Wes Groleau wrote: > On 05-27-2011 03:55, Tom Wetmore wrote: >> You were right and should have stuck by your guns!! > > I may be misremembering, but I think the disagreement was that someone, > perhaps Bob, thought that the UID would be a universal ID for a _person_ > which could be used for that person in all databases. Others stated, > no, it distinguihses a _record_ from other records that otherwise might > seem identical. > >> I don't know where the others got the interpretation that a UID should >> be unique only within one database. This is certainly not a GEDCOM rule > > I don't recall seeing that interpretation. > PAF's help files specifically state that; Legacy's help files /imply/ it. From PAF's help file: PAF 5.0 assigns each record in the .paf file a unique record serial number. Like a record identification number (RIN), a Unique record serial number identifies each individual in a .paf file. Unlike a RIN, a unique record serial number is unique worldwide. This mean that each individual in each of the .paf files has a different number than all the other individuals in all other .paf files that are made worldwide. (para) The number does not change if you export a GEDCOM... Legacy's help file mentioned (UIDs) indirectly under Intellishare: One person in the group is designated the "Keeper of the Records" (Keeper for short). This person keeps the master Family File. Legacy automatically marks all the individual records in the Master Family File with a serial number that uniquely identifies each person ... Note that both refer to it not as an ID but as a record serial number.* That distinction virtually guarantees it has to be within a single database; as an analogy, Xcel and QPro wouldn't know whether line 564 in "Monthly Expenses" came before line 564 in "Appliance Repair", so why should and how would PAF/Legacy/whatever know whether RIN 534 in "McCOY.paf" came before or after RID 534 in "Hatfield.fdb"? Cheryl *Which, yes, inspires one to wonder why they didn't CALL it a URN or USRN or USN ...