Forms are great. But....!

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

Forms are great. But....!

Colin Spiller

Hi

I like gramps, and I like Forms. Forms are a great way of entering data. But only when they work correctly!

I'm using GRAMPS: 5.0.1
Python: 3.6.9 (default, Apr 18 2020, 01:56:04) [...
BSDDB: 6.1.0 (5, 3, 28)
sqlite: 3.22.0 (2.6.0)
LANG: en_GB.UTF-8
OS: Linux
Distribution: 4.15.0-101-generic

As this is UBUNTU 18.04.4 LTS that's the latest release and my databases are therefore BSDDB.

I have 6600 individuals and 3700 families.

BUT! Mixed in amongst this lot are many corruptions. Take this chap James Holloway see https://drive.google.com/open?id=1o_2mHTWqvgDgdRRMkYm2ZvBCVPmunvjV

I found him age 11 in the 1911 census and used the form to enter the family some time ago.

Now I've come to look at him again, the form looks like this https://drive.google.com/open?id=1KjLxiBHQwle3SRCcdlL7YJCdWg6pXnFj

It's a totally different family! Quite a valid family but James Holloway is nowhere to be seen in it And I have lots of other similar examples.

No I haven't aborted gramps in the middle of an update. Yes I have used "Check and repair database".(No issues - well 1 missing image which I'll fix in a minute).  Yes I have rebuilt secondary indexes. Yes I have rebuilt reference maps [What's a reference map?] Yes I have backed up, created a new tree and imported the backup. And the problem still exists. Yes I'm getting really frustrated with the whole thing! By the way, the 'strange' census form is still correctly linked to the people shown in it, as well as to James and his family.


I have two problems.

1. How do I fix it? I can delete all the entries in the strange 1911 census form and re-enter both families again linking to the correct people, but this empty 1911 form is still linked to James. And re-entering them creates duplicate 1911 census events for everyone I re-enter. And my original census form which includes James H must still exist somewhere,

2. How do I locate all the other instances where this has happened?  And I know here are quite a few because I keep tripping over them.


Is it possible somehow to produce a report or similar showing say Form, name and the address from the form? Or something. Anything. But even then, how to fix it?

Help needed before I junk the whole lot please.

Colin


-- 
Colin Spiller
[hidden email]


--
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: Forms are great. But....!

GRAMPS - User mailing list
On 22/05/2020 21:12, Colin Spiller wrote:
And my original census form which includes James H must still exist somewhere

All the form information is stored in event and event reference attributes.  You can look at these using the standard Gramps editors to check that the information is still there.

I expect that the problem is caused when two form events share the same citation.  This citation is passed to the form editor and the wrong event is chosen.

If this is the case, then I'll fix the form gramplet.  The way the editor is invoked changed slightly when I introduced the form functionality into the previous census addon.


Nick.




--
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: Forms are great. But....!

prculley
In reply to this post by Colin Spiller
To understand what may have occurred, you have to understand how forms are attached to your db.

  1. A form is connected to a Source (US 1940 Census) by giving that source the 'Form' attribute with the value 'US1940' which identifies which Gramps form to use.
  2. When you create a form, this creates an Event, which is filled with the place and date at the top of the form.
  3. It also attaches a Citation (which refers to the original source) to the Event.
  4. When you originally fill out a form, you pick one or more Person, Families etc. Doing so, attaches that event to the associated people or families.
  5. The actual form data ends up in what we call the EventRef attributes, which are visible when you first edit the person, and then from the bottom 'Events' tab, edit the Event (census, in this case), and look at the top part of the 'Event Reference Editor' 'Attributes' tab.
And then when you want to look at a form again:
  1. When you select a person in the active view, the Form Gramplet looks at the Events attached to him, and any Events attached to his immediate Families.
  2. The Gramplet then examines the Events, looking at Citations attached to the Event.
  3. For each Citation, it then looks at the Source the Citation refers to to see if it has the 'Form' attribute and the form type.
  4. It then knows what type of form to show.
  5. When you select and edit a form, the Form Editor looks at the Citation and tracks it back to the appropriate Event, and then back to the people and families where it gets the Event References to fill out the form.

