Picking apart the tangled mess of fact and fiction that is Williams_AndersForWikiTree.ged

Question

I've spent the last 2 days doing some major surgery on some profiles from this import, separating husbands from wives and sorting out the children. The typical problem is that the name of a person's spouse is correct, but it's the wrong person. Some details of each person are correct but they were never a couple.

My next operation is to slice and dice [[Nilsdotter_Brunell_Hustru-1|Sara Nilsdotter Brunell]]. She is from Brunskog, Värmland. I am not interested in her, but will correct her husband and leave some notes for some other volunteer.

Comments

Gerry have you checked on "Litteratur" on Värmlandsrötter looks like we have transcribed church books... and also on CD

• Brunskog Husförhörslängd 1759-63 Rolf Brodin
• Brunskog födde, vigde, döde 1688-1930 Sven Haage
• Brunskogs församlingsbok 1688-1720 Sven Haage
• Brunskogs Husförhörslängder 1760-1799 Sven Haage
• Brunskogs och Mangskogs befolkning 1503-1761 Sven Haage

 

I tried to read Brunskog EI:1 (1691-1763) Image 7 / page 7 (AID: v5051.b7.s7, NAD: SE/VA/13065) and for me this is difficult.....

Anbytarforum has changed URL so I added the new URL to transcription help on category Brunskog (S)

1. A good family tree on internet about her
   1. Has links to anbytarforum that mention her
   2. Sources in this tree
      1. Mangskog d 1706-1709, 1712-1737, Mangskog församling, (SVAR C:1, 55591).
      2. Mangskog f. 1706-1708, 1712-1730, Mangskog församling, (SVAR C:1, 55591).
      3. Brunskog v 1691-1710, 1712-1763, Brunskog församling, (SVAR E:1, 1691-1763).
         

---

I'm also chomping away on this infamous gedcom when I come across it, doing roughly the same thing as you, Gerry - sourcing and correcting the ones that are relevantly connected to people/parishes I'm working on and familiar with, cutting the rest adrift. In some cases I recycle a singleton that was imported without dates or place names - there'll always be a place for someone with an Olsdotter or Andersson LNAB somewhere.

I have come to think of this beast as the You_Know_Who.ged

---

It wasn't my intention to re-ignite the flame war over whether or not to have gedcom imports. I have spoken my piece in that debate in the past. Rather, I was letting the community know that I had removed & corrected wrong information from this profile, but I wasn't going to research it any further -- maybe someone would be interested or related.

Magnus, I think it would be appropriate for you to consider the points raised here:

http://www.wikitree.com/g2g/81792/gedcoms-make-a-mess?show=81833#a81833

and here:

http://www.wikitree.com/g2g/81792/gedcoms-make-a-mess?show=81844#a81844

If not for the benevolence of Anders Williams, so many profiles would never have been added. As for counting sources -- each profile from this gedcom that I have looked at has several. So that would pass your test with flying colours. The mess is not because of sources, but from totally bogus linkages.

Blame it on the error report! A single "child born before father" error opened a can of worms that took me 3 days to untangle.

 

---

Re Gerry :-)

That is the cool thing with the Project database Error report you feel that you have one big family tree...

I have generated 30 generations report on http://www.sdms.si:92/default.htm and then I connect to places Eva is doing nice genealogy...

I think we need to have a GEDCOM debate. WikiTree is not a static platform and now with the Database Error project we get more focus on quality and also we get numbers on problems inside WikiTree.

Is WikiTree a place to store unsourced GEDCOM files or is it people doing genealogy together..... I would like to see more focus quantity that is important...

For me WikiTree is a place to work together doing genealogy and learn from each other not a place to upload GEDCOM files and jump out the back..... I am really impressed of you trying to understand the church books of Brunskog EI:1 (1691-1763) Image 7 / page 7 (AID: v5051.b7.s7, NAD: SE/VA/13065) I tried to do a screen dump of the page but did find what part of the page Sara Nilsdotter Brunell was mentioned on ;-) 

---

Hwa? Second couple from bottom right. Marriage record. Thanks for bothering with screen dumps, Magnus. That's one aspect I have little patience with.

