Modern Software Experience


Family Tree Maker 2008

Family Tree Maker 2008, a complete rebuild of the product, accomplished something entirely unexpected: it made Family Tree Maker Classic look good.


Family Tree Maker 2008 was a disastrous release. Family Tree Maker was never a particular good program to begin with. Its poor database design, arbitrary limitations and crummy GEDCOM support have caused its unfortunate users many problems.
Family Tree Maker 2008, a complete rebuild of the product, accomplished something entirely unexpected: it made Family Tree Maker Classic look good.

Family Tree Maker 2008 is so awful, that many who tried Family Tree Maker 2008 switched back to Family Tree Maker 16, to an even earlier FTM Classic version or simply decided to switch to a competing product in utter disgust.


The Family Tree Maker 2008 review concluded that FTM 2008 is a downgrade. I still consider the downgrade verdict for FTM2008 to be one of my milder conclusions. Family Tree Maker 2008 won the Worst Genealogy Product of 2007 award with overwhelming ease.

Apparently, my FTM 2008 verdict has some legal backing. As I was reading up through the posts on’s own Family Tree Maker blog, I came across a comment by Collin Patrick, who notes that was advertised as a FTM Upgrade. This contravenes the Trade Discription [sic] Law here in the UK, it’s not an upgrade, and that he got the U.K. distributor to refund his money because of that.

Users tend to prefer the program they already know, but FTM 2008 is not much like FTM 16 at all, and it is so bad, that my Legacy 7 review even suggested that current Family Tree Maker 16 users should not pay for a downgrade to Family Tree Maker 2008, but would be better of upgrading to Legacy Family Tree 7 instead.

Saying that stability and performance are development priorities is a roundabout way of admitting that it is as unstable as a generously filled water balloon balanced on a sharp hot needle and underperforms like a lethargic snail in a slow motion video.


Family Tree Maker 2008 is a snailtacular, unbelievably unstable, incomplete and unreliable piece of work - and you don’t even need to take my word for that. That is’s own opinion!

As I reported in The Family Tree Maker 2008 Service Pack 3 Beta Experience already, I caught Benjamin Nettesheim,’s Brand Manager for Family Tree Maker admitting that

Unfortunately there are too many defects that remain in the current product. We believe we are making significant progress to address these. Additionally, a number of people have had the understandable impression that things missing in FTM 2008 were deemed unnecessary. In truth, these features are valuable but simply have not been re-written into the product yet. Our top development priorities are stability, performance, and missing features.

This is the official word by’s own Brand Manager for Family Tree Maker, posted on the official Family Tree Maker blog, admitting that the product has too many defects, is still unfinished, and that stability and performance are development priorities. Saying that stability and performance are development priorities is a roundabout way of admitting that it is as unstable as a generously filled water balloon balanced on a sharp hot needle and underperforms like a lethargic snail in a slow motion video.

Whether this public admission about FTM 2008’s deplorable product quality has anything to do with him leaving is not known. All we know is that Russell Hannig took over as Brand Manager.

damage control

News about Family Tree Maker 2009 started with an email announcing that Family Tree Maker 2009 is a free upgrade for registered Family Tree Maker 2008 users. That seems to be a damage control measure.

Most new functionality is stuff that should have been in FTM 2008 already, and several promised features are still nothing but promised features, so it is not unreasonable to think of Family Tree Maker 2009 as Family Tree Maker Service Pack 4. The name change seems to be a marketing decision to get a new stock keeping unit (SKU) out for the impending Holiday Season.

Those who have suffered Family Tree Maker 2008 know that they deserve a lot more fixes than FTM 2008 SP3 just for having put up with this monstrosity. All those unfortunate users could have been counted on to complain loudly if did not come through with the necessary fixes, just because it decided to change the version number on them.
Besides, still has a promise to fulfil…

Family Tree Maker 2009

Family Tree Maker 2008 Second Edition

I really found it hard to not think of Family Tree Maker 2009 as Family Tree Maker 2008 Service Pack 4 or Family Tree Maker 2008 Second Edition. I do not believe a product really deserves the public attention of a full review just because some marketroid decided to change the product name or version number.
I cannot shake the impression - one that Ancestry has been giving many customers for years now - that this version change was deemed necessary merely because it was the season for a version-number change, and is not based on any particular major product change.

Family Tree Maker 2008 Service Pack 4

It sure made me wonder how different FTM 2009 is from what Family Tree Maker 2008 Service Pack 4 should be. Well, I figured out the answer to that question before I finished and posted this review (see Family Tree Maker 2009 Technology) and was not positively impressed.

If you have not read that article, just know that the little bit of research I did sure confirmed my impression that FTM 2009 is really FTM 2008 SP4 with another label slapped on it. By the time I decided to make the usual brief technology look into a separate article, I had already spent considerable time on this review, and therefore decided to finish it.


Saying that FTM 2008 prompted some initial customer dissatisfaction is like calling the American Civil War a minor provincial labour dispute.


Russell Hannig posted the Family Tree Maker 2009 product announcement on 2008 Aug 28. He immediately followed that with another post reminding owners of FTM 2008 that their upgrade to FTM 2009 is free. His remarks that There was some initial customer dissatisfaction with the functionality in that release [FTM 2008] is a promising candidate for an Understatement of the Year award.

His less than forthright remark concerning FTM 2008’s quality is the first disappointment in that post. Saying that FTM 2008 prompted some initial customer dissatisfaction is like calling the American Civil War a minor provincial labour dispute.

The second disappointment was the entire FTM 2009 upgrade debacle.’s spectacular failure to just do it turned the promised free upgrade from a damage control measure into an underwhelming and frustrating experience, the Family Tree Maker 2009 Upgrade Debacle.


The FTM 2009 announcement lists many improvements. The list presented there does not just list improvements over FTM 2008 Service Pack 3, but all improvements over the original FTM 2008 release. The announcement marks all features new to Family Tree Maker 2009 with a star, but does not compare the FTM 2009 feature set with FTM 16.

