South is just about two years old now. A lot has happened. There’s lot of documentation now. South is a tool that helps you migrate your data schemas. A quick show of hands shows that almost everyone had heard of South.
After a first initial simple version, people started to use it. Publicity on blogs ensued and even more people started to use it. And feature creep of the good kind followed: more functionality!
Originally you had steps you’d go through like “add table”, “delete table” and “we’ll call some custom step that you wrote”. Later a much more automatic system was found. Automatic change detection: you’d “freeze” the current ORM configuration so that you can later compare to it and generate automatic change instructions.
In the last release there are some new features like pre/post migrate signals that allow you to customize your migrations more. And the speed improved.
What’ll happen in the future. Probably multi-db support, but with some manual steps as the use cases are so varied. Aargh, and what to do about non-relational databases? Renames are probably coming.
South itself won’t end up in core. You can think of different implementations, so blessing one variant in the core? No. But base parts of South might end up in the core so that migration tools (both South and others) can more easily work with it.
(Aargh, Andrew speaks English like a machine gun speaks bullets, so I have probably missed some parts he said.)
(About the photo: 1990 school trip to Berlin)
My name is Reinout van Rees and I program in Python, I live in the Netherlands, I cycle recumbent bikes and I have a model railway.
Most of my website content is in my weblog. You can keep up to date by subscribing to the automatic feeds (for instance with Google reader):