Folktunefinder

FolkTuneFinder version 4 out now

On Monday 9th January 2011, version 4 of FolkTuneFinder.com went public. It was a bit of a journey getting to this point. I run FolkTuneFinder in my spare time. Mostly it runs itself without intervention. I keep an eye on things, monitor spam (except for a recent occasion when it got out of hand) and answer mail. There is a fair amount of programming behind it, and I’ve re-written the search engine a few times over the years.

Read more →

FolkTuneFinder: Why are the search results so different to what I typed?

When searching FolkTuneFinder, you may find search results that you don’t agree with or can’t understand. You may think tune has nothing to do with your query, or the highlighted notes bear no relevance to what you typed. Here’s why. The thing about folk tunes is that they’ve survived in the aural tradition, in many cases for quite a long time. A good tune spreads because people like it, and different parts of a tune may appeal to different people.

Read more →

FolkTuneFinder index building used to be expensive

The first version of FolkTuneFinder was written in a combination of Java and PHP. I was still working out the best way to do melodic indexing, and the index build process was parallelised. The job ran across 14 Apple Xserves, made available to me by my university. That was back in 2008. These days it runs in a single virtual machine … somewhere.

Read more →

Recent downtime on FolkTuneFinder.com

I started FolkTuneFinder as a student project back in 2008. I’d done websites for a few years before, but this was the first serious one with any kind of heavy lifting or interesting behaviour. Over the years I added features that allowed people to interact, such as the commenting and FolkTuneFinder blogs, which has been surprisingly popular. I have always had a very small problem with spam: I received perhaps a small handful of blog posts a month, which was fine to deal with.

Read more →