Modern Software Experience



version seven

On 2007 Dec 30, Wholly Genes (WG) newsletter announced the availability of The Master Genealogist (TMG) version 7. This is the first major upgrade in three years. Version six was announced on 2004.Dec 20.

TMG version 7 follows TMG version 6.12.

Windows version requirement

The most significant news for version seven it that TMG will no longer run on Windows 98 or Windows NT 4. This basic information is not in either the product page or the user guide, but hidden near the bottom of the What’s New page.

The minimum Windows version for TMG 7 is Windows 2000. TMG 7 will run on Windows 2000, Windows XP and Windows Vista. TMG 7 does not require Microsoft .NET. It uses the FoxPro run-time, included in the TMG setup.

Wholly Genes describes TMG 7 as native to Windows Vista, a claim which I consider to be confusing and misleading at best.

Wholly Vista claim

native or certified?

In their announcement of TMG 7, Wholly Genes describes TMG 7 as native to Windows Vista, a claim which I consider to be confusing and misleading at best. Drivers may be native to Vista, applications may be certified for Vista.

no Vista logo

Microsoft has a Vista logo program through which solution providers, such as software vendors, can obtain a Certified for Windows Vista or Works with Windows Vista logo.

The Master Genealogist 7 does not sport either logo. That implies that either TMG 7 does not meet the Vista logo requirements, or that Wholly never even applied for a logo. Personally, I think it is both.

Microsoft may get annoyed over misleading compatibility claims, so I strongly advice Wholly Genes to rephrase their marketing literature and just claim that TMG 7 works on Vista.


By the way, with a bit of fiddling, TMG 6 will work on Vista too. Wholly Genes has a free guide on installing TMG 6 on Vista.

I do not recommend Windows Vista and have not tried TMG 7 on Vista. I have tried TMG 7 on Windows XP Service Pack 2 with all subsequent patches.

wholly inaccurate

Truth is that Wholly Genes’s claim that TMG 7 is native to Windows Vista is as inaccurate as possible. TMG 7 does not even support the Vista look and feel. The truth is that TMG is not even a native Windows application.


FoxPro run-time

TMG is not a native Windows application at all, but a Visual FoxPro application. TMG is built on top of Visual FoxPro and requires the FoxPro run-time to work.

TMG 6 was built on top of Visual FoxPro 7 (VPF7), TMG 7 on top of Visual FoxPro 9 (VFP9).

TMG’s increased operating system demands are only partially to blame on the increased demands of the newer FoxPro run-time. The Visual FoxPro 9 run-time does not support Windows 95 or Windows NT 4, but it does still support Windows 98 and Windows ME. Wholly Genes has not documented why, but although the VFP9 run-time does support Windows 98, TMG 7 does not.

FoxPro end-of-life

Microsoft has stopped development on Visual FoxPro and that has serious effects on genealogy software (see Microsoft kills FoxPro and The Effect of FoxPro’s Death on Genealogy). Wholly Genes needs to consider its options and communicate what they are going to do. So far, no such message has been forthcoming.

Wholly Genes might just as well follows Legacy’s example and offer their lesser edition for free.

two editions

Wholly Genes offers TMG in two editions, known as the Silver and the Gold Edition. The Gold edition includes a printed manual [2008-01-28: not anymore, see update at bottom] and several report formats and features not available in the Silver edition. If you have a look at the feature comparison chart by Wholly Genes itself, you will notice that the Silver Edition lacks features that you probably do not want to do without.

The Gold Edition is the full product. The Silver Edition is the same product, but with some features left out. Now, I would love to say that the Silver Edition is the Gold Editions minus some nice-to-have advanced features many researchers can do without, but that is not case. The Silver Edition lacks basic must-have features.

Silver edition users must not only do without most wall charts and without direct export to word processor formats. Those can be considered nice-to-have luxury features, but Silver editions users must also do without HTML output, and be satisfied with reports that have neither a table of contents nor an index.

Wholly Genes has crippled the Silver edition so much, that it is insufficient for anyone who is remotely serious about genealogical research. The Silver Edition is arguably inferior to all the free genealogical software that does offer HTML output and full reports. Wholly Genes might just as well follows Legacy’s example and offer their lesser edition for free.

two flavours

Wholly Genes is an American company, and TMG was originally designed for the American market. Explicit support for the UK was first added to version 5. Since then, TMG has been available in two flavours, an USA flavour and a UK flavour. Now, it is perfectly possible to use TMG if your ancestors are not from either country, but if they are from the USA or the UK, you may like the program better because of its inclusion of USA and UK specific features.

multi-language support

Wholly Genes claims that TMG supports ten different languages, but that is not correct, it actually supports eight different languages in support of ten locales. The languages are Afrikaans, Dutch, Danish, English, French, German, Italian, and Norwegian (both Bokmål and Nynorsk).

When you select anything but the English-English (en-EN) or the English-American (en-US) locale, a dialog box pops up warning you that TMG is sold as an English-language only product. I gave the Dutch menus a quick glance and they looked good to me. A random report I tried was - apart from user-written notes - in Dutch too, but there is no Dutch help file.

locale differences

There is no English help file either. There is just one help file, and it is in Amglish. The setup program for the UK flavour allows installation in several languages, but it defaults to Amglish because it does not support English.