Probably more than you ever wanted to know about how Forms work.

In order for the attached form to get confused like this, it seems likely the 'Reference Maps' which track backwords from objects to the other objects that refer to them may be corrupted. 

If you Edit the census Event attached to the person, and look at the References tab, it should list the person(s) from the census, if the list doesn't include the starting person, the 'Reference Maps' are corrupted.

If check & repair, or Rebuild Reference Maps did not fix the issue, I would suggest exporting your db to Gramps XML, starting a new tree and importing it again.  This should also have the effect of recreating the Reference maps.

I cannot recall any specific bugs in Gramps 5.0.1 that would cause this, although a LOT of bugs have fixed since then.  You may want to consider an upgrade https://github.com/gramps-project/gramps/releases

Paul C.

On Fri, May 22, 2020 at 3:30 PM Colin Spiller <[hidden email]> wrote:

Hi

I like gramps, and I like Forms. Forms are a great way of entering data. But only when they work correctly!

I'm using GRAMPS: 5.0.1
Python: 3.6.9 (default, Apr 18 2020, 01:56:04) [...
BSDDB: 6.1.0 (5, 3, 28)
sqlite: 3.22.0 (2.6.0)
LANG: en_GB.UTF-8
OS: Linux
Distribution: 4.15.0-101-generic

As this is UBUNTU 18.04.4 LTS that's the latest release and my databases are therefore BSDDB.

I have 6600 individuals and 3700 families.

BUT! Mixed in amongst this lot are many corruptions. Take this chap James Holloway see https://drive.google.com/open?id=1o_2mHTWqvgDgdRRMkYm2ZvBCVPmunvjV

I found him age 11 in the 1911 census and used the form to enter the family some time ago.

Now I've come to look at him again, the form looks like this https://drive.google.com/open?id=1KjLxiBHQwle3SRCcdlL7YJCdWg6pXnFj

It's a totally different family! Quite a valid family but James Holloway is nowhere to be seen in it And I have lots of other similar examples.

No I haven't aborted gramps in the middle of an update. Yes I have used "Check and repair database".(No issues - well 1 missing image which I'll fix in a minute).  Yes I have rebuilt secondary indexes. Yes I have rebuilt reference maps [What's a reference map?] Yes I have backed up, created a new tree and imported the backup. And the problem still exists. Yes I'm getting really frustrated with the whole thing! By the way, the 'strange' census form is still correctly linked to the people shown in it, as well as to James and his family.


I have two problems.

1. How do I fix it? I can delete all the entries in the strange 1911 census form and re-enter both families again linking to the correct people, but this empty 1911 form is still linked to James. And re-entering them creates duplicate 1911 census events for everyone I re-enter. And my original census form which includes James H must still exist somewhere,

2. How do I locate all the other instances where this has happened?  And I know here are quite a few because I keep tripping over them.


Is it possible somehow to produce a report or similar showing say Form, name and the address from the form? Or something. Anything. But even then, how to fix it?

Help needed before I junk the whole lot please.

Colin


-- 
Colin Spiller
[hidden email]
--
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: Forms are great. But....!

Colin Spiller

Thank you for your reply. And thanks to Dave as well - sorry I forgot to say.


Complicated isn't it. I'm going to re-read this and study it again later.

The quick answer is that the census events themselves reference the correct persons.

https://drive.google.com/open?id=17FqNUT0XRLqnOewIIXGJ41o2yY8hq_nZ

So the Attributes and the References all seem fine. Just links to the wrong Form.

Colin


On 22/05/2020 22:37, Paul Culley wrote:
To understand what may have occurred, you have to understand how forms are attached to your db.

