* Import ged from Ancestry to FH, general advice needed

[GrahamJWood]

Due to a recent family marriage I have a family tree on Ancestry which I would ultimately like to merge with my much larger tree in Family Historian, the export from Ancestry produces a 42000 line ged file . An initial look at the Gedcom file in FH showed me that this was not going to be straight forwards, with multiple error, warnings and ultimately corrections.

The family relationships look OK, as do census sources (with page references etc) other sources look less well represented such as birth Marriage and Death indexes which seem to have no page numbers.

Is there any general advice about how to proceed with this? Are there other things I should watch out for? Are there any plugins that could help?

[Jane]

Have you read through the advice in the Knowledge Base?
how_to:import_from_ancestry|> Import from Ancestry website

[tatewise]

Graham, if your Profile is correct and you are still using FH V5 then an upgrade to FH V6.2 would be beneficial, as stated in the advice Jane mentioned.

There are some Plugins mentioned there that may also help.

However, Ancestry uses a 'novel' approach to recording Sources and the initial FH Project derived from its Gedcom may need significant work to align it with your way of working in FH. Only after doing that should you consider merging. Take anything downloaded from Ancestry with a large pinch of salt, and ratify all Sources.

If you encounter any specific problems, do ask for advice.

[GrahamJWood]

Thanks Jane & Mike, I have now read the KB, I am using 6.2 and have now updated my profile. I will start working through the advice.

[GrahamJWood]

I couldn't spot any specific mention of 6.2 in the Ancestry import KB advice (I don't have FTM), does 6.2 change anything relevant?

[tatewise]

It says: "ƒh V6.1 offers significantly enhanced features for FTM migrants, many of which also benefit Ancestry migrants".

Both FTM and Ancestry use many custom facts, but FH V6 has an Extended Set of Facts to accommodate them. See glossary:work_with_fact_sets#custom_facts_from_other_products|> Custom Facts from other Products.

Both FTM and Ancestry prefer to link source details directly to Citations rather than in separate Source records. The new View Media buttons in FH V6 allow Media linked to both Citations and Sources to be easily displayed.

See how_to:about:version_6.x.y|> Family Historian Version 6.2.5 under New in 6.1.4 (Feb 2016) ~ Improved GEDCOM Import, especially from Family Tree Maker and Ancestry GEDCOM files and associated links.

[GrahamJWood]

Mike, I have followed the instructions up to installing the "import from Ancestry or FTM plugin", when I ran this I got an error fhSetValueAsDate(ptr,datItem) failed INDI1.MARR.DATE in record ID 1, followed by may other function failures on the same record field
Any guesses as to what may be wrong?

This is individual 1 from the ged file (name redacted)

0 @P1@ INDI
1 BIRT
2 DATE Jun 1957
2 PLAC Cleveland, Yorkshire, Northumberland/Westmorland
2 SOUR @S-1300701919@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: Jun 1957 Birth place: Cleveland, Yorkshire, Northumberland/Westmorland
3 _APID 1,8782::93466383
1 NAME xxxx x /xxxxxxn/
2 SOUR @S-1300702136@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Marriage date: Mar 1983 Marriage place: Scarborough, Yorkshire North Riding
3 _APID 1,8753::25557563
2 SOUR @S-1300701919@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: Jun 1957 Birth place: Cleveland, Yorkshire, Northumberland/Westmorland
3 _APID 1,8782::93466383
1 MARR
2 DATE Mar 1983
2 PLAC Scarborough, Yorkshire North Riding
2 SOUR @S-1300702136@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Marriage date: Mar 1983 Marriage place: Scarborough, Yorkshire North Riding
3 _APID 1,8753::25557563
1 SEX F
1 FAMC @F2@
1 FAMS @F1@


Regards
Graham
p.s. it would be really helpful of the product had selectable text in error messages so it can be copied and pasted into forums etc.

[tatewise]

The primary problem is that 1 MARR is NOT a valid Individual fact, and I suspect if you investigate the All tab for that person it will be a UDF identfied with a star bullet.

1 MARR for Marriage Event should only appear in Family records.
I don't believe even Ancestry misuses that tag like that, so have you any explanation how it got there?

It is probably not now necessary to run that Plugin and the advice needs updating in line with the how_to:import_from_family_tree_maker|> Import from Family Tree Maker (FTM) that says "In the latest ƒh V6.2 many of the problems below have been resolved as described in What’s New in Version 6.1 without the need for Quick Fix Plugins."

