I've begun reading Stephen Hero again, this time taking extra-special notice of the way in which it was transcribed from manuscript to print. Because it was an unfinished text, Joyce left all kinds of cross-outs and extra notes on its pages, and the publisher did its best to represent these things with some limited indicators ([ crossed out text ], « inserted text », *misc. footnotes), and it got me thinking about typing my notebooks again.
Around 2006, I began to put my notebooks in a simple text format that was little more than typing page markers and content. By summer 2007 I had given up on this, opting instead to scan the notebooks in their entirety. This provided me with a nice reference when the real notebook was not nearby, but it was still lacking certain special faculties that make computer documents invaluable: searching for text.
Since my recent experiments with XML and XSLT, my appreciation of markup has increased a little. In theory, I could put the entirety of my notebooks in markup, then write a program to search for them, including specific things like all instances of "Equinox" written in blue ink, all crossed-out passages containing cursive, or looking for all text placed written landscape-style on the page. This would be invaluable, and even more so as an interface to bring up search results in terms of the page where the items were found—I could display a scan of the page!
This is a very attractive proposition and I'd like to explore it a little further. I've begun looking online at documents that describe such systems, usually in conjunction with ancient manuscripts. Keep an eye on the bookmarks in the blog's sidebar (or just look here) if you're interested, too.
