DATES misinterpreted in Scandinavian records.
Scandinavian records generally (I am not sure on the exact list of countries this would include) have the dates written as Day/Month. Thus 9/5 should be interpreted as the 9th of May NOT the 5th of September. I am seeing that the Source Linker is reading them WRONG. In the RECORD view the date is interpreted/displayed correctly. Thus the INDEXING was done correctly. The source linker should NOT change the record and interpret the date the wrong way. Computer automation should prevent errors not cause them. Here is an example of what I am talking about.
https://www.familysearch.org/ark:/61903/1:1:QLGN-WNGS You will see in this record view Karna Larsen birthdate is listed as 9 May 1914. If you open up the image you see Karna near the bottom with the birth written 9/5 1914. The record screen thus shows the correct date.
here is the source linker for this record — see on the left hand side of the screen that the source linker is getting the date wrong. It is writing out the date as the 5 Sept 1914. You can see that her brother Karl Larsen has the same problem. He was — according to the image — born the 1/4 1909 — which should be interpreted as the 1 April 1909 and source linker is written is birth as the 4th of January
How come the record viewer gets it right and the source linker gets it wrong? This needs to be fixed as soon as possible. Using the source linker to add new people will get their birthdates wrong. Using the source linker to decide if a record is a match is confusing. Please consider a fix.
Comments
-
As another user adding comments, I haven't seen this before and if this is supposed to be an improvement, this is badly done. If you go to the record viewer link given above, there is a ^ over in the right margin of the date display. Click on that and you will see that both dates are displayed. The same mark is by the place and clicking there shows that the record actually has three possibilities for the place name.
Assuming the record was originally indexed with correct information, why was the record edited to add incorrect information and make this a multiple choice exercise? Since the source linker cannot handle more than one choice for a data item and so has to choose one of the possibilities at random, this is even more of a problem.
0 -
oh I hadn't ever noticed the ^ choices for dates and places. Interesting. I notice that hovering on the ^ gives the text "view edit history". Does this mean that in some point the 5th of Sept was saved with the record but that it was changed/altered (by a person? by a computer?) to the correct 9th of May? I'm still not understanding how the error is happening. But I still maintain that if if the RECORD VIEWER knows the right date how come the SOURCE LINKER doesn't?
1 -
Can the record viewer algorithm use the same sort of clues that a human would use? Namely 1) it's a Scandinavian record and there is a certain way they right their dates and 2) while there is ambiguity in this particular date (if you didn't know the Scandinavian date conventions) but there's plenty others on the page to look to to decide how dates are being written on the page.
0