Gramps Workflow Ideas

classic Classic list List threaded Threaded
24 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Gramps Workflow Ideas

Steve Youngs

Hi,

I'm looking at moving my research across to Gramps and I'm looking for ideas on a good workflow. My research is 99% UK based so I'll use UK examples below.

Let's take for example an 1891 census for England, RG12 piece 1562 folio 64 page 07, schedule 38. A fairly typical residence, if there is such a thing. Let's assume that this is the only information I have at this point in my research.

I can use the Forms gramplet to transcribe the information from this schedule. That get's me basic People records, with just names, and a shared census Event record with linked Media and a Citation records.

 

Q1. What is a good strategy for the Reference field in the census Form? RG12-1562-64-07-38 would work for this example, but often a family splits across folios and/or pages. I think institutions are slightly different.

Q2. Is there a reason why the piece, folio, page and schedule numbers are not included as structured data in the form?

Q3. For married / widowed females would you set the name type to "Married Name" (I appreciate it's not a certain conclusion based on a census image, just as a male could have changed their name)

Q4. Is there anyway to populate the form data from an external source (e.g. www.freecen.org), given a appropriate reference? e.g. by screen scraping, REST API, ....

 

Of course the census entry provides us with lots more information about these individuals: birth date, where born, marriage status, children, occupation, residence, ...

 

Q5. Is a good workflow to create Events, Families, Places for all of this, adding the Citation created by the Forms gramplet?

e.g. for James Goff birth about 1840, occupation Groom,

for Mary Goff, birth about 1841

a Family record from James and Mary

children with father set to James Goff

a residence event shared by the people

....

Q5. Is there any way to automate this from the information already entered to the Forms gramplet? e.g. right click a form cell and have an entry "create event..."

Q6. From the citation, for example in the Event editor for a birth event, what's the easiest way to display the census form again?

 

My typical strategy is then to take this low probability census information and use other sources to increase the confidence

  Is James the father of the children?

  Is Mary the mother of some / all the children?

  Were James and Mary married?

  etc. etc.

 

Thanks in advance for your help and ideas.

Steve



_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

GRAMPS - User mailing list
Have you looked at the 2007 wiki webpage
Recording UK Census data - Gramps

I haven't read it closely and any dozen year old article would probably benefit from being updated... particularly a UK topic by a UK Gramps user, such as yourself.

Some seems out-of-date. Things like Census add-on had been supplanted by the Form Add-on in version 4.2.x

You've asked a LOT of questions for any single maillist posting... it's a bit overwhelming. Plus, there are many opinions about recording Census data that have been repeatedly discussed -- without reaching a concensus. 

How about searching the wiki & the maillist archive to see what is there? Maybe you can help consolidate the existing information on the wiki and help expand the topic.

Google the user guide wiki:
  census site:gramps-project.org/wiki

Search "UK census" on the Gramps-Users maillist archive:

-Brian

On Sun, Apr 28, 2019 at 6:31, Steve Youngs

Hi,

I'm looking at moving my research across to Gramps and I'm looking for ideas on a good workflow. My research is 99% UK based so I'll use UK examples below.

Let's take for example an 1891 census for England, RG12 piece 1562 folio 64 page 07, schedule 38. A fairly typical residence, if there is such a thing. Let's assume that this is the only information I have at this point in my research.

I can use the Forms gramplet to transcribe the information from this schedule. That get's me basic People records, with just names, and a shared census Event record with linked Media and a Citation records.

 

Q1. What is a good strategy for the Reference field in the census Form? RG12-1562-64-07-38 would work for this example, but often a family splits across folios and/or pages. I think institutions are slightly different.

Q2. Is there a reason why the piece, folio, page and schedule numbers are not included as structured data in the form?

Q3. For married / widowed females would you set the name type to "Married Name" (I appreciate it's not a certain conclusion based on a census image, just as a male could have changed their name)

Q4. Is there anyway to populate the form data from an external source (e.g. www.freecen.org), given a appropriate reference? e.g. by screen scraping, REST API, ....

 

Of course the census entry provides us with lots more information about these individuals: birth date, where born, marriage status, children, occupation, residence, ...

 

Q5. Is a good workflow to create Events, Families, Places for all of this, adding the Citation created by the Forms gramplet?

e.g. for James Goff birth about 1840, occupation Groom,

for Mary Goff, birth about 1841

a Family record from James and Mary

children with father set to James Goff

a residence event shared by the people

....

