Ancestry.com takes another step away from its genealogical roots…

Ancestry.com takes another step away from its genealogical roots…

We could see it coming…back in March of 2017, one of our first blog posts was about Ancestry.com’s new tool “We’re Related” (We’re Related app is a lot less frivolous than it first appears). It was a bit of a “hot take” about how it was less silly than it seemed and how it could be very powerful if it’s expanded to a tool that is predictive of your matches.

We’re Related is making suppositions based (apparently) on an algorithm that can draw the line between what you know, and what it guesses is true, to build a potential line for you. If this technology is ever leveraged against some of my brick walls instead a gimmick like linking me to Blake Shelton, Ancestry might really be on to something.

Before we take any victory laps…we have to admit, we were incredibly naive. We never guessed that Ancestry would take this powerful technology and use it to take it’s worst, most frustrating feature, and make it much more dangerous.

The new feature is the “Potential Father/Mother” suggestion, and I’m going to let Carolynn ni Lochlainn detail all the challenges of this new tool, and the risks, in her SPOT ON “From Paper to People” Podcast #27 (From Paper to People: What I Hate About New). Please listen, but her upshot is that this feature is an easy way for those new to genealogy to quickly build out their trees, and the tool forces you to create the ancestor without any sources attached.

One of the biggest drawbacks of Ancestry is the Public Trees that are so often inaccurate, and are often built solely on other people’s unsourced trees. Now, it’s a certainty that these trees are going to start to mushroom, and by design have NO citations attached to the new ancestor.

Ever wonder why Ancestry has delivered even more accurate admixture and even prettier graphs, but none of the tools needed to do serious genealogical research? It’s because there’s no additional revenue from genealogical tools.

The good news is that we as serious users can avoid the downfalls, and use the predictive part of this feature to do the research for us, but we must immediately attach the citations to any newly added ancestor. We, as a community, can also make sure we NEVER use a Member Tree to support a fact. You can link the Member Tree ancestor to yours, but make sure all facts are unselected before you link them. They will see your additional work, and you them, but you will not perpetuate their unsourced facts.

This slideshow requires JavaScript.

But, Ancestry.com isn’t packed full of serious hobbyists/professionals and “Potential Parent” is going to take the problem of Member Trees and make it explode it beyond what we could have imagined. At some point, the tree feature in Ancestry is going to be unusable. Ancestry.com will continue to be a great source of primary research, but it will be nothing more than a data repository for those of us who are serious about this work.

And, back to our naivety…the most frustrating thing is that we should have known better. Again, going back to our vaults, we saw right away that AncestryDNA is here to support genealogy ONLY because it’s a good way to gather DNA tests (Dancing with the Devil: The Tradeoffs of Modern Genealogical Research). Once Ancestry realized that pushing pretty graphs and “ethnicity” was the best way to sell more tests, they pivoted and met their true goal with these tests: the largest DNA database that will generate a tremendous amount of revenue from drug companies, etc. who can leverage your tests to understand how their drugs might work. Ancestry now (or soon will) make more money from monetizing your DNA than it does from supporting our genealogical work.

Screen Shot 2018-09-16 at 9.28.32 AM
How did the public records “Reclaim the Records” paid to get show up here, for paid members only?

Ever wonder why Ancestry has delivered even more accurate admixture and even prettier graphs, but none of the tools needed to do serious genealogical research? It’s because there’s no additional revenue from genealogical tools, but putting more effort into the graphs will drive more people to test, which will grow the database, and grow the revenue stream.

As a community we have to get ready to accept that Ancestry is not a partner in our work, and is not in business to support us or our needs. They exist to generate revenue, and as long as that interest and ours intersect, we’re good, but as they make more money from other streams they are going to sacrifice our needs to focus on revenue. You’re already seeing that with things like “Potential Parents”, more admixture, and their new collections consisting of public records gathered at great expense by groups like Reclaim The Records and putting them behind the paywall.

