I'm still holding out for a stored text stream. I base this hope on the fact that if you search for a word in a Topaz file, the search results show the hits in the context of the surrounding text stream, just like a MOBI file. It seems like a lot of work to recreate a text stream from the glyphs for search results.
In comparison, searching a PDF file shows the results in the context of the document. If Topaz were modeled on PDF, it seems like the search results would be shown in the same fashion.
G
|