Q5. Is there any way to automate this from the information already entered to the Forms gramplet? e.g. right click a form cell and have an entry "create event..."

Q6. From the citation, for example in the Event editor for a birth event, what's the easiest way to display the census form again?

 

My typical strategy is then to take this low probability census information and use other sources to increase the confidence

  Is James the father of the children?

  Is Mary the mother of some / all the children?

  Were James and Mary married?

  etc. etc.

 

Thanks in advance for your help and ideas.

Steve

_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org

Untitled Download Attachment
Untitled (268 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Brad Rogers
In reply to this post by Steve Youngs
On Sun, 28 Apr 2019 11:10:56 +0000 (UTC)
Steve Youngs <[hidden email]> wrote:

Hello Steve,

As with much of Gramps/genealogy, there is not one 'right' answer.  I,
and no doubt others, will offer up methods.  It's then up to you to
choose.  The main thing, obviously, is to be consistent.

>Q1. What is a good strategy for the Reference field in the census Form?
>RG12-1562-64-07-38 would work for this example, but often a family
>splits across folios and/or pages. I think institutions are slightly

If a family splits across pages, I enter each part of the family
separately.  Thus, if Husband, wife and first child are at
RG12-1562-64-07, and remainder of family are at RG12-1562-64-08, then
that's the two refs I use.  Note that I normally don't bother entering
the entry number (col.1 on the forms).  I only use it where two
different families in my tree are on the same page - a fairly rare
occurrence (for my tree, at any rate).

>Q2. Is there a reason why the piece, folio, page and schedule numbers
>are not included as structured data in the form?

That's a question for the forms creator - Nick Hall, I believe.  My
apologies to both Nick and the real creator, if I'm wrong.

>Q3. For married / widowed females would you set the name type to
>"Married Name" (I appreciate it's not a certain conclusion based on a

Until I'm certain, I enter the surname in brackets, thus (Surname).
Only when I know the maiden name do I change that.  Further, I always
use maiden name as the preferred name.

>Q4. Is there anyway to populate the form data from an external source
>(e.g. www.freecen.org), given a appropriate reference? e.g. by screen

No.

>Of course the census entry provides us with lots more information about
>these individuals: birth date, where born, marriage status, children,
>occupation, residence, ...

All of which are, in reality, conjecture;  Birth dates (well,
calculated birth years) in particular, are only as good as the person
giving the info.  Even then, if they were lied to, the data will be
wrong.  Furthermore, the calculated year is, most often, off by one
year.  That is, one year later than the actual birth year.

>Q5. Is a good workflow to create Events, Families, Places for all of
>this, adding the Citation created by the Forms gramplet?

As you can imagine from my answer to Q4, I don't enter events for birth
marriage, occupation, etc. based on census info.  In part, because the
census event replaces or combines (take your pick) the residence &
occupation events.  Also, as previously mentioned, much of everything
else is conjecture.

>Q5. Is there any way to automate this from the information already
>entered to the Forms gramplet? e.g. right click a form cell and have an

I suspect that would require a change in the gramplet API.

BTW, that's two Q5s   :-)

>Q6. From the citation, for example in the Event editor for a birth
>event, what's the easiest way to display the census form again?

The only way to display the data entry form is to go back to person view.
By selecting source citation, references etc. from an event window to get
to the census event, you'll get a similar census event window open.  All
the necessary data is there (Attributes), of course.  It's just that the
appearance differs.

--
 Regards  _
         / )           "The blindingly obvious is
        / _)rad        never immediately apparent"
Black man got a lot of problems, but he don't mind throwing a brick
White Riot - The Clash


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org

attachment0 (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Ron Johnson
In reply to this post by Steve Youngs
On 4/28/19 6:10 AM, Steve Youngs wrote:

Hi,

I'm looking at moving my research across to Gramps and I'm looking for ideas on a good workflow. My research is 99% UK based so I'll use UK examples below.

Let's take for example an 1891 census for England, RG12 piece 1562 folio 64 page 07, schedule 38. A fairly typical residence, if there is such a thing. Let's assume that this is the only information I have at this point in my research.

I can use the Forms gramplet to transcribe the information from this schedule. That get's me basic People records, with just names, and a shared census Event record with linked Media and a Citation records.

