Jump to content
blomble

Primary Tag designations disappeared after "Validate"

Recommended Posts

Primary Tag designations disappeared from one dataset (of 20) after running "Validate File Integrity"

 

In my experience "Validate File Integrity" will often find a few errors after a merge.

In this instance it came up with over 1000 errors and subsequently all Primary Tag designations (other than Parent/Child tags) disappeared from my Dataset 9:

 

All the data appears to be present but there is not an asterisk in sight.

 

I now have to reset all the tags on 677 names. See you next week, unless someone has a quick fix!

 

Mark

Share this post


Link to post
Share on other sites

Previous comments have suggested that it is always wise to do an Optimize immediately following any VFI. Further, if a VFI ever gives errors it has always been stressed that you should do another VFI followed by Optimize pair until the VFI shows no errors. Previous explanations of the reason for this have been that fixing some errors now make it possible to identify further errors, so continued VFI/Optimize pairs are required.

Share this post


Link to post
Share on other sites
Previous comments have suggested that it is always wise to do an Optimize immediately following any VFI. Further, if a VFI ever gives errors it has always been stressed that you should do another VFI followed by Optimize pair until the VFI shows no errors. Previous explanations of the reason for this have been that fixing some errors now make it possible to identify further errors, so continued VFI/Optimize pairs are required.

 

 

Thanks mjh,

 

I missed the posting re: Repeat VFI/Optimize pairs.

 

note to developers

If the above is 'best procedure', perhaps there should be a 'Suggest You Optimize now' warning after VFI ?

 

Mark

Share this post


Link to post
Share on other sites
Hi Mark,

 

Did that fix your Primary Tag designations?

 

 

Hi Michael,

 

No - I'm currently fixing them by hand - and I'm REALLY missing some kind of Macro function!

 

BTW - I do Backup, Validate & Optimise after every merge, but I hadn't thought of repeating the VFI.

 

Unfortunately, as the problem only affected one data set, I did not notice the issue until well after more data had been entered.

 

Mark

Share this post


Link to post
Share on other sites

Mark,

 

No - I'm currently fixing them by hand
Gosh, sorry to hear that, not much fun I am sure.

 

- and I'm REALLY missing some kind of Macro function!
Yes, I checked the TMG Utility Help files and can find no capability there. Although I do not use them, I have heard of separate programs that allow you to assign a series of key actions to a special key that then works in other programs.

 

I assume you can identify the specific people and either make them a Focus Group or filter for them in the Project Explorer so you can deal with just these people. Further, if the loss of Primary only affects certain tag groups you might filter the Details window to only display those tag types to reduce the clutter while doing the "clean up". (Of course I assume you know that only one tag among multiples in a given tag group can be Primary.) Of course, while it is usually desirable to have a primary tag in the Birth, Name, and Mother/Father relationship, and Death tag groups, depending upon what you are doing it may not be a problem for Other tag types. Of course I understand the desire to "clean up" the dataset when you notice the problem.

 

Unfortunately I can think of no easy way to identify people who have tag groups where no tag is marked primary. Maybe some other user will have an idea.

 

You said you do "Backup, Validate & Optimise after every merge". Hmmm... I would recommend doing the VFI/Optimize pair first (repeating the pair only if VFI shows errors) and then do the backup. You probably don't want to backup/save a file that might contain errors.

 

the problem only affected one data set
Very strange. Can you think of anything you did that is unique to that dataset? Is this the only dataset that you have been doing merge with? And what was the source and nature of the merged data? This is very unusual and should not happen.

Share this post


Link to post
Share on other sites
Mark,

 

the problem only affected one data set
Very strange. Can you think of anything you did that is unique to that dataset? Is this the only dataset that you have been doing merge with? And what was the source and nature of the merged data? This is very unusual and should not happen.

 

Many thanks for your time on this, Michael.

 

The really strange thing is that the affected dataset is one that is rarely edited and was not involved in any merge. It is an almost complete branch that was imported a few years ago. I have struggled to think of any different process that I used recently, but nothing comes to mind.

 

My process is that I have 19 datasets that contain individual trees of the family name. Another dataset contains all the 'lonesome' individuals, and one dataset called 'temp' that I use in the merging process.

I use Terry Riegel's recommended method of merging: Copy Focus Group to final Dataset / Merge / Move Focus Group to Dataset 'temp' / Delete Dataset 'temp'. Optimise & VFI.

 

I have done it this way for many years without a problem. I think I will be throwing in a few more VFI/Optimise pairs after this issue!

