrong spouse, wrong marriage place.

    The Pedigree Resource File has 2 entries: one with right birth year and State, right death date, wrong spouse. The other is the same as the AF entry except the County of death is at least spelled correctly (still wrong location there).

    No amount of merge-tinkering will wind up with correct parents and place of death. A doer of good might choose the correct spouse listed once if the IGI data is in the overall database.

    The wrong parents listed did not have a known child by the same name as the above subject, and the wrong spouse is not known to have married a person by the subject's surname, so at least a 'split' function is not required in this instance.

    However, there are certainly items in the above databases that merge different persons by more or less the same name.

    ReplyDelete
  2. Thank-you for this post, AI. I've told my DearREADERS this is a MUST READ.

    ReplyDelete
  3. Dear Geolover,

    Perhaps I mis-communicated. In NFS today Combine mushes together multiple sources.

    In the NFS of tomorrow, sources (such as the IGI, AF, and PRF) are extracted and exist independently of the conclusion tree. Sources are associated with conclusions in the tree. But the tree can be "fixed" independent of information in the sources.

    Merging is used to "fix" multiple copies of the same person in the tree.

    Keep in mind that in "our tree," "fixed" is determined by the community. It happens like this: You or me applies "your fix" or "my fix." If no one objects, it becomes "our fix." If someone objects, they undo the change and engages the fixer in a discussion. When consensus is reached, someone makes "our fix."

    Once a discussion is underway, it is bad form to apply or re-apply "your fix" or "my fix." Doing so can result in temporary suspension.

    When consensus can not be reached, the decision goes to an arbitrator who decides "our fix."

    I believe that FamilySearch product managers are operating under the assumption that in any disagreement, genealogical practice and standards dictate a best "our fix." I don't believe that. It may be necessary to allow an appeal to a panel of experts who rule by majority vote.

    -- The Insider

    ReplyDelete
  4. Dear AI, Thank you for trying to clarify any confusion on my part.

    I posted the above genealogical mess because I am confident (abstractly, not having access to nFS Tree) that there are several copies of the target person in the Tree, probably all with wrong spouse, wrong parents, wrong marriage place, wrong place of death. Hence merging (which I do understand) will diminish server load but probably not increase truthiness.

    This combination for this person is on the web in myriad copies. Only a few copies have succumbed to a decades-long correction campaign and show correct spouse and known parent. Two I have installed give supporting evidence.

    My point, perhaps poorly made, is that the fundamental (derivative, secondary and lacking reference to where the assertions are copied from) source materials for the nFS Tree do not give much guidance for accuracy.

    I am glad that the correction-of-conclusion process is heading in a more streamlined and rational direction. The Elephant in the Living Room is that countless numbers of rites have been performed based on erroneous genealogical accounts. Consequently there are persons who have an attachment to the mistaken assertions over and above the sheer weight of error-riddled trees on the web and their mistaken sources.

    It is to be hoped that a robust correction process will be able to prevent the 'wiki-war' phenomenon.

    ReplyDelete
  5. The one thing I have been waiting for is for the FHL films to be available online. This would simplify my life in so many ways.
    I, too, don't have much faith in certain IGI records and would be deeply disheartened to see my ancestor's erroneous marriage information carved even deeper into stone. Several people had to come together to correct an IGI perpetuated misstatement of the facts. It would be troubling if this work were ignored.

    ReplyDelete
  6. Why should a "consensus" opinion determine the correct answer to some puzzles?? GIGO still rules!!

    All the published info, including that of a HIGHLY respected professional researcher on a particular ancestor of mine assumes that he was remarried to a 2nd wife, back in the old country--

    Wrong!!! -- because I am a direct descendant of the couple, I KNOW that the first wife was actually the surviving spouse and I have a copy of her 2nd marriage here in the USA in her old age to show that she did not die decades earlier, as implied in the FHS films, etc...

    That is part of the problem of working in highly inbred ethnic groups... the same names/surnames can apply to many different people in closely spaced times and places...

    ReplyDelete
  7. Dear Anti-Consensus,

    If you have evidence and can communicate it clearly, then I wouldn't worry until we see how this thing unfolds.

    See my Labor Day article for more information about how it will--hopefully--work.

    -- The Insider

    ReplyDelete
  8. As hard as it was for me to sort parents, children, spouses and second spouses I will NOT put my tree "out there" for others to edit. There are several wrong versions "out there" right now, in fact they far out number the correct version. Would my tree be overruled by quantity as opposed to quality? I won't take that chance. I don't see this as being a good solution to the ancestor tree. It only lets more people have an opinion whether or not it is fact. I will leave my tree public at ancesty.com and hope that researchers will find it and do their own checking to see if those are their relatives. Examples are a man with his wife's parents as his, a woman married to her son, several children born after the husband's death but carrying his last name, children born when the mother is 3 years old, in a list of children born every 2 years one child is born in a state that the family never lived in the same year as another child that does belong to the family and so on. Too many people just aren't careful. But they will be invited to "correct" my tree? I don't think so!

    ReplyDelete