How to run two trees concurrently

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

How to run two trees concurrently

sturdy
Hi all,

I've managed to paint myself into a corner and need some help. I think I need to run two side-by-side copies of Gramps 3.4.7 on Mint 16 so that I can compare the same person (~400 persons), events, media, etc. in two trees before merging the two trees. I believe this is possible by installing a second Gramps and using the grampshome env var. My intent is to delete unwanted objects in either tree prior to merging. I think the Gramps CLI can be used but I would appreciate a little help with the details. Alternatively, I've thought of installing a second Gramps in a VM which may be the safer method.

Best regards,
Sturdy
Reply | Threaded
Open this post in threaded view
|

Re: How to run two trees concurrently

enno
Hi Sturdy,
> I've managed to paint myself into a corner and need some help. I think I
> need to run two side-by-side copies of Gramps 3.4.7 on Mint 16 so that I can
> compare the same person (~400 persons), events, media, etc. in two trees
> before merging the two trees. I believe this is possible by installing a
> second Gramps and using the grampshome env var. My intent is to delete
> unwanted objects in either tree prior to merging. I think the Gramps CLI can
> be used but I would appreciate a little help with the details.
> Alternatively, I've thought of installing a second Gramps in a VM which may
> be the safer method.
You can simply start a 2nd copy of Gramps from the menu. When it's
configured to load the last used tree, like I have that here, the 2nd
copy will find the 1st tree locked, and give you the chance to open or
create another one. I'm running 2 copies while typing this, and the only
thing you may need to be careful with is changing preferences that are
written to Gramps' INI files. You will probably be safe when you don't
change preferences, and take care to close the 2nd copy first, so that
the database that you load when Gramps starts is not changed. This
proves to work here.

Whether this is an efficient way depends on the nature of your trees. If
there is a large overlap, it looks like a lot of work to me, and that is
also true if the trees have different origins. If both were originally
the same, but changes have diverted, it may be easier to export person
and other screens to CSV files, and compare those, or even compare
GEDCOMs instead. That will not work if same persons have different IDs,
which is likely the case when both trees have completely different origins.

In the latter case, I often use PAF (which runs fine in Wine) or
RootsMagic in a Windows XP virtual box. Both have duplicate merging
tools that work way faster than the one in Gramps, and the one in
RootsMagic has the same sort of ranking as Gramps. The free version is
fine for this.

regards,

Enno


------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
Reply | Threaded
Open this post in threaded view
|

Re: How to run two trees concurrently

sturdy
Hi Enno,

Thanks for the quick come-back. Surprisingly simple! A second copy seems to run okay if treated as you said. I've received the 'locked' message many times and never really tried to continue with a second tree. I'll experiment but I believe the new media files will need to go into a new folder under the current relative path.

My second tree is in Ancestry and is a large update of the tree already in Gramps. Unfortunately it is not just a _simple_ merge. I have worked out a method to download from Ancestry including media and import thru FTM into Gramps but I let this get out of control so can see no way except to compare people and merge each person. Probably over 400 people but half may be new. I won't know until I can do the compare.

Thanks for the assist...
Sturdy
Reply | Threaded
Open this post in threaded view
|

Re: How to run two trees concurrently

enno
Hi Sturdy,
> Thanks for the quick come-back. Surprisingly simple! A second copy seems to
> run okay if treated as you said. I've received the 'locked' message many
> times and never really tried to continue with a second tree. I'll experiment
> but I believe the new media files will need to go into a new folder under
> the current relative path.
You're welcome. I didn't think about the media yet, since I don't manage
much there, and I assume that what you got from Ancestry is new, and can
therefore easily be added to the existing path.
> My second tree is in Ancestry and is a large update of the tree already in
> Gramps. Unfortunately it is not just a _simple_ merge. I have worked out a
> method to download from Ancestry including media and import thru FTM into
> Gramps but I let this get out of control so can see no way except to compare
> people and merge each person. Probably over 400 people but half may be new.
> I won't know until I can do the compare.
I have exactly the same problem here. I subscribed to Ancestry after
Nick found some interesting records for me there, and found that using
the green leaf feature can easy lead to adding lots of new sources, and
persons too, without tracking everything. So now, I have more than 23
000 persons to compare against my Gramps database, and that's impossible
without the right tools.

Knowing that you can run FTM, and therefore must still have Windows
somewhere, I recommend to download the free version of RootsMagic, and
then import both trees in there. It runs duplicate checks way faster
than FTM, of which I tested the free German version today, and it can
quickly show what data differs for each person pair, using colour codes.
IMO, it is way more exact than FTM, because I found that the latter
assigns scores of 1000 to matches that are absolutely not exactly the
same, and requires an extra click to find out what the differences are.
RootsMagic shows them right away, so you can make notes of which persons
to copy from one tree to the other, or select persons to copy right away
if you can run RootsMagic in a virtual box next to Gramps in Linux, like
I do. RootsMagic can also create reports for this, but those may be
restricted to paying users. And if that is the case, you may better use
PAF for this, which runs fine in Wine, so that you don't need a virtual
box for that.