Some of the new features not in FTM 2008 that Ancestry.comc laims for FTM 2009 are

There is more, but one major and much asked-for feature, the ability to create books, is still missing. The announcement includes a second list of features; a list of features that promises to release later. That list includes Book Building. It also promises source templates (like RootsMagic 3 and Legacy 7 already offer).

The misnamed Ahnentafel reports refers to a report that is supposed to be an ahnentafel, but is not.

Book Building

Because made an unconditional promise to provide the Book Building feature to all FTM 2008 customers, it will probably have to provide all FTM 2008 customers a free and unconditional upgrade to FTM 2009, see The Family Tree Maker Book Building Promise for details.

FTM Data CDs

Another important feature that is still missing is the ability to read FTM data CDs. It is not just weird that FTM 2009 still cannot read these CDs, but quite surprising that does not include the free Family Archive Viewer on the FTM 2009 CD-ROM either.


I described the Family Tree Maker 2009 Upgrade Debacle in a separate article. This review focuses on the FTM 2009 product itself. The big question is how it compares to FTM 2008. Have its many issues been solved, and is it finally a worthy successor to FTM 16?

I received an FTM 2009 CD-ROM on 2008 Sep 30. It is not a box, and there is no printed manual, it is just a CD-ROM. There is not even a jewel case, and used to paper sleeve that contains it as the envelop. Clearly not the full product, but a upgrade package.


A CD-ROM can hold some 700 MB, the FTM 2009 CD-ROM contains some 300 MB, and that is quite a lot for a genealogy program.

The FTM 2009 CD-ROM is an AutoStart CD-ROM. The AutoStart program has a big "install now" button, but it is not the installation program. It is a Flash-based menu. There are two tabs on the upper right side, "Install" and "Extras". The Install tab has just the "install now" button, the Extras tab has links to the manual in PDF format, training tutorials and some commercial links you do not need.


The AutoStart program is a Flash program. Using Flash when a native Windows application will do does not make sense. You do not need to have Flash installed, the CD contains the necessary Flash player, but why use Flash at all?

The folks sure seem a bit Flash-obsessed. Not only is the start-up program a Flash program, the movies are in Flash Video format (*.flv) too, and that does not make sense either. Family Tree Maker is a Windows program. The movies should be a in Windows Media Player-supported format like AVI or WMV.
These Flash movies are included twice. They are included in the cabinet file the installer uses and in the movies subdirectory.

tutorial movies

The movies are installed in the Movies subdirectory of the application directory, so you do not need the CD again to watch the movies. You can view those movies using the Tutorial_PC.exe application in the application directory. If you choose Help | About you’ll see that it actually Adobe’s Flash Player 9.

The movies are each just a few minutes long .These videos aim to give a beginner a quick overview of the program, so don’t expect to learn a lot. The videos are somewhat misleading; if the only information you had were these videos, you might think that FTM 2009 performs all its actions instantaneous, and that is certainly not the case.


Microsoft .NET

Family Tree Maker 2008 and 2009 are a Microsoft .NET programs. You need to have .NET installed before you install FTM. The FTM 2009 CD-ROM includes an setup for Microsoft .NET, but it is Microsoft .NET 2.0. Version 2.0 was introduced in 2006, the current version is Microsoft .NET 3.5 SP1. The version on the CD-ROM is 2.0.50727.42, that is version 2.0 without either Service Pack 1 (version 2.0.50727.1433) or Service Pack 2 (version 2.0.50727.3053).

I recommend installing the latest version .NET from the Microsoft site, and to apply all service packs.


If you already have Microsoft..NET installed (and you probably do), the setup program goes straight for installation of FTM itself. This is a regular InstallShield set-up program. It defaults to installing FTM 2009 in %ProgramFiles%\Family Tree Maker 2009, which is unlikely to overwrite any existing FTM installation, but you can still change that to any directory you fancy.

There are no other installation options. The setup process takes a few minutes, and once it is done, it offers to start the program. The resulting program directory is about 442 MB large. The movies mentioned above take up about 70 MB, and you really don’t need to see those twice, so you could delete to save some space.

shortcut icon

The setup program creates a desktop icon, whether you like that or not. Weirdly, the icon it shows is not taken from FTM.exe, but from %SystemRoot%\Installer\{27711CB0-26B3-4D99-88A9-4E4D60C34850}\FTM.exe1_02A859E4B7134F329B688E8F60173A19.exe. That is so wrong. You should change the shortcut to take its icon from the FTM.EXE file.

Windows Vista

This review describes my experiences with FTM 2009 on Windows XP Service Pack 3, but I bought a Windows Vista machine before I finished it, and decided to try installing it. Vista’s rather annoying User Account Control (UAC) issued a warning because it is an AutoStart CD and because setup.exe tried to access my computer, but I experienced no other problems installing it. I did notice that the CD’s start-up program is a Flash 8 program, while the previously mentioned Flash player for the movies is Flash version 9.

installation tip

If you have suffered FTM 2008, you probably want to play it safe and not want switch from 2008 to 2009 until you are satisfied that FTM 2009 actually works. Just keep FTM 2008 installed and install FTM 2009 into another directory.

You can even run them side-by-side. FTM 2008 and 2009 are Microsoft .NET applications, and Microsoft .NET allows running multiple different versions of the same program at the same time.

starting up

FTM 2009 tries to connect to the Internet as it starts up, even before it shows the splash screen. I complained about this before, and it is disappointing that this reprehensible behaviour has still not been fixed.

The about box shows that this is FTM version

Family Tree Maker 16 starts in 2 or 3 seconds, Family Tree Maker 2008 needs 30 seconds and Family Tree Maker 2009 needs 2m50s - That’s close to three minutes just to start up!

painfully slow

Program start-up is still painfully slow. It took about two or three minute for FTM to actually show a screen. Many users were hoping that FTM 2009 would start faster than FTM 2008 does, but my first impression is that it is actually considerably slower than FTM 2008. Switching from one "workspace" to another takes minutes too. It is as if no one in the FTM development team cares about performance at all.

