- From: Dr. Olaf Hoffmann <dr.o.hoffmann@gmx.de>
- Date: Wed, 29 Sep 2021 11:05:05 +0200
- To: public-epub3@w3.org
- Cc: bernd@bernd-eichhorn.de
Hello, just a comment (from a german author): To provide language information with DublinCore:language ist required for all EPUBs. This already implicates the language, mainly used in the book. But in the OPF-document as well as in content elements in XHTML and SVG you can provide language information with xml:lang attribute as well. Especially as an author you can use the xml:lang attribute to provide language information for DublinCore:title and several other metadata-elements. Of course, if you own a copy of an EPUB (and a licence does not dissallow this), you can add such metadata yourself as well (but only for own use, never try to distribute modified EPUBs in any way). To provide information about the type of title you can additionally use in EPUB3 the EPUB:meta element with a refines attribute and an attribute property="title-type" to indicate the type of title. For a collection at least you can indicate the type 'collection'. With property="display-seq" for all such titles you can provide information about the order or importance of titles. To add information about all authors of texts within the book is no problems, they can be mentioned as contributors, if they do not belong directly to the creators of the EPUB. To provide information about included works/titles (including for example licence information about used images and graphics) in a machine readable way in the OPF-document seems indeed difficult or even impossible. Currently unfortunately the OPF-document does not allow structured meta- information as possible for example in the SVG:metadata element with typically RDF, Dublin core terms, some other XML with nested elements. But authors can use in EPUB3 an OPF:link element to point to an additional XML-document containing such advanced data (RDF, Dublin Core, ONIX, MARC etc). But presumably no specific EPUB viewing program will expose such machine readable data to the audience. I tested several programs (mostly browser extensions) to get an impression about their EPUB (2 and 3) capabilities to avoid major desasters for the audience. Interpretation both of required or optional features has a brought distribution under such programs in general. It is not defined, what programs like calibre have to do with such meta- information in the OPF-document in general. Calibre still has a focus on the outdated EPUB2, therefore it might still ignore advanced options possible with EPUB3. Indeed, concerning some features calibe 4 and later still seem to have regressions, but I think for metadata interpretation there was not much change in the past years. Some other programs may have some updates concerning this. Unfortunately due to changes at Mozilla/Gecko/Firefox the most advanced EPUB presentation programms I know have regressions in newer versions as well. In the past the extension EPUBReader 1.x had a relatively good presentation of EPUB2 metadata, better than calibre, firefox-compatible current versions have no at all anymore. I think Lucidor/Lucifox have some advanced presentation for metadata from EPUB2 and EPUB3. Continue to have fun with EPUB nevertheless!
Received on Wednesday, 29 September 2021 09:05:22 UTC