Person ID codes

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

Person ID codes

Tony Hacche
When I imported a previously created tree (GEDCOM formatted) into Gramps, all the person IDs were of the format Pnnn.

I have just added a couple of new people to the tree and their IDs are of the format Innn.

I was just curious as to the different format?

Thanks,

Tony

--
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: Person ID codes

victorengel
Are the IDs from the source of the GEDCOM? Was it a different source for each round? BTW, I've noticed that ancestry.com changes the ID from export to export. It's apparently a sequential number generated at time of export. Geni, on the other hand, seems to use unique ids that remain constant unless there is a profile merge or something like that.

Victor

On Thu, Oct 10, 2019 at 2:32 AM Tony Hacche <[hidden email]> wrote:
When I imported a previously created tree (GEDCOM formatted) into Gramps, all the person IDs were of the format Pnnn.

I have just added a couple of new people to the tree and their IDs are of the format Innn.

I was just curious as to the different format?

Thanks,

Tony

--
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: Person ID codes

Tony Hacche
Victor,

> Are the IDs from the source of the GEDCOM?

Oh, good point. I had not considered that the IDs were from the GEDCOM file. I just assumed they were created by Gramps on import. Looking at the original GEDCOM file I can see that these are indeed where they are from. Mystery solved ;-)

> Was it a different source for each round?

No, the IDs with the format Innn came from when I added a couple of people into me tree using the Gramps "Add a new person" dialogue box. I left Gramps to create the IDs and that's the format that was automatically used.

Thanks,

Tony

--
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: Person ID codes

Nick Hall
On 10/10/2019 09:03, Tony Hacche wrote:
> No, the IDs with the format Innn came from when I added a couple of people into me tree using the Gramps "Add a new person" dialogue box. I left Gramps to create the IDs and that's the format that was automatically used.

The Gramps format is defined in the preferences.

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: Person ID codes

Tony Hacche
Nick,

> The Gramps format is defined in the preferences.

So they are. Just had a look and I have a clash between the format of the imported people and places with both beginning with "P".

Next job will be to change the format of the IDs for the imported people to "Innnn".

Any gotchas to be wary of when I do this?

Thanks,

Tony

--
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: Person ID codes

Patrick Gerlier
To all:

Does it mean that all IDs are in the same namespace? Persons, places,
events, … I think it matters mainly when BSDDB is the storage backend.
It should not be a problem with SQLite and other DB engines, unless
tables use "abusively" blob records.

Patrick

Le 10/10/2019 à 14:07, Tony Hacche a écrit :

> Nick,
>
>> The Gramps format is defined in the preferences.
> So they are. Just had a look and I have a clash between the format of the imported people and places with both beginning with "P".
>
> Next job will be to change the format of the IDs for the imported people to "Innnn".
>
> Any gotchas to be wary of when I do this?
>
> Thanks,
>
> Tony
>



--
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: Person ID codes

Nick Hall
In reply to this post by Tony Hacche
On 10/10/2019 13:07, Tony Hacche wrote:
The Gramps format is defined in the preferences.
So they are. Just had a look and I have a clash between the format of the imported people and places with both beginning with "P".

Next job will be to change the format of the IDs for the imported people to "Innnn".

Any gotchas to be wary of when I do this?

Be careful when using the Reorder IDs tool with a SQLite backend.  The BSDDB backend happened to return objects in creation date order, despite the order not being defined.  The result was that the IDs were assigned a number close to before but with gaps being filled.  This is not the case with SQLite.

We have a fix, but it isn't in the latest version yet.

As always, make a backup before you make any major changes.


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: Person ID codes

Tony Hacche
Nick,

> Be careful when using the Reorder IDs tool with a SQLite backend.  
> The BSDDB backend happened to return objects in creation date
> order, despite the order not being defined.  The result was that
> the IDs were assigned a number close to before but with gaps being
> filled. This is not the case with SQLite.

My tree isn't large so I was planning on a manual change of the people IDs by editing them in the person dialogue. Would that be OK do you think?

Thanks,

Tony

--
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: Person ID codes

Nick Hall
In reply to this post by Patrick Gerlier
On 10/10/2019 13:14, Patrick Gerlier wrote:
> Does it mean that all IDs are in the same namespace? Persons, places,
> events, … I think it matters mainly when BSDDB is the storage backend.
> It should not be a problem with SQLite and other DB engines, unless
> tables use "abusively" blob records.

No.  You can use the same ID for different objects.

The Gramps ID field is indexed in both the BSDDB and SQLite backend, but
the unique constraint is enforced in the application layer not in the
database.  In SQLite we duplicate the ID in to a separate column for
indexing.


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: Person ID codes

Nick Hall
In reply to this post by Tony Hacche
On 10/10/2019 14:14, Tony Hacche wrote:
> My tree isn't large so I was planning on a manual change of the people IDs by editing them in the person dialogue. Would that be OK do you think?

Yes.




--
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: Person ID codes

Dave Scheipers
In reply to this post by Nick Hall
Actually Nick, the warning is reversed. SQLite reorders according to
the record creation.   On the BSDDB backend the ID numbers are
scrambled. PaulC just put the fix in for 5.1.2.

https://gramps-project.org/bugs/view.php?id=10641.

On Thu, Oct 10, 2019 at 9:00 AM Nick Hall <[hidden email]> wrote:

>
> On 10/10/2019 13:07, Tony Hacche wrote:
>
> The Gramps format is defined in the preferences.
>
> So they are. Just had a look and I have a clash between the format of the imported people and places with both beginning with "P".
>
> Next job will be to change the format of the IDs for the imported people to "Innnn".
>
> Any gotchas to be wary of when I do this?
>
> Be careful when using the Reorder IDs tool with a SQLite backend.  The BSDDB backend happened to return objects in creation date order, despite the order not being defined.  The result was that the IDs were assigned a number close to before but with gaps being filled.  This is not the case with SQLite.
>
> We have a fix, but it isn't in the latest version yet.
>
> As always, make a backup before you make any major changes.
>
>
> Nick.
>
>
> --
> 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