> Finally, while I liked the article a lot, I disagree with the premise that projects _must_ be finished.

To be fair, the author admits that including maintenance, a project is never really finished.

Instead, he says that a project should be _published_. Preferrably, on GitHub :P