Q) I was looking at the cached version of an HTML document, but the links from the first page appear to be broken. What gives?
A) Unfortunately, HTML doesn't have a notion of document per se, and there is no easy way to know which of the links from the initial page are part of the document and which are not. Implementing support for "composite documents" will provide a facility to specify which resources are part of a document (athough it doesn't address the problem of determining which these are automatically).

Q) Are the documents in the DocuBase repository backed up?
A) Yup.

Q) What guarantees do I have that the documents and collections will persist and be available?
A) Well, this is a research project, so we only guarantee "best efforts". It lives on a well-maintained server, and nothing will disappear overnight.