The support for English is decidedly half-hearted. Wholly Genes’s idea of supporting English is adding the correct English spelling for a few hundred words to their Amglish dictionary, which is decidedly the wrong way round.

The differences between the UK flavour and the USA flavour is in such things slightly different settings and installation of a few English timelines. I find it strange that Wholly Genes offers this as two different flavours, instead of just offering single product, with pre-set option collections.

CD-ROM vs. download

Now, if you have decided to buy this program, have decided on an edition, and have decided on a flavour, you still have to decide between a download and box with a CD-ROM. The physical CD-ROM option is slightly more expensive than the download-only option, but the CD-ROM includes The Universal British Directory of Great Britain 1791 from Archive CD Books, and if you opt for the Gold edition on CD-ROM, you also get a printed manual.

All in all, I’d say that the UK flavour of the Gold Edition on CD-ROM is your best option, if you decide buy it, but I am not warmly recommending this monstrosity.


side by side

The TMG installation program is made with the Wise installation wizard. It does not bother to try and detect a TMG 6 installation, and defaults to a directory with a 7 in the name. So if you have TMG 6 already, TMG 7 will not default to installing over it, but to installing along side it. That allows you to run TMG 6 and TMG 7 at the same time.

restart demand

A full installation requires a bit less than a quarter gigabyte.

The installation program disappoints. Once it is done, it claims that The system must be restarted to complete the installation.. That seems nonsense to me. TMG is just a FoxPro application, and why would the mere installation of a FoxPro application require a complete system restart? I believe the restart to completely unnecessary and therefore to be in direct violation of Microsoft’s installation guidelines.

I clicked the Cancel (restart) instead of the OK button and then clicked the desktop icon it had created. The TMG 7 install does not provide an option to choose whether you want that desktop icon or not, it just creates it.

Once you click it, you get to see a progress bar, with the text Configuring The Master Genealogist v7 above it, then another dialog box you can’t even switch away from, and finally a dialog asking you to choose between running the trial version or unlocking the full version. You make a choice, and then get the Welcome to the Master Genealogist dialog box.

All this worked fine without having restarted, and that seems to confirm my suspicion that the demand to restart is bogus.

TMG 7 does not read TMG 6 projects directly.

converting from version 6

If you are using TMG 6, do not expect an automatic or flawless conversion to version 7. You would expect to install TMG and continue working, but that does not work. TMG 7 does not read TMG 6 projects directly. To access your TMG 6 projects from TMG 7, you must create a backup in TMG 6 and then restore from that backup in TMG 7.

Once a project has been converted to version 7, it cannot be converted back to TMG 6, so you will be happy you made those backups. It seemed to work for the empty project I created in TMG 6, but that may be a border case, and it is not officially supported.

TMG 7 did not pop-up any warning about the project version, but just went ahead and read it. There was no message that is was upgrading a version 6 database to version 7 format. On a happier note, when I tried to restore the version 6 database without changing project name, TMG did warn that I was about to restore a backup over existing data and asked me whether I was sure I wanted to continue. The restore proceeded without errors.

When you decide to backup, the program does not just go ahead and do it, but presents you with yet another lengthy series of dialogs to click through, where you have enable lots check boxes to make sure it actually performs a backup at all.


That the backup defaults to disk drive A: feels like a throwback to the sneaker nets of the 1980s, but that is the least of the annoyances. When you decide to backup, the program does not just go ahead and do it, but presents you with yet another lengthy series of dialogs to click through, where you have enable lots check boxes to make sure it actually performs a backup at all. It could have done a complete project backup in the time it takes you to click through this user-unfriendly mess confirming that you really want to do a backup.


I advise anyone who is switching from version 6 to 7 to make those backups before installing version 7. When I tried to run TMG 6 again, and merely opened its sample project, the FoxPro run-time popped up Abort, Retry, Ignore message boxes. I even got Unknown member CNTIMAGE. 1042 OPENDATASETS - Abort, Retry, Ignore pop-ups when I decided to create an empty project.
Trying to think of something simpler for the program to do that it could not possibly fail at, File| Exit was the only thing that came to mind quickly.

It is possible to continue by choosing to ignore the program errors, but these continual program errors do not give me a warm fuzzy feeling about the code quality. What really worries me though is that googling for UNKNOWN MEMBER CNTIMAGE brings up a complaint from 2004 about the same message in version 5.15. This is version 6.12, and this version 5 defect still isn’t fixed?

sample database

I was very interested to know whether Wholly Genes had improved its import routines at all, but I started with a look at the sample project, to get an overall feel for this version.

This sample project is a small database, just 63 people and 73 relationships. Unsurprisingly, TMG opened it quickly enough, but it started to do something immediately after opening, and became unresponsive, in fact even so unresponsive that Window changed the title bar to include the infamous this program is not responding text. I waited a few seconds more.

editions and flavours

Message Manager

Once TMG was finished doing whatever it was doing, a Tips and Hints windows appeared, and the program responded normally again. It also tried to connect to the Internet for no apparent reason. As TMG had not indicated that was going to connect or why, I initially denied the access, and that did not seem to matter. I then closed TMG, restarted it, opened the sample project again, allowed Internet access this time, and then got the Message Manager screen.