I was hoping that this dramatic performance decrease was particular to the very first start-up only, that FTM 2009 was using its first run as a kind of extended set up, and that subsequent runs would be faster. I was severely disappointed to find that every start-up was this slow.

I did some quick measurements. Family Tree Maker 16 starts in 2 or 3 seconds, Family Tree Maker 2008 needs 30 seconds and Family Tree Maker 2009 needs 2m50s - That’s close to three minutes just to start up!



Family Tree Maker 2009 looks just like Family Tree Maker 2008. When you run them side by side, it is hard to tell the difference. The one difference that is always visible in the upper left corner is the icon; the FTM 2008 icon is a stylised tree, the FTM 2009 icon is a single leaf.

If you quickly browse through the various tabs ("workspaces"), the most noticeable difference is that the Publish tabs sports more chart types.


Family Maker 2009 retains the " workspaces" introduced in FTM 2008. I did not like it in FTM 2008, and I do not like it now. The "workspaces" menu is above the normal menu. The choices you have there look like buttons, but work like tabs. When you select one such tab, you get, well, a different tab. Each tab has its own menu and window below it. The interface would be more intuitive if these "workspaces" tabs did not just act like, but also looked liked tabs.


The FTM user interface is not flexible, but rigid. Other programs let you open multiple windows and show two trees next to each other, but FTM cannot handle more than one view on one database.

Most tabs (" workspaces" ) share the same three-pane layout. That gives a sense of consistency, but it feels artificial, as if someone decided that this visual consistency is more important the actual usability of each tab.

Perhaps I just do not grasp the blinding brilliance of the fixed three-pane design, but I often like to move all extraneous stuff out of the way and maximise whatever I am working on, and the ability to do so is missing.
The tree in the people view will show up to five generations. It could easily display six if it were possible to collapse the left and right panes.


FTM will shows only one database at a time and uses fixed windows and menus to display it. That still isn’t uncommon, but it remains weird to see a new program present such a limited, inflexible and completely uncustomisable interface.

FTM has a fairly straightforward interface yet still manages to confuse a bit, because its top-level tabs are styled like buttons and referred to as "workspaces". The user interface becomes easier to work with once you recognise these button-style "workspaces" as the top-level tabs they really are.

If you think that FTM 2008 has an un-Windows feel to it, know that I agree with you. The major problem is that the tabbed "workspace" interface is modal, while Windows programs should be modeless.
Each tab is a mode in which you can do just a few things, if you want to do something else, you need to switch to another mode first. That’s not just annoying, it is against Windows user interface guidelines.

Another problem is that FTM does not have a single program menu as it should, but different menus for each "workspace" tab, and that these tab menus are modal. You cannot choose a menu item and automatically switch to the tab associated with that menu time, because that menu item you need will just not be there.

You must first switch to the right tab before you can even choose the menu item you want.
That is not just annoying - and sure to continually break your flow when you try to get some work done - it is also an implicit demand to learn this modal menu system by heart, as you have to select the right tab before you can choose the menu item you want.

The idiosyncratic interface takes up quite some screen space and is unnecessarily busy. That busy interface creates an initial impression of a feature-rich program, but makes you click more often than necessary.
The user interface would be greatly improvised already if it simply offered a single global menu like other Windows programs do.

You would expect a relatively new product to feature a modern ribbon interface, like Legacy Charting, but instead you get something That’s an awkward throwback to the days of modal user interfaces.


The modal tabbed interface seems like a poor knock-off of the ribbon interface used in Office 2007 by someone who fails understand it. In the ribbon interface, tabs along the top of the ribbon let you choose between different ribbons, much like an top level menu lets you pick different menus. The menu is ribbon-style instead of pull-down style, there isn’t a ribbon and a pull-down style menu too. The display style has changed, but there is still just one menu, and it is still available throughout the entire program. The tabs along the top of the menu let you choose between the various ribbon tabs. Those tabs apply to the ribbon only, do not change the main window and and do not imply different program modes.

The first genealogy program to use a ribbon is Legacy Charting. You can download a free trial to try what it is like. It takes some getting used to, but it is still just a menu, it does not change the main window when you select another tab and you do not need to switch modes to get to another menu. There is simply one ribbon for the entire program.


It is not just the start-up that is slow. Switching from the "Plan" tab to the "People" tab took about five minutes. I guess that is because FTM 2009 defaulted to loading the last project I loaded into FTM 2008, and that happened to be the married1200 test file I had created for Genealogical Record: Most Married. One guy marrying 1200 times is a rather extreme test case. Still, FTM 2008 was unresponsive for about three minutes, while FTM 2009 took about five minutes to switch workspace. I found that it needs those five minutes the first time of each session only, and that subsequent switches take about a second. Still, starting up and displaying this database takes about eight minutes (three for program start-up and five for the database). That is seriously pathetic performance.

GEDCOM import

To really put a program through its paces, I need some data to work with. Now that FTM 2009 had surprised with its stunningly slow start-up, I was very curious to see how it would perform the GEDCOM import test.


I started by importing the 1 MB GEDCOM. Family Tree Maker is one the many genealogy programs that annoy with their idiosyncratic GEDCOM import, but it isn’t as bad as some other programs. I quite like that it lets me pick a file and a target location before hitting the "Continue" button. That makes it easier to time the import.

not accurate

Importing the 1 MG GEDCOM took 55 seconds. Once it is done importing, Family Tree Maker displays a dialog with some statistics, including the import time. I’ve reported before (in Family Tree Maker 2008 Service Pack 3) that the time it displays there is not accurate. FTM 2009 is still less than honest about its performance. It takes 55 seconds for the dialog to appear, yet FTM claims an import time of 48 seconds.

FTM 2009 is twice as fast as FTM 2008. Well, half as slow.

import speed

FTM claims to have imported 14.346 records in 48,1 seconds, which amounts to an import speed of 298 records per seconds. It actually imported 4862 INDI records in 55 seconds, so I calculate an import speed of 88,40 INDI records per seconds.

