I've been meaning to write about this ever since I got one, but other things intervened. I happened to be in the area of the NYPL Science and Business Library one day a while back while the Espresso Book Machine (from On Demand Books) was still in residence, and stopped in to get myself a copy of Edward Abbot's Flatland (copyright free text here).
The Internet Archive has scanned the text of the 1884 edition from the University of Toronto for Microshaft, er, Microsoft, or at least it so appears from the rights paragraph. It's a little confusing, so I'll let you intrepret it:
Digitized for Microsoft Corporation byt he Internet Archive in 2007. From the University of Toronto. May be used for non-commercial, personal, research, or educational purposes, or any fair use. May not be indexed in a commercial service.
Apparently, you can used the text itself, which has fallen out of copyright, but not this particular scanned image. Is copyright getting more insane, or is it just me?
The book itself is, well, perfectly serviceable for what it is: nicer than a paperback, not quite as nice as a trade edition, of which it's about the same size. My biggest complaint is the paper, which is just copier paper, and not particularly nice copy paper. It's a bright enough white, but the texture leaves something to be desired. That's probably just me being picky, though. In the future, I hope they'll be using acid-free paper to combat the problem of deteriorating books. Just having them in an on-line archive is not enough. (Why? Two words: electromagnetic pulses. Good-bye accumulated knowledge of mankind, the equivalent of the burning of the original Library of Alexandria.) The covers are nice and heavy and a coated matte finish. How long the whole thing lasts is still up in the air; I'll keep you posted. The look of the book inside, of course, depends on the quality of the digitization, which was okay, but not perfect. You get a sense of the original's wear and tear without the actual thing, and with older books, the benefit of some nice hand-typesetting with beautiful type. The Digitized by Microsoft ® watermark on the bottom of every page sort of spoils that, though.
The reason for using copier paper is that this is mostly what the Espresso Book Machine is: two copiers in tandem, one a heavy-duty duplex B&W for the text, the other a heavy-duty duplexing color copier for the covers. In between is a mini-production line that fans, cuts, collates, sands the binding edge, applies glue and attaches the cover, then delivers the whole package to another trimmer—and this is where it fell apart.
My cover is a little mangled because the final trimmer, I suspect, was showing some wear and tear and the delivery system was not quite accurate enough to get the book block and cover positioned quite right. The guy who was operating it popped the book out with minimal damaged and I took it home and finished trimming the cover with an Xacto. This is was version 1.5 of the machine, and there are still some bugs to work out, obviously, but all in all, I think it's a very successful experiment.
If you'd like to see the specs, there's a PDF of the information sheet I picked up, here.
Comments