I added an alternative place name starting from 1 march 2022.
In the reference editor this always resolves to the alternative name, even if the date is clearly before the alternative. This is the same in the places category. All is OK in the person editor.
Making reading and understanding a problem.
Is this normal behavior?
In the event editor, the place always displays using today’s date. When the event is saved then Gramps can have the place name displayed based upon the event’s date.
Yes. If the primary/preferred Place name does not have a date defined and an Alternative Name does have a date, then the Alternative date will take precedence.
since the “To” and “Before” date range/span are functionally the same… they are both affected by by the selections in the Limits tab of Preferences. There is NO unbounded range.
similarly the “From” and “After” date range/span are functionally the same… they are also both affected by by the selections in the Limits tab of Preferences.
So you have to put in (absurd) end values in a full “From - To” instead.
so instead of a “before 1950”/“to 1950”, I put in “From 0 1 to 1950”
and instead of a “after 1950”/“from 1950”, I put in “From 1950 to 3000”
Never realised this so detailed, Just assumed it was for ‘born before’ etc.
it protests with using 0 (zero), but 1 is accepted in my case. (just tried)
1 Like
The limits set in preferences for ‘before’, ‘after’ and ‘about’ do NOT AFFECT place records. ‘before 1941’ in a place record would include all dates from the start of time up until 1941. ‘after 1941’ would include all dates after 1941 until today.
LIMITS SET IN PREFERENCES DO NOT APPLY TO PLACE RECORDS!!!
Additional things to keep in mind:
When a date is entered with just the year, Gramps sees it as 1 January of that year. ‘after 1941’ is actually ‘after 1 January 1941’. Likewise, if just the month and year are entered, Gramps sees it as day 1 of that month.
An Event with no date will display the place name with the earliest possible date. Think, what place name would display on 1/1/0001.
does from 1 jan 1900 include 1 jan 1900?
I tried this: ABC from 1 jan 1900. Then made a birth on 1 jan 1900 and it didn’t register.
I thought after 1 jan should start at the 2nd, but from should start at the first.
Similar for until (include boundary) and before (exclude boudary)
No! It would be true starting on 2 Jan. If you use ‘from 1 Jan 1900 to 2000’ then yes, it will be true. Do not ask me why. An open-ended ‘from’ with no closing ‘to’ is a fairly new Gramps ability.
If you set a place date ‘after 1941’ it starts from 1 Jan 1941 including the first. If you set the place date ‘after 1 Jan 1941’ it starts on 2 Jan 1941.
I try to avoid using ‘before’ in the place records. Instead, I will leave the last entry in either the Enclosed by tab or the Alternative Names tab with no date information.
I like to use a town in Connecticut as an example of how I set up a place record.
The place Name and Alternatives.
And its Enclosed by records.
Notice that there are entries with no date information in the last entries for both names and enclosed by tabs. This is for the events that may have no dates. My default catch all.
Giving these results
Remember, that all place records within the hierarchy of a place can have their own name changes and other enclosed by records. The goal is to set up records so Gramps can navigate an event’s date path back up through the hierarchy.
And no one but you sees the place records. The goal is to have the place in a person’s event list display correctly and that those relatives born in the same place, regardless of name changes, can all be found together, in both the small village or the large city.
I must disagree. When using discussing the Display Name feature for Places it only affect the output rather than the records themselves. (At least in 5.2 version. And since the Place rework stalled, it seems unlikely to be different in newer releases.)
In this example, Sharon does not even have any Alternative Names. But the Primary Name has a founding date. And if the date of the Event does not fall on that date, then Gramps displays a ?
instead of the city name. And the entire township of Hickory (1833-1974) was incorporated into the borough of Hermitage in 1976. And the USA primary is “USA” with “From 4 Jul 1776” but when there is no date for an event, it chooses to use the “British Colonies” with “to 4 Jul 1776”
The name Sharon will only display when the event occurs on 6 Oct 1841. There is no ‘after’ included. And if the place never had a name change then no date information is required.