It is disappointing to see continuing this dishonest practice of presenting incorrect import speed claims, even after it has been publicly exposed already. That said, the import time itself is a pleasant surprise. It isn’t stellar performance, but it sure is considerably faster than FTM 2008 SP3.
FTM 2009 is twice as fast as FTM 2008. Well, half as slow.

FTM 2009 creates an import log and it still reports the same ten errors as before. It is just faster than FTM 2008. Like FTM 2008, the import is just an import into the program, it does not resolve the place names against the place authority. That is something you can do after the import.


The 100k INDI GEDCOM is one of those tests that help separate the toys from the tools. FTM 2008 managed to import it, but took more than three hours to do so. FTM 2009 needs a bit more than 2 hours and a quarter. The import box disappears after 2h16m31s, the summary dialog appears some 13 seconds later, after 2h16m44s. However, at that point the import is not done yet. The People tab displays a tree, but FTM takes almost another one and half minute to display the index.

FTM’s dialog claims an import time of 2h16m26,441 seconds (8186), but the actual import time is 2h18m01s (8281s), that is more than one percent more. That same dialog box claims an import speed of 51 records per seconds, while I calculate an import speed of (100.067 INDI ÷ 8281 s =) 12,08 INDI records per seconds.

That is slow, but not as slow as FTM 2008, which manages only 8.62 INDI record per second. The improvement is slightly more than 40 % (11605 s ÷ 8281 s = 1,4014). So, for large files the improvement isn’t as dramatic as for small files, but it is still considerable.

Although an improvement of 40 % is considerable, it hardly affects FTM’s ranking. FTM 2009 is faster than Heredis 10 now, but it is still one of the very slowest programs. Despite this improvement, FTM 2009 is still solidly in the bottom layer of the geneapack, and still performs worse than TMG 7.

After a year of tweaking, FTM 2009 is still more than five times as slow as FTM 16. That is an elephantine engineering embarrassment.

What matters to most long-time FTM users is that, despite the improvement over FTM 2008, FTM 2009’s import speed is still only 18% of that FTM 16. After a year of tweaking, FTM 2009 is still more than five times as slow as FTM 16. That is an elephantine engineering embarrassment.
FTM is still far away from becoming a Confucius Cup contestant.

log file

The import log file claims 194 errors. That is less than FTM 2008 SP3 claimed for the same file. FTM 2008 labelled valid PAF-specific tags such as the _ITALIC tag as errors. FTM 2009 no longer makes that mistake.

FTM continues to live to up its reputation for inflexibility regarding dates, as all 194 ostensible errors concern dates. Some are real errors, such as "2 par 1829" instead of "2 Apr 1829", but a lot of dates that FTM complains about are merely dual dates such as "1853 / 1854", or dates that specify a range using FROM and TO or specify alternatives using OR. Although FTM is one of the oldest programs around, and has version number 18 by now, and was even completely rewritten not too long ago, FTM still doesn’t feature proper support for GEDCOM dates.

import log quality

It may be hard to find the record each reported date belongs to, but it is a reasonable import log. The log file lists the line number, the error and the (ostensibly) erroneous date. There is an import summary at the end - but without the import time and calculated speed.

The major issue I have with this log file is that FTM presents its own limitations as errors, and thus ends up claiming errors when the GEDCOM file is fine. Here is another reminder to that it is not okay to list an error when there is none. When the problem is an FTM limitation, the log file should identify it as an FTM limitation, not claim an error in the file.

missing info

I wonder why the import log does not highlight which persons or persons the (ostensibly) erroneous data is associated with. It is perhaps somewhat unreasonable to demand it in general, because whether a GEDCOM reader knows this depends on how that readers works.

Do not misunderstand my wonderment. I am not wondering why FTM uses a GEDCOM reader that isn’t aware which INDI record the date belong to, I wonder why FTM uses a GEDCOM reader that is aware of the relationship, yet still doesn’t bother to list this useful information in the log file.


Once FTM is done importing, it shows the imported data on the People tab. It also shows all date errors on in the Plan tab, on a subtab titled "Current Tree". All those errors have been added as Tasks you should resolve, complete with the GEDCOM line number and the name of the individual the dates are associated with. Obviously then, the GEDCOM import knows, but is just not listing that information in the import log.

This unimpressive performance is on a 2,7 GHz machine with 2 GB of RAM. I shudder to think how poor the performance would be on a machine with 1 GB of RAM.

GEDCOM export


FTM 2008 takes 7m48 to export the 100k database to a GEDCOM file. While the export is running, the Windows Task Manager shows FTM’s memory usage increasing to 605 MB, more than half a gigabyte!

FTM 2009 exports the database in 6m52s and its memory usage climbs to 570 MB. It is better, but still more than half a gigabyte, and nothing to be proud of. This unimpressive performance is on a 2,7 GHz machine with 2 GB of RAM. I shudder to think how poor the performance would be on a machine with 1 GB of RAM.

Once the export is done, memory usage does not return to previous levels. Even half an hour later, when .NET’s garbage collector has had ample opportunity to run, the program is still claiming all that memory for no apparent reason at all. This is unprofessional programming, plain and simple.


It is not just the exporting routine that underwhelms. The GEDCOM file it produces does not impress either. It is more than a year ago already that I remarked (in An Early Look At FTM 2008 Beta) that the ostensible GEDCOM files are not really GEDCOM files, because FTM does not even get the GEDCOM header right. Fixing that mistake should take just a few minutes, but more than a year later, it still has not been fixed.

The HEAD.SOUR.VERS field should contain the FTM version number, with a maximum length of 15 characters. Instead, it contains the string Family Tree Maker ( The full name in that field isn’t a version number, and it not needed either, as it is in HEAD.NAME field already. The string FTM 2009 writes is twice the maximum length. That’s a direct violation of the GEDCOM specification. As a result, applications that do respect the GEDCOM specification may not be able to read it.


The introductory movies mention the ability to mark up notes with bold, italics, underline, and even colour. I wondered how that would export to the GEDCOM format, which does not support this. A quick test provided the answer: all that mark-up is lost on export.


