The Gene Genie Blog

Advertising - the elephant in the room

Ryan O'Neill — Wed, 10 May 2017 21:58:18 GMT

Advertisers have overstepped the mark, like a pushy person who has stepped too close, they have invaded our personal space. Just like that person, they are also in your face and they don't seem to care that they've made you uncomfortable. Rather than moderate their behaviour when people are upset with them, they'll just go undercover, stalking you and ducking into cover when you turn around.

This is the state of the web in 2017 and it's one of the reasons GeneGenie will not be accepting advertising or using any advertising networks. To illustrate my point, I'd like to walk through some of the issues with one of the businesses I use personally.

Like many family historians I have an account with the behemoth that is Ancestry.com. Considering the huge amount of documents they hold I'd say the service is fairly priced and the best out there.

The problem with Ancestry is that their site is infested with tracking code that;

Slows the site down.
Sends my browsing habits to advertising networks.
Breaks the site so it is unusable.
Is a potential security risk.

Ancestry charge me a fairly hefty subscription but that doesn't seem to be enough for them, they also seem to be selling my browsing habits to advertising networks (I'm pretty certain they are not doing this for free). There is no way to opt out of this tracking code and it's on every page.

To see what I mean, take a look at the screenshot below which was taken when the Ancestry home page loaded. All of the pink / red text lines are blocked network requests to advertising networks (and some non advertising site related scripts). I can see OutBrain, Google, Twitter and TVSquared among many others. Outbrain are the outfit that monetise web pages with those nasty click bait titles similar to 'You'll never guess what happened next, number 3 will SHOCK you').

If you want to see this data for yourself, follow the steps at the bottom of this post.

So every time you load a page, all of those businesses (and more) get to know what you are looking at and for how long. Unless of course you use an ad-blocker. In that case, the site just won't work in places. For example it's not possible to access your messages in the site whilst the ad-blocker is on because it can't distinguish the adverts from the real code that is used to operate the web site.

So it seems even huge businesses like Ancestry need advertising to survive, even when you've paid for the service in advance. It might be simple greed though. If the money from selling your customer data makes the difference between being a profitable business or going under, perhaps your business model is wrong.

An open and transparent business

Which brings me to the point of this post. It's ludicrous that we need to say 'we don't sell your data'. It should be the reverse, if a business sells your data it should be in plain language and made obvious, not hidden in a 3000 word licence agreement.

To be completely clear about this; We (GeneGenie.com) promise to treat you (our customers) fairly and honestly. We will not share your data with anyone and will only ever use the data we have to troubleshoot issues with our systems. We'll encrypt any sensitive data and are happy for external reviews to ensure compliance.

Seeing for yourself

I apologise, but these steps are necessarily technical. If it's too confusing, commandeer a nearby child.

Start Google Chrome and install an ad-blocker (I recommend UBlock Origin AND Adblock Plus. Now press F12, this will show the developer tools where you'll be able to see what's going on. Click the Network link near the middle top and then type Ancestry.com in to the address bar.

You should now see a list of failed requests in the debug console.

New design, new developers...

Ryan O'Neill — Fri, 07 Oct 2016 19:25:23 GMT

Since the last blog post we've had quite a push on software development, open sourced the work we are doing, started the new website, chosen a winner for our logo design, filed trademarks, had people trying to defraud us and got very deep into the weird world of genealogical date formats.

The logo winner

We chose this, hope you like it. You can see the finalists for competition logo design here. Something we learned, if you ever file a trademark, watch out for the scammers (see below).

Open sourcing the family tree importer

Before we can show a family tree on the website, we need to be able to write software capable of reading family tree files.

We started off by basing our efforts on an old open source project that had been gathering dust on the web. Although this meant that we got a great start, we also had a few bugs and some missing functionality that we really thought was needed.

We are open sourcing (sharing) this code with everyone (customers, the public and competitors) as we will with all of the work we do.

During October 2016 we've had a lot of help from open source from contributors and are expecting to be able to use the software to import family trees onto a test web site in early November.

If you are of a nerdy / geeky persuasion or are just plain curious, you can see the innards of our creations at The Gene Genie Project.

The oddness of the French Revolutionary Calendar

Family tree files can have their dates formatted according to a few different calendars, one of which is The French Revolutionary Calendar. This calendar ran from 1793 to 1805 and for 18 days in Paris during 1871. It has 30 days per month, 10 days per week and the months are named for the predominant weather / agricultural activity during that period. For example, the winter months are named Snowy, Rainy and Windy.