  1. A form is connected to a Source (US 1940 Census) by giving that source the 'Form' attribute with the value 'US1940' which identifies which Gramps form to use.
  2. When you create a form, this creates an Event, which is filled with the place and date at the top of the form.
  3. It also attaches a Citation (which refers to the original source) to the Event.
  4. When you originally fill out a form, you pick one or more Person, Families etc. Doing so, attaches that event to the associated people or families.
  5. The actual form data ends up in what we call the EventRef attributes, which are visible when you first edit the person, and then from the bottom 'Events' tab, edit the Event (census, in this case), and look at the top part of the 'Event Reference Editor' 'Attributes' tab.
And then when you want to look at a form again:
  1. When you select a person in the active view, the Form Gramplet looks at the Events attached to him, and any Events attached to his immediate Families.
  2. The Gramplet then examines the Events, looking at Citations attached to the Event.
  3. For each Citation, it then looks at the Source the Citation refers to to see if it has the 'Form' attribute and the form type.
  4. It then knows what type of form to show.
  5. When you select and edit a form, the Form Editor looks at the Citation and tracks it back to the appropriate Event, and then back to the people and families where it gets the Event References to fill out the form.

Probably more than you ever wanted to know about how Forms work.

In order for the attached form to get confused like this, it seems likely the 'Reference Maps' which track backwords from objects to the other objects that refer to them may be corrupted. 

If you Edit the census Event attached to the person, and look at the References tab, it should list the person(s) from the census, if the list doesn't include the starting person, the 'Reference Maps' are corrupted.

If check & repair, or Rebuild Reference Maps did not fix the issue, I would suggest exporting your db to Gramps XML, starting a new tree and importing it again.  This should also have the effect of recreating the Reference maps.

I cannot recall any specific bugs in Gramps 5.0.1 that would cause this, although a LOT of bugs have fixed since then.  You may want to consider an upgrade https://github.com/gramps-project/gramps/releases

Paul C.

On Fri, May 22, 2020 at 3:30 PM Colin Spiller <[hidden email]> wrote:

Hi

I like gramps, and I like Forms. Forms are a great way of entering data. But only when they work correctly!

I'm using GRAMPS: 5.0.1
Python: 3.6.9 (default, Apr 18 2020, 01:56:04) [...
BSDDB: 6.1.0 (5, 3, 28)
sqlite: 3.22.0 (2.6.0)
LANG: en_GB.UTF-8
OS: Linux
Distribution: 4.15.0-101-generic

As this is UBUNTU 18.04.4 LTS that's the latest release and my databases are therefore BSDDB.

I have 6600 individuals and 3700 families.

BUT! Mixed in amongst this lot are many corruptions. Take this chap James Holloway see https://drive.google.com/open?id=1o_2mHTWqvgDgdRRMkYm2ZvBCVPmunvjV

I found him age 11 in the 1911 census and used the form to enter the family some time ago.

Now I've come to look at him again, the form looks like this https://drive.google.com/open?id=1KjLxiBHQwle3SRCcdlL7YJCdWg6pXnFj

It's a totally different family! Quite a valid family but James Holloway is nowhere to be seen in it And I have lots of other similar examples.

No I haven't aborted gramps in the middle of an update. Yes I have used "Check and repair database".(No issues - well 1 missing image which I'll fix in a minute).  Yes I have rebuilt secondary indexes. Yes I have rebuilt reference maps [What's a reference map?] Yes I have backed up, created a new tree and imported the backup. And the problem still exists. Yes I'm getting really frustrated with the whole thing! By the way, the 'strange' census form is still correctly linked to the people shown in it, as well as to James and his family.


I have two problems.

1. How do I fix it? I can delete all the entries in the strange 1911 census form and re-enter both families again linking to the correct people, but this empty 1911 form is still linked to James. And re-entering them creates duplicate 1911 census events for everyone I re-enter. And my original census form which includes James H must still exist somewhere,

2. How do I locate all the other instances where this has happened?  And I know here are quite a few because I keep tripping over them.