---

Thanks

I do a try.... saw that Wikimedia have the possibility to annotate in pictures.... would be magic to be able to transcribe pictures in WikiTree

See Commons:Image_annotations

Examples:

1. Berliner_kongress.jpg
2. Wikimania_2012_Group_Photograph-0001.jpg
3. Vitraux_de_la_basilique_Notre-Dame,_Gen%C3%A8ve_23.jpg

Answer 1

I am on your side.... the mentioned gedcom is not serious when you look at what is uploaded in Sweden is my feeling. Unsourced and a taste of bitter Ancestry Green leaves genealogy were connecting is done by guessing not sources. 

We need to stop import gedcoms without sources for every singel person

1. If Wikitree has an honour code that everything should be sourced why is it ok to upload gedcom files were not each singel person has sources
    
2. Lesson learned many people who upload unsourced GEDCOMs don't clean them and add sources ==> maybe have date on created profiles from GEDCOMs that you have 6 months to "clean" this profile after that sorry this profile can't be part of WikITree its not good enough...   
    
3. GEDCOM uploaded profiles are mostly a mess and nothing readable, We should do Genealogy in WikiTree not storing unsourced GEDCOMs 

 

Comments

Well the gedcom in case was imported back in 2011, I guess rules weren't as strict then. I do hope it couldn't happen again.

For my own part I made a few small experiments with gedcom import and then decided to enter people manually. Gedcom cleanup was as much work and less fun. I think I saw somewhere that you did a similar test-and-discard.

---

Sorry Eva it's still a mess... 

And I guess we will spend 1000 of hours cleaning up just the mess the gedcom in the title did with Swedish sources...

It takes 10 minutes to upload and a life time to fix - WikiTree Unsourced GEDCOM imports 2016

Below a chart I did last week checking if GEDCOM are more sourced today and the answer is NO NO NO I guess with more people on internet and more people duplicating unsourced trees on Ancestry this will go crazy.

1. I checked 130+ uploaded GEDCOMS during 2 days (not imported just uploaded)
2. One was good sourced - 1148 sources on 111 profiles link 
3. Nine had more sources than profiles created
4. For me good genealogy then you have at least 10-20 sources per profile. Below a green line everything below that line has less sources than profiles defined

Google Chart

 

My understanding of GEDCOM inside WikiTree

1. Everyone who works seriously on Wikitree feels Gedcoms uploads are a mess.
    
2. The only "positive" people I have seen are people doing genealogy where you have nearly no sources and an unsourced Ancestry family tree is a breadcrumb they don't want to miss....
   
   My feeling in Sweden with all our good genealogy sources I feel we don't trust a source if its not from a church book and a good reference to Arkiv Digital or SVAR. References to Ancestry and GID to old version of scanned church books are not good enough.... a GEDCOM like the one mentioned above you never trust as it feels like it's not serious genealogy more green leaves clicking and getting excited that you have roots to 1560 in a country called Sweden and even better if it's a Swedish King.....

 

Chart with number of profiles and number of identified sources in uploaded GEDCOMs during 2 days

If people uploaded well researched gedcom files then the proportion should be red lines should be 10-20 times higher than the blue...

We have a big big conflict with WikiTree being a single World Wide Family tree and then import gedcom files from places they tell people please duplicate profiles from each other.....

What needs to be done according to me as unsourced GEDCOMS will not be less. 

1. I miss some kind of task force inside WikiTree were you identify problems and look into solutions right now G2G is some kind of undefined pipe to WikiTree development. Aleš Trtnik is doing magic things using his magic power outside WikiTree as that is the easiest/fastest way...
    