Focus on the other more general purpose Plugins and the Extended Set for facts.

On further inspection, that does not even look like a Gedcom snippet that has been imported into FH.
It would not have an Individual record id of @P1@ but would be @I1@ and the SOURce links would not include hyphens and just be S followed by a number, e.g. 2 SOUR @S1234@. So has that Gedcom been used via File > Project Window > New Project to create a new project as advised at the start of how_to:import_from_ancestry#quick_fix_plugins|> Quick Fix Plugins and the log file reviewed?

BTW: You can take a screenshot of anything and use the Attachments tab below the Edit Post box.
See how_to:about#problem_reporting|> Problem Reporting.

[GrahamJWood]

Mike,
it was the original ged file downloaded from Ancestry
With 6.2.5 does the advice about not moving data to notes fields still apply or should I just restore defaults and import?
Regards
Graham

[GrahamJWood]

Mike,
with no new fact sets and Unrecognised data not moved to comments I get this in imported extensions:-
Individual Record 59: Converted tag '_MILT' to attribute 'Military Service' for about 20 records
I also get quite a lot of exception reports starting with :-
Family Historian Exception Report - Wednesday 02 Aug 2017
=========================================================

Lines marked "EXCLUDED:" indicate data which has not
been loaded, and explain why it wasn't loaded.

Indented lines marked "EXCLUDED BRANCH LINE:" indicate
data lines which are not necessarily invalid in
themselves but which have had to be excluded as a
consequence of excluding other lines.

Lines marked "INFO ONLY:" indicate data that has
been successfully loaded, but where special action
was needed to load it (e.g. GEDCOM errors detected
and automatically corrected by Family Historian).

---------------------------------------------------------

Record Type=Individual. Gedcom Id=P1. Record Number=2.
l.37 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.38 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE Mar 1983"
l.39 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Scarborough, Yorkshire North Riding"
l.40 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 SOUR @S-1300702136@"
l.41 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt"
l.42 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 DATA "
l.43 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "4 TEXT Marriage date: Mar 1983 Marriage place: Scarborough, Yorkshire North Riding"

Record Type=Individual. Gedcom Id=P2. Record Number=3.
l.99 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.100 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE Sep 1951"
l.101 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Scunthorpe, Lincolnshire"
l.102 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 SOUR @S-1300702136@"
l.103 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt"
l.104 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 DATA "
l.105 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "4 TEXT Marriage date: Sep 1951 Marriage place: Scunthorpe, Lincolnshire"

Record Type=Individual. Gedcom Id=P3. Record Number=4.
l.151 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.152 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE Sep 1951"
l.153 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Scunthorpe, Lincolnshire"
l.154 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 SOUR @S-1300702136@"
l.155 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt"
l.156 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 DATA "
l.157 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "4 TEXT Marriage date: Sep 1951 Marriage place: Scunthorpe, Lincolnshire"

Record Type=Individual. Gedcom Id=P6. Record Number=6.
l.213 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.214 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE Mar 1983"
l.215 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Scarborough, Yorkshire North Riding"
l.216 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 SOUR @S-1300702136@"
l.217 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt"
l.218 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "3 DATA "
l.219 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "4 TEXT Marriage date: Mar 1983 Marriage place: Scarborough, Yorkshire North Riding"

Record Type=Individual. Gedcom Id=P7. Record Number=7.
l.224 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 BIRT Upper End"

Record Type=Individual. Gedcom Id=P8. Record Number=8.
l.302 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.303 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE Sep 1952"
l.304 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Willesden, Middlesex"
l.305 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 SOUR @S-1300702136@"

I then get a whole load of RESI tagged data which is the Census info for example
Record Type=Individual. Gedcom Id=P18. Record Number=18.
l.693 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 RESI Relation to Head: Head"
l.702 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 RESI Age: 8Relation: Daughter"
l.708 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 RESI Relationship to Head: Servant"
l.714 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 RESI Marital Status: MarriedRelation to Head of House: Wife"

The census related ones are the bulk of the lines in the exception log.
Looking at the data in FH citation dates and census data looks like the main issues
Regards
Graham

[GrahamJWood]

Mike,
On the census data citations (See attached image) each census year has only one source, but the source reference is correct for this method, but the date is missing, as is the assessment. possible steps are:
Step 1 fill in the date given the source
Step 2 fill in the assessment to "Primary evidence"

Then the more difficult step:
Step 3 re-engineer with one source per household per census and all the data captured as can be done by Ancestral sources when entering from original data.
Step 4 download and attach the image of the page to the new source

Are there any plugin tools to use the data in the file to achieve this any or all of these steps?
Regards
Graham

[tatewise]

Graham, now you have a Project and Exception Report you have a better handle on things.
Despite being a market leader, Ancestry is very poor at implementing Gedcom correctly.