 

Q1. What is a good strategy for the Reference field in the census Form? RG12-1562-64-07-38 would work for this example, but often a family splits across folios and/or pages. I think institutions are slightly different.


A future researcher might not know what RG12-1562-64-07-38 mean.  Thus, I'd make the Citation "Volume/Page" line say something like this ( which is from one of my own citations):
Whitton, Whitton, Durham / Page 4 / No. 16  - Reference RG12/4054

For a citation that spans multiple pages, I'd do:
Whitton, Whitton, Durham / Pages 4-5 / No. 16  - Reference RG12/4054


Q2. Is there a reason why the piece, folio, page and schedule numbers are not included as structured data in the form?


I think Gramps citations follow the GEDCOM standard.

Q3. For married / widowed females would you set the name type to "Married Name" (I appreciate it's not a certain conclusion based on a census image, just as a male could have changed their name)


Yes.  (Until such time as I discovered her maiden name.)

Q4. Is there anyway to populate the form data from an external source (e.g. www.freecen.org), given a appropriate reference? e.g. by screen scraping, REST API, ....

 

Of course the census entry provides us with lots more information about these individuals: birth date, where born, marriage status, children, occupation, residence, ...

 

Q5. Is a good workflow to create Events, Families, Places for all of this, adding the Citation created by the Forms gramplet?

e.g. for James Goff birth about 1840, occupation Groom,

for Mary Goff, birth about 1841

a Family record from James and Mary

children with father set to James Goff

a residence event shared by the people


I don't use Forms.  Thus, in situations like this, I
1. Create (or edit) the first person
2. Create a Citation from the Census, along with the image.
3. Put the Citation in the clipboard.
4. Create all reasonably deducible events for that Person, citing them from the citation in the clipboard.  (Ones that can be shared also get copied to the clipboard.)
5. If this is a new Family, create it, cite it, etc.
6. Create (or edit) the next person.
7. Create all individual Events (remembering to cite them) and dragging shared events from the clipboard.
8. Update the Family object as necessary.
8. Go to step 6 until all people are added.

....

Q5. Is there any way to automate this from the information already entered to the Forms gramplet? e.g. right click a form cell and have an entry "create event..."

Q6. From the citation, for example in the Event editor for a birth event, what's the easiest way to display the census form again?

 

My typical strategy is then to take this low probability census information and use other sources to increase the confidence

  Is James the father of the children?

  Is Mary the mother of some / all the children?

  Were James and Mary married?

  etc. etc.

 

Thanks in advance for your help and ideas.

Steve




--
Angular momentum makes the world go 'round.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

ACProctor
In reply to this post by Steve Youngs
Re: Q1: TNA has a comprehensive scheme for cataloguing all the materials it holds, and they publish recommendations for how it should be used to cite their materials at Citing Documents. For the censuses, this amounts to using the departmental code, series number, piece number, and book number (for 1841) for the specific census "item"; and then internal identifiers of folio and page number to identify a specific page within that item. For instance, ‘HO 107/11/12, folio 12, page 19’ in 1841, or ‘RG 9/2460, fo. 43, p. 27’ in 1861. Ranges of consecutive page numbers are conventionally represented as n1n2, where the separator is a n-dash character. NB: this scheme also applies to BMD registrations.

Tony (2nd attempt due to daft message size limits)

On 28/04/2019 12:10, Steve Youngs wrote:

Hi,

I'm looking at moving my research across to Gramps and I'm looking for ideas on a good workflow. My research is 99% UK based so I'll use UK examples below.

Let's take for example an 1891 census for England, RG12 piece 1562 folio 64 page 07, schedule 38. A fairly typical residence, if there is such a thing. Let's assume that this is the only information I have at this point in my research.

I can use the Forms gramplet to transcribe the information from this schedule. That get's me basic People records, with just names, and a shared census Event record with linked Media and a Citation records.

<snip>



_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Paul Womack-2
In reply to this post by Brad Rogers
I appreciate that everybody can make their own choices about how to record in Gramps, but since you're offering advice, I will comment.

To treat census data as mere "conjecture" seems extreme - all records can be wrong, all records are written by fallible or potentially deceitful humans. Applying this principle consistently would lead to rejecting all records, and finding another hobby.

In practise I treat genealogy from a scientific approach, i.e. All family trees (genealogical databases) are the best approximation to absolute truth one can make, supporting by the evidence one has. The process of research is one of improving the approximation.

