On Fri, 27 May 2011 00:55:01 -0700 (PDT), Tom Wetmore<[email protected]> wrote: > >>On Thursday, May 26, 2011 11:31:45 PM UTC-4, Bob Melson wrote: >>> >>> Y'all'd have to check the archives for the down'n'dirty, but Wes and I >>> and I don't remember who all else had a discussion of this very thing >>> (_UID, >>> _UUID, _GUID) some time back. I contended then that UUID (Universal >>> Unique ID) and it's close kin, UID and GUID, meant that the number >>> assigned to Cousin Mortimer should be the same universally - here, >>> there, >>> wherever it appears. Not so. The "universe" is the machine where the >>> data appears and on which the software resides by which the UUID is >>> generated and it's only there that it's guaranteed to be unique. Take >>> that same data to another machine or other software and, guess what?, >>> the >>> Universal Unique ID will be different. Or, say, Mort's data changes - >>> ta DA, you may (or may not) get yet another UUID. >>> >>> The only use I can see for UUIDs is in determining the origin of a >>> particular record - if I publish Mort's data and include the UUID and >>> some time later find that Snively Whiplash has published the exact same >>> data with the exact same UUID and has claimed it to be his own, then >>> I'll know what to think about ol' Snively, won't I? >>> >>> Swell Ol' Bob >>> >>Bob, >> >>You were right and should have stuck by your guns!! >> >>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 >>since GEDCOM doesn't even have the UID concept. Their interpretation >>can only be coming from some vendor's misinterpretation, so to treat that >>misinterpretation as if it were a rule or the way things should be is >>wrong. >> >>A UUID is intended to be unique for all time and place. If a vendor says >>they use UID's where the U means "universal" and they don't support this >>then they don't support UID. A program should not alter a UID value upon >>import. If it does, though, who really cares, since there's no way to >>take advantage of UID's in genealogy. So no wonder it doesn't matter to >>anyone (yet). >> >>Tom Seems to me at least one requirement for universality is that the same data input on different machines results in the same output on those machines. This is absolutely not true - in my experience - when dealing with UUIDs. Worse yet, not only does the same data NOT result in identical UUIDs when input on different machines running different software, it doesn't even result in identical UUIDs when input on different machines using _identical_ software. This would, IMO, support the contention that uniqueness and universality is restricted to a single machine and application on that machine. Bob -- Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas ----- The greatest tyrannies are always perpetrated in the name of the noblest causes -- Thomas Paine

On 05-27-2011 10:53, Bob Melson wrote: > Seems to me at least one requirement for universality is that the same data > input on different machines results in the same output on those machines. > This is absolutely not true - in my experience - when dealing with UUIDs. > Worse yet, not only does the same data NOT result in identical UUIDs when > input on different machines running different software, it doesn't even > result in identical UUIDs when input on different machines using > _identical_ software. This would, IMO, support the contention that > uniqueness and universality is restricted to a single machine and > application on that machine. UUID = Universal Unique ID. If two machines generate the same string, then they have blown the Unique criteria. It is not intended to be a tag that is common to distinct items that happen to be identical. That would be a checksum. :-) -- Wes Groleau There are two types of people in the world … http://Ideas.Lang-Learn.us/barrett?itemid=1157

On Friday 27 May 2011 18:22, Wes Groleau ([email protected]) opined: > On 05-27-2011 10:53, Bob Melson wrote: >> Seems to me at least one requirement for universality is that the same >> data input on different machines results in the same output on those >> machines. This is absolutely not true - in my experience - when dealing >> with UUIDs. Worse yet, not only does the same data NOT result in >> identical UUIDs when input on different machines running different >> software, it doesn't even result in identical UUIDs when input on >> different machines using >> _identical_ software. This would, IMO, support the contention that >> uniqueness and universality is restricted to a single machine and >> application on that machine. > > UUID = Universal Unique ID. If two machines generate the same string, > then they have blown the Unique criteria. It is not intended to be a > tag that is common to distinct items that happen to be identical. > That would be a checksum. :-) > > -- > Wes Groleau > > There are two types of people in the world … > http://Ideas.Lang-Learn.us/barrett?itemid=1157 I'll answer this and the one immediately previous with this. To start, I don't want to re-ignite the previous discussion regarding *IDs, aside from saying that it seems to me that a {Globally|Universally} Unique ID should indeed be unique everywhere; given the identical input even on different machines; the resulting ID should, or so it seems to me, be identical but unique to that input. This is not "blowing" the unique criteria, any more than identical checksums derived from identical strings on different machines "blows the checksum criteria". Matter of fact, I think a checksum would be a helluva lot more useful than a *ID when you come right down to it. All that said, it's more than likely that I have a faulty understanding of what "globally unique" or "universally unique" actually mean. Based strictly on the meanings of the words, though ... The ID produced on my machine uniquely identifies a record ON my machine but is otherwise of no value and, to my mind, appears to be redundant as there are other record identifiers that "uniquely" identify that record. That same record on another machine will produce another unique ID, different from the one produced on my machine and valid only on the machine producing it. Go to a third (or a fourth or an Nth) machine with an identical record and you'll get a 3d or 4th or Nth ID, different from all others and, IMO, valueless for identifying the record anywhere except on the machine on which the ID is produced. The end result is that we have N records with N IDs, all unique, and none of 'em (the IDs) useful for any discernible purpose. So, will SOMEbody please 'splain me this thing called a globally/universally unique ID and its place in the grand scheme of things? Stumped Ol' Bob -- Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas ----- The greatest tyrannies are always perpetrated in the name of the noblest causes -- Thomas Paine