I love postgres to death, and they do amazing work. It's certainly the best database I've ever used, and that's before thinking about postGIS and pg-routing. T

I love postgres to death, and they do amazing work. It's certainly the best database I've ever used, and that's before thinking about postGIS and pg-routing. Their documentation is top notch and if you sit in their IRC channel you'll see a lot of people helping each other, including core devs answering basic questions to really involved ones. It's really an amazing project in almost every regard. I recommend it to anyone who needs a database in almost every situation as your first one. If you outgrow pg's capabilities, including full-text search, then you're probably also having capacity planning and infrastructure issues that need to be addressed too when moving to more specialized software. However, it drives me bonkers that they don't use a bug tracker because their current mailing-list system "works." I think it makes it hard * to find the status or history of an issue ("just use the mailing list search!" and then try to piece together and separate multiple threads * to allow younger developers (which they want to have talk to all the other devs before doing anything at all) * to look for open issues (which they claim don't really exist because things are fixed quickly, but that's not always the case) * to make sure issues don't get lost (which they do), I've also heard it said that bug trackers attract low quality issues, which, even if true, I don't see why that's such a veto issue. They actually have a quasi-bug-tracker for things like their hacker fest projects where you can find an issue and offer suggestions on it.