For example, if a person has ever been alive, they MUST have a birth event. How precisely and reliably the date or location of that birth event is known will hopefully improve over time.

  BugBear


>Of course the census entry provides us with lots more information about
>these individuals: birth date, where born, marriage status, children,
>occupation, residence, ...

All of which are, in reality, conjecture;  Birth dates (well,
calculated birth years) in particular, are only as good as the person
giving the info.  Even then, if they were lied to, the data will be
wrong.  Furthermore, the calculated year is, most often, off by one
year.  That is, one year later than the actual birth year.

>Q5. Is a good workflow to create Events, Families, Places for all of
>this, adding the Citation created by the Forms gramplet?

As you can imagine from my answer to Q4, I don't enter events for birth
marriage, occupation, etc. based on census info.  In part, because the
census event replaces or combines (take your pick) the residence &
occupation events.  Also, as previously mentioned, much of everything
else is conjecture.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Peter Merchant
In reply to this post by ACProctor
Ouch, I can see that I have a lot of retrospective work to do to record things that other people can recreate. My practice has been to rename all downloaded items  as date-event-person i.e. '1891-census-william Merchant' and thus losing the other identification details. I'll have to find them again and perhaps edit the pic to add the text to it.

But thanks to all for the interesting discussion here.

Peter M.

On 29/04/2019 09:46, Tony Proctor wrote:
Re: Q1: TNA has a comprehensive scheme for cataloguing all the materials it holds, and they publish recommendations for how it should be used to cite their materials at Citing Documents. For the censuses, this amounts to using the departmental code, series number, piece number, and book number (for 1841) for the specific census "item"; and then internal identifiers of folio and page number to identify a specific page within that item. For instance, ‘HO 107/11/12, folio 12, page 19’ in 1841, or ‘RG 9/2460, fo. 43, p. 27’ in 1861. Ranges of consecutive page numbers are conventionally represented as n1n2, where the separator is a n-dash character. NB: this scheme also applies to BMD registrations.



_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Brad Rogers
In reply to this post by Paul Womack-2
On Mon, 29 Apr 2019 10:20:17 +0100
Paul Womack <[hidden email]> wrote:

Hello Paul,

>To treat census data as mere "conjecture" seems extreme - all records
>can

Obviously, I didn't make myself clear;  I was referring specifically to
births there.  Almost inevitably, as calculated from census data, it will
be wrong by one year.  So, I don't add a birth event based on census
data.  I *do*, however, use that data as a starting point to locate a
birth or baptismal (if before registration began) record.

The cynical me says that, TBH, pretty much every record _is_ conjecture
anyway;

Example 1
I know of people that, according to their birth certificate were born
one or two months after the actual event.  Why?  So the parents didn't
have to pay the late registration fee.

Example 2
I have a marriage certificate relating to one man who I can find no other
references to, whilst the woman seems to be involved soon after with a
man of a totally different name.  Even the children of the married
couple have a different (third) surname.  So, is that one man, known
by two (or three) different names, or two men?  Thus far, I don't know.
I suspect I never will.

Example 3
My own birth.  Nobody checked the time when I was born.  It was 'about
midnight'.  Was I born on the 18th or the 19th?  My birth certificate
records the date as the 19th, but no one /really/ knows.

Example 4
A chap in my tree is one of two people with that name, born in the same
place about 18 months apart.  As it happens, it's one of the men in
example 2 - the one that doesn't have a marriage event, so I have no
father's name as a guide.  So, which of the births is the correct one
for my man?  No idea.

I know that, ultimately, we *have* to follow the paper, but we should
all remember that:
A mother is a matter of fact.  A father is a matter of faith.

>For example, if a person has ever been alive, they MUST have a birth
>event. How precisely and reliably the date or location of that birth

Clearly, and I don't pretend otherwise.  I just don't clutter my tree
with vague events.  Gramps is where I detail the 'facts', as you say, as
best we can determine them.

