This project has moved and is read-only. For the latest updates, please go here.

Support for The Master Genealogist?

Aug 14, 2013 at 9:55 AM
Edited Aug 15, 2013 at 7:43 PM
Alerted to your app by Peter at Lost Cousins and like the reports you have.

Sadly I notice TMG isn't one of your supported sources and wondered if you had plans to include it?

I can export a GEDCOM from TMG and load that into FTAnalyzer, but some tag types don't map well.

TMG offers custom tags for each census, which can be date constrained and hence validated on entry, which I've extensively used. But down side is that rather than regular GEDCOM 1 CENS/2 DATE being created, I'm seeing an event of the type 1 EVEN/2 TYPE Census 1851/2 DATE, which FTAnalyzer doesn't recognise as a Census record.

Happy to provide sample GEDCOM off list and/or assist in any future testing,


Aug 14, 2013 at 12:47 PM
Edited Aug 14, 2013 at 1:15 PM
The program should support any application that adheres to the GEDCOM 5.5 or 5.5.1 standard. I've not used TMG so I don't know what level of adherence to the standards they go in for. The fact it didn't appear on the list was not because it isn't supported, but because I didn't find it in a search of modern software.

It sounds like they haven't used the established standard and chose to use the custom tag EVEN (event). This all too often happened with older programs many of which are really showing their age. That said I will try to support it if you can give me a sample GEDCOM.

Update: I just did a search and oh dear. Their webpage says it needs a "Pentium 166 with 64 Megs of RAM" which is a spec that is relevant to around the mid 90s ie: about 15-18 years out of date, the first 166s were launched in June 1995. From other parts of the site they are clearly trying to update their software but the website is incredibly dated.

It may therefore be a struggle as their blurb says "we support GEDCOM but..." So it may be they only support a limited subset of GEDCOM data or an older version. When you export does it give any hint what version of GEDCOM is supported?

I'll try my best but no guarantees sorry. Any information or sample data you can give to make the process easier would be appreciated. eg: a sample GEDCOM extract showing a variety of the CENSUS records covering different years.
Aug 14, 2013 at 4:14 PM
Thanks for the speedy response. The other way to interpret "...needs a "Pentium 166 with 64 Megs of RAM"" would be they think their code is very light weight :-) Parts of their website may be in need of a refresh, but the product is very much current and has a loyal following. Version 8 first shipped in Dec2011 and is now at v8.08 (and no I don't work for them).

Even my somewhat elderly version of TMG does export to GEDCOM 5.5 and I've never had issues with the files so created. The nub of the issue is probably custom tags, which I see you're not adverse to proposing. TMG does have a simple "Census" tag, but I choose to use the decade specific custom tags they provide.

As I initially indicated, I'm more than happy to help and provide sample data, but as I said, off list, please. I'm new to and still finding my way around and not sure how to do that right now.

Aug 14, 2013 at 5:05 PM
Ah good to hear it supports GEDCOM 5.5 the issue then is the custom tags vs the default GEDCOM CENS tag. I support the use of custom tags for recording entries on Lost Cousins so it should be possible to support these custom tags.

I've contacted you directly with my contact details so if you could email me a sample GEDCOM with examples of the different years in use that would allow me to attenpt to modify the program to copy with these custom facts.
Aug 14, 2013 at 6:08 PM
Thanks and sample GEDCOM sent.

Creating a custom tag for Lost Cousins in TMG is very easy to do and I've already done a minimal test against your v2.0.1.2 with no issues
Aug 15, 2013 at 7:46 PM
Thanks for the changes you made in v2.0.1.3 which is now handling the custom census tags (or at last the ones I use) from TMG.
Aug 23, 2013 at 9:12 AM
Just to add another "vote" for TMG users (and another bounce from Peter's newsletter).

I am using the UK Gold 8.08 version which I think is the latest release.

For censuses I copied the standard census tag and created a different one for each census (I even have one entry for 1831!).

In the GEDCOM export section mine reads: 1 EVEN 2 TYPE Census 1881 which Martin can probably translate better than I can.

Martin, I hadn't noticed the validation year option, thanks, you learn something new about TMG every day!

If I can assist at all, please let me know.

Regards - Terry
Aug 23, 2013 at 9:41 AM
Edited Aug 23, 2013 at 10:03 AM
Terry, welcome.

That 1 EVEN 2 TYPE Census 1881 etc looks fine and the census year specific tags were a standard feature of the UK editions of TMG. I envy you to having useful 1831 data. FT Analyzer is now able to handle those without issue (although I'm not sure about the 1831 !!!).

I'm using a somewhat older version of TMG and did have a minor issue when trying to create the new "Lost Cousins" tag in TMG.

For my first attempt, I copied the existing tag "Event-Misc" (via Tools -> Master Tag Type list). Whilst I named the copied tag as "Lost Cousins", I found that on export in my GEDCOM that it appeared as 1 EVEN 2 TYPE Event-Misc1, even though I'd set the label to be "Lost Cousins". I stress this may be an issue with my old version of TMG or some other form of user error on my part.

