Hi RIchard, all,
This is actually a big problem for us. Luckily, it only manifests in a small number of horses, one of which I owned in her brief career. (She was called Catskill and was, erm, not very good).
Anyway, long and short, our original processing code created a unique id out of horse name and country code. Sadly, this is not unique, because another horse with the same name/origin can be registered some years later – as in the cases of Catskill and Herculean.
The good news is that we have a means of separating these out now (there are about 200 we’ve identified in the last 15 years, so it’s not a huge deal, but it’s obviously not insignificant either).
We are testing this on a separate specially created development instance and, once we’re happy it’s doing what it is supposed to, we will move it to the main development environment. Once I’m happy it’s working there, we’ll promote it to live.
Looking at a fortnight or so for the fix. This problem has actually been in situ since we started, and only came to light recently when Catskill ran. It is, as far as I know, the only incorrect data in our entire database, which runs to millions of rows, and I’m obviously extremely keen to resolve this.
So, please bear with us in the short term. I very much hope to have this sorted by the end of the month.