In a recent lengthy comment to one of my blogs, the commentator said, in part, "I think one of the biggest issues for FamilySearch Family Tree is the ease of using it, or rather, the lack thereof." In reading the entire comment, it appears that the basis for the opinion was primarily in two areas:
the lack of an adequate merging system and the users inability to add multiple sources at the same time.
First of all, the merge system on FamilySearch.org's Family Tree is not yet fully functional. The limitations imposed by its connection with the New.FamilySearch.org (NFS) database is not fully resolved and may not be resolved for many months. I would assume that there is little incentive to "fix" the merge function until the rest of the issues with NFS are resolved. Meanwhile, Family Tree is a rapidly evolving program with new features appearing almost weekly.
The issue of adding multiple sources is related to another issue, adding individuals in groups by uploading GEDCOM files. Both of these issues raise serious concerns with the goal of maintaining the accuracy of the data in the program and at the same time reducing the number of duplicate entries. The main issue is adding duplicate people to the program, but the proliferation of duplicate source citations, photos, stories and other items is also a concern.
From an absolute standpoint, the question is whether or not the program should allow a reasonable number of duplicates within specific tolerances or try for a zero tolerance level. Is it better to allow duplicates or risk the loss of some valuable information that may accompany multiple copies of the same entry?
With regard to uploading GEDCOM files, the answer is simple. The program allows for only one individual entry per person. There is a built-in zero duplicate tolerance. Hence, the process of adding a GEDCOM has been slowed by making the user examine each potential entry to determine no duplication is involved.
The issue of adding multiple sources in more problematic. There are good arguments for allowing multiple sources to be attached to an individual. But at the same time, there are also arguments that the whole Family Tree should require careful one-by-one entries to maximize the possibility that the entries are carefully considered and as accurate as possible. With sources there are many possibilities for errors. The source may not apply to the target individual or it may be totally unreliable. Limiting the method for adding sources raises the chance that these concerns will be addressed on an individual basis.
It is my opinion that the Family Tree should be built a carefully as possible from the start. It already suffers from errors introduced by the incorporation of previous unreliable databases, let's not make it worse by allowing wholesale uploading of information. Family Tree is not a program where you can dump your interim research, it is designed to be the ultimate archive of our collective knowledge about our ancestors and as accurate and complete as possible.