Quote:
Originally Posted by Faterson
The trouble is when that file is extremely long, like dozens of iPad screens long. Unfortunately, there are many mangled e-books like that out there. 
Marvin currently uses filename+percentage inside the file, which should be pretty much the same as byte-count. For brief or moderately long chapters (sections = files), the sync results seem to be pretty precise. It's those extra-large sections=files that are troublesome; they may contain several chapters per section/file... It's a mistake on the part of the e-book publisher, but regular Marvin users will be unable to fix it manually. |
Well, percentage is an approximate number based on a simple calculation based on the size of the file, not the contents. In ePub a file should not be greater than about 300K (actually a bit less) to work on some of the readers out there since the entire file must be loaded into ram. Using approximations is unlikely to yield good results as noted by some users, but as you say it is close most of the time. If there are a lot of images in the file then the references to these images can throw the percentage off. Of course, I am not talking about the display number, just the stored number used like a bookmark to find the location to resume reading.
Dale