It's a very literal calendar and would appeal to nerds normally, except that it's a special case that is hardly ever used (a lot of work for not much gain). We need to make sure our software imports all of this data without loss and can compare against dates in the Roman, Julian, Gregorian and Hebrew calendars.

I know of no-one who uses this format and I'm not sure many people will at all, ever. We still need it though if we are to be taken seriously and ensure that family trees are loaded correctly.

Scam trademarks

If you ever need to register a trademark for your business, beware. There are relatively cheap, easy and official ways of filing most government documents online and they all seem to attract their share of copycat scam artists.

After recently filing a trademark registration (and paying the fee) online via the proper government website we received two almost identical invoices (£800 and £900) by post from Eastern Europe in immaculate English. Someone, probably from the UK has set up an automated system that monitors the trademark registrations register and sends lookalike invoices in the hope they'll get paid. I'm impressed at the amount of work scammers will put in to their schemes to avoid honest work.

The small print stated they were not invoices, but 'solicitations' to be listed in their private database. They even had their own terms and conditions that suggested if you tried to reverse a payment (after finding out you had been scammed I imagine) that they would not honour it.

If you are filing government documents online and the site does not end in something like your national variant of .gov.uk then stop and ask for someone else's opinion.

As with all 'paperwork' we receive of this variety, it was put to good use for getting our log burner going nicely.

Logo contest

Ryan O'Neill — Wed, 17 Aug 2016 17:02:28 GMT

While I've been working behind the scenes with the database, Christina has been running a design competition for our first logo.

We'd love your feedback on the ideas generated. So if you'd like to see the designs and vote for one, click the contest link at the bottom.

We chose to use one of the 'design contest' formats because we wanted to see what ideas we'd get as opposed to having a single design agency present a limited set of ideas. We also don't have the budget for a full on design process right now, we will probably revisit the design when we do.

Update: We did previously have a link here to the design competition, but I've removed it because they sent a snotty email asking us to (we are not on-topic enough for them). I wouldn't normally bother, but I don't want to give them any good publicity.

So if you need a logo designing, don't go to DesignHill.com.

Roadmap

Ryan O'Neill — Fri, 05 Aug 2016 07:42:35 GMT

These are the current ideas we have, there will be more and some of these may get dropped. We may choose to develop them out of order too. If you'd like to be kept up to date please signup to the beta invite. Right now, we don't know what will be paid for and what will be free. Most things will be free, the paid for options are likely to be where we are using a lot of storage or processing power and need to cover our costs.

If you have any ideas, let us know!

To keep with the theme of genealogy we'll be calling each milestone a 'generation'.

Generation 1 - Basic trees

Site design (mobile friendly) - Hopefully we'll be able to invite the public to participate in the selection of our logo and design.
Database - ArangoDB, a hybrid graph database which is well suited to the data we'll be storing.
User accounts - Create end user accounts and the whole account management process.
Error and performance tracking - Background task, we need to see if anything is going wrong as things grow.
Tree import - Ability to import test sets of family tree data.
Tree creation - Create family trees visually using the web site.
Tree export - Export your current tree to a standard data file.
Search - Basic at first but also identifying what types of search we'll be adding (place, names etc.).
Search engine friendly pages - Outputting the tree in a format that can be found and read easily.
Infrastructure review (backups, availability) - Making sure all of your data is held in duplicate, in different places.
Feature voting - A small sub system for registered users to vote for the features they want.
Open source licence - Subject to finding a licence that works for the business we'll be open sourcing as much as we can.

Generation 2 - Apps everywhere

Application Programming Interface (API) - We'll have developed the API in the first generation and will look at opening it up to registered users here. This will enable the development of...
Mobile apps - These will use the API to access the same database that the web uses. Initially only family tree editing.
Custom domain names / white label - Your own website, without any mention of GeneGenie.com but run for you as a service.

Generation 3 - DNA

DNA profile matching - Matching against other trees with your uploaded DNA profiles.
Visualisations - Making it easy to find potential genetic matches without needing to understand the science.

Generation X - The future

Four dimensional intelligence - This is where it all started, making the research easier. I'll explain in a future post.
Location intelligence - Using Machine Learning to figure out place names.
Redesign - The initial web site will have grown a lot, do we need a makeover?
Peer research - Big topic for another post.
Visualisations (more) - Another big topic.
Image analysis - Photo face matching and intelligence.
Localisation - The site, but not in English.
Cross platform server - More of a design goal, this is for nerds really.
Automated tree synchronisation (import / export) - Make sure your research stays in sync with the other web sites out there.
Email delivery - Review our email provider to ensure as much as possible gets through.
Plugins - Are plugins needed if we already have an API?