Its welcome message makes it clear that the TMG Message Manager is similar to Legacy’s news and update check. The Message Manager defaults to checking for new messages every seven days, but is configurable from a menu.

I advise TMG to always display a brief message before trying to connect, so that we know why our firewalls display an alert, and to provide the user with the option to turn the automatic connection attempt off before the first connection attempt is made.

user interface


TMG version 7 will look very familiar to version 6 users. You can run them next to each other, and it is hard to tell the difference. They look just the same.

All that can be said in favour of TMG 7’s needlessly complex and annoyingly unintuitive user interface is that it will be familiar to those who struggled to learn TMG 6’s user interface.

add a person

I tried to add a new person as you do in many other programs, by clicking or right-clicking the obvious location on screen, but even version 7 of TMG still isn’t intuitive like that. Wholly Genes continues to ignore Windows user interface design guidelines. All that can be said in favour of TMG 7’s needlessly complex and annoyingly unintuitive user interface is that it will be familiar to those who struggled to learn TMG 6’s user interface.

Browsing through the Wholly Genes forums, I learned that there are issues with the add button in version 7.0. It is certainly noteworthy that a program has defects in basic functionality like that, but it is still possible to add persons. It isn’t the defects that bug me, it is the broken by non-design user interface.


It is very disappointing to note that after three years of development, the convoluted user interface, easily user complaint number one, has not improved at all. If anything, it has worsened. They fixed the embarrassing window resize issue (where everything, including all fonts changed sized if you resized the window), but did nothing to the overall interface, and have not issued a patch for TMG 6 users either.

TMG version 7 still suffers from the Cue Cards clutter added in version 5.

Cue Cards clutter

TMG version 7 still suffers from the Cue Cards clutter added in version 5. The so-called Cue Cards are literally another layer on top of the basic interface. You get ostensibly helpful windows that not only clutter up your screen even more, but are sure to annoy you by breaking your workflow. Generally, their brief content manages to be completely unhelpful, even downright insulting, by stating nothing beyond the most obvious. It would be better to add tooltips, more context-sensitive help, or perhaps provide some starter videos. These so-called Cue Cards only worsened an already overloaded interface.

Beginner Buttons

Instead of finally fixing TMG by overhauling its interface, Wholly Genes has cluttered up the Person View with what they call, I kid you not, Beginner Buttons. Apparently, the folks at Wholly Genes believe it to be a better idea to label you as a beginner for not having memorised their idiosyncratic interface, than to actually fix that interface.

And as if the Cue Cards were not bad enough, there are now are Reminder Popup Windows to, eh, help you make sense of the existing mess.

If TMG had a normal, usable user interface to begin with, no Cue Cards, Beginner Buttons and Reminder Popup Windows would be necessary.

Reminder Pop-Up Windows

If TMG had a normal, usable user interface to begin with, no Cue Cards, Beginner Buttons and Reminder Popup Windows would be necessary. The problem is not that new users do not understand the convoluted menus, the problem is that the needlessly convoluted menus get in the way of using the program’s features.

The problem is not that users have trouble understanding the program, but that Wholly Genes has not exerted itself to understand its users. Users are often shy to suggest more than small improvements. Wholly Genes’s competitors will be happy if Wholly Genes continues to kid itself that following up a few such suggestions makes them oh-so responsive to user wishes and needs. What the users really want; what TMG really needs, is a complete overhaul of this poor excuse for a user interface that constantly gets in the way of using the program’s features.

new features

I mentioned one new feature already: the Message Manager is new to version 7. Most changes since version 6 are little things such as the ability to sort on columns that could not be sorted before, and more preference options. I am not going to try and mention every little change, but do want to highlight a few features.

If there is one reason to recommend TMG despite its many serious flaws and shortcomings, its complex but flexible reporting capabilities must be it.


The latest version offers yet more control over report narrative than before. Control over the narrative produced by reporting is one area in which TMG really shines and it has become even better. If there is one reason to recommend TMG despite its many serious flaws and shortcomings, its complex but flexible reporting capabilities must be it.

image annotation

TMG now offers the ability to annotate images – of course, like the popular FotoTagger utility, without modifying the original image.

Associates Window

TMG’s best new feature may be the Associates Windows – a window that shows everyone who is in any way, through any event, related to the focus person. It may not seem a feature with a direct application, but it is quite practical to see what other persons appeared on the same census form, lived in the same house, whom they were marriage witnesses for, etcetera. It may help you discover relations and inspire research directions.

Check for Duplicates

Then again, the one feature that was still lacking in version 6 and may now convince you to upgrade to version 7 is Check for Duplicates While Adding People, with the matching rules configurable through a preference page.

bug fixes

There is a whole list of bug fixes included in version 7. I am happy that Wholly Genes is acknowledging those defects, and I applaud them for publishing this list. What bothers me is that Wholly Genes appears to take the position that you should pay for the upgrade to get those fixes.

