Exporting, and reimporting changes

[jimlambeth 0]

I've search the forums and not found this scenario. In fact, in my years as an amateur genealogist - I've never seen a clear description of this (either specific to a piece of software - or at a higher, conceptual level). I think it's best explained by example:

 

- I wish to export my TMG data to another program, for a couple reasons:

1) Another program has a convienent search mechanism (for example, Ancestry.com)

2) Another program is more portable (an old example was taking my Treo to the library, rather than the whole laptop ...)

 

- At some point, I wish to merge CHANGES back into my main TMG project, and file.

 

How the heck do I do this? I'm not particularly into doing a full "find duplicate people" against the 6000+ (x2?) people in my data set each time. Ugly, time consuming, and prone to stooper induced errors.

 

I feel like there's something basic that I'm not "getting" here. I'm not even sure my questions are right ... but here at stab at it:

- Can I, somehow, delete/remove the non-changed records before reimporting/merging the "updated" data set?

- Is there something automated (perhaps using those darn reference numbers) that would make merging back in more automated? People have tried explaining this to me before (I think) - but I just didn't understand.

 

I once had a Palm program that flagged changed records ... had I been able to Delete the non-changed records (I couldn't) ... I'd have exactly what I'm looking for here. Basically, can I pair down the import file to just chaged records, or can I better automate a mass merge of large, similiar files.

 

I can't be the only one the wishes to occassionally push data out to another program, and then bring it back. TMG is my home base - but not my only tool.

 

What am I missing here? I'm probably going to need a remedial explanation. Referrals to reading material gladly accepted!

[Roy_Delos_Reyes 0]

I've search the forums and not found this scenario. In fact, in my years as an amateur genealogist - I've never seen a clear description of this (either specific to a piece of software - or at a higher, conceptual level). I think it's best explained by example:

 

- I wish to export my TMG data to another program, for a couple reasons:

1) Another program has a convienent search mechanism (for example, Ancestry.com)

2) Another program is more portable (an old example was taking my Treo to the library, rather than the whole laptop ...)

 

- At some point, I wish to merge CHANGES back into my main TMG project, and file.

 

How the heck do I do this? I'm not particularly into doing a full "find duplicate people" against the 6000+ (x2?) people in my data set each time. Ugly, time consuming, and prone to stooper induced errors.

 

I feel like there's something basic that I'm not "getting" here. I'm not even sure my questions are right ... but here at stab at it:

- Can I, somehow, delete/remove the non-changed records before reimporting/merging the "updated" data set?

- Is there something automated (perhaps using those darn reference numbers) that would make merging back in more automated? People have tried explaining this to me before (I think) - but I just didn't understand.

 

I once had a Palm program that flagged changed records ... had I been able to Delete the non-changed records (I couldn't) ... I'd have exactly what I'm looking for here. Basically, can I pair down the import file to just chaged records, or can I better automate a mass merge of large, similiar files.

 

I can't be the only one the wishes to occassionally push data out to another program, and then bring it back. TMG is my home base - but not my only tool.

 

What am I missing here? I'm probably going to need a remedial explanation. Referrals to reading material gladly accepted!

 

 

The only program I've ever worked with that did something like this was in a medical lab I worked for... and the cost ran into the mid six figures!

[Terry Reigel 0]

Basically, you can't do that.

 

First, you can't export all your data from TMG to GEDCOM unless you use only the most basic features of TMG, and even then details like source definitions will get pretty scrambled. If you accept that you will only get a partial export, you can export and then do processing in another program. But when you try to re-import lots of records will be different, even those you never changed in the other program, because of the losses in the export/import process will mean you are getting back different records than you sent.

 

Next, TMG does not have the ability to compare two data sets record-by-record to find differences. And, if it did, it would find lots of false changes for the reasons I above. Keep in mind that there are many different kinds of records involved - names, events (and within them, dates, sort dates, places, memos, and citations), sources, repositories - to name the more common ones. I can't imagine how the results of such a comparison could be presented to user in a way that would be usable.

 

I would suggest this process:

 

1. Make your export, and keep track of who you made changes to.

 

2. When you are ready to reconcile the differences, open both programs side-by-side and use copy and paste to transfer new information back to TMG. This way, it gets entered "correctly" - that is to your data entry standards - in TMG, which will not happen if you import it directly. If you really want to import it, import it and leave it as a separate data set, and use Copy Persons to copy any new people you found. There really is no way to copy specific facts or source information except 1) the copy and paste process for each field, or 2) copy the person and then merge the two copies and manually eliminate the resulting duplicate tags.

[Michael Hannah 0]

I agree with Terry that this is an extremely difficult problem, and basically you can't get there from here. However, I am one who likes to "think outside the box" and would offer the following as a possible aid in this process. If you are using Windows and have the Office products, then you might make use of the feature in MS Word that identifies differences in documents. Perhaps you might do the following:

 

Recognizing that an export will lose a lot of data, first export and then import the export into a new project.

Next run an MS Word report, perhaps the Individual Narrative, on all people using all tags, and including both excluded and sensitive data.

Now you modify your exported data in the other program.

Then you import that modified data into TMG as a new project.

Run the exact same MS Word report as before.

Use the MS Word feature to compare the two documents, and then do the copy/paste that Terry suggests to enter the new/changed data into TMG.

 

As for taking your data to the library, you might look at the companion products for TMG (GedStar or Pocket Genealogist) that fit the actual TMG databases onto portable devices. But even when I had my full TMG data on my Palm I found that I preferred to make separate notes of the source information that I found and do the actual modification of TMG back in my office. Now I have my highly portable OQO with the full TMG with me at the library and I still often find it better to make separate full source notes and do the actual TMG data entry back home. Just my preferred way to do the work.

 

Hope this gives you ideas,

[Terry Reigel 0]

Clever way to exclude the losses from export-import, Michael. And your method does reduce the records examined to those which appear in a narrative report, including source notes if you include them in the report.

 

With 6000 people at issue that would be a huge narrative to process, but it might be possible to focus the reports on some subset to make them more manageable.

[jimlambeth 0]

Clever way to exclude the losses from export-import, Michael. And your method does reduce the records examined to those which appear in a narrative report, including source notes if you include them in the report.

 

With 6000 people at issue that would be a huge narrative to process, but it might be possible to focus the reports on some subset to make them more manageable.

 

 

Thank you both. It does seem like finding an easy way to isolate the individuals with changes is the only angle, and that would be something at the GEDCOM level, or in the other program. I had wondered about some from of text compare, Michael. I do understand the data integrity issues ... I've messed that up several times. Fortunately I've had good backups each time - or have tried something in a test instance.

 

Best of all though - I don't feel like I don't "get it" any more. Mosting I can't figure it out because it's not terribly practical.

 

Thanks!

[Terry Reigel 0]

You're welcome.

 

Terry

[Michael Hannah 0]

If you are doing GEDCOM compares, these files are just plain text files and various programs exist to compare such files, such as "diff".

 

Glad we could give you ideas, and keep you from looking for something that is not there!