The export dialog is clean and simple. Its biggest user interface mistake is that it defaults to FTM format instead of GEDCOM. All the options, mostly some privatising options, are on one dialog.

character encoding

The big problem is that the major option that should be on that dialog is still completely absent. A GEDCOM export should offer a choice of character encodings, but FTM does not do so. FTM 2009 always exports all GEDCOM files as "ANSI" (Windows ANSI, Window code page 1252).

FTM 2008 supports neither UTF-8 nor ANSEL and FTM 2009 still does not support either UTF-8 or ANSEL.

FTM 2009 is a Microsoft .NET program. Microsoft .NET is a Unicode-based system that runs on top of Unicode-based Windows. Yet somehow, FTM 2009 doesn’t support Unicode.


FTM 2009 is a Microsoft .NET program. Microsoft .NET is a Unicode-based system that runs on top of Unicode-based Windows. Yet somehow, FTM 2009 doesn’t support Unicode.

It is not just the GEDCOM export that fails to support either UTF-8 or UTF-16. The GEDCOM import does not support UTF-8 or UTF-16 either.

It is not just GEDCOM import and export. If you copy and paste anything but Windows ANSI into FTM 2008, FTM replaces it with question marks. FTM 2008 SP3 and FTM 2009 seem to accept all character now, but that is still not good enough.

I did some quick experiments, and found that FTM 2009 now supports Unicode in notes, but not in names. If I enter 孔 夫 子 (Confucius) as a name and then later load the database again, it comes out as \u5b54 \u592b \u5b50.

Additionally recall that FTM’s GEDCOM export always uses Windows ANSI. It supports just the 256 characters in Windows ANSI, and everything else will be still turned into question marks.

vendor lock-in

Fixing the database save and restore is not good enough.

Letting you enter data that you cannot get out again, is a classic vendor lock-in scenario.
Until FTM finally supports UTF-8 export, FTM should warn users that FTM allows input of any character, but will not export those characters to GEDCOM. The absence of both the UTF-8 export and the warning about FTM’s export limitation is unacceptable.

Until FTM support UTF-8 export, your only hope of getting all data again is a third-party program that imports FTM 2008/2009 databases directly. As far as I know, no program offers direct FTM 2008 import yet.

import mangling

By the way, FTM should refuse to import ANSEL and UTF-8 GEDCOM files because it is incapable of handling them, but it does not admit to this limitation. Instead, it pretends to be capable of reading them, but bluntly imports these GEDCOM files in exactly the same way it imports Windows ANSI-encoded GEDCOM files, thus mangling your data upon import, with no way to restore it.

This is worse than TMG 7. At least TMG admits that it cannot import an UTF-8 file. Wholly Genes "solution", manually changing the header to say that the file is ANSI encoding, is very bad advice indeed, and produces the same erroneous results, but at least you’re made aware that there is a program limitation. FTM 2009 pretends everything is just hunky dory.


data file

I did not time it exactly, but I noticed that the 100K INDI GEDCOM import spent about ten minutes writing the data file. That sure made me curious about the database size. It is 172.944.384 bytes. That is largish, but hardly explains the slow database write. It is also about the same size as the database that FTM 2008 produced.


I had already noticed that FTM 2009 automatically opened my last FTM 2008 project. I wondered whether FTM 2008 would read an FTM 2009 database. When I closed FTM 2009 to try and open the database with FTM 2008, I had to sit through its auto-backup of the 170 MB file.

Many genealogy programs offer to make backup upon exit, and backing up automatically is a safe default. I agree with that default, and would not mind having to sit through it every time I exit the program if it was speedy, but FTM 2009’s backup routine is so slow that it really annoying. Luckily, you can turn the auto-backup feature off.

FTM 2008 compatibility

I do not like FTM opening the last project I was working on either. It wouldn’t matter much if FTM was blazingly fast, but when it takes five minutes to display a file containing just 1201 persons, it really matters that FTM does not try to guess what I am about to do, but leaves me, the user, in control. There’s already a list of recently opened projects on the Plan tab, and that is enough. There appears to be no way to turn this misbehaviour off.

FTM 2008 had no problem reading the FTM 2009, and, even after changing a little thing, FTM 2009 had no problem reading the FTM 2008-modified file. It sure seems that FTM 2008 and FTM 2009 use the same file format.

FTM 16 compatibility 

The FTM 2008/2009 file format is different from the FTM 16 file format. FTM 2009 will import FTM 16 files (*.FTW) and FTM 16 auto-backup files (*.FBK), as these are exactly the same format. It cannot import compressed FTM backup files (*.FBC). This is a rather silly limitation, as it would be trivial for to include the FBC decompression code in FTM 2009. It should in fact have been in FTM 2008 already, and that FTM 2009 still does not include it is surprising.

Many genealogy program compress their backups using ZIP and their backup files are merely ZIP files with another extension, but the *.FBC format is proprietary. So, to import it into FTM 2009, you need to import the backup into FTM 16, and then export the FTM 16 database into FTM 2009.

There were many problems with import from FTM 16 in FTM 2008, and most seem to have been fixed in FTM 2009, but not all.
If you have a name like Albert (Bob) Chris Daniels in FTM 16, then FTM 2009 will import the entire name, and split it so that it will sort under C, not D. That is arguably the correct way to do split the name, and you should have written Albert Chris (Bob) Daniels if you want FTM to sort the name under D. After all, the call name within parentheses signals the end of the given name. However, if you export the first example from an FTM 16 database to GEDCOM and then import that GEDCOM into FTM 2009, it will be sorted under D too. Whatever you consider to be the correct way of splitting, the two different import methods do give the same results, and either one or the other is wrong.

memory usage

After restarting FTM 2009, loading the 100k INDI GEDCOM and merely visiting all the tabs, the Task Manager showed FTM to be using 418 MB. The GEDCOM file is about 38 MB and GEDCOM is an inefficient format, yet FTM is using eleven times as much. That is excessive memory consumption.