I firmly believe that TMG should provide a free Service Pack to existing TMG 6.12 users. Wholly Genes should sell version 7 because their customers desire the new version 7 features, not because it is the only way for existing customers to get version 6 defects fixed. Besides, existing Windows 98 and Windows NT 4 customers cannot even run version 7.

GEDCOM import

refuses GEDCOM

I started my GEDCOM import testing with royalfam.ged, a small file of just 3 ½ MB.

TMG 7 refused to import it: The UTF-8 GEDCOM D:\GEDCOMs\royalfam.ged can’t be imported. Please see the GEDCOM Import help topic..

no UTF-8 support

The help button on the import dialog brings up the help file, but not the right topic. After a little digging through the help file, I found this message:

If you attempt to import a GEDCOM that was exported using the UFT-8 (Unicode) character set, you will get an error message referring to a 'missing' or 'wrong' header or to the fact that this is a UFT-8 GEDCOM. At the present time, TMG does not support Unicode and cannot import these GEDCOMs..

What? Unicode became part of GEDCOM with the GEDCOM version 5.3 draft, released on 1993 Nov 4, and the official GEDCOM version 5.5 specification was released on 1996 Jan 12. That is a dozen years ago.

help file does not help

The help files continues See the following FAQ for a work-around:

Pardon me? What’s this? Why am I directed to an URL? Why are the few lines of text I find at that URL not in the help file itself? What kind of help is that? What if I encounter this TMG limitation in an environment without Internet access? I am supposed to drive to the nearest Internet cafe just because Wholly Genes did not want to list a major program limitation in their help file?

I am supposed to drive to the nearest Internet cafe just because Wholly Genes did not want to list a major program limitation in their help file?

damage your data

If you follow the URL, you arrive at a topic buried deep in Wholly Genes’ forum, where it tells you to open the GEDCOM file in NotePad, change the character encoding to ANSI and then save he file again as Windows ANSI, and to simply ignore the warning regarding the loss of information by clicking OK.

Pardon me again? Ignore the information loss? Wholly Genes tells its users to simply ignore it? Don’t worry about it, just do it? Never mind that TMG can’t import a standard GEDCOM file, just damage your data and all will be fine? What the Fake?

Import speed is perhaps better expressed in minutes needed per megabyte than megabytes processed per minute.

another GEDCOM

The GEDCOM import does not impress at all, not even for the GEDCOMs that TMG does support. I tried Steeves.ged, a Brother’s Keeper 5.2 GEDCOM file encoded in Windows ANSI. The GEDCOM import apparently starts the processing by checking the length of all the GEDCOM lines, and takes several seconds per megabyte to do so. Only after that first pass do you get to name the imported project and choose a directory.

Import is definitely on the slow side. So slow, that TMG does not provide import statistics when it is done. Import speed is perhaps better expressed in minutes needed per megabyte than megabytes processed per minute.

error pop-up

Once the import is done, TMG suggests reviewing the master place list, and offers to show the import log. I decide to view the import log. The log appears, but that is not all that appears. Together with the import log appears an Abort, Retry, Ignore dialog box, stating Menu has not been defined with DEFINE POPUP. 34 FRMMEMO.OLETER1.LOSTFOCUS.

That isn’t a TMG error message, that is a FoxPro runtime error message, telling me that the code fouled up. When I decide to ignore the programming error, I just get the same error but 39 instead 34, 44 instead of 34, 47 instead of 34. These numbers are apparently line numbers in some module in the TMG FoxPro Basic program, and for each of these lines, FoxPro reports an error. When I decide to Retry, the number does not increase; FoxPro keeps trying the same TMG program line, and every time it tries, the code fails. I decide to keep down the I key, as a quick way to keep pressing the Ignore button. After perhaps several dozen pop-ups, the import continues.

The import files notes a few warnings. The messages are clear, and do seem to concern GEDCOM mistakes of Brother’s Keeper. The messages in the import list include line numbers, a clear message and the line itself.

The import listing does not provide statistics. I estimate that it took TMG about 15 minutes to import a 5 MB file.

import speed

total experience

To get a good idea of import speed, I timed the import of HundredThousand.ged, a GEDCOM with just over 100.000 individuals in it. This is an ANSEL GEDCOM created with PAF

It takes TMG a while to start displaying a progress bar. While this progress bar is being displayed, TMG is otherwise unresponsive, and does not even repaint it own windows. The progress bar seems be updated about once a second, and soon I am looking at a progress bar against a the background of a completely white window.

Unresponsiveness during import is a problem with all but the most carefully coded genealogy programs.

It takes about ten minutes for this progress bar to finish. That is not the import time, but how long it takes TMG to find the length of the longest line in the file. I confirm the project name and file location to start the actual import.

Export continues without run-time errors. It takes 1h25m before I am allowed to click okay to display the listing file. TMG then takes about another minute for some validation, bringing the total import time to one hour, 26 minutes and 12 seconds.


