CompuServe Messages

Rich, We do data capture for the print industry as well, and I belong to an orginazation that runs its own small press, so I have some idea of costs/efforts on that side as well. (Obviously I'm only addressing text concerns here, with illustrations/images, but not real multimedia) Cost per K of data capture is about the same (we charge about a buck/thousand bytes, proofed and tagged). Tagging for a typesetter vs a retrieval engine is about the same level of effort. Novels runs3-500,000 bytes, a reference book something over a million. OTOH, CDROM _updates_ seem to run about 10 times that–3 to 10 million bytes added or seriously changed. Design seems to be a similar effort per "work" several days to mess with typefaces, chapter headings, indexes, SGML markups, book sizes, hypertext jump tags, and so forth. I would also say that jacket copy is pretty similar except that a book doesn't need a user's manual<g>. Mastering the CDROM is a one time charge vs. a per page cost for making printing plates. Breakeven seems to be around 200 pages, and as we all know per copy cost is much lower for the CDROM. One statement of yours I would like to clarify however is author's revenues: Authors certainly get (or could get) a bigger chunk of gross revenues, but there's so freaking much more on a CDROM (half a million standard pages is a _lot_ of novels) that the amount paid in royalties per megabyte-copy will probably be lower. Whether this helps a particular author seems situation dependant–If I buy the complete works of Edward Veeblefester rather than just the 2 novels I would buy in paper, he's probably ahead pretty seriously. OTOH, the author of a single novel restamped (neologism alert) in "A history of the Detective Story" probably loses. Frank