I have a primary RM DB that I use for my publishing and exporting via GEDCOM to my TNG websites. I have cousins that send data do me, and I create a seperate RM database with whatever they send me and standardize names and places. This makes it easier to drag and drop the people and minimize duplication. In the process I create a source that I attach to everyone in the submitted database so when I drag and drop them I have a record of where they came from.
For whatever reason at some point in V11 an oddity was created that has generated some whacky GEDCOM results.
My gut feel is it had to do with a citation duplication removal. One of my drag/drops had for some reason a large number blank source/Citations. The end result was people from drag/drop A that duplicated drag/drop B ended up with multiple sources that were duplicated.
Since I’d had serious issue is this process in the past I made a backup, merged sources. checked all the counts, and all looked well so I merged all citations, checked all the counts, and all looked well.
So from the “all looked well” standpoint I try to export a GEDCOM to upload to my TNG sites. A GEDCOM file is create that feels fairly complete, but the progress dialog closes abnormally at 0% complete.
After several passes I create a new RM DB with the GEDCOM file and there seems to have been a complete export EXCEPT for some oddities. There are twice as many citations in the new file as the old, and there are 20 times the media links.
I’m honestly not able to follow all the weirdness you have described, and I am not a TNG user. But I have one suggestion to consider. Namely, GEDCOM does not support RM’s de-duplicated citations. So when you make GEDCOM, all the de-duplicated citations will be re-duplicated. RM really has no other choice. And the re-duplication of citations will result in the re-duplication of media links associated with those re-duplicated citations.
If this sounds strange, remember that RM7 did not support de-duplication of duplicate citations. So in effect, a GEDCOM produced by RM7 produced duplicated citations because the duplicate citations couldn’t be de-duplicated in the first place. So a GEDCOM produced from RM11 after re-duplication of citations should have the same number of citations as would have been produced from RM7.
Even though I’m not a TNG users, when I export RM11 GEDCOM for use in GedSite, the re-duplicated citations look in GedSite exactly as I would expect. There are two things different about my situation. One is that I am not importing GEDCOM from other researchers. The other is that none of my sources or citations are blank. But that’s easier to control because I’m doing all the data entry by hand rather than by GEDCOM import.
Sorry I can’t be more help.
Thanks Jerry, that’s actually very helpful. I was not aware of the reduplication of citation and media links. That explains a lot, so that leaves me down to a UI issue with the export status dialog that I can live with. The export dialog used to show counts for all the various record types, but now shows counts of only people, family, and sources then closes. This gives the impression the export is incomplete.
I’ve been paying extra attention to that due to the other issues I’ve been having with the export. One of the files from a cousin I had imported had special characters (copy/pasted from Facebook messages) in notes fields and that was causing the export to crash. After I fixed that I found out that the SSN fact I’d brought over from Ancestry was creating a null fact type and THAT was causing a different crash. So I’ve been looking at status dialog box a LOT lately and apparently just noticed an issue that’s been there for some time.
The counts I’m talking about have nothing to do with TNG and are actually from the RM databases. My original, and one that I created from the exported GEDCOM. I’ve been trying to figure out what’s happening BEFORE going to TNG as that’s visible to the public. I guess I can do that now, but the extreme nature of the reduplication process still seems odd to me.
Have you tried running db tools? Not sure if orphans of some sort would cause a Gedcom issue and maybe TNG does not like.
Maybe related to duplicate citiations / sources that have blank fields? just throwing thoughts out there since I can NOT see that db
First off TNG has nothing to do with the issue, the issues have been in the creation of the GEDCOM, not the consumption. Until earlier today I haven’t even tried importing into TNG because I couldn’t determine if the GEDCOM was complete and syntactically correct. I determined it was “correct” but with wacky numbers for Citations and MediaLinks when I created a blank RM DB and imported the GEDCOM I exported from my master into it.
As far as DB tools I run them at least once a session and they never find (or fix) anything wrong. They miss a LOT of foriegn key issues which is why I wrote my own tool to check them.
I’ll still need to research further as there may be latent issues in the DB left over from the massive media issue I have last year when I hit the merge citation bug and had to do some massive manual cleanup and lots months of work.
sorry if my reply was confusing – my point was if something is wrong with the GEDCOM that could potentially cause issues when TNG imports it. I have used TNG since TNG11/12 (thru tng15) but do not currently have an active website – I do have a local server though
I had an media issue last year after running the media tool when 11.0 came out late summer.
Could not fix in RM – I replicate the issue 5 times and reported to RM and moved on – the technician blamed everything but the program. I spent longer trying to document the issue than I did fixing it with Sqlie depsite over 500 media record being impacted. This occured after moving files to sub dir (the filenames did not change). But clearly the cleanup tool / fix media broke things as I confirmed no issue were apparent previously
