Jostein and Greg,
Thanks for the always-welcome feedback on concrete library needs.
I can see 2 different use cases there:
- production of distribution formats (DAISY 2.02, DAISY 3)
- production of archival / interchange documents (DTBook)
These conversion needs look totally legitimate to me.
As always, going from a permissive grammar like HTML (as used in EPUB3) to a stricter grammar like DTBook is not easy and not always predictable. The EPUB 3 ToC can be leveraged to get the basic structure, but the html-to-dtbook part often has to make assumptions on the structure of the input, making the conversion quite brittle when used with different input sources.
Going from EPUB3 to DAISY 2.02 is probably an easier task to tackle, although I'm not sure about the exact needs there. I would assume that by the time you can expect to receive EPUB 3 with Media Overlays from publishers (i.e. not so soon) device support will be there. I then assume that you're thinking of text-only EPUB 3 input. Do you intend to convert them to text-only DAISY 2.02 books ? Or text+audio DAISY 2.02 books with TTS ?
I think the production of text+audio DAISY 2.02 can fit in the TTS-based production development that will start this year. Production of text-only DAISY 2.02 might be started as an off-track development, if it is a high priority. In any case, I understand that producing DAISY 2.02 is still very demanded.
On the project management side, as you know there are still pretty big items in the current charter's objectives that will consume most of our current resources; these are legitimate and sorely wanted too (e.g. TTS-based production). We can always play with priorities and do some little off-track development, but at the end the charter needs to be respected. That said, as always, if an organization wants to start and take the reins of the development of a particular conversion, we'll fully support it.