--
 Regards  _
         / )           "The blindingly obvious is
        / _)rad        never immediately apparent"
We are the chosen
Changed - Judgement Centre


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org

attachment0 (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

ACProctor
In reply to this post by Peter Merchant
Formal citations would use the references as defined by the associated archive, Peter -- also identifying the specific archive, of course -- but usually supplemented with written details such as personal names, dates, district/locality, etc. This would be Elizabeth Shown Mills' recommendation, and the extra information would serve as a backup if the reference was in error. I know many of us UK folk take shortcuts because these codes are very well-defined, and our census search engines have to accommodate a search-by-code. Also, because those recommendations may seem too verbose.  :-)

However, note that there's a difference between how you organise your materials (images, documents, etc), and how you index them. Your suggested file naming is a very common approach, but it falls down for items that mention multiple names, places, and dates (e.g. probate documents). I am sorry to say that I recently had a rant about this (https://parallax-viewpoint.blogspot.com/2019/01/organising-more-resources.html) because a number of writers with archival experience -- who should know better -- were recommending it, when it clearly clashed with their own approach with physical items in an archive. As a person with a strong IT background, I had to put the record straight.

Tony

On 29/04/2019 10:57, Peter Merchant wrote:
Ouch, I can see that I have a lot of retrospective work to do to record things that other people can recreate. My practice has been to rename all downloaded items  as date-event-person i.e. '1891-census-william Merchant' and thus losing the other identification details. I'll have to find them again and perhaps edit the pic to add the text to it.

But thanks to all for the interesting discussion here.

Peter M.

On 29/04/2019 09:46, Tony Proctor wrote:
Re: Q1: TNA has a comprehensive scheme for cataloguing all the materials it holds, and they publish recommendations for how it should be used to cite their materials at Citing Documents. For the censuses, this amounts to using the departmental code, series number, piece number, and book number (for 1841) for the specific census "item"; and then internal identifiers of folio and page number to identify a specific page within that item. For instance, ‘HO 107/11/12, folio 12, page 19’ in 1841, or ‘RG 9/2460, fo. 43, p. 27’ in 1861. Ranges of consecutive page numbers are conventionally represented as n1n2, where the separator is a n-dash character. NB: this scheme also applies to BMD registrations.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Ron Johnson
In reply to this post by Brad Rogers
On 4/29/19 5:10 AM, Brad Rogers wrote:
On Mon, 29 Apr 2019 10:20:17 +0100
Paul Womack [hidden email] wrote:

Hello Paul,

To treat census data as mere "conjecture" seems extreme - all records
can
Obviously, I didn't make myself clear;  I was referring specifically to
births there.  Almost inevitably, as calculated from census data, it will
be wrong by one year.  

This is why the "About" date modifier exists.

So, I don't add a birth event based on census
data.  

Well, you should, with an "About" date modifier.

I *do*, however, use that data as a starting point to locate a
birth or baptismal (if before registration began) record.



--
Angular momentum makes the world go 'round.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Brad Rogers
On Mon, 29 Apr 2019 07:37:15 -0500
Ron Johnson <[hidden email]> wrote:

Hello Ron,

>Well, *you should*, with an "About" date modifier.

No.....

I *could*.

I _choose_ not to.

--
 Regards  _
         / )           "The blindingly obvious is
        / _)rad        never immediately apparent"
They really dig me man, and I dig them
To Be Someone (Didn't We Have A Nice Time) - The Jam


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org

attachment0 (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

John Paton
In reply to this post by Brad Rogers
On 28/04/2019 14:00, Brad Rogers wrote:
> All of which are, in reality, conjecture; Birth dates (well,
> calculated birth years) in particular, are only as good as the person
> giving the info. Even then, if they were lied to, the data will be
> wrong. Furthermore, the calculated year is, most often, off by one
> year. That is, one year later than the actual birth year.

And even furthermore - in the UK Census for 1841 (at least in Scotland)
the ages of adults (ie over 15) were *supposed* to be rounded down to
the nearest 5 years *by the enumerator* - some enumerators  did, some
didn't, some may have rounded up. So a person who gave their age as 29
should have had the age rounded down to 25 when the information from the
shedules was transferred to the Enumerators book. If the enumerator
didn't round down or up then all of what Brad says above is still true
but for the 1841 census there is the further issue that ages given as
exact multiples of 5 are often going to be incorrrect.

