appended is a "first pass" at a p.g. categorization scheme.

unlike the "tagging" models, i'm using here a "folder" model.

inside each folder will be a bunch of alias "files" that point to
the e-texts proper.  this allows you to put one e-text into
multiple folders, by simply generating another alias file.
the possibility of nesting folders is also present.

as i've said, i think the cataloging tool needs to give people
an ability to rule out unwanted classes of e-texts, including:

first cut -- by language
second cut -- human genome
third cut -- audio
fourth cut -- video
fifth cut -- magazines
sixth cut -- copyrighted
seventh cut -- reference
eighth cut -- religious
ninth cut -- poetry
tenth cut -- plays
eleventh cut -- short story collections
twelfth cut -- anthologies

and once a person has ruled out whatever they don't want,
the downloading of the other e-texts can go automatically.

-bowerbird

p.s. 

books

top-100 titles

selected top-100 titles in html

fiction

nonfiction

reference
  dictionary
  encyclopedia
  thesaurus
  quotations

poetry

plays

short story collections

anthologies

religious

heavily illustrated
  beatrix potter

children's
  beatrix potter

cookbooks

how-to guides

magazines
  punch/punchinello
  scientific american

items of interest that are not text and not html

audio
  music
  human-read books
  text-to-speech books

video

copyrighted work

human genome project

languages other than english
  french
  german
  italian
  ...