Regarding moving data to notes fields, it only really affects the Death Cause details, and they are they least of your worries.

I would have also expected a lot of UDF reports for the 3 _APID tags, et al.
Use the how_to:handling_unrecognised_data_fields|> Handling Uncategorised Data Fields (UDF) techniques to find and adjust all UDF.

Did not know that Ancestry misused 1 MARR like that. It is unbelievable.
You will have to manually substitute correct Family Marriage Events. Sorry!

Likewise, the misuse of 1 RESI for 1 CENS events.
Even if there were any tools or Plugins to cope with those, the results still would not match the way AS does it with a full transcript, and the Media needs to be linked manually for each household anyway.
So I suspect they all need to be entered using AS as you do now, and then all the redundant 1 RESI and census year Source records need to be deleted.

The Change Any Fact Tag Plugin should help with deleting both 1 MARR and 1 RESI redundant facts.

If anyone else has had to cope with importing from Ancestry they may offer further advice.

[GrahamJWood]

Mike,
On the marriage events see screenshot below, the fact is correct but the citation has no information at all (despite the fact that this information is viewable on Ancestry including district, page etc. This is the raw output in the Ancestry ged file in notebook++ for this individual (too long for screenshot) Followed by the relevant spousal family entry
0 @P13@ INDI
1 BIRT
2 DATE 13 Apr 1897
2 PLAC Sheffield, Yorkshire West Riding, United Kingdom
2 SOUR @S-1300698920@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: 13 Apr 1897 Birth place: Death date: Dec 1980 Death place: Scunthorpe, Humberside, England
3 _APID 1,7579::46255602
2 SOUR @S-1300698734@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: Jun 1897 Birth place: Yorkshire West Riding, United Kingdom
3 _APID 1,8912::25727569
2 SOUR @S-1300698414@
3 PAGE Class: RG13; Piece: 4382; Folio: 94; Page: 32.
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Sheffield, Yorkshire, England Residence date: 1901 Residence place: Attercliffe cum Darnall, Yorkshire, England
3 _APID 1,7814::27904437
2 SOUR @S-1300695133@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Carbrook, Yorkshire, England Residence date: 02 Apr 1911 Residence place: Sheffield, Yorkshire-West Riding, England
3 _APID 1,2352::31118142
1 DEAT
2 DATE Dec 1980
2 PLAC Scunthorpe, Humberside, England
2 SOUR @S-1300698920@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: 13 Apr 1897 Birth place: Death date: Dec 1980 Death place: Scunthorpe, Humberside, England
3 _APID 1,7579::46255602
1 NAME Horace /Osbeldiston/
2 SOUR @S-1300702136@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Marriage date: Sep 1923 Marriage place: Sheffield, Yorkshire West Riding
3 _APID 1,8753::60915507
2 SOUR @S-1300698920@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: 13 Apr 1897 Birth place: Death date: Dec 1980 Death place: Scunthorpe, Humberside, England
3 _APID 1,7579::46255602
2 SOUR @S-1300698414@
3 PAGE Class: RG13; Piece: 4382; Folio: 94; Page: 32.
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Sheffield, Yorkshire, England Residence date: 1901 Residence place: Attercliffe cum Darnall, Yorkshire, England
3 _APID 1,7814::27904437
2 SOUR @S-1300695133@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Carbrook, Yorkshire, England Residence date: 02 Apr 1911 Residence place: Sheffield, Yorkshire-West Riding, England
3 _APID 1,2352::31118142
1 RESI Relation to Head: Son
2 DATE 1901
2 PLAC Attercliffe cum Darnall, Yorkshire, England
2 SOUR @S-1300698414@
3 PAGE Class: RG13; Piece: 4382; Folio: 94; Page: 32.
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Sheffield, Yorkshire, England Residence date: 1901 Residence place: Attercliffe cum Darnall, Yorkshire, England
3 _APID 1,7814::27904437
1 RESI Relation to Head of House: Step Son
2 DATE 02 Apr 1911
2 PLAC Sheffield, Yorkshire-West Riding, England
2 SOUR @S-1300695133@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: abt 1898 Birth place: Carbrook, Yorkshire, England Residence date: 02 Apr 1911 Residence place: Sheffield, Yorkshire-West Riding, England
3 _APID 1,2352::31118142
1 MARR
2 DATE Sep 1923
2 PLAC Sheffield, Yorkshire West Riding
2 SOUR @S-1300702136@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Marriage date: Sep 1923 Marriage place: Sheffield, Yorkshire West Riding
3 _APID 1,8753::60915507
1 SEX M
1 NAME Horace /Osbaldiston/
2 SOUR @S-1300698734@
3 NOTE http://search.ancestry.co.uk/cgi-bin/ss ... try&gss=pt
3 DATA
4 TEXT Birth date: Jun 1897 Birth place: Yorkshire West Riding, United Kingdom
3 _APID 1,8912::25727569
1 SOUR @S-1300695677@
2 PAGE Ancestry Family Trees
2 DATA
3 TEXT http://trees.ancestry.com/pt/AMTCitatio ... 619&pid=13
1 OBJE
2 FILE http://trees.ancestry.com/rd?f=image&gu ... 619&pid=13
2 FORM jpg
2 TITL Granddad O
1 OBJE
2 FILE http://trees.ancestry.com/rd?f=image&gu ... 619&pid=13
2 FORM jpg
2 TITL 1948 Grandma Granddad O Scarb
1 FAMC @F7@
1 FAMS @F5@