It seems the import has improved since version 6. I managed to import this file into version 6 just once, and it took many hours. Moreover, on every other try, the import would stall on some FoxPro run-time error. The version 7 import happened without any run-time errors. This suggests that Wholly Genes has taken some time to debug their GEDCOM import code.

By the way, the error I encountered with the other file does not seem an import error to me, but just coding error in the pop-ups introduced in this version. The imported file looked fine.


I do think that an import should not require me to click once I’ve started it, and that requiring me to do so not once but twice is a user interface blunder. Still, I’ve discounted the 12 seconds to more than compensate for the time it took me to click several times. That leaves an actual import time of one hour and 26 minutes, and that is 5160 seconds.

Hundred thousand individuals in somewhat more than 5000 seconds is a bit less than 20 individuals per second. That is very slow.

import listing

The import listing consists mostly of the oft-repeated warning that TMG does not recognise PAF’s _AKA tag, and a few warnings that TMG does not recognise the _PRIMARY tag. That is acceptable, as these are PAF extensions. TMG also fails to recognise the EMAIL tag, but then again, PAF should not have included it in a GEDCOM 5.5 file, as the EMAIL tag is a GEDCOM 5.5.1 feature.

primary parents

The import listing complains about individuals with multiple parent couples. It is perfectly normally for someone to have multiple parent couples. For example, when someone remarries and the new couple adopts the kids from the first marriage, those kids have multiple parent couples. TMG can handle that just fine. TMG’s warning is not that having multiple parent relationships is wrong, but that I should label one of these as the primary. That may be good advice, but the import would not need to produce that warning if it had understood PAF’s _PRIMARY tag.

Somehow, I expected version 7 of a program that has been around so long, and is advertised with the phrase the one that does it all to import a PAF GEDCOM without a hitch.


Based on just the import listing, it is tempting to conclude that import failed for non-standard tags only. That would make the import correct yet somewhat disappointing. Somehow, I expected version 7 of a program that has been around so long, and is advertised with the phrase the one that does it all to import a PAF GEDCOM without a hitch.

database changed

When I browse through the imported data in TMG, I soon notice that the quality of the import is less than the import listing would lead you to believe. There are additional issues, not listed in the import listing file.

TMG does not support the _UID tag included in PAF files. As _UID is a PAF extension, it is perfectly valid for TMG to ignore it. Instead, TMG added comments to my database, something a program should never do without my explicit permission.

inconsistent handling

It bothers me that TMG puts warnings about the unrecognised _AKA tag in the import listing, but does not put warnings about the unrecognised _UID tag in there. This inconsistent handling of unrecognised tags is worrisome.

Unknown GEDCOM tag: CREM

One TMG-created comment I noticed is Unknown GEDCOM tag: CREM. TMG does not complain that CREM is misplaced, it simply complains that the tag is unknown. I did a double-take when I saw that one. A quick check of the GEDCOM 5.5.spec confirms that CREM is a valid GEDCOM 5.5 tag, which was added in GEDCOM 5.4. It is a valid tag.


TMG is technically correct that EMAIL is not a GEDCOM 5.5 tag. Whatever I think of the practice, most GEDCOM import routines accept GEDCOM 5.5.1 tags in a GEDCOM 5.5 file. So, after noticing that it does not import CREM correctly either, I have to wonder whether it really supports GEDCOM 5.5 or perhaps merely GEDCOM 5.3? Some tests might answer that question, but I decided to move on and have a look at the GEDCOM output instead.

GEDCOM export


I do not like TMG’s Wizards. There simply are too many screens with too many options when you just want to do something. You cannot simply choose to export to GEDCOM, perhaps specify a file name and sit back. No, you must click through eight (!) tabs before TMG starts exporting. The various tabs offer all kind of options, and some will appreciate that, but I am strongly in favour of just a few basic options and a just do it button on the first wizard screen.

TMG’s export routine is not fast. In fact, it is downright slow.


I notice that the GEDCOM header list the TMG version as VERS v 7.0. You might think that it should be VERS 7.0 instead, and that is certainly best practice, but the GEDCOM spec allows any version-identifying string of up to 15 characters. The header is a bit silly but correct.

After import into and export from TMG, all the CREM tags have gone up in smoke.

TMG comments gone

I browse through the TMG GEDCOM to see what has become of the comments that TMG added to my file. These are nowhere to be found.

After import into and export from TMG, all the CREM tags have gone up in smoke.

I tested with a small filed that contained one individual, with a birth, death and cremation event. TMG changes the cremation into a comment upon import, and when you export again, not even that comment remains.

Upon import, TMG had changed each PAF _UID into a comment. When I created an ahnenlist to evaluate the HTML output, these comments disgraced the result, but when it generated a GEDCOM, the comments did not appear.

invalid DATE tags

leading zeroes

I also notice that the DATE tag for a CHAN tag is always 36 characters (34 characters followed by CR/LF) long. A correct CHAN.DATE tag consists of four character for the CHAN tag, one space, 10 or 11 character for an exact date, and the line terminator.

