There are at present two extreme models for public distribution of scientific information on the internet. The first is an electronic shadow of paper journal publication. It concerns documents which may be or have been published via a traditional journal. These documents are placed by their authors on centralized preprint servers or personal "home pages". Even if they will never be traditionally published, they retain identifiable marks from tradition, notably: 1) They are (relatively) immutable. Once distributed, they are expected to remain constant in their content and location, up to possible small changes. 2) They have a definite authorship. There is a small set of persons who can be held accountable for the content of the document and should be credited for the information contributed. The second extreme is an information resource of a more mutable, and community-generated, character. This extreme is typified by databases, software archives, and the loose, often encyclopedic, compendia of knowledge and opinion of interest to specialized communities, the ubiquitous lists of answers to "Frequently Asked Questions".