Like with subtitles, it appears that diacritics (characters not in ASCII set) are vastly disregarded in the community.
SInce I bought the T3 I noticed a big difference between them in what non-English glyphs are concerned.
I got the T1 fonts and confirmed that these glyphs are there, and T1 can display
correctly the title of the books that have these diacritics, yet the content is still
wrongly displayed (
ie with ? instead of the actual character). The T3 has no issue at all, the books converted with calibre are displayed (almost - I am a perfectionist) perfectly.
The only simple solution I was using so long was to add a custon font definition to CSS in calibre, to point to an existing font, like amasis. An ebook so converted looks ok on all ereaders I tested (those that do not have the required font, eg amasis, simply ignored that style). Unlike external fonts (on SD or internal flash memory) the existing fonts have preserved their path across PRS generations - at least this is what I found.
The question (yes, there is a question, too) would be if someone managed to use the internal fonts for non-English characters with no hacking at all (some settings on calibre, some settings in HTML, whatever). And yes, I have read severeal hundreds of threads and internet pages, but I might have overlooked some.
Thanks for attention and big thanks for any good answer