Is it possible somehow to produce a report or similar showing say Form, name and the address from the form? Or something. Anything. But even then, how to fix it?

Help needed before I junk the whole lot please.

Colin


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


--
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: Forms are great. But....!

Colin Spiller
In reply to this post by prculley

Hi. I'm back.

I've upgraded to 5.1.2 as Paul suggested and now I've lost my gramplets. The Form gramplet - and the new Combined Relationship View - among others are nowhere to be seen. Lots of others, strangely. Hey ho!

However, I now have more information on my James Holloway problem. In the People category, there's a Citations tab, and James has 3 major source entries in it:

1911 Census for England and Wales - Findmypast    This was used for James' birth event

    <No Citation>

Baptism Register

    <No Citation>

1911 Census RG14        This was used for James' Census event. Inconsistent, wasn't I!

    1911 35 Elswick Avenue    [repeated 8 times]

This last was the reference I added to the census entry for the 'correct' family, but when you open it up, the References tab has a gazillion entries (sorry, I exaggerate. But only slightly 33 entries in total). Mostly People but 7 Events. Several different families.

There's another screenshot at https://drive.google.com/open?id=1p8lJipFY6O5CWl3mxGMFgU9Fzg0eaycH

I thought the Repeated 8 times might match the persons in the 1911 census with James, but here are only 6 recorded. They all appear to refer to the same Citation C5083


I can see much more of what's happened, but it still doesn't make sense to me.

Colin



On 22/05/2020 22:37, Paul Culley wrote:
To understand what may have occurred, you have to understand how forms are attached to your db.

  1. A form is connected to a Source (US 1940 Census) by giving that source the 'Form' attribute with the value 'US1940' which identifies which Gramps form to use.
  2. When you create a form, this creates an Event, which is filled with the place and date at the top of the form.
  3. It also attaches a Citation (which refers to the original source) to the Event.
  4. When you originally fill out a form, you pick one or more Person, Families etc. Doing so, attaches that event to the associated people or families.
  5. The actual form data ends up in what we call the EventRef attributes, which are visible when you first edit the person, and then from the bottom 'Events' tab, edit the Event (census, in this case), and look at the top part of the 'Event Reference Editor' 'Attributes' tab.
And then when you want to look at a form again:
  1. When you select a person in the active view, the Form Gramplet looks at the Events attached to him, and any Events attached to his immediate Families.
  2. The Gramplet then examines the Events, looking at Citations attached to the Event.
  3. For each Citation, it then looks at the Source the Citation refers to to see if it has the 'Form' attribute and the form type.
  4. It then knows what type of form to show.
  5. When you select and edit a form, the Form Editor looks at the Citation and tracks it back to the appropriate Event, and then back to the people and families where it gets the Event References to fill out the form.

Probably more than you ever wanted to know about how Forms work.

In order for the attached form to get confused like this, it seems likely the 'Reference Maps' which track backwords from objects to the other objects that refer to them may be corrupted. 

If you Edit the census Event attached to the person, and look at the References tab, it should list the person(s) from the census, if the list doesn't include the starting person, the 'Reference Maps' are corrupted.

If check & repair, or Rebuild Reference Maps did not fix the issue, I would suggest exporting your db to Gramps XML, starting a new tree and importing it again.  This should also have the effect of recreating the Reference maps.

I cannot recall any specific bugs in Gramps 5.0.1 that would cause this, although a LOT of bugs have fixed since then.  You may want to consider an upgrade https://github.com/gramps-project/gramps/releases

Paul C.

On Fri, May 22, 2020 at 3:30 PM Colin Spiller <[hidden email]> wrote:

Hi

I like gramps, and I like Forms. Forms are a great way of entering data. But only when they work correctly!