Here is my Pennsylvania record. I use the date the Articles of Confederation went into effect for the name change and then to be enclosed by USA. Yes, there is a gap between 4 Jul 1776 no longer enclosed by the UK/Great Britain record and then being enclosed by the USA.
Mercer and Allegheny Counties never had a name change and only ever existed in Pennsylvania.
Here is my Hermitage record.
Here is my record for Sharon. There are no dates on the name field or alternative names.
I do not understand how Townships interact with anything. We do not have them in New England. And the Sharon Wikipedia page did not make reference to any township. If you tell me, I’ll add it the Sharon hierarchy.
You should look at the Genealogical Map of the Counties (pdf) from the Pennsylvania Historical and Museum Commission. The borders changed frequently.
Where Allegheny is now was originally (unofficially) considered part of Lancaster before 1755. Subdivided to be part of Cumberland by 1780 then it was broken out into Westmoreland county. When northwestern portion of Pennsylvania was acquired from a 1784 treaty with the Six Nations, Allegheny was created in 1790 from part of it and part of Westmoreland. Which was subdivided to create Mercer (among other counties) around 1802. And then the southern portion of Mercer and the north portion of Beaver county became Lawrence county (my home county) in 1849.
But with regard to Sharon and Hickory/Heritage, they only existed in Allegheny and then Mercer. Any record for these two places never occured with these names in Lancaster, Cumberland or Westmoreland counties.
And while these counties may of shrank or changed borderlines, their names never changed and remained inside the borders of Pennsylvania.
This must be a bug. “After 1941” should mean after 1941‑12‑31 23:59:59 (or 00:00 on the next day), including all days after that point.
-
From 1941 should mean from 1941‑01‑01, including that day.
-
After 1941 should mean from 1942‑01‑01, including 1942‑01‑01.
-
After 1941‑January should mean from 1942‑02‑01 and later.
In other words:
-
“From ” must include the written day.
-
“After ” must exclude the written day.
-
“From ” must include the entire month.
-
“After ” must include every day after that month.
-
“After ” must exclude all days in that year and include only days after the last day of that year.
This is basic chronological logic.
‘After 1941’ should therefore mean after 31 Dec 1941.
GEDCOM’s interpretation (‘after 1 Jan 1941’) causes overlap between “1941” and “After 1941”, which is chronologically impossible.
LDS temple‑workflow logic ≠ historical chronology.
Gramps inherits this theological scheduling convention, which leads to incorrect historical periods.
—
At some point we need to stop treating GEDCOM as if it were a neutral technical standard. It isn’t. It is a theological workflow format that has been adopted as if it were a historical data model — and that blind adoption keeps producing errors like this.
GEDCOM’s date semantics were designed for a very specific internal process, not for historical chronology. When software follows those semantics uncritically, we end up with results like:
-
“After 1941” starting on 1 Jan 1941,
-
overlapping ranges between “1941” and “After 1941”,
-
and interval boundaries that make no chronological sense.
This isn’t a Gramps problem — it’s a GEDCOM inheritance problem.
And as long as genealogy software continues to follow GEDCOM’s conventions without questioning them, historically incorrect periods will continue to propagate.
If we want historically valid timelines, we have to stop treating GEDCOM as a universal truth and start treating it as what it actually is: a legacy format with built‑in assumptions that do not match historical reality.
—
Note: The following explanation was first written in Norwegian and then translated into English with assistance from Microsoft Copilot.
I would like to thank everyone in the discussion for helping me conclude
that my approach to “Places” is correct in that they do not have dates.
“Events” have dates, “Places” have Lat:Lon.
I use Alternative Names purely as a list of names through time as an
aide memoire.
If you cannot define somewhere by area within a small radius around a
physical point then it is an “Administrative/Academic Concept” ie the
Roman Empire, the British Empire, the United States of America.
The main reason I initially took this approach was the UK Registration
system of “Registration Districts” for Births, Marriages and Deaths
which has been in a constant state of change since it was enacted in
1837 trying to use these as physical locators of events for say a heat
map would cause chaos.
phil