 

Unfortunately the loss of the Birth Tags primary designation means that the 'sort by date functions' do not work, so I do have to fix it all. (BTW - EVERY primary tag disappeared within that one dataset, except the parent/child ones)

 

Mark

Share this post


Link to post
Share on other sites
BTW - EVERY primary tag disappeared within that one dataset, except the parent/child ones
How very strange. Sorry, but I am out of ideas. Maybe one of the developers or beta testers will comment?

Share this post


Link to post
Share on other sites

Mark,

 

Check out the Adjust Birth and Death feature in TMG Utility. In particular, read step 2 in the instructions. The feature has an option to set the primary designation for birth and/or death events. It may help you in this case. THe likely outcome is that some people would have to be fixed by hand (cases where a person had two or more events in the birth group or in the death group) but it would take care of the majority of people automatically.

Share this post


Link to post
Share on other sites

Thanks John,

Unfortunately I have just finished editing manually. Your method might have helped, but all other events had to be toggled (census etc.,), so I would have had to find each individual anyway. The lack of a birth date actually made that easier on the picklist.

Share this post


Link to post
Share on other sites

Well, it's 3 years later & exactly the same problem has occurred - complete loss of all primary tag designations, but only within one Data Set of twenty. This again occurred after a "Validate" that showed 1700+ errors (usually 4-30).

I still have no explanation, & no en masse method of fixing the issue, but have used TMG utilities to fix Births & Deaths. All others have to be corrected individually.

 

Mark

Share this post


Link to post
Share on other sites

Hi Mark, I'm still here.

 

Wow! What a bummer!!

 

You said before that you would do more VFI's and Optimizes. Have you been doing them regularly? Do you always continue to do them until you get no errors? Do you have a recent backup? Maybe it would be quicker to use the backup than resetting everything?

 

Seems you have a lot (20?) of datasets in your production project. Do you actually need them to be in the project with the dataset that you are regularly editing? Maybe all those datasets you are not changing you could keep in a separate project for reference, with a good set of backups. You do know that you can have two copies of TMG running at the same time? One can be your project of these reference datasets, and the other the project where you are making changes.

 

Just a thought to consider.

Share this post


Link to post
Share on other sites

Thanks for holding :)

A once every 3 years issue is going to be really tough to pin down.

 

I've been very well behaved with my data management over the past few years.

I always validate after every merge (usually twice) until I get zero errors. I do make daily backups, but it has to be a real disaster for those to be called upon - lose the day's work or repair the errors. Perhaps what I really need is a timed backup that works in the background.

I did resort to using a recent backup last time this happened, but when I made the same Merge / Validate manoeuvre, I found the same issue. Maybe something odd is occurring in the Data Set that may be dormant until the next Validate.

 

re: the multiple data sets

I have a one name study in the database, with a number of established lines. Different researchers are supplying the data in their particular area of interest and I find it much easier to manage if I keep these data sets separate. This does mean that my data has a lot more "Copy / Merge / Validate / Delete" activity than most, & perhaps that is part of the problem.

(The down side of the above is that each time I publish, I have to merge all the lines into a "Combo" line for Second Site to read.)

Share this post


Link to post
Share on other sites
A once every 3 years issue is going to be really tough to pin down.
Agreed, but at least by reporting it maybe the developers will notice and review what might be happening with hopes of fixing it in a subsequent version.

 

re: the multiple data sets
From comments I have seen many users that do one-name studies have taken to containing everything in one main data set in one project. They keep the various "lines" from different researchers identified using one or more Flags. By filtering on the Flags they can either restrict the Project Explorer to only view a subset of people of interest or produce reports for only a certain subset of people.

 

If the same person in your data is likely to be in multiple lines (from multiple researchers), then a separate Flag for each line/researcher may make the most sense. That way the same person can have Flags set to 'Y' for multiple researcher Flags. If overlap is minimal, then a single Flag with multiple possible values may work better. Even with separate Flags, you can always filter a List of People report for multiple conditions based on multiple Flags and set another temporary Flag using the report's Secondary Output.

 

I agree that Validate should not be doing what it seems to be doing, but a single project containing a single data set might minimize the likelihood. And this single data set might minimize your own overall work by minimizing your "Copy / Merge / Validate / Delete" operations.

 

If you do decide to merge to one data set, first create all necessary Flags in every data set with 'N' as the default value. Then use a separate List of People report in each data set to change the appropriate Flag value for all its people to 'Y'. Only then would you merge all data sets.

 

Hopefully some users actively doing one-name studies will also comment.

Share this post


Link to post
Share on other sites

Thanks Michael,

I came to the same conclusion and merged everything into one data set (unfortunately before I read your posting about the flags).

However, Terry's TMG tips has a very good tutorial on how to add the flags post-merge, so all turned out well.

 

Mark

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×