I'm using GRAMPS: 5.0.1
Python: 3.6.9 (default, Apr 18 2020, 01:56:04) [...
BSDDB: 6.1.0 (5, 3, 28)
sqlite: 3.22.0 (2.6.0)
LANG: en_GB.UTF-8
OS: Linux
Distribution: 4.15.0-101-generic

As this is UBUNTU 18.04.4 LTS that's the latest release and my databases are therefore BSDDB.

I have 6600 individuals and 3700 families.

BUT! Mixed in amongst this lot are many corruptions. Take this chap James Holloway see https://drive.google.com/open?id=1o_2mHTWqvgDgdRRMkYm2ZvBCVPmunvjV

I found him age 11 in the 1911 census and used the form to enter the family some time ago.

Now I've come to look at him again, the form looks like this https://drive.google.com/open?id=1KjLxiBHQwle3SRCcdlL7YJCdWg6pXnFj

It's a totally different family! Quite a valid family but James Holloway is nowhere to be seen in it And I have lots of other similar examples.

No I haven't aborted gramps in the middle of an update. Yes I have used "Check and repair database".(No issues - well 1 missing image which I'll fix in a minute).  Yes I have rebuilt secondary indexes. Yes I have rebuilt reference maps [What's a reference map?] Yes I have backed up, created a new tree and imported the backup. And the problem still exists. Yes I'm getting really frustrated with the whole thing! By the way, the 'strange' census form is still correctly linked to the people shown in it, as well as to James and his family.


I have two problems.

1. How do I fix it? I can delete all the entries in the strange 1911 census form and re-enter both families again linking to the correct people, but this empty 1911 form is still linked to James. And re-entering them creates duplicate 1911 census events for everyone I re-enter. And my original census form which includes James H must still exist somewhere,

2. How do I locate all the other instances where this has happened?  And I know here are quite a few because I keep tripping over them.


Is it possible somehow to produce a report or similar showing say Form, name and the address from the form? Or something. Anything. But even then, how to fix it?

Help needed before I junk the whole lot please.

Colin


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


--
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: Forms are great. But....!

Dave Scheipers
I've upgraded to 5.1.2 as Paul suggested and now I've lost my gramplets. The Form gramplet - and the new Combined Relationship View - among others are nowhere to be seen. Lots of others, strangely. 

Each version registers their own gramplets and addons.  Go to the Menu >> Edit >> Preferences to install the addons. You may have to uncheck the "Do not ask about previously notified addons"


On Wed, May 27, 2020 at 4:37 PM Colin Spiller <[hidden email]> wrote:

Hi. I'm back.

I've upgraded to 5.1.2 as Paul suggested and now I've lost my gramplets. The Form gramplet - and the new Combined Relationship View - among others are nowhere to be seen. Lots of others, strangely. Hey ho!

However, I now have more information on my James Holloway problem. In the People category, there's a Citations tab, and James has 3 major source entries in it:

1911 Census for England and Wales - Findmypast    This was used for James' birth event

    <No Citation>

Baptism Register

    <No Citation>

1911 Census RG14        This was used for James' Census event. Inconsistent, wasn't I!

    1911 35 Elswick Avenue    [repeated 8 times]

This last was the reference I added to the census entry for the 'correct' family, but when you open it up, the References tab has a gazillion entries (sorry, I exaggerate. But only slightly 33 entries in total). Mostly People but 7 Events. Several different families.

There's another screenshot at https://drive.google.com/open?id=1p8lJipFY6O5CWl3mxGMFgU9Fzg0eaycH

I thought the Repeated 8 times might match the persons in the 1911 census with James, but here are only 6 recorded. They all appear to refer to the same Citation C5083


I can see much more of what's happened, but it still doesn't make sense to me.

Colin



On 22/05/2020 22:37, Paul Culley wrote:
To understand what may have occurred, you have to understand how forms are attached to your db.