John Paton




_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Ron Johnson
On 4/29/19 8:11 AM, John Paton wrote:

> On 28/04/2019 14:00, Brad Rogers wrote:
>> All of which are, in reality, conjecture; Birth dates (well, calculated
>> birth years) in particular, are only as good as the person giving the
>> info. Even then, if they were lied to, the data will be wrong.
>> Furthermore, the calculated year is, most often, off by one year. That
>> is, one year later than the actual birth year.
>
> And even furthermore - in the UK Census for 1841 (at least in Scotland)
> the ages of adults (ie over 15) were *supposed* to be rounded down to the
> nearest 5 years *by the enumerator* - some enumerators  did, some didn't,
> some may have rounded up. So a person who gave their age as 29 should have
> had the age rounded down to 25 when the information from the shedules was
> transferred to the Enumerators book. If the enumerator didn't round down
> or up then all of what Brad says above is still true but for the 1841
> census there is the further issue that ages given as exact multiples of 5
> are often going to be incorrrect.

I'd love to know the justification for that rule...

--
Angular momentum makes the world go 'round.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

John Paton
On 29/04/2019 14:41, Ron Johnson wrote:

> On 4/29/19 8:11 AM, John Paton wrote:
>> On 28/04/2019 14:00, Brad Rogers wrote:
>>> All of which are, in reality, conjecture; Birth dates (well,
>>> calculated birth years) in particular, are only as good as the
>>> person giving the info. Even then, if they were lied to, the data
>>> will be wrong. Furthermore, the calculated year is, most often, off
>>> by one year. That is, one year later than the actual birth year.
>>
>> And even furthermore - in the UK Census for 1841 (at least in
>> Scotland) the ages of adults (ie over 15) were *supposed* to be
>> rounded down to the nearest 5 years *by the enumerator* - some
>> enumerators  did, some didn't, some may have rounded up. So a person
>> who gave their age as 29 should have had the age rounded down to 25
>> when the information from the shedules was transferred to the
>> Enumerators book. If the enumerator didn't round down or up then all
>> of what Brad says above is still true but for the 1841 census there
>> is the further issue that ages given as exact multiples of 5 are
>> often going to be incorrrect.
>
> I'd love to know the justification for that rule...
>
I suppose they were doing the Census for purposes decided by the
government of the day - didn't give a thought to the problems it might
cause us today - most inconsiderate of them :).

I know that it applied in Scotland - Scotlands People website is clear
on that. I expect that it applied throughout the UK for the 1841 Census
- it was after all a UK Government Census and the SNP weren't around
then. From 1851 onwards 'exact' ages were recorded.

John




_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

ACProctor
In reply to this post by Ron Johnson
The rounding down of ages over 15 -- which also happened in England and Wales (or was supposed to) -- was for statistical reasons. They were used to
get a rough profile of adult ages.

When working out a birth-date from a census -- or any document that gives an age on a particular date -- then it's not always one year out. Ignoring
the rounding issue, the Canadian censuses changed between age-at-last birthday and age-at-next birthday (I cannot recall which way the change was
now), and all census returns might have an age given in months, weeks, or even days, in which case the calculated range of possible birth-dates would
be narrower.

Some time ago, I created a little tool to help genealogists calculate a range of possible birth-dates given a date-of-recording and an age, possibly
given in different units (see http://parallax-viewpoint.blogspot.com/2015/05/measurement-tools.html#BirthDateCalc). This also deals with the UK
rounding issue, and the Canadian issue. The range can then be represented using a GEDCOM date of BET rather than ABT. I hope someone finds this useful.

Tony

On 29/04/2019 13:37, Ron Johnson wrote:

> On 4/29/19 5:10 AM, Brad Rogers wrote:
>> On Mon, 29 Apr 2019 10:20:17 +0100
>> Paul Womack<[hidden email]>  wrote:
>>
>> <snip>
>> Obviously, I didn't make myself clear;  I was referring specifically to
>> births there.  Almost inevitably, as calculated from census data, it will
>> be wrong by one year.
>
> This is why the "About" date modifier exists.
>
<snip>


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

GRAMPS - User mailing list
The logical reason Canada would change to age at next birthday is that they'd be at least that age by the time the data was tabulated... or being referenced for budget decisions.

But I wonder if logic applies in Canadian government processes any more than it does in the US bureaucracy?

-Brian

"Goverment. If you think the problems we create are bad, wait until you see our solutions." - a Despair.com deMotivational poster.



On Mon, Apr 29, 2019 at 11:39, Tony Proctor
The rounding down of ages over 15 -- which also happened in England and Wales (or was supposed to) -- was for statistical reasons. They were used to
get a rough profile of adult ages.