The genealogy features of Ancestry are still there, for now, but the bad Member Trees we suffer through today are likely going to be remembered as the golden age of online genealogy research.

 

There’s no other way to put it – MacKiev’s Family Tree Maker is garbage

There’s no other way to put it – MacKiev’s Family Tree Maker is garbage

One quick point, as always, we receive no financial benefit or consideration for any product or service we review/recommend/discuss here. Everything we discuss is our opinion alone, and we talk about it because we use it.

It’s been a year since MacKiev’s Family Tree Maker dominated this blog, as they struggled to deliver their first version since they took over from Ancestry, and this is what we’ve learned: MacKiev’s Family Tree Maker is garbage.

It’s strong language, but their product continues to be literally dangerous to your data. FamilySync has been a COMPLETE disaster since the moment it was promised (then delivered 4 months late), with MacKiev choosing a synchronization strategy that creates data corruption. Companies like RootMagic have delivered sync without issue, and without risk. But nearly a year since MacKiev finally delivered their FamilySync, it’s still buggy and unreliable. It’s dangerous to your work. It’s garbage.

We’re starting to think that it would have been better to let Ancestry kill the product off. It would have been a hard year, but at least we wouldn’t have wasted that year hoping that MacKiev could actually create/support software.

First off, let me say that I have a LOT of experience working with software delivery…with both commercial products and deploying/supporting in-house developed software. I’ve been doing it for 25 years. And if this was product was deployed in the large corporate environment I currently manage for a Fortune 50 company, we’d pull it out. At all costs. We’d never support this horrible effort, with so little partnership from the vendor.

And, for you loyal readers, you might be asking why we’re using FTM anyways. Didn’t we give away free copies of RootMagic to readers who’d paid for the FTM upgrade last August when MacKiev couldn’t get their act together? We did. And we still use/love RootsMagic…but…

Screen Shot 2018-08-31 at 10.23.36 PM
“Thank you for stopping your work and spending an hour backing up your data, but our TreeSync is so fragile it’s best you put off more research for another 2 hours or so because we don’t know how to design state-aware data synchronization”

The way RM manages citations just isn’t workable for how we support facts in our trees. For example, we created a custom citation in RootsMagic for the 1900 US Census for Roman and Mary Jones and copied the source to each of the 3 Roman’s facts supported by the citation (Name, Birth, and Residence). When we run a TreeSync with Ancestry, everything went off as expected, but it created 3 separate copies of the same citations…one for each fact. Additionally, there’s no central place to manage/edit/view all sources for a tree, which makes it VERY hard to update citations, etc.

So, we still use RootsMagic for our 60+ speculative trees, but we went back to using FTM for our main, public tree.

Since going back to Family Tree Maker, it’s been one disaster after another. First, there was the months of “Orange” sync status late in 2017 (which we missed, luckily, because we’d kicked them to the curb). About 6 months ago we had to start over and re-download the tree from Ancestry, which destroyed all of our source citations. After two months of work, the database corrupted, and we started a cycle of restoring databases, and getting about a month’s use of Family Tree Maker, and then hitting corruption. We have to restore, and repeat the process.

It really seems that this corruption is happening during FamilySync, and if that’s completely inexcusable. Their sync process HAS to be robust enough to not commit records until the system has no risk of corruption. I’ve worked with data replication since 1997 and every tool has a non-destructive method of committing data, and backing out changes if there’s failure/corruption. Plus, RootsMagic has figured this all out…we sync constantly, and there’s never red/orange/green OR corruption. Just repeated success.

The issue is just with MacKiev.

We’ve figured out how to mitigate the risk of FTM corrupting our data by doing constant syncs (change a record, sync, change another record, sync, etc.) and by reviewing the sync reports each time. For example, the last time we had sync stop working, we noticed that the marriage record for Felice’s side of the family was causing changes in the birth record attached to Rick’s grandmother. By deleting both facts we were able to sync successfully and then re-add the facts, and not have to resort to a restore…but I’m only comfortable doing this because I have 20+ years working with/troubleshooting database issues.