  1. A form is connected to a Source (US 1940 Census) by giving that source the 'Form' attribute with the value 'US1940' which identifies which Gramps form to use.
  2. When you create a form, this creates an Event, which is filled with the place and date at the top of the form.
  3. It also attaches a Citation (which refers to the original source) to the Event.
  4. When you originally fill out a form, you pick one or more Person, Families etc. Doing so, attaches that event to the associated people or families.
  5. The actual form data ends up in what we call the EventRef attributes, which are visible when you first edit the person, and then from the bottom 'Events' tab, edit the Event (census, in this case), and look at the top part of the 'Event Reference Editor' 'Attributes' tab.
And then when you want to look at a form again:
  1. When you select a person in the active view, the Form Gramplet looks at the Events attached to him, and any Events attached to his immediate Families.
  2. The Gramplet then examines the Events, looking at Citations attached to the Event.
  3. For each Citation, it then looks at the Source the Citation refers to to see if it has the 'Form' attribute and the form type.
  4. It then knows what type of form to show.
  5. When you select and edit a form, the Form Editor looks at the Citation and tracks it back to the appropriate Event, and then back to the people and families where it gets the Event References to fill out the form.

Probably more than you ever wanted to know about how Forms work.

In order for the attached form to get confused like this, it seems likely the 'Reference Maps' which track backwords from objects to the other objects that refer to them may be corrupted. 

If you Edit the census Event attached to the person, and look at the References tab, it should list the person(s) from the census, if the list doesn't include the starting person, the 'Reference Maps' are corrupted.

If check & repair, or Rebuild Reference Maps did not fix the issue, I would suggest exporting your db to Gramps XML, starting a new tree and importing it again.  This should also have the effect of recreating the Reference maps.

I cannot recall any specific bugs in Gramps 5.0.1 that would cause this, although a LOT of bugs have fixed since then.  You may want to consider an upgrade https://github.com/gramps-project/gramps/releases

Paul C.

On Fri, May 22, 2020 at 3:30 PM Colin Spiller <[hidden email]> wrote:

Hi

I like gramps, and I like Forms. Forms are a great way of entering data. But only when they work correctly!

I'm using GRAMPS: 5.0.1
Python: 3.6.9 (default, Apr 18 2020, 01:56:04) [...
BSDDB: 6.1.0 (5, 3, 28)
sqlite: 3.22.0 (2.6.0)
LANG: en_GB.UTF-8
OS: Linux
Distribution: 4.15.0-101-generic

As this is UBUNTU 18.04.4 LTS that's the latest release and my databases are therefore BSDDB.

I have 6600 individuals and 3700 families.

BUT! Mixed in amongst this lot are many corruptions. Take this chap James Holloway see https://drive.google.com/open?id=1o_2mHTWqvgDgdRRMkYm2ZvBCVPmunvjV

I found him age 11 in the 1911 census and used the form to enter the family some time ago.

Now I've come to look at him again, the form looks like this https://drive.google.com/open?id=1KjLxiBHQwle3SRCcdlL7YJCdWg6pXnFj

It's a totally different family! Quite a valid family but James Holloway is nowhere to be seen in it And I have lots of other similar examples.

No I haven't aborted gramps in the middle of an update. Yes I have used "Check and repair database".(No issues - well 1 missing image which I'll fix in a minute).  Yes I have rebuilt secondary indexes. Yes I have rebuilt reference maps [What's a reference map?] Yes I have backed up, created a new tree and imported the backup. And the problem still exists. Yes I'm getting really frustrated with the whole thing! By the way, the 'strange' census form is still correctly linked to the people shown in it, as well as to James and his family.


I have two problems.

1. How do I fix it? I can delete all the entries in the strange 1911 census form and re-enter both families again linking to the correct people, but this empty 1911 form is still linked to James. And re-entering them creates duplicate 1911 census events for everyone I re-enter. And my original census form which includes James H must still exist somewhere,

2. How do I locate all the other instances where this has happened?  And I know here are quite a few because I keep tripping over them.


Is it possible somehow to produce a report or similar showing say Form, name and the address from the form? Or something. Anything. But even then, how to fix it?

Help needed before I junk the whole lot please.

Colin


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