When working out a birth-date from a census -- or any document that gives an age on a particular date -- then it's not always one year out. Ignoring
the rounding issue, the Canadian censuses changed between age-at-last birthday and age-at-next birthday (I cannot recall which way the change was
now), and all census returns might have an age given in months, weeks, or even days, in which case the calculated range of possible birth-dates would
be narrower.

Some time ago, I created a little tool to help genealogists calculate a range of possible birth-dates given a date-of-recording and an age, possibly
given in different units (see http://parallax-viewpoint.blogspot.com/2015/05/measurement-tools.html#BirthDateCalc). This also deals with the UK
rounding issue, and the Canadian issue. The range can then be represented using a GEDCOM date of BET rather than ABT. I hope someone finds this useful.

Tony

On 29/04/2019 13:37, Ron Johnson wrote:

> On 4/29/19 5:10 AM, Brad Rogers wrote:
>> On Mon, 29 Apr 2019 10:20:17 +0100
>> Paul Womack<[hidden email]>  wrote:
>>
>> <snip>
>> Obviously, I didn't make myself clear;  I was referring specifically to
>> births there.  Almost inevitably, as calculated from census data, it will
>> be wrong by one year.
>
> This is why the "About" date modifier exists.
>
<snip>



_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Peter Merchant
In reply to this post by ACProctor
I also have been in IT since before it was known as IT, albeit in communications technology rather than databases. My take on your argument is that data must be useable to the end user, whoever that might be. Yes in a database we need an index that might be the set of not-quite random digits that form the  source identification, but they could also be my date-event-person as this is also unique.  For whoever uses this in the future, going through my files by family folder and date-event-name filename will make a lot more sense than a folder for 'source' containing a lot of randomly (apol.)  named files.

I would ideally like to have this set of digits associated with my identification to verify that I have found the correct record. Either I could edit it into the picture, or add it as metadata, or have an extremely long filename.

Also I suggest (not proven) that searching the Internet/source for a date/event/person will find the record whereas searching for  40365_293708-00225 is unlikely to find this source.

Am I  considering this all wrong? I must read your blog and the Eliz.Sh. Mills stuff and see if I can find common ground.

regards,
Peter M.





On 29/04/2019 11:23, Tony Proctor wrote:
Formal citations would use the references as defined by the associated archive, Peter -- also identifying the specific archive, of course -- but usually supplemented with written details such as personal names, dates, district/locality, etc. This would be Elizabeth Shown Mills' recommendation, and the extra information would serve as a backup if the reference was in error. I know many of us UK folk take shortcuts because these codes are very well-defined, and our census search engines have to accommodate a search-by-code. Also, because those recommendations may seem too verbose.  :-)

However, note that there's a difference between how you organise your materials (images, documents, etc), and how you index them. Your suggested file naming is a very common approach, but it falls down for items that mention multiple names, places, and dates (e.g. probate documents). I am sorry to say that I recently had a rant about this (https://parallax-viewpoint.blogspot.com/2019/01/organising-more-resources.html) because a number of writers with archival experience -- who should know better -- were recommending it, when it clearly clashed with their own approach with physical items in an archive. As a person with a strong IT background, I had to put the record straight.

Tony

On 29/04/2019 10:57, Peter Merchant wrote:
Ouch, I can see that I have a lot of retrospective work to do to record things that other people can recreate. My practice has been to rename all downloaded items  as date-event-person i.e. '1891-census-william Merchant' and thus losing the other identification details. I'll have to find them again and perhaps edit the pic to add the text to it.

But thanks to all for the interesting discussion here.

Peter M.

On 29/04/2019 09:46, Tony Proctor wrote:
Re: Q1: TNA has a comprehensive scheme for cataloguing all the materials it holds, and they publish recommendations for how it should be used to cite their materials at Citing Documents. For the censuses, this amounts to using the departmental code, series number, piece number, and book number (for 1841) for the specific census "item"; and then internal identifiers of folio and page number to identify a specific page within that item. For instance, ‘HO 107/11/12, folio 12, page 19’ in 1841, or ‘RG 9/2460, fo. 43, p. 27’ in 1861. Ranges of consecutive page numbers are conventionally represented as n1n2, where the separator is a n-dash character. NB: this scheme also applies to BMD registrations.



_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

ACProctor
Let me clarify a little, Peter. The codes that the TNA document are archival ones, used for organisation of their materials. They are not digital
identifiers, and the fact that UK census search engines can accept them as search criteria is (I believe) stipulated by TNA, but it wasn't part of
their specific design.

