I have to think about what you are saying. We are coming to the info with different expectations. I'm not immediately sure how to make the hierarchy more visi

I have to think about what you are saying. We are coming to the info with different expectations. I'm not immediately sure how to make the hierarchy more visible with this data. The source data has no single entrypoint except arbitrarily chosen, and can be (or seem, practically) infinite in all directions, with arbitrary relationships at any point -- maybe something like a very large wiki with a different internal structure (more at http//:onemodel.org ). A structure could maybe be chosen for presentation purposes though. But the "article" you refer to was made as an exception and is not the way information in my ~"knowledge base" is typically structured. Each page I see as not a document but as an entity with properties and also a set of references to whatever else relates. More about my outline comments elsewhere in this discussion, where I can export chunks of it as individual outline documents, to some # of levels from any point, but conceptually it does not lend itself to paragraphs or documents, as much as ideas and relations. I'm not explaining it as well as I'd like -- again, there is more at that web site about the intentions of my software that this comes from. It's like a big mind map (with future plans to hang more data and code, on all the nodes.) Or maybe each page could be seen as a paragraph, broken into bullets which each can link to details on that thing, if someone is interested. Or it is a set of topic sentences, with contents underneath each sentence or link being the full paragraph, for the interested -- like where I wrote in this discussion about enabling skimming. I don't want to stifle discussion, if you have questions or more suggestions that can in some way account for the above, or improve on it... Thanks.