User Interface


Browsing through the tabs, I noticed that sources do not display correctly. Several sources start with a bunch of tab characters. That is a trick I use to make them sort above other sources in PAF, and it gets sources to the top of the list in FTM too. Neither PAF nor FTM shows these tabs in the list of sources, both skip the tabs and then display the actual title. PAF does that correctly, FTM does not. After a few characters, FTM displays ellipsis and That’s it, the rest of the title is not shown.

This is a harmless but annoying defect. The mistake seems to be that FTM starts displaying at the first actual character, yet includes the tabs it skipped in its calculation of the string width, and when this erroneously calculated string exceed the width of the left pane, it displays ellipsis to show the rest doesn’t fit. For some sources FTM 2009 shows just one or two characters, and for others it even show nothing but the ellipsis.

The defect does not harm the actual data, but does affect the usability of the source pane.

Surely I am not the only one using this sources sorting trick? How come this defect, that was present in FTM 2008 already, is still present in FTM 2009? Oh well, it is a problem for just a few sources, and the defect is easily avoided by removing the tabs (you just lose the desired sorting).

Working with sources in FTM is just not practical. When you try to work with sources, you constantly find yourself looking at an hourglass cursor instead.

lethargic response

The bigger problem is FTM’s lethargic response. When you click a source it easily takes five or ten second for the source citations window to update. Meanwhile, you can do nothing. When you click on a source citation, the display flickers while the update of the detail window takes another second. Working with sources in FTM is just not practical. When you try to work with sources, you constantly find yourself looking at an hourglass cursor instead.

Working with sources is not the only thing that is slow. Starting up takes minutes. Fifteen seconds have gone by already when it finally displays the splash screen with a busy bar. It takes a minute and forty seconds to display the Task Pane of the Plan tab. Switching between tabs takes one or two seconds. The list of thing you have to wait for goes on and on.

A few things perform well. When I select a person from the index pane, the tree pane updates almost instantly. When I click a place in the places pane, the map pane shows a map in about a second.


The map displays pretty fast, and the place recognition seems reasonably smart. It seems to shows the highest level it recognises; if there is spelling mistake in the town name, it will still show the province. It even seems to ignore misspelling of a province as long as the city and country are spelled correctly; It will show Chicago, Cook, Illinois, United States of America even if Cook is missing and Illinois is spelled wrong.

One limitation of the mapping function is that it does not know about historic place names. One problem I encountered is that it will not highlight the place but just the municipality if you include both. It will only highlight the place if you remove the municipality, so I am not really enthusiastic about this feature yet.

Resolve Place Name

When you’ve just loaded a file, the place index will show question marks in front of each name. Only when you resolve a place name does that question mark disappear. I randomly picked "Fremont, Newaygo, Michigan, United States of America" as a place to resolve. That’s a perfectly specified name, so there is nothing to resolve really. Family Tree Maker displays a dialog box that lets me pick between a bunch of places called Fremont, including several in Michigan, but only one in Newaygo. It defaults to that one. That sounds nearly perfect. The only thing they could do to improve this interface is use some colours to highlight what matches and what doesn’t.

A serious problem with FTM’s place resolving is that it involves choosing the "Replace" button, which changes a perfectly specified place name into one that abbreviates the country name. I am sorry to have to point that out, but that is plain wrong. It may make sense to use "USA" in reports, instead of spelling it out all the time, but place resolution should always use the full, correct name for every part of the place name.

Oh well, it is just a copy of an old database. That old database contains plenty of typos and even several blunders such as the wrong province for a place name. Overall, the place recognition seems pretty good, and the place resolution dialog is a simple one that keeps the user in control, but you just don’t want to resolve 3993 places by hand.

…an actual resolution performance of less than one place name per second.

Resolve All Place Names

Family Tree Maker has a "Resolve All Place Names..." menu item. When you choose it, FTM suggests that you make a back up first. The "Resolve All Place Names" dialog initially displayed place names and a progress bar, but soon became completely white and stayed that way until resolution of place names was finished. Resolving 3993 place names took 14m29s (869s). That is rather slow, it is just 4½ place name per second, and that is without taking into account the place names it did not resolve. Browsing the place index, I estimate the number of resolved places to be perhaps 20 %, and that would imply an actual resolution performance of less than one place name per second. That is both poor resolution and slow performance.

Ambt Doetinchem

Once the resolving process is done, the dialog shows a list of the places that have not been resolved, along with possible matches. On the top of my list is the suggestion of "Doetinchem" for "Ambt Doetinchem". The match it suggests happens to be correct, but resolving it the FTM way, by actually replacing "Ambt Doetinchem" with "Doetinchem" is incorrect, "Ambt Doetinchem" is the historically correct name and using "Doetinchem" instead is wrong. It is not just a mere name change, but a loss of information. Back then, there was both "Ambt Doetinchem" and "Stad Doetinchem".


FTM does not recognise "Emlichheim, Niedersachen, Germany" or "Emlichheim, Bentheim, Niedersachen, Germany". I discover that it wants "Emlichheim, Graftschap Bentheim, Niedersachen, Germany". Never mind whether it should include "Graftschap". What really upsets me is that its suggestion is not "Emlichheim, Graftschap Bentheim, Niedersachen, Germany" at all, but just "Germany"; if you were to blindly accept the suggested match, you would lose "Emlichheim" and "Niedersachsen" entirely - you would lose valuable information!

This is not some minor spelling mistake for some small hamlet no one ever heard about,… It is like replacing "New York" with "New-york".

The so-called place authority also wants to replace the place name "’s-Gravenhage" (notice the single quote and lower case S the place name starts with) with "S-Gravenhage" (no single quote, and an uppercase S). The name in’s database is wrong, and you can only "resolve" the place name by accepting’s misspelling. This is not some minor spelling mistake for some small hamlet no one ever heard about, ’s-Gravenhage is the official name of the town also referred to as Den Haag and internationally known as The Hague. It is like replacing "New York" with "New-york". Actually, it is worse, as The Hague is a seat of government, and New York is not. It really is like replacing "Washington, D.C." with "washington DC". This is a serious blunder that reflects a lack of testing.


