Lets say you have a census with information about a person.
You download the relevant pages only.
Is those pages downloaded the source,
or
Are the hole census the source and the spesific pages more of a citation?
And the web site itself you got it from is the repository. (In my case, https://www.digitalarkivet.no)
It would depend on how you have organized your Census records. You do not say which country so I’ll use the U.S. as my example.
Census could be the source for all census records wherever found. All other information goes into the citation. Year, Country, State, …
U.S. Census and U.K. Census could be two sources with the rest of the relevant information going into the citations. Year, State, …
U.S. 1880 Census and U.S. 1900 Census with each year becoming a source. This is my method. The Citation becomes State, County, Locale, District, Page, Line or Family.
But you can create more sources adding the State to each year’s census. U.S. 1880 Census for Massachusetts and U.S. 1880 Census for New York
This becomes a balance of Sources versus Citations. Fewer Sources would have more Citations. More Sources would each have fewer Citations.
The Source’s title plus the Citation’s volume/page line should have everything needed to find the record again. Where you make the break that becomes the Source with the remainder becoming the Citation is your choice. As I said, I make the break at the census by country and year for the Source portion of the record.
And before you commit yourself to a particular style, try printing some of the Gramps reports that contain endnotes so see if you like how your sources and citations appear there.
I’ve been using the Form (formerly Census) gramplet for census data for
many years and have adopted an approach that fits with it. So the
source is the census for a particular place and year (such as U.S.
Federal Census for 1940 or Iowa State Census for 1925) and I have a
unique citation for each family (or person in the case of a census like
the 1915 Iowa census which used a separate card for each person). The
image(s) of the page(s) are added to the gallery for the citation. This
means a page may be attached to multiple citations if I create events
for more than one family on the page. See
https://gramps-project.org/wiki/index.php?title=Form_Addons for more
information on the gramplet.
I haven’t determined yet how I should use repositories and haven’t made
any use of that feature so far. As I learn more and have time, I’ll
probably go back and deal with that.
Allen Crider
As I can see on this site, it’s the digitalized version of the Norway National Archives. So repository should be the Norway National Archives and it’s website url the reference to that repository, especially url to the source document you are citing. With Reference Type: Electronic.
In France, censuses was made nation wide every 5 years. So I’d created some France Census, 18xx and I’d created citations for each page and each family on a page I used from a census. But like @GeorgeWilmes said result was not what I want.
So, practically censuses was realized by towns the same first quarter of these years, I splitted my by-year sources into by-town and by-year sources. It’s nicer to see in that form and easier to link repository to each source (more than one hundred archives repositories in France where censuses are dispatched in), and reports are more pleasant to see. But I still use a page and a family on that page as base to my citation records.
The reason for sources, is to show the reader where they can get to see the document etc that you are quoting.
So the Repository should tell the reader where to go. Website, Library, Aunt Maria’s Cupboard etc.
The Source should identify the book, document series, photograph album etc.
The Citation should identify the page or document or picture in a way that is unambiguous and foolproof. Archives have their own way of uniquely identifying documents, which is, in my opinion, over the top, but also a good place to start.
Seems like many of you do for example:
Source: 1900 Census (Oslo)
Citation: One page from it (If its one page for the house/apartment itself and then another for people in it, those two should probably be the same citation then?)
People/Events: link all related events and people to that citation. (So things share citation)
After some testing, most reports dont show citations or sources at all, and the only one I tested that show media attatched to them is the Narrated web one, and that one only show the media of Citations, not the ones that is on Sources.
Related note, is there some good third party reports/web reports?
Yes, not decded if I will have “Norwegian National Archives” as Repository or “Digitalarkivet” as one yet as they are different web pages.
After looking in gramps to understand what you mean, do you mean that link to the spesific Cesus on their site goes in to the “Call Number” field here?
I have noticed there is no “internet” tab on Sources so i cant put it there, and Digitalarkivet do not allow you to download more than 100 pages at the time so cant just add the hole thing as media. I dont know if its point to have the hole census downloaded either for that matter.
I will download the spesific pages that have info I use tho (rather than link to the spesific pages)
link to Digitalarkivet home page goes in the internet tab in the Repository thing.
Also, Digitalarkivet is a Archive but also a web site, wonder if I should make a new repository type that is Electronic Archive or just have it as Archive.
Norway did every 10 years, but some places had local one in between those some times.
So you try to do one citation per family even if two of them are at the same page?
I think I will now have one censius as one Source and not lump them together per year, aka most years it will be for example one for Oslo and another for Ă…lesund the same year.
Ah, ok.
You could export your data to GEDCOM and use any software or web site that will import it. In that case, your source and citation information will be stored in the GEDCOM fields for TITL, AUTH, PAGE, DATE etc. according to how to arranged them in Gramps. So again, it is a good idea to try any other solutions using a sample of your data before you put a lot of effort into data entry.
Are you sure ? normaly media are shown for all objects (citation, sources)
here is an example for the next gramps5.2:
http://noiraud.allowed.org/Gramps52/en_GB/src/9/9/b39fe3f390e30bd2b99.html
I just checked again and dont know how I missed it but you are correct that it do show it, dont know how I missed it first time.
But media citation is showed right on the event/person without clicking the citation but on Source you have to spesifically click the source to see it.
Sorry about that.
BTW on that one, the one for 5.2, I like some of the changes I see 
Wouldnt it then be problems with media and stuff?
Also with third party I ment plugins too
I’ve taken this position reading Evidence Explained wich indicates that our readers must know where to access the original record. To do that I named the physical archive center in the repository record. In the source record, the reference to that repository is electronic and the call number is the web site call number/reference. Here for a Belgium birth register found on FamilySearch web site:
Citation of this source looks like that, URL to FamilySearch record is stored into a citation attribute:
Yes when I’m working on a family basis (“foyer” is the french word for a fireplace, a family home place):
When I’m working on a name basis (in a whole small village) I’ve only one citation entry for that research:
