[houshuang]

I've improved some other aspects of the system, and made it easier to install (still really hard unfortunately), but not really worked on this. Since I export both Skim notes and Kindle notes to my wiki, it wasn't ever very important. Just would have been neat.

I tried doing what that other link is trying, my problem is that in converting the PDFs to text, the text would often be slightly mangled, page numbers would end up in the middle of the page etc, so the clippings wouldn't match perfectly with Skim. Skim also won't search or create highlights across page gaps... So I was trying to do something quite fancy to get across this with hinged on collapsing all the text, but making note of where the page gaps where, doing a fuzzy search of the clipping, expanding etc. But I could never find a fuzzy algorithm for Ruby that worked (the ones I found all focused on matching a word from a dictionary, not matching a substring and finding its location).

Unfortunately Amazon doesn't seem at all interested in expanding the functionality of the Kindle, unless it makes you buy more books from Amazon.com. A friend of mine is using a Nook with Android to read PDFs, it has Skim on the device. Although I love e-ink screens I think an integrated app that let's you read the PDFs on various devices and keeps track of annotations etc is the future...

Stian