Ah, I should have guessed it would be a destructive process. Still something I'll keep in mind but I have a particular weakness about destroying old books, even if intellectually I understand 1) it's not a valuable book and 2) there are other copies out there and 3) it's worth more to me in PDF form.
Non-destructive scanning is very much an option, but it's slower, more expensive and requires a bulky rig. It's probably only worth the effort if you're a serious data hoarder.
If you have a handful of precious books that you really can't bear to cut up, commercial non-destructive scanning services are available at ~$30 per book.
Only $30? (No missing 0?) That's on average (assuming a range of novels to textbooks thay makes the average...) only doubling the book cost to get a copy in another format.
Not ideal (IMO buying the physical book should entitle you to a zero marginal cost PDF copy too, which some publishers do adhere to) but certainly better than I expected.
I completely understand that sentiment! I definitely have books that I keep because they are valuable to me both as a book and as the information they contain.
Textbooks, are by their nature, more 'conduits of information' than works of art. (TAOCP not withstanding). Also they get regularly re-issued in order to be resold to students so their knowledge is more widely dispersed than many books.
When I was at Google I visited the book scanning operation a couple of times and they had these machines that would photograph a page, dewarp, and some of them even turn the page (some required a person to turn the pages). While I would love to have such a capability, the only way I could justify it would be to have texts I needed to add to my data set that were not mine to destroy.
My work-around for this issue so far has been to find and buy a second copy of a book on the used market to digitize, best of both worlds for me in that situation.
The process used by the Internet Archive and Google for digitizing books is nondestructive; the Archive open-sourced the plans for their "Scribe" book-scanning machine, and others have made downloadable plans for laser-cuttable versions. It's slower than running a destroyed book through a sheeet-fed scanner (typically you get on the order of ten two-page scans per minute), but it's suitable even for rare books.