Screen Shot 2018-09-01 at 2.10.45 PM
Five easy steps to have your data corrupted anyways!!

But the real nail in the coffin is how MacKiev has chosen to deal with their corruption issues. Instead of architecting a proper deployment, or fixing their code, or building in better error trapping, they turn it back on the users of their product to protect themselves from Family Tree Maker’s failures. They are insisting that you add the following steps to every FamilySync:

  • Backup your database (32 minutes for our 3700 person tree)
  • Compress your database (6 min.)
  • Wait for Green conditions (0-240 min.)

Recently, we did a day’s worth of tree building and citing/sourcing (6 hours) and then we had to stop our work and take nearly an hour to sync. Then, we waited for 2 hours for MacKiev’s FamilySync  to go back to Green. Then, the database corrupted anyways, and we lost all 6 hours of work and had to repeat it.

Screen Shot 2018-08-17 at 9.40.26 AM
When you see this, you’re screwed…they corrupted your database and you lost everything you did since your last backup!!

If we back up every hour, we risk less data, but we cut our productivity in half (work an hour, back up for an hour)…assuming we don’t have to wait for Green.

In the mean time, I switch over the RootsMagic and work on some speculative side project, with regular FamilySync’s, while FTM is dead in the water waiting to “Go Green”.

As much as I have invested in Family Tree Maker, and as much as it is the tool that really meets my reporting needs the best, We’re starting to think that it would have been better to let Ancestry kill the product off. It would have been a hard year, but at least we wouldn’t have wasted that year hoping that MacKiev could actually create/support software.

TEST DRIVE: My 48 hours with the new Family Tree Maker 2017

TEST DRIVE: My 48 hours with the new Family Tree Maker 2017

I finally was notified Friday night of my eligibility participate in MacKiev’s 48-hour beta test of FTM2017. While I have been, and will continue to be, critical of the process (Mackiev’s latest update engenders even less confidence, puts 2017 release 3 weeks behind with no firm date for release) I was largely happy with the new software, and can’t wait for it to go live.

Listen to their advice: Use a practice/non-critical tree for your testing.

I have over 20 years experience testing software releases (MacKiev 2 weeks late with Family Tree Maker 2017 release, still “getting close”), so I’ve been through this before, however for this beta test I didn’t do an exhaustive breakdown of every feature or even attempt to use the new features. I just kept it simple, focused on how to sync with my existing FTM 3 trees that were linked to Ancestry.com, and went through a few generations of new ancestors to a speculative tree I’d chosen to test with.

That brings me to my first impression. Listen to their advice: Use a practice/non-critical tree for your testing. My concern wasn’t about data loss, and I don’t think there’s a reason to be concerned about that, but since my larger trees have multiple owners/editors/viewers, if I had to re-upload them Screen Shot 2017-04-30 at 3.04.24 PM.pngand reassign those it would be difficult. I feel like it was good I was concerned about that as it relates to a test, and like it will be less of a concern when we’re not limited to a short beta.

Other than that, I feel that the interface was easy to understand, and as a long-time user of Family Tree Maker there were no surprises. Due to a confidentiality agreement I agreed to I won’t go into detail about look/feel/placement of things in the application, but I think it’s safe to say I felt like it was not much of a learning curve going from FTM 3 to FTM 2017. My impression was that overall, FTM 2017 felt more modern, updated and refreshed.

My impression is that this is a mature, (nearly?) ready for production release software package that will be a welcome refresh for FTM users. I have some complaints, but since I can’t yet discuss features or how/if various features have changed, I can’t go into them until Family Tree Maker is released to the public. Generally I’ll say that given MacKiev’s spotty rollout of this product, and some of the complaints I can’t yet detail, Family Tree Maker 2017 is likely to keep me satisfied in the short-to-medium term while I start to research alternatives just in-case this is as good as it gets.