I notice two GEDCOM violations in TMG’s CHAN.DATE tags. First of all, TMG uses leading zeroes in dates (e.g. 08 Oct 2005), whereas the GEDCOM spec explicitly states that CHANGE_DATE is of type DATE_EXACT and must have a length of 10 or 11. The change date would never have a length of 10 if GEDCOM demanded leading zeroes, nor does the standard ever provide an example with leading zeroes. GEDCOM defines DATE_EXACT as consisting of a DAY, MONTH and GREG_YEAR, in that order, with a DAY part that is either 1 or 2 characters long.

All that Wholly Genes can mount in their defence is that the specification does not explicitly forbid leading zeroes, but that hardly seems enough reason to produce GEDCOMs with values that some readers will not expect.

A search and replace command to search for DATE 0 and replace it by DATE has 58.367 results.

trailing spaces

The second violation is that the TMG GEDCOM contains 18 spaces after each change date. That makes the value following the CHAN.DATE tag 29 instead of 11 characters long. The GEDCOM spec does not allow any spaces there.

The use of trailing spaces in the DATE field is a direct violation of the spec, and that means that other applications may reject such files as not being GEDCOM files. Realistically, the use of trailing spaces is unlikely to pose a problem with any but extremely inflexible GEDCOM import routines.

Still, a quick search/replace in a hex editor shows there to be 100.068 stretches of 18 characters each, one for each INDI record, and a hundred thousand times 18 bytes totals 1,8 MB of wasted space.

You’d expect about one third of the change dates to have a leading zero, but there are many more dates than just change dates. Remarkably, the superfluous chain of 18 spaces occurs after change dates only, but the leading zeroes appear on more dates.

HTML output

not obvious

It is not immediately obvious how to generate web pages. There is no option to export to HTML in the Export Wizard. There is Web menu, but it essentially just a bunch of links. There is no Create Website menu item. Instead, TMG supports HTML as one of an output format for its reports. That gives you all the same reporting capabilities for a web site as for other output formats.


I decided to try a simple ahnenlist (which TMG calls an ahnentafel report). The report options contain an HTML tab where you can customise the output a bit by directly specifying a title or a background colour, but I go with the defaults.

no style sheets

The output files always have the three-letter *.HTM extension, not a four-letter *.HTML extension. Although they are more than ten years old already, TMG still does not support style sheets yet. Because of that, I had already lowered my expectations to old-style hardcoded HTML with a few hyperlinks, yet TMG’s output still managed to disappoint. What I saw made me cringe.

TMG’s so-called HTML output carries only the vaguest resemblance to actual HTML.

not HTML

Where do I begin? Although computer-generated, the output does not validate as HTML at all. I did not expect XHTML 1.1 or XHTML 1.0 Strict, but it is not even XHTML 1.0 Transitional or HTML 4.01. It is not even valid HTML 3.2 or HTML 2. TMG’s so-called HTML output carries only the vaguest resemblance to actual HTML.

errors, errors, errors

There is no doctype. There is no character set declaration. Open and close tags do not mach, there is even a </font> close tag although there is no <font> open tag anywhere on the entire page. The tags are uppercase instead of lowercase. Tags are placed illegally (<center> within <h2>).

It is one big hopeless mess. I wondered for a second why the W3C validator complained about illegal placement of the <body> tag, while it came directly after the <head> tag as it should, but then I noticed that the <html> tag itself is missing…

To describe TMG’s HTML output as a proof-of-concept code, hastily cobbled together more than a decade ago, and never looked at since would be sheer kindness. It is the very antithesis of professional HTML. Wholly Genes should thank browsers for their amazing ability to make some sense of it and display TMG’s output at all.

It is easy to miss in the maze of menus that there is a way to combine multiple reports into a single website. The Book Manager menu items lets you combine multiple reports into a book, and those reports may be HTML output.

Then again, with output like this, you should not use TMG for HTML output, but pick an alternative, any alternative – its output couldn’t possibly be worse. Many programs will generate a website from a GEDCOM file, and John Cardinal’s Second Site will generate it directly from TMG’s FoxPro database.

master place list

TMG has a master place list. However, TMG’s place name import is rather disappointing if you have not carefully structured your place names already.

filling columns backwards

TMG’s place name import does not seem to have improved; it simply looks at your comma-separated place name field, and starts filling its columns backwards.

Thus, if you have just a place name, TMG will sort it into the country column. If you have a place name and a province, the province will be sorted into the country column, and the place name in the province. If you have a place name and a country, the place name will be sorted in the province column.

Although this must have been suggested by many users, TMG still does not use a master list of countries and their provinces to improve the import results.


That the master place list labels the province column as state is a bit jarring, but does not affect the program’s capabilities. And the import limitations discussed so far are not all bad. Those limitations may make you add provinces and countries to all your place names, so that you do not have to suffer these limitations again the next time you import into TMG.

county handling

A real problem is the inclusion of county, or rather how TMG handles it. Many countries have regions between the city and province levels, but more often than not, their use in addresses is not just rare, but perceived as pedantic or plain wrong. Moreover, many American genealogists do not include these regions in their place names either. Yet whenever you do not include the region, TMG puts your place name in the region column….

