There are so far two books I have liked so much that I typed them in to make them freely available. I just finished the second one: Kai Lung Unrolls His Mat, a quite singular work of fantasy by Ernest Bramah first printed in 1927.
The other one is Why don't we learn from history? by B. H. Liddell Hart.
3 comments:
Any reason you couldn't use a scanner and OCR it in?
I've never seen an OCR process that worked very well. Even with very crisp copy in a known font, error rates of 1 bad character in 20 are common; expect a lot of tedious cleanup. My book is ancient and yellowed, I would have to destroy the book to scan it, and it still might not scan at all in this condition.
There's another factor: I enjoy typing. It's sort of like a dynamic meditation for me: I go into a flow state (see Flow by Mihaly Csikszentmihalyi) and try to balance the conflicting needs of high keying rates and low error rates. While proofing the final section of this book, I found only about one error per page.
The diybookscanner.org guys typically use digital cameras and commercial OCR software like Abbyy Fine Reader or Omnipage. No spine cracking needed, and it's a whole heck of a lot easier than typing the whole thing in.
Post a Comment