lifecycle of projects begin in cathedral, creates platform for explosive creative input, small cycle to user goals for releases, big cycle to user goals (reliability) over time, reduces the usefulness of the project for self-motivated creative exploration (the 'let's re-factor all this unfinished v2.0 syndrome). Linux pushes past this by farming the user effectiveness measures out to the distributions returning Linux to a platform for creative collaboration. The ability of projects to effectively protect the creative input platform is important (because it is one of the crucial contributions of FLOSS). But then again running software is vital as well. In all these situations the motivations of the people, and the alignment to the goals of the project, and the coordination mechanisms, is vital. But there are ways to handle it (eg paying people to do 'scutt-work') when motivations, structure, coordination mechanisms align, there should be an explosion of contribution.