You can correct the place names after import, but you will have to do that again and again every time you import. TMG does not import correctly, and does not remember your corrections either. TMG’s simplistic approach may be acceptable for small databases, but is entirely unpractical for large ones.

PAF error

The first thing I noticed on the Master Place List was a bunch of &#00063 entries. That is a question mark (?) encoded as a character entity reference. That is not TMG’s fault, but PAF’s fault. This way of coding characters is unnecessary but allowed in UTF-8 output, PAF makes the mistake of using it in ANSEL output.

double-dotted i

One place name includes a double-dotted ï. PAF correctly encoded this in ANSEL as E8 69 (Combining Diaeresis followed by Latin Small Letter I), but TMG did not import it correctly. TMG changed the double-dotted ï into a Diaeresis followed by a Latin Small Letter I - two characters instead of one.

This is very surprising. The Letter Small Letter I with Diaeresis, to use its Unicode name, is so common that it is part of the 256 characters in the Windows ANSI character set (it is number 239, hex code EF).

There is really no excuse to not support it, but what bothers me even more than not supporting it, is that the import listing made no mention of this TMG failure.

TMG direct import

direct import

Direct import from genealogy databases has always been popular. The popularity of direct import declined as GEDCOM became popular. However, GEDCOM has limitations too, and direct import is simpler to use.

import, not export

TMG supports direct import from many competing product. This feature is aimed at getting your data into TMG, not at getting it out again. It is direct import only, direct export is not supported, and TMG’s file format is not documented.

Wholly Genes used to document the TMG file format, but seems to have stopped doing so about ten years ago.

undocumented file format

Wholly Genes used to document the TMG file format, but seems to have stopped doing so about ten years ago. There does not even seem to be a link to that old document on their site anymore, but if you google around, you may figure it out, or happen across the File Structures for the Master Genealogist thread in the Other Downloads subtopic of the News and Information topic of their forum.

The file contains tmfstr.rtf, a 1998 May document describing the TMG 3.6 file format.

commercial product

Wholly Genes offers TMG’s direct import module as a separate product. They call this GenBridge. Once you know how unreliable and slow TMG’s import is, it comes as no surprise that only a few small vendors without the resources to develop import themselves licensed it. For many years, not one large commercial vendor was willing to risk their reputation on GenBridge - until’s Family Tree Maker 2008.

Family Tree Maker 2008

Family Tree Maker 2008’s failure to import a PAF database (see FTM review) through GenBridge does not impress, but did not surprise me either, as I had wasted quite some time already attempting to import various files into TMG 6 already.

What really surprised is when I saw complaints (.e.g. on the blog) about FTM 2008 not importing TMG correctly. Complaints that are confirmed by FTM 2008’s Service Pack 1 which includes Resolution of issues pertaining to import of The Master Genealogist files.
If there is one format that TMG’s import module should have no problem handling, it is it’s own format. Oh well, that was FTM 2008, this is TMG, with the very latest version of Wholly Genes’ import code.

what matters

That Wholly Genes offers TMG’s direct import module as a commercial product does not matter to the TMG user. All that matters is how well TMG’s direct import works.

To find out, I asked TMG to directly import the original PAF file that I created the hundred thousand individuals GEDCOM from.

direct import speed

TMG generates an import log, which ends with the sentence Import completed in 6 hrs, 22 mins, 13 secs.. You really need to add 20 to 40 seconds to that for initial database generation before the start of the import, clicking through import wizards, and clicking through all the windows that pop-up once the import is done.

Let’s call it a round 6h22m30s. That is 382,5 minutes, is 22950 seconds.

100.067 / 382,5 = 261,6 individuals per minute.

100.067 / 22950 = 4,36 individuals per second.

The direct import is more than 4 times as slow as TMG’s already snailtacular GEDCOM import.


This is extremely slow. The direct import is more than 4 times as slow as TMG’s already snailtacular GEDCOM import. A direct import can take advantage of the indexes in the file to import faster, so it hard to understand why the direct import is so snailtacular.

import errors?

The import listing does not report any errors or import limitation.

The direct import does recognise PAF’s a.k.a. field (_AKA tag) and imports it as Name-Var.

However, when I looked for the place with the double-dotted ï to see how that faired, I was surprised to notice that the master place list looks different.

The Latin Small Letter I with Diaeresis was imported correctly, but it was not where I expected it, because the whole list looked different. Using GEDCOM import, a place string such Hannover, Germany was interpreted as a State (province) and a Country, using direct import, it is interpreted as a Detail and a City. Sigh.

On the bright side, the direct PAF import did not mess up the Latin Small Letter I with Diaeresis.

duplicate people?

Most genealogy software does not fare to well when I ask it to suggest possible duplicates in a file of hundred thousand lineage-linked individuals. The typical result is a list so huge that it becomes impractical to take these suggestions seriously.

I wondered how TMG would do. There are quite a few options, and that may be the key to getting a manageably small list of highly likely matches. There are definitely more options than I’ve seen on other programs. Still, for my initial try, I stuck with the defaults Wholly Genes had plugged in. The 100k INDI database is an old file and I now know for a fact that there are duplicates in there.

