[saxondawg]

With heuristic processing, Calibre moves from excellent to scary excellent in document conversion. I realize it's a rather experimental feature, and won't be effective in every situation. But my initial observation is that's it's doing a much better job of detecting things like hard returned blank lines between paragraphs--my pet peeve with e-book files. Calibre has done a pretty good job of catching these for some time, but it seems to be more thorough with heuristic processing. (Note: I had noticed that Calibre was good about eliminating the extra ^p code between paragraphs, but when the ^l (line break, I think) was used, Calibre didn't pick that up--think that may have been fixed).

It's amazing how far Calibre has come in becoming a program that pretty well marks up an e-book document to be anything you want it to be. I can now manage my Kindle collections in Calibre. And I have (accurate enough) page numbers for all my books. These are things Amazon should have provided, and Calibre is filling the gap. I intend to make a donation to the writers of this program, and echo those who have reminded us to do so in another thread. This is one of my favorite programs, and someone deserves to make some money from it.