2. Having an honour code telling that we use sources is not serious if you can upload profiles without sources...
    
3. Today we say it's ok to upload 100 profiles with not a single source change that to all uploaded profiles need to have at least one source prefered is that every fact has at least one evidence  
   1. change the GEDCOM import to be skip as default
   2. skip "grayed" if no sources to a profile ==> you can't import the person f you have no sources...
   3. when creating a WikiTree profile from a Gedcom add a GEDCOMImport template 
      1. Add a function like Watchlist Unsourced to find profiles in your Watchlist with the GEDCOMImport template ==> you find those profiles easier....
      2. Add errors in the Database Error project to find profiles with the template ==> profile is just imported and not cleaned.... 

---

I joined WikiTree a little over a month ago, and uploaded my GEDCOM with about 500 people.  Nearly all of them have sources on MY database, but when they were uploaded, most of those sources were turned into some seemingly random numbers I don't understand.  Maybe it was something I did when I sent it, and I'm slowly trying to put them back.  Just wanted you to know that some unsourced profiles may not be intentional.

---

That is very similar to my experience with the import of a sourced gedcom, Bob. I was more cautious, though - a much smaller import.

---

Eva Bob .... do you have any profiles you remember who had this error....

I assume there is a lot of problems with export and import so its no simple equation.  

On the GEDCOM upload page my understanding is you can see how many SOUR tags you have in a profile. As one source can be used on more profiles its not an exact measurement.... and also some software add the source to a Note tag....

Below how a source looks like in a Gedcom

0 @S-472261927@ SOUR
1 REPO @R-997111160@
1 TITL 1940 United States Federal Census
1 AUTH Ancestry.com
1 PUBL Ancestry.com Operations, Inc.
1 _APID 1,2442::0

I think it's no easy solution to this but it feels like a growing problem....   

---

Thanks, Magnus.  Look at 'Amalia Louise Hedvig Cronsioe (1831 - 1910)', for example.  Just one of very many like it. There are strange numbers as sources.

---

Thanks Bob

If we check that gedcom upload it was 795 individuals and 136 sources and earliest person born 1660 

A fast look at the GEDCOM file I feel that the Ancestry export of data is the problem it is not much in the file....

 

My quick analysis...

If we check the raw file and Amalia we have

0 @I393@ INDI her unique number in file
1 _UID 33656EC8937A4D4BB8367FC7A0E5A2DCAA74
1 NAME Amalia Louise Hedvig /CRONSIOE/
2 SOUR @S2@ using source S2
1 SEX F
1 BIRT
2 DATE 23 JAN 1831
2 PLAC Malmö Garnisonsfsg, Malmö, Malmöhus län
2 SOUR @S488@ using source S488 
1 DEAT
2 DATE 1 JAN 1910
2 PLAC Malmö (M)
1 FAMS @F61@
 

If we check the sources S2 and S488

0 @S2@ SOUR
1 TITL GF documents
1 AUTH Genealogiska Föreningen, Riksförening för Släkstsforskning

0 @S488@ SOUR
1 TEXT S-686485590

And this is what the GEDCOM import created...

The WikiTree profile  Cronsioe-1

 

So if we should import more Sources I think the problem is that the EXPORT need to export all sources in the GEDCOM....

A guess is that GEDCOM is not high prio for Ancestry... I have the client FTM and a subscription and I am not impressed how fast new functions are delivered for Ancestry.....  they have last year sold the FTM client to a foreign company so what will happen in the future is difficult to say....

My conclusion from this one sample If we have problems that GEDCOM export is weak then its another argument not using GEDCOM import...
 

EDIT Bob when I check Ancestry I assume this is Amalia
http://person.ancestry.com/tree/78082985/person/40378484886/facts

Then we just also have 2 sources at Ancestry..... if you have another exampel please let me know....

Answer 2

The most irksome aspect of this Frankensteinian monster of a gedcom and other similar ged-concoctions of parts from different familytrees is when WikiTree turns up as a source for the erroneous connections somewhere else on the Web, as I see from time to time.