I noticed many Frisian place names in the dialog box. Spelling of Frisian names is an issue, most have both a Dutch and Frisian spelling. I generally use the Dutch spelling, simply because that is the official one. I have used Frisian spelling occasionally ("Ferwert" instead of "Ferwerd"), and I am additionally sure to have a few spelling mistakes, but these two issues do not seem enough to explain the many Frisian place names in the dialog box. It is almost as if’s database doesn’t include the Friesland province at all. I sure did not misspell Baarderadeel, yet FTM does not recognise "Baarderadeel, Friesland, Netherlands" as anything but "Friesland, Netherlands". Perhaps the seemingly poor recognition of Frisian names is related to FTM poor support for municipalities.

Web Search

The Web Search tab does not work until you register Family Tree Maker 2009, and wants you to log in every time you use the search tab, even if you just want to search RootsWeb or Google. That’s ridiculous. has no need to track my usage of this feature.


spell check

The spell check supports multiple languages and locales. It is a disappointing that a product sold internationally defaults to Amglish instead of English, but the bigger problem with the spell check was revealed in the article on FTM technology; all the spelling databases are several years old.

consistency check

I consider the consistency check one of the most useful features of genealogy programs, and like to see what it makes of my test files and my own database. Alas, while Legacy 7 surprised with the most extensive consistency checks yet, FTM 2009 does not provide consistency checks at all.

chart types

FTM 2009 adds a few chart types not found in FTM 2008. Family Tree Maker now includes a hourglass chart, a bowtie chart (hourglass chart on its side) and my favourite, the fan chart. The unimaginatively titled Family Tree Chart is a stylistic variation on the fan chart, in which all the boxes are horizontal instead of rotated.

I love fan charts because they provide a good overview of your ancestral research. Ever since Millennia released Legacy Charting, which loads databases in mere seconds, I’ve been using it frequently to quickly load a database and get a quick visual overview.

I tried creating a fan chart in FTM, and was struck by how long it took to get see it - for a file with just one person in it, it took about ten seconds. It only took that long the first time I asked for it, the second time it was less than a second, but I still don’t like having to wait for something as simple as a chart for a practically empty database. The chart includes the text "name: " in empty boxes, and there seems to be no way to get rid of either that text or the boxes, so I am particularly underwhelmed by this. And, pardon me for saying this, but the FTM fan chart is plain ugly.

This is such an obvious design blunder, and so apparent in charts of five and six generations already, that I dare say that the fan chart has been rushed out of the door without any testing to speak of.

fan chart

For a brief while I liked that there seemed to be no practical limit to the number of generations that FTM’s fan chart can handle - when I asked for 25 generations, it cheerfully accepted the request - but it soon became apparent that FTM’s cheery acceptance of the request was not founded on its capability to handle it, but on the programmer’s failure to imagine or explore the limitations of the code.

Some fifteen minutes later, the Windows Task Manager showed FTM to be using more than a gigabyte of RAM - for an essentially empty database!
Meanwhile, FTM 2009 continued to show one of those indeterminate progress bars that give no hint of the actual progress whatsoever. The Windows user interface guidelines are clear that a program should show a determinate progress bar if possible. After almost eighteen minutes, the program failed and just disappeared from the screen. I had to suffer FTM’s painfully slow start-up again to continue testing.

Legacy Charting limits fan charts to 12 generations. I’d like to see 14 or 16 generations, but at least Millennia’s programmers bothered to know the limitations of their code and Legacy Charting does not let you specify a size it cannot handle.

I tried moving up to 12 generations in FTM, one generations at a time, but FTM was clearly struggling to do eight, and nine generations was quite obviously already more than the code could handle comfortably. Fairly low generation counts even revealed a serious design error. With higher counts the lines between the boxes became a larger, and the boxes became smaller, practically ensuring that you will not be able to read the names in those boxes, not even when the chart is printed professionally. This is such an obvious design blunder, and so apparent in charts of just five and six generations already, that I dare say that the fan chart has been rushed out of the door without any testing to speak of.

search and replace

One PAF feature I find very convenient is its global search and replace. FTM 2009 has this feature, but it unusably slow. PAF takes seconds, FTM 2009 needs many minutes. When I tried to stop a replace operation after a few minutes, FTM 2009 apparently hung, and I decided to kill and restart it.



FTM 2009 is built with a lot of third-party components. The separate Family Tree Maker 2009 Technology. article discusses the technologies used to build FTM 2009.

The quick summary is that FTM 2009 is still using the same components, even the same versions, as FTM 2008. In plain English: FTM 2009 really is FTM 2008 Service Pack 4.

Mentioned above, but worth repeating here is that FTM 2009 still does not support Unicode, UTF-8 or ANSEL, and that it does not read *.FBC compressed backup files or FTM data CDs.

The FTM 2009 CD-ROM contains a setup for .NET, but it is an old version, and I recommend installing the latest version and updates.

There apparently is a plug-in interface for FTM 2008/2009, but information on that interface does not seem to be available. Even information on how to get that information does not seem to be available. I gave Mike Ward, the PR manager for plenty of time to answer the question, as I asked it several months ago, but he apparently does not consider FTM worthy of third-party developer attention. I do not know why not, but I do know that not enthusiastically providing information on developing plug-ins and promoting the development of such plug-ins invites speculation that does not to want stimulate third-party support right now because it is already planning to kill the FTM product entirely.


user interface

FTM’s user interface seems designed by someone who thinks that user interface design is about being different. Although built with modern technology, including handy user interface components, the user interface does not feel modern, but distinctly un-Windows.

The FTM 2009 interface is needlessly modal. That FTM 2008 introduced this interface was bad. That FTM 2009 still sports the same broken interface is worse.

FTM’s usability is further impaired by serious performance issues.


Although import performance has improved since FTM 2008, it is still embarrassingly slow compared to FTM 16. Its start-up time is long enough to make a cup of coffee, but the biggest issue is that the user interface contains so many delays that it simply isn’t a truly interactive program anymore. This is a stop and wait program, so much so that it is not just annoying, but close to unusable.

FTM 2009’s memory usage is not modest. If your system does not have plenty of RAM for this inefficient program to work in, you will experience even worse delays than I did.