A journey of a thousand miles...

Ryan O'Neill — Sun, 10 Jul 2016 18:02:00 GMT

I've begun a new venture and have already taken the first steps in what I hope will shake up the business of online genealogy (family tree research).

If you mentally groaned with 'Not another family tree website', I understand your thoughts well, there are a lot of money grabbing, badly designed, badly run and downright dishonest businesses out there and that's why I think it needs shaking up (there are some very good ones too, it's not all terrible).

These are my initial aims although they may change as the business grows (if they do, I'll be open and honest about any changes).

Goals

Your site - An easy to use system that helps you build your family tree and put it online quickly so that your tree is at the top of the search results and can be found easily by distant cousins.
Face matching - Who is that person in the back of the family photo? The more people that contribute their family photos, the more the system can offer potential matches.
DNA matching - Proving or disproving a link to another person by uploading your DNA data.
Peer Research - Need a document from an archive on the other side of the country or world but you just can't get there? We want to make it easy for you to find someone who lives near the archive your family records are in so that for a reasonable charge, they'll do the leg work for you.
Four dimensional profiling - Finding links between facts. This is a huge topic that I want to cover in detail when it's underway.
You are not the product - I'm fed up being treated with disdain by bad businesses who use my content to build their site without making it clear that's what is going on. Sure, someone needs to pay for hosting, development and staff costs but let's talk about it up front. Ideally I'd like to have a few optional extras that can be purchased that will help cover the free usage.
Community based - I'm scratching my own itch first (wanting to find out more about my family) but I'd also like to build what other people want. To that end, we'll be creating a ~~User Voice account~~ (the free and low cost accounts have gone, we'll have to manage feature lists another way) where features can be voted on and I can see what to look at next.
Open source - This one will take some thinking about but I'd like to share the source code for the systems I build, for many reasons but mainly security, scientific progress and gaining the trust of users. The licence will probably allow use by non-profit and individuals with a commercial option for businesses. Everyone who commits source code that is accepted would retain their copyright but provide a non-exclusive licence so that it can be used in the site.
Open - All of your research is public (except information on living people and their direct ancestors) so that it feeds into other research. Your family can see what you are doing and the search engines can direct cousins your way. If you want to keep your family research private, then I'd ask you to think again. We all gain by collaboration and if you want to copy my pictures, I've no problem with that (as long as you provide attribution). If you really want a private family tree then there will be a paid option for that.
Multiple devices - Want to access your tree on your iPhone, Android or Windows device? We'll have apps for them all and if you have a device that is unsupported, then the API can be used to build your own client software.
API - A bit technical, but we'll be building the site with a public Application Programming Interface (API) from day one which will enable opening it up to all kinds of devices and custom software.
Free - The big one. I've been advised not to make my work free but I really do think that with the right balance of free and (optional) paid features it can work for everyone involved. There won't be any nagging 'upgrade now' popups. In fact, there won't be any nasty popups at all.
Free of adverts and other menaces - Advertising slows a site down tremendously, ruins the end user experience and can be used to inject malicious code into your browser. I hate it, so I'm not having any tracking of users and you won't see adverts.

Don't get me wrong, I'm not not a hippy trying to give everything away for free. This needs to work both ways (I can build a business and you get what you want whilst being treated decently). I think the best way of making that happen is to be open and make the optional extras valuable enough so that the business costs are met.

I'm effectively making the business plan open source by writing this up. Anyone can copy these ideas and that's partly the point. If they do a better a job than I would then we've pushed genealogy research forward.

Creating a web site, setting up SSL (the padlock in the corner) and finding a host could take months not so long ago. Now, I can get all that and more done in a day but there is still a huge amount of effort required to get the ball rolling.

For example, this blog was very quick to setup but still required attention to detail (domain redirecting, SSL, caching, security, hosting) and if I missed any of the puzzle pieces it could potentially add days of confusion and look sloppy.

Looking at the amount of work required it's easy to feel intimidated. Just to write up my current set of ideas will take most of a day.

I've always wanted to do something big and that I would be proud of, I think this is it.

As Magnus Magnusson used to say, I've started, so I'll finish.

Image Lao Tzu A journey of a thousand miles begins with a single step by BK is licensed under CC BY 2.0