0 @F5@ FAM
1 HUSB @P13@
1 WIFE @P12@
1 CHIL @P20@
2 _FREL Natural
2 _MREL Natural
1 CHIL @P2@
2 _FREL Natural
2 _MREL Natural
1 CHIL @P21@
2 _FREL Natural
2 _MREL Natural
1 MARR
2 DATE 29 September 1923
2 PLAC Christ Church, Attercliffe, Sheffield, Yorkshire West Riding, England
2 SOUR @S-1300702136@
3 _APID 1,8753::60891044

You are of course right about the 3 _APID tags everywhere!

I think as a matter of policy Ancestry is using searches to pick up the data for some citations rather than correctly storing it in the ged file. This enables them to lock people in. Based on the search information included in notes fields (first few highlighted in green) it should be possible to generate a better ged file, but I suspect this would be a lot of work (akin to the plugin that gets images mentioned in the guidance).

I will keep investigating, until I fully understand what is going on here, I guess there may be commercial difficulties which are stopping Calico from putting the effort into a full extraction of data. Currently on my main tree I have been double entering on Ancestry and FH as I find the hints and data availability useful on Ancestry, but I like to have full control and good diagrams which FH gives me. If anybody knows that this is going to be improved in a future release I would be grateful for a heads up before I put a lot of effort in.
Regards
Graham

[tatewise]

OK, it seems that the Individual record 1 MARR event is redundant and can be deleted without loss of information.
I suspect the Note simply refers to the GRO Marriage Index that can be easily found knowing the Names, Date, and Place.
The Family record 1 MARR event gives a full Date and Address, so whoever entered that presumably has a Marriage Certificate or it is family knowledge.

Are all the other 1 MARR entries very similar with both Individual and Family record entries?

You are probably correct about creating a better Gedcom file for Sources providing that you have a subscription account for Ancestry in order to access the http:// links to online Source documents.

If your objective is for FH to hold your master database, and for Ancestry to provide hints based on that data, then there is another way avoiding double entry that may be worth investigating.

Use the Export Gedcom File Plugin with the (ANC) Ancestry Family Tree mode and upload that exported Gedcom to Ancestry. However, each time you do that I guess there will be a lot of hints for sources you have already found, but you will know which branches need researching and can ignore all the others.

It is also possible to export to RootsMagic in a similar way and use its TreeShare with Ancestry feature.

[GrahamJWood]

MIke,
looking at the ones I posted, it is worse than that, there are actually two marriages erroneously recorded for this person, one from the marriage index, the other from family information (see screen shot) This might explain why one is against the individual as there is no spouse linked, so what ancestry has done here is logical. I need to fix any entries like this on Ancestry before exporting, the exception report may be a help in doing this. I will repost when the data is fixed.