Online resources should, ideally, be treated as "digital archives", and the providers should define permanent digital identifiers that could be used
to locate the same materials over and again, directly and without having to always go through a search engine . Being forced to use a search engine to
find the same digital material as you found last time leaves you open to databases having changed with new records added or contents merged/divided.
Digital identifiers could also distinguish transcriptions from images, and even different versions of each (e.g. if an error is fixed in one of them).
Sorry ... another rant over ... more details at https://parallax-viewpoint.blogspot.com/2016/12/impermanent-links.html. Maybe one day archives and IT
people will have a similar approach.

Tony

On 30/04/2019 09:33, Peter Merchant wrote:

> I also have been in IT since before it was known as IT, albeit in communications technology rather than databases. My take on your argument is that
> data must be useable to the end user, whoever that might be. Yes in a database we need an index that might be the set of not-quite random digits
> that form the  source identification, but they could also be my date-event-person as this is also unique.  For whoever uses this in the future,
> going through my files by family folder and date-event-name filename will make a lot more sense than a folder for 'source' containing a lot of
> randomly (apol.)  named files.
>
> I would ideally like to have this set of digits associated with my identification to verify that I have found the correct record. Either I could
> edit it into the picture, or add it as metadata, or have an extremely long filename.
>
> Also I suggest (not proven) that searching the Internet/source for a date/event/person will find the record whereas searching for 
> 40365_293708-00225 is unlikely to find this source.
>
> Am I  considering this all wrong? I must read your blog and the Eliz.Sh. Mills stuff and see if I can find common ground.
>
> regards,
> Peter M.
>
>


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Ron Johnson
In reply to this post by Brad Rogers
On 4/29/19 7:52 AM, Brad Rogers wrote:

> On Mon, 29 Apr 2019 07:37:15 -0500
> Ron Johnson <[hidden email]> wrote:
>
> Hello Ron,
>
>> Well, *you should*, with an "About" date modifier.
> No.....
>
> I *could*.
>
> I _choose_ not to.

I wrote "should", not "must".


--
Angular momentum makes the world go 'round.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Gramps Workflow Ideas

Ron Johnson
In reply to this post by ACProctor

Everything you write is true, but doesn't it require all researchers to know
about these digital identifiers beforehand?

On 4/30/19 3:51 AM, Tony Proctor wrote:

> Let me clarify a little, Peter. The codes that the TNA document are
> archival ones, used for organisation of their materials. They are not
> digital identifiers, and the fact that UK census search engines can accept
> them as search criteria is (I believe) stipulated by TNA, but it wasn't
> part of their specific design.
>
> Online resources should, ideally, be treated as "digital archives", and
> the providers should define permanent digital identifiers that could be
> used to locate the same materials over and again, directly and without
> having to always go through a search engine . Being forced to use a search
> engine to find the same digital material as you found last time leaves you
> open to databases having changed with new records added or contents
> merged/divided. Digital identifiers could also distinguish transcriptions
> from images, and even different versions of each (e.g. if an error is
> fixed in one of them). Sorry ... another rant over ... more details at
> https://parallax-viewpoint.blogspot.com/2016/12/impermanent-links.html.
> Maybe one day archives and IT people will have a similar approach.
>
> Tony
>
> On 30/04/2019 09:33, Peter Merchant wrote:
>> I also have been in IT since before it was known as IT, albeit in
>> communications technology rather than databases. My take on your argument
>> is that data must be useable to the end user, whoever that might be. Yes
>> in a database we need an index that might be the set of not-quite random
>> digits that form the  source identification, but they could also be my
>> date-event-person as this is also unique.  For whoever uses this in the
>> future, going through my files by family folder and date-event-name
>> filename will make a lot more sense than a folder for 'source' containing
>> a lot of randomly (apol.)  named files.
>>
>> I would ideally like to have this set of digits associated with my
>> identification to verify that I have found the correct record. Either I
>> could edit it into the picture, or add it as metadata, or have an
>> extremely long filename.
>>
>> Also I suggest (not proven) that searching the Internet/source for a
>> date/event/person will find the record whereas searching for 
>> 40365_293708-00225 is unlikely to find this source.
>>
>> Am I  considering this all wrong? I must read your blog and the Eliz.Sh.
>> Mills stuff and see if I can find common ground.
>>
>> regards,
>> Peter M.

--
Angular momentum makes the world go 'round.


_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
12