I did not have any problem using the ADD (rather than copy) button, set the label to "Lost Cousins" and changed the default Style to "UK Standard Place" since that's where the bulk of my research is (but you're not going to use the Address fields on this tag, so probably not significant). I left the Tag Group as "Other event".

The new "Lost Cousins" tag will then show up on the Tag List and all you need to do is add a valid LC year, 1841, 1881 or 1911 for UK enumerations to get the full benefit of the FTAnalyzer Lost Cousins report.

On export the new LC tag should show up in your GEDCOM as 1 EVEN 2 TYPE Lost Cousins 2 DATE 1881 Suggest you confirm it is working as expected after entering just one or two new Lost Cousins tags.
Aug 24, 2013 at 8:40 AM
Edited Aug 25, 2013 at 8:49 AM
Thanks Martin, I will give that a try and report back.

The 1831 census that I have, I came across by accident whilst reading a book that listed which areas had what records and for Wormingford in Essex a large chunk of the 1831 census survived and is typed up onto microfiche at ERO.

I have been on LC for a few years and unfortunately have never made contact with anyone, but I have a fairly complete tree so I am not sure that I am missing anything.

ftanalyzer is a very useful tool (thanks to the author) so I am happy to spend some time tidying up my immediate family data using it.


OK, update: Using the "TAG copy" method appears to work fine on TMG 8.08. Do I need to change the abbreviation from the default "los"?
Aug 24, 2013 at 7:11 PM

I left my abbreviation also as the default of "los." As far as I'm aware the abbreviation isn't significant.

The key for FT Analyser seems to be whether the GEDCOM exported has those three magic lines
2 TYPE Lost Cousins
2 DATE yyyy

where yyyy is 1841, 1881 or 1911, the three years that LC supports from the UK Census

Aug 25, 2013 at 9:41 AM
Hmm, I copied the Event-Misc it seems to create the following:

2 TYPE Event-Misc1
2 DATE 1911
2 NOTE Lost cousin entry

Even though when looking at the TAG itself there is no mention of Event-misc1.

ADDing rather than COPYing works:
2 TYPE Lost Cousins
2 DATE 1911

however with a possible odd effect, on my Grandfather's entry it shows an orange block in ftanalyzer for 1911 and he has a CENSUS 1911 tag as well. He also has a CENSUS 1901 entry and that shows as red.

My G Grandfather has LOST COUSIN and CENSUS tags for 1911 and 1881 and he shows as all green.

OK, I have looked at the GED file and the CENSUS 1901 & 1911 are missing from my export probably as they are not noted as PRIMARY tags. I can change the export for that, but I am assuming that will compound the problem when a CENSUS tag overrules the LOST COUSIN tag in the CENSUS RESULT REPORT so it will always show as green and not orange?

Does it matter where the LOST COUSIN tag is in relation to the date of the CENSUS entry? A LOST COUSIN tag implies that there is a CENSUS entry as well.

If I have misunderstood how ftanalyzer is handling the data I apologise in advance.

Regards - Terry
Aug 25, 2013 at 10:07 AM
Edited Aug 25, 2013 at 10:12 AM
Hi Terry,

The orange block means that the Lost Cousins tag is being recognised correctly but it is not seeing a census record for that person.

The program works by accumulating a list of facts for an individual or family from your GEDCOM file at no point does any fact "overrule" any other fact, they are all still there. As long as your family tree program exports that data to the GEDCOM then FTAnalyzer will import that data and store and process all the facts.

The report you are looking at is designed to show green when all is well. For a regular census year eg: 1901 this is simply that you have entered a census. For a Lost Cousins census year eg: 1911 this has two conditions first you have entered a census and second you have tagged the person as entered on Lost Cousins.

So the program will look at both the census fact and Lost Cousins fact for 1911 and show the following colours:
  • RED - No census fact, No Lost Cousins fact (you need to find this person on a census or enter a census fact for them, try double clicking to auto search)
  • ORANGE - No census fact, Yes Lost Cousins fact (highlights you are missing a census entry)
  • YELLOW - Yes census fact, No Lost Cousins fact (highlights you need to enter the census on Lost Cousins and tag them as entered)
  • GREEN - Yes census fact, Yes Lost Cousins fact (yay you have done everything for this person, for that census year)
So the colours show you all possible combinations of the two facts for that census year. So you can see at a glance missing data. Does this make more sense?

PS. It also shows green if you have a census entry but it isn't a Lost Cousins year so there is no Lost Cousins fact to be entered.

NB. The data errors tab will also show if you've made an error and tagged someone as having a been entered on Lost cousins for a non census year or a non Lost Cousins year.
Aug 25, 2013 at 10:20 AM
Excellent thanks, that makes perfect sense.

I am having an off topic conversation with Martin about the/my TMG output as it has served to complicate things by my exported ged not containing some of the census tags, which I am guessing is down to my export selection, but I can't yet see where at the moment, as I elected to export all tags.

I did wonder why I was seeing so many red blocks when I know that I have the censuses entered for most of my tree.

Thanks again - Terry
Aug 25, 2013 at 10:31 AM
Glad to been of assistance. It would be very useful if either/both of you could knock up some instructions for TMG once you get things sorted, so that other users of that program could know what to do in future.