Originally Posted by paulpeer
Not only the metadata cannot be read. The main problem is the table of contents of the device: imagine you have a reader with 50 books in Tamil, or Hindi). In the TOC you'll see a beautiful list of 50 rows of question marks. If you want to open a certain book, you will have to try them all and hope the one you are looking for isn't the 50th ...
There are two different things being discussed here:
1. The table of contents of the book. This will indeed use the device font which may or may not support the appropriate characters. A way around this is to have an "internal" TOC in the book itself, implemented with hyperlinks.
2. The list of books on the device - ie the "library". This may display metadata from the book or it may simply display the filename, depending on the device.