Lorna, your point about cross-referencing two websites and two databases via the TMG ID stimulates some other ideas for your 'some year I will rethink the whole thing'. Internal to RootsMagic is a program assigned UID or 'universal ID'. It is a 64 bit number generated from system time and other system numbers and randomization which makes the probability infinitesimally small that any two persons created on any two computers will have the same UID. It is exported to GEDCOM and I have passed it through an Ancestry upload and download. While invisible and inaccessible through the application, its export and SQLite open up some possibilities. Externally and supported by both Legacy and RootsMagic is the FamilySearch Family Tree ID. If two persons in two different databases match to the same person in FSFT, then they will have a common FSFT ID. This is visible but uneditable in RootsMagic and could be another key for cross linking between your two websites and to FSFT. Maybe John Cardinal is looking at these possibilities for Second Site. Tom > On Sep 24, 2014, at 9:58 PM, LornaMoa <lornamoa@gmail.com> wrote: > > Thanks Tom > >>So you cannot combine two TMG projects with overlapping TMG IDs in a common RootsMagic database and have all the RINs match all the TMG_IDs. I wonder if you can even do that in TMG. To keep common RINs, yet keep the persons separate, they must be in separate databases. << > > Fully understood. > And yes, TMG-IDs are unique within a database but there's nothing stopping us in TMG from using the same id in another database should we wish too. > > The linked projects I'm talking about are currently in two databases (with some of the individuals having the same id between the two db) because the end results required are quite separate and my ingenuity within TMG ran out when trying to keep them only in one AND to show both real parents and DNA "parents". > Quite apart from which, many of the dna project participants are of no interest to my personal research, even if they do turn up as dna matches to the One Name Study people. > > My main database already contains all of my own family research, PLUS two one name studies, and I happily produce the separate One Name Study webpages from my main database using flags and contains 10s of 1000s of people from many years of research. > But the dna website was a step too far given I needed (well wanted) to show a tested person both with his actual pedigree and with his DNA lineage. > http://dnasurnames.info if you are at all curious. > > I cheat within the dna database so I can show both sets of "parents" for selected people. > The ones I show on the haplogroup charts are the dna haplogroup branch "person", with a "son" being the patriarch of the line associated with the Y/mtdna value, and the "child" being the tested person. > The latter two are duplicated in the database but linked using custom two principal tags. > In the Lineages aka Surnames section the duplicated people above show up with normal family relationships, and usually a link to where else they are published (linked by TMG-ID and smarts in SSite), eg one name study or my personal family webpages. > > Yes, a RM Gedcom output to SSite would probably work, but for the above site, as it currently is configured, only if I can control some of the RINs/TMG_IDs, hence my question about whether your utility would also work on one person in isolation to change the RIN as opposed over the whole database. > > Some year I may well rethink the whole thing, but for now, my energy is directed elsewhere. > > > Lorna Henderson > http://lornahen.com >