GEDCOM support

FTM 2009 still isn’t honest about its own import performance, the GEDCOM import still lacks proper date support. The import log is not as informative as it could be, and still presents FTM limitation as ostensible errors in your GEDCOM file.

The GEDCOM export is extremely RAM hungry, and the FTM does not seem to release the RAM once the export is done. The GEDCOM export does not support either ANSEL or UTF-8. FTM writes ANSI encoded GEDCOM files only, despite the fact that the GEDCOM standard does not allow that encoding. FTM 2009 still does not write correct GEDCOM headers.

The GEDCOM import does not support ANSEL or UTF-8 either, but FTM does not admit its limitation. Instead, FTM pretends to be able to read the file, but actually imports it as if it was encoded in Windows ANSI, thus mangling your data in the process, without even a warning!

Because FTM 2009 mangled the ANSEL-encoded 100k INDI GEDCOM on import, I consider that import to have failed. That is what the strike-though in the table below means. I wasn’t that strict when I did the FTM 2008 review, but I am that strict now. The timing for Family Tree Maker are not stricken through, as Family Tree Maker 16 does support ANSEL.


I commented upon this in the FTM 2008 review already, but it remains a weird accomplishment that FTM 2009, although a Microsoft .NET program, still does not support Unicode. Pervasive Unicode support is one of the major features of Microsoft. NET, and the built-in Unicode support is probably one of the reasons picked this platform. Yet somehow the FTM programmers managed to create a program that lacks Unicode support. FTM 2009 is slightly better than FTM 2008, you can use Unicode in notes, but it still limits its export to the 256 characters in the Windows ANSI character set.


Family Tree Maker 2009 is not as unstable as Family Tree Maker 2008. The various service packs have improved stability, but it is still a slow and memory hungry application.


FTM 2009 features some new chart types, but these seem to have been rushed out of the door. FTM accepts fan chart sizes it cannot handle. The code is slow, memory hungry and eventually crashes. Moreover, the fan chart is not just ugly, but almost unusable; when you ask for six generations or more, the lines between the boxes take up more space then boxes that have to contain the name and other details.

Display of sources is still defective. FTM still lacks basic features such as a consistency check. That the program defaults to Amglish instead of English is a minor detail, but that the spelling databases are several years old is not. It has an auto-backup feature, but it is so annoyingly slow that you want to turn it off. A similar issue is that it automatically opens your last open project when you start FTM, even if you explicitly closed that project before exiting.

place name authority

The place mapping support is not what it should be. FTM has a so-called place authority, but the database is not up to date and the place resolution code still gets confused by a municipality between a place name and a province. The place name resolution process is slow, and the suggested matches often lose important information. It wants to abbreviate country names, and even suggests changing good place names into bad ones. This so-called place name authority is not useless, but you should approach this wannabe authority with a healthy dose of scepticism.

FTM 2009 is still so poor, shows so little real improvement over FTM 2008, that it is hard to hold any hope for the future.


This successor to FTM 2008 still isn’t feature complete. It still lacks the Book Building feature and data CD reading capabilities of FTM 16. FTM Classic compatibility still needs work and its GEDCOM capabilities are still poor. The state of more than one feature suggests a lack of design and a lack of testing. It is also worrisome to note that several defects reported for FTM 2008 have still not been fixed a year later.


If you consider features only, then FTM 2008 was FTM 16 after a featurerectomy. If you take compatibility, stability, and performance into account, the FTM 2008 release was nothing short of softicide.

FTM 2009 is hardly different from FTM 2008. Not only do FTM 2008 and FTM 2009 apparently use exactly the same the database format, FTM 2009 even feels like FTM 2008 Service Pack 4 with another label slapped on it. A look at the technology inside FTM 2009 confirmed that impression.

defective downgrade

To call FTM 2008 a defective downgrade is kind judgement. FTM 2009 is slightly better; it is less defective and not as unstable as FTM 2008, but it is still slow, feature-incomplete, memory hungry, and awkward to use. FTM 2008 is bad and FTM 2009 is not much better. It is still an utterly unrecommendable downgrade from FM 16. Its overall quality is still plain unacceptable.

I do not recommend holding out for improvements, but switching away from FTM to either Legacy or RootsMagic. FTM 2009 is still so poor, shows so little real improvement over FTM 2008, that it is hard to hold any hope for the future.

bonus tip: FTM 16 still available

I do not recommend FTM 16, but it is still way better than either FTM 2008 or FTM 2009, and you can still buy it if you want to. is not going out of way to make sure you know this, but they are still selling FTM 16. You can in fact still buy FTM 16 through the store. has discounted its price from US$ 39,95 (about € 28,00) to US$ 29,95 (about € 21,00).

The Who do you think are software package is Family Tree Maker 16 UK Edition and includes Who you think you are video clips and celebrity trees of the first two series. Several webstores sell it for about £ 12,00 (about € 15,00).


2009-01-28 special note to Wikipedia editors

Thank you for linking to this review. The hard link you used will break one day. Please use the permalink.

2009-05-01 Family Tree Maker 2009 Service Pack 1 has released Service Pack 1, see Family Tree Maker 2009 Service Pack 1 Available.

2010-04-24 ahnentafel

Added a few words about the abusage of ahnentafel.

import speed

Family Tree Maker 16 (16.0.350)

time in seconds361509
INDI per second135,0666,31
bytes per second29.330,4225.711,99

Family Tree Maker 2008 SP3 (

time in seconds112 11.605
INDI per second43,418,62
bytes per second9.427,63 3.343,33

Family Tree Maker 2009 (

time in seconds558.281
INDI per second88,4012,08
bytes per second19.198,094.685,35

product details

productFamily Tree Maker
websiteFamily Tree Maker
priceUS$ 39,95
requirementWindows XP or Vista, with .NET
noterequires CD-ROM drive to install.
VerdictDowngrade from FTM 16


Family Tree Maker 16

Family Tree Maker 2008

Family Tree Maker 2009

Ancestry blog

also mentioned