Quote:
Originally Posted by twowheels
Page numbers (screen number) on an ebook reader would be relative to font & margin sizes, so they can't be used to refer to a location in the book when talking to somebody else about a location in the book.
|
Both ePub and Kindle use byte counts for locations. Adobe's method with ePubs create psudo-pages, which appear to be about a standard page length. Amazon chose to use 128-byte chunks, presumably because they were not like pages. This was a huge human factors mistake. If they had chosen 1000-byte to 2000-byte chunks, locations would have been "page like", and they could still exactly identify a location by adding the page offset.
Mathematically, the two schemes are very similar. The Adobe scheme is actually a bit of a mess (1.5KB chunks of the
compressed files, rounded to a whole number of pages per chapter). But it looks like real pages. The choice of 128-byte chunks by Amazon make the offset a small number, and it in general about as good mathematically as any other chunk size. But readers would have been much happier with a larger size nearer to the pages they are used to.
In any case, since none of these schemes refer back to an external source (e.g. the actual page numbrs in a physical book) they can't be compared between ebook formats. The PDF format does give real pages, but that makes it a less good reflowable format than AZW/MOBI and ePub.