I haven't chosen to put my FH ged on Ancestry because it has a lot more info and photos than the online copy, and I don't want to risk publishing that information, like you I don't entirely trust their safeguards
Regards
Graham

[mjashby]

Graham,

The thing to understand about both Ancestry Tree and Family Tree Maker is that neither product uses; and nor do they understand the concept of Family Events/Facts. To both products, all Events/Facts are Individual so, in the case of Marriages, both parties have individual Marriage Events and the software uses a 'tagging' system which enables the user to select the appropriate spouse/partner which needs to have that same Individual Fact/Event. One of the major problems with this comes you export a GEDCOM from either product, as there will be no Family Facts/Events to directly link the spouses, although this isn't always apparent as they are often 'brought together' when GEDCOMs are imported by their parental links to the same children.

So, in your first marriage entry you have a full marriage entry tagging both spouses in your Ancestry Tree, whereas the second (GRO) Fact is only matched to one spouse. I suspect that when you matched the GRO Source, either you or the Ancestry system didn't relate that Individual Source to the other Marriage entry. This relatively easy to correct on Ancestry:

- Select the Source and click 'View'
- Select 'Associated Facts'
- Click the '+' against the Marriage Fact you want to be associated with and '-' against any Fact Event you want to disassociate from the Source.
- Delete the unnecessary Fact once it has been completely isolated from from any Source.
- Select the Spouse and see if any similar changes need to be made.

It's a tedious approach to correction, but hope it helps.

Mervyn

[tatewise]

Graham, you are effectively confirming what I said, that there are two Marriage Events records: one for the Individual, and one for the Family couple. The Individual one for the Marriage Index can only give the registration quarter & district, BUT they are both the same Marriage, AND yes fixing them in Ancestry would be a good idea.

The Export Gedcom File technique does NOT transfer any Media images to Ancestry, and it is feasible to filter the details as explained in how_to:exporting_gedcom_with_multimedia#export_to_website_without_sensitive_data|> Export to Website Without Sensitive Data.

If the main purpose of obtaining Ancestry hints is to discover Census & BMD records, then there are alternative methods.
FH V6 offers hints in the Focus Window for FindMyPast and MyHeritage as explained in how_to:internet_data_matches|> Internet Data Matches. That advice cross-refers to two Plugins that help search for records missing from your database and currently support Ancestry, FindMyPast, FamilySearch & MyHeritage. None of those involve uploading any of your data to any online trees.

BTW: I have to disagree with Mervyn regarding Family Events/Facts. Both Ancestry & FTM most definitely do understand Family Facts. Any Gedcom that I upload to Ancestry with Family Marriage Events are shown in Ancestry with each Marriage Event linking the two spouses exactly as in FH. When the Gedcom is downloaded from Ancestry it has a Family record holding each Marriage Event as confirmed by Graham's earlier posting that shows the downloaded Individual and Family records, each holding one Marriage Event. It is only the Individual Marriage Event that is abnormal.

When adding a Fact in Ancestry for Marriage it asks who the spouse is in order to create the correct Family Event. It is not obvious how the Individual Marriage Event can be added as it does not appear on the standard list, so I suspect it may have been created by this particular user as a Custom Event.

[GrahamJWood]

Mervyn,
Thanks for the advice, I did already know how to fix the Ancestry marriages (I did it the other way around by looking at the fact I wanted to keep and associating the source, then deleting the redundant fact), normally I would have spotted this sooner, but I wasn't giving this tree my full attention until it became formally attached to mine through the marriage of my daughter last weekend. Much of this tree was initially created by my daughter's mother-in-law, who hasn't been using Ancestry as long as I have and is less familiar with it's many opportunities to duplicate data.

Mike,
The only online subscription I have is Ancestry (worldwide), so I haven't used the internet matching, it would be great if Ancestry were an option, but I am not aware of any moves in that direction at the moment, if anybody else is please let me know, I would be happy to help in any early trials.
Regards
Graham

[GrahamJWood]

I have found another type of anomaly, If a birth has entry has a description (see screenshot attached) this throws up the following exception:

Record Type=Individual. Gedcom Id=P7. Record Number=7.
l.224 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1 BIRT Upper End"

From this entry in the Ancestry raw gedcom:
0 @P7@ INDI
1 BIRT Upper End
2 DATE 10 January 1928
2 PLAC Shipton Under Wychwood, Oxfordshire, England

Which is imported as:
1 BIRT
2 DATE 10 JAN 1928
2 PLAC Shipton Under Wychwood, Oxfordshire, England
2 _UNCAT Upper End

I don't know if this is legitimate Gedcom or if "Upper End" should be transferred as 2 NOTE Upper End
This is there because of Ancestry's failure to handle "address" detail separately to "place".
Regards
Graham

[tatewise]

The two Plugins mentioned, for searching for missing records, work well with an Ancestry subscription.

