Let me try to generalize this debate a little. There are two "paradigms" in ebook formats. Rendered and reflowable. Examples of rendered formats are PDF, PS, DJVU, DVI, TXT, any collection of image files like CBR/CBZ. Examples of reflowable formats include (HTML+CSS, LRF, FB2, RTF, any XML based ebook format). Within reflowable formats there is the concept of semantic information. Some formats support more semantic information, some less.
Within reflowable formats, a format that supports fmore semantic information is superior to one that does not, so for example this makes HTML+CSS > LRF. To get an idea of what semantic information is, read my previous posts.
Now for me, this thread is fundamentally about the suitability of rendered ebook formats as a general purpose ebook format. PDF being the prime example of a rendered ebook format.
My position is basically that, for general ebooks, one can achieve enough typographical control using reflowable formats that the advantages of having semantic information make them superior to rendered formats.
|