Quote:
Originally Posted by stewacide
Something like MobiPocket Creator is doing a whole lot of clean-up afterwards - things like filtering out page headings and numbers, de-hyphenation, image cropping, etc. - which would be impossible in real time, at least on current ereader hardware. Yes, you can do a simple reflow without this extra processing, but it looks/reads like crap, see: Calibre's PDF>ePub output.
|
Actually it doesn't look like crap. (EDIT: well maybe Calibre's efforts do, as mentioned in a post above) but I used an ereader that does PDF reflow for a year or more before I got the Kobo. It worked well, and made it possible to enjoy reading PDFs. So, I am not speaking theoretically, it can be done, it has been done, very successfully.
Quote:
Originally Posted by stewacide
And of course a simple scanned PDF, which I think is largely what people want to read, is impossible to do any of this with anyway, at least without the additional step of OCRing which introduces a whole extra bunch of problems.
|
It is quite clear from the discussion - and I am sure I mentioned it somewhere - that we are NOT talking about scanned image PDFs.