Product Review: Speechpad.com transcription services

Screen Shot 2017-04-04 at 9.18.04 AM

Note: I receive no financial benefit for reviewing/endorsing any of the products/services on this site. All reviews are based on my experiences and may not apply the same for everyone.

One of the first pieces of advice I received when I started tracking my family history was: stop chasing the pieces of paper, get a tape recorder, and sit down with your family and start recording their recollections of your family’s history. (Thanks Tony Burroughs! Black Roots by Tony Burroughs)

As I talked about in an earlier post (read: How to: Getting started researching your family tree), I bought a Sony digital recorder and I’ve used it to record many conversations. At least two of the people I’ve interviewed are no longer with us, so I have some of the only formal oral history from them on-record. But, if you’ve followed the advice and conducted these recordings, what do you do with them? Transcribing them by hand isn’t practical, and ultimately you won’t get around to it…trust me, I type an accurate 80 wpm and I never could do it for more than a few minutes.

Speechpad.com was super easy and straightforward to use, and the output was exactly what I hoped it would be.

I knew I wanted an online transcription service, and since I’d converted my interviews to .mp3 format I figured it would easy to upload and convert. What I didn’t know was how accurate/useful the output would be. As it turns out Speechpad.com was super easy and straightforward to use, and the output was exactly what I hoped it would be.

Process

I visited the site and setup an account, which was all straightforward. When you login you’re taken to the “Upload” screen, which again was pretty straightforward.

Screen Shot 2017-03-28 at 9.35.14 AM

Screen Shot 2017-03-28 at 9.33.04 AM

Once the .mp3 file(s) are uploaded, you’re given the option to select how quickly you want the transcript. The prices go up with the urgency, as you’d expect. When I’ve used the service I’ve always selected the “1 Week” option, and it’s never taken more than 2-3 days to get completed.

Screen Shot 2017-03-28 at 9.34.29 AM

You can add a “verbatim” option for $.25/min. that will display all of the stutters, repeated words, filler words, etc. that the standard transcript will remove. Looking at this option (Verbatim Option) you can get a great picture of what the final results will look like. I’ve never used this option since the standard transcript has always worked for my needs.

Screen Shot 2017-04-04 at 10.31.59 AM

Once the transcript is complete you’ll receive an email, and you can download it in several formats (text/RTF/HTML/Word).

Pricing

Basic transcripts (all I’ve ever ordered) start at about $55/hour, with options such as verbatim or rush delivery available as add-ons.

Results

When you download your transcript, you’ll get a file that looks like this:

Screen Shot 2017-04-04 at 10.39.07 AM

What I’ve found is that the transcript is very accurate, and what’s missing/inaccurate is largely spellings of family names or when someone mumbles. To transform the transcript into a usable form, I will save a copy with “Edited” appended to the file, and listen to the recording while I clean up the file. I often have the kids with me for interviews, so there will be several conversations going on or questions from a 5 year old that get picked up by the microphone. I’ll take out irrelevant portions, correct spellings, add in what I hear from the parts flagged [inaudable], and change the font while also adding line numbers.

 

Screen Shot 2017-04-04 at 10.46.41 AM

I also add an introduction explaining the provenance of the transcript, so that years from now if this document falls in the hands of a researcher, they know some of the history behind it. Finally, I’ll copy my edited version of the transcript with “Public” appended, and remove any private or sensitive information (addresses of living relatives, etc.). I convert that Public file to .pdf, and now I can cite this interview by page and line number, and attach it to my public trees!

My first upload was converted to .mp3 incorrectly, and I had a long set of discussions with the Speechpad.com Customer Support team trying to resolve the issue. Even though it was my fault, despite me arguing that it wasn’t, they were completely helpful and went above and beyond to make me happy. They even refunded the original transcription fee and discounted my resubmission…despite it all being my fault! The number of emails they sent until I was fully satisfied was above-and-beyond, and quite impressive.

All in all this is a great service, a reasonable price, easy to use, great output, and fantastic customer service. I’ve found the last transcription service I’ll ever use!