To my regret, the diff in Gramps master

https://www.gramps-project.org/wiki/index.php?title=GEPS_009:_Import_Export_Merge

can not be used for this, because it works with database handles
(internal IDs) to compare, and even when changed to GEDCOM IDs it would
not help, because I know that Ancestry uses different ones. So what we
really need as a first step is a database diff that uses the duplicate
search algorithm to find persons that are equal or similar, and then
shows where things changed. Adding merge is yet another step, that the
sites do quite good I think. Gramps merging is not very smart, because
it duplicates all events, even when they are exactly the same.

Long ago, there was a Dutch made tool called GEDcompare, but the latest
version that I tried was very annoying, because it gave all sorts of
error messages during import of my GEDCOM files, which I think were
perfect, GEDCOM standard wise. And another program which was advised to
me here

http://genealogy.stackexchange.com/questions/3121/which-tools-are-available-for-gedcom-comparison

was not able to process the database that I had then, let alone the one
I have now.

And another program that I found, called geddiff, is just a useless
prototype. :-(

regards,

Enno


------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
Reply | Threaded
Open this post in threaded view
|

Re: How to run two trees concurrently

sturdy

On 03/19/2014 08:05 PM, Enno Borgsteede wrote:
> Hi Sturdy,
>> Thanks for the quick come-back. Surprisingly simple! A second copy seems to
>> run okay if treated as you said. I've received the 'locked' message many
>> times and never really tried to continue with a second tree. I'll experiment
>> but I believe the new media files will need to go into a new folder under
>> the current relative path.
> You're welcome. I didn't think about the media yet, since I don't manage
> much there, and I assume that what you got from Ancestry is new, and can
> therefore easily be added to the existing path.
Actually, I compounded the problem and have a combination of old that
needs updating and also many new people, events, media, etc.

>> My second tree is in Ancestry and is a large update of the tree already in
>> Gramps. Unfortunately it is not just a _simple_ merge. I have worked out a
>> method to download from Ancestry including media and import thru FTM into
>> Gramps but I let this get out of control so can see no way except to compare
>> people and merge each person. Probably over 400 people but half may be new.
>> I won't know until I can do the compare.
> I have exactly the same problem here. I subscribed to Ancestry after
> Nick found some interesting records for me there, and found that using
> the green leaf feature can easy lead to adding lots of new sources, and
> persons too, without tracking everything. So now, I have more than 23
> 000 persons to compare against my Gramps database, and that's impossible
> without the right tools.
Well, now I don't feel quite so bad since I only have a few hundred to
update. :-)

> Knowing that you can run FTM, and therefore must still have Windows
> somewhere, I recommend to download the free version of RootsMagic, and
> then import both trees in there. It runs duplicate checks way faster
> than FTM, of which I tested the free German version today, and it can
> quickly show what data differs for each person pair, using colour codes.
> IMO, it is way more exact than FTM, because I found that the latter
> assigns scores of 1000 to matches that are absolutely not exactly the
> same, and requires an extra click to find out what the differences are.
> RootsMagic shows them right away, so you can make notes of which persons
> to copy from one tree to the other, or select persons to copy right away
> if you can run RootsMagic in a virtual box next to Gramps in Linux, like
> I do. RootsMagic can also create reports for this, but those may be
> restricted to paying users. And if that is the case, you may better use
> PAF for this, which runs fine in Wine, so that you don't need a virtual
> box for that.
I currently run FTM in a Win8 vm. FTM 2014 (not 2012) seems to be the
only practical way to get the media from Ancestry. But this adds to my
update problem. FTM 2014 seems to handle media well (FTM 2012 was a
disaster).

I'll download Rootsmagic and give it a look. My current thought
(untried) is to run the old tree and a second update tree of new and
updated persons side-by-side and look for same events which I think can
then be removed from persons in the new tree before merging with the old
tree. Unfortunately, Gramps hides most data behind layers so some extra
clicks may be required. There is also a lot of intermediate processing
required to go from FTM Win => Gramps Linux. And there seems to be a
couple of remaining gedcom issues. I have a mostly accurate list of
about 275 names that require the update/merge. Others not on the list
will be new names and should import with sources, events, media, etc. In
the last two days I processed 20 from the list and because of the little
green leaves, added 61 more names! I'm doing family lines so when I
update/merge from the list, the new persons should all be linked
automatically.
> Gramps merging is not very smart, because
> it duplicates all events, even when they are exactly the same.
True, but at least this method is safe and doesn't lose data.

> Long ago, there was a Dutch made tool called GEDcompare, but the latest
> version that I tried was very annoying, because it gave all sorts of
> error messages during import of my GEDCOM files, which I think were
> perfect, GEDCOM standard wise. And another program which was advised to
> me here
>
> http://genealogy.stackexchange.com/questions/3121/which-tools-are-available-for-gedcom-comparison
>
> was not able to process the database that I had then, let alone the one
> I have now.
I think a useful gedcom compare would be almost impossible with the
number of interpretations. I regularly look at gedcoms from Ancestry,
FTM, GenJ and Gramps. A Few months ago I put the Gramps Example tree
thru each and the exports were all different. Once imported the gedcom
produced the same tree so that implies the db may be more comparable
than the gedcom. Or maybe to compare to tree, first put each into a db
then compare.

Thanks again. I'm going to process this group as a test so will let you
know how this works out.

Sturdy

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users