I have started and abandoned RSS aggregator projects 4 times in the last 4 years. The first was in C++ and MFC, then Python, another Python, then Scala. Certai

I have started and abandoned RSS aggregator projects 4 times in the last 4 years. The first was in C++ and MFC, then Python, another Python, then Scala. Certain things about feed aggregation that you might think would be cool wind up not really working out well in my opinion. It's not until you try to aggregate together thousands of blogs that you realize that the best blogs are very much cults of personality, and can't be easy clustered together like mainstream media as in Google News. You'd also be amazed at how poorly implemented simple markup specifications like RSS 2.0 can be. I forget what law it is that says the easier it is to implement something, the worse the prevailing implementations are because of barriers to entry, but I think it's very much true. Thirdly, I'm very throughly convinced that the browser is the best environment for consuming hypertext, not special desktop applications. Therefore apps like NetNewsWire get it wrong in my view. These aren't instances of business failure, just reasons I abandoned things and opinions I came to have by writing code and seeing what happened.