The Birth anomaly is another Ancestry misuse of Gedcom that does NOT allow values on same line as Event tags.
You may well find this for lots of types of events.
Whether FH puts it in 2 _UNCAT or 2 NOTE depends on Tools > Preferences > File Load/Save and Move invalid data into note fields where possible.
Try importing with that option ticked and unticked to see the differences.
The advantage of putting into 2 _UNCAT is it allows a custom Plugin to move the _UNCAT values to Address in most cases, or wherever is appropriate for each type of fact.

[GrahamJWood]

Marriage event exceptions.
I have now resolved all of these (48 across 1120 people), having two marriage events for one actual marriage is an issue, but not what causes the exception. The exception is caused by not selecting the spouse in the marriage event, this is done automatically in Ancestry for certain marriage sources and not for others
Example exception
Record Type=Individual. Gedcom Id=P888. Record Number=888.
l.31268 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "1 MARR "
l.31269 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 DATE January 1914"
l.31270 - INFO ONLY: Loaded uncategorised data (non-GEDCOM): "2 PLAC Plomesgate, Suffolk"
l.31323 - INFO ONLY: Detected & fixed field format error (data loaded using Uncategorised Data Field): "1

If a spouse is not selected, Ancestry associates the event with an individual record rather than a family record. This is a different interpretation of the GEDCOM rules to that which FH makes (I am not expert enough to know which are correct, possibly both). FH would create a family record with an unknown spouse. If the spouse is not known one can still be created in Ancestry(I used the list of available surnames on the page if it was a FreeBMD based before 1915)

[tatewise]

I can categorically state that a Marriage Event (1 MARR) in an Individual record is invalid GEDCOM.

If you wish you can look it up via fhugdownloads:contents:gedcom_standard_release_5.5|> Utility ~ GEDCOM Standard Release 5.5 starting on Page 30 it lists Family Events that includes MARR followed by Individual Attributes and Individual Events on Page 31 that omits MARR.

Also there you will see that Attributes such as EDUC and OCCU can have a <value> on same line, but Events such as BIRT and DEAT can have Y in special circumstances, but otherwise nothing.

I have investigated many genealogy products as part of my Export Gedcom File Plugin, and FH has the most compliant implementation of GEDCOM 5.5. There are also several GEDCOM Validators you could find via a Google search, but beware there are two definitions in popular use: GEDCOM 5.5 and GEDCOM Draft 5.5.1. They are very similar, but a few key differences can impact successful migration between products using the different versions.

[GrahamJWood]

Mike,
Excellent! You cited your evidence, Ancestry has got its marriage events wrong.
Regards
Graham

[AdrianBruce]

... Ancestry has got its marriage events wrong. ...

What's probably happened (oh dear, I'm defending Ancestry...) is that if you think about the FreeBMD type marriage indexes, they don't index a pair of people, but one of the parties and give some other possibilities for their spouse. So if you were to apply such an index to your tree on Ancestry, and the said person's other spouse isn't already there, then the marriage can only be applied as an individual event to the one party.

Even when you add the other spouse in, I think it may stay as an individual event on one party and an individual event on t'other party.

Individual marriages events are, of course, as Mike says, not valid GEDCOM.

So when Ancestry produces its GEDCOM under these circumstances, it creates individual marriages events in GEDCOM - naughty!

I imagine that if you already had the two parties as married in your Ancestry tree, then I think that the FreeBMD type index is applied to the couple's marriage event and this would be valid GEDCOM. Maybe. Possibly. If so, this would account for why it's not a hugely prevalent issue when downloading GEDCOMs from Ancestry. Unless, of course, you know better...

[GrahamJWood]

Adrian,
please read my earlier post. I have removed all 48 Marriage related errors in Ancestry by ensuring there were two people linked to each marriage event by selecting a spouse (if necessary create one first), it did not then generate erroneous Marriage tags in the Individual record.
Regards
Graham