I found this new feature to be no better than those in other genealogy programs. With any reasonable-sized database, coming up with suggestions takes way too long, and the number of results is simply excessive. Perhaps, with appropriate tweaking of the settings, TMG might produce a more manageable result list.


There is the integrated to-do list. There is support for DNA data.


A timeline is a fairly simply feature, that is gaining in popularity. Several other programs support timelines, and you may have seen it on some genealogical web sites too.

TMG comes with a bunch of ready-made timelines, and you can you can choose which ones it should use. I do not get very enthusiastic about canned timelines, but TMG includes a dialog to create your own list, to type in the events you consider important for your genealogy.

quality issues


Wholly Genes’s own forum is actually a good place to get an idea of the many quality issues this product has. There are messages about little things like menu check marks not working and more serious things like FoxPro Fatal Error, Shutdown While Editing Details.

What worries me are the messages about defects that have apparently not been fixed yet although they were reported years ago, for previous versions of TMG.

structure version missing

Every program has defects. What worries me are the messages about defects that have apparently not been fixed yet although they were reported years ago, for previous versions of TMG. For example, the structure version missing is a message that appears on start-up for no apparent reason, after which you much restart TMG. This cryptic message seems to be related to TMG’s less than perfect ability to read and write its own project format. It was apparently first reported in 2005 for version 6.07, and is still being reported for version 7. Obviously, I am not the only one still experiencing defects that were first reported years ago.


Wholly Genes native to Windows Vista claims is wholly inaccurate.

misleading marketing

wholly inaccurate

Wholly Genes native to Windows Vista claims is wholly inaccurate. TMG does not have any Vista certification and simply isn’t a native Windows program.

does it all

Wholly Genes’s announcement once again presents TMG as a the program that, as they like to say, does it all, but it cannot even handle an ANSEL GEDCOM, and its ostensible HTML output does not exactly reduce obtaining another program for web publication to a superfluous luxury.

meeting the bar

The TMG 7 announcement is titled THE BAR IS RAISED AGAIN FOR FAMILY HISTORY SOFTWARE, in all-caps. Apparently, Wholly Genes believes that TMG jumps a high bar that products of other vendors cannot reach, that their product is a standard of excellence that others can only aspire to.

It is perhaps slightly more accurate to say that even this very latest version of TMG still does not meet the lowest bar for family history software. Although TMG 7 actually demands a Unicode-based version of Windows, it still does not support Unicode at all.

inadequate product


Wholly Genes could try to explain the lack of Unicode support as a FoxPro issue, but no one is forcing them to continue to build on FoxPro (au contraire), and they cannot blame FoxPro for TMG’s tormenting lack of performance. FoxPro has a well-deserved reputation as a fast desktop database, so TMG’s snailtacularity is quite an accomplishment.

GEDCOM import

The quality of the GEDCOM import is poor. TMG refuses to import UTF-8 GEDCOM files, and its ANSEL GEDCOM import cannot be relied upon. TMG does not recognise basic GEDCOM 5.5 tags such as CREM, the handling of unrecognised tags is inconsistent, and the import log it creates is incomplete.

The import listing is incomplete. TMG fails to list all import failures in there. You need to examine the database to find out.

The quality of GEDCOM export is not much better.

HTML output

The ostensible HTML output is below amateur level. If you decide to buy TMG despite it other shortcomings, you will probably want to buy Second Site sooner or later. Wholly Genes knows this, and does not even mind implicitly contradicting their the one that does it all line by selling it through their own site…

user interface

The user interface has not improved, but is even worse than before. It is a magnificent example of how not to do it. It unintuitive, cluttered, annoying and literally insulting.

quality assurance

Quality assurance appears to be a low priority at TMG. I soon experienced a defect first reported for version 5, and others experienced defects first reported for version 6.


Apparently, Wholly Genes is infatuated with features over everything. They seem focused on having yet more features, but unconcerned about either the quality or usability of existing ones. Basic functionality is absent or defective. Data is lost, and databases are messed up. Code quality is poor and long-standing defects remain unfixed. The result is one of the slowest, least reliable, hardest to use and overall worst genealogy programs ever made.


2008-01-28: no more printed manual

The TMG Newsletter reports that, effective immediately, TMG will no longer ship with a printed users guide. The guide is still included as a PDF on the CD-ROM, and available as a free download from the site. The price for the slimmed down package has dropped from $79.95 to $ 69.95. Shipping cost and taxes drop accordingly.

The newsletter additionally warns that resetting defaults will reset not just program settings, but custom reminders as well.

2010-04-24 ahnentafel

Corrected abusage of ahnentafel.

2011-05-23 TMG 8 Beta

TMG 8 Public Beta is a Quick Look at the just released TMG 8 Beta.

product details

productThe Master Genealogist
companyWholly Genes
websiteWholly Genes: The Master Genealogist
priceUS$ 69.95 (Gold Edition)
requirementWindows 2000 or later
noteBased on FoxPro, a discontinued platform.
VerdictPoor Product with featuritis.
RatingBelow par.