(This discussion is placed outside our Scriptorium to make more easy to invite subdomain wikis, mailing lists and users from Commons to help achieve a consensus in a central place)
Currently, the language categories for books and digitized texts is a very big mess in Commons. There is no clear way to browse all available books in a given language without being lost in book covers, individual book pages, PDF files in a given language that does not contain anything useful to Wikisources (e.g. some Wikimedia promotional folders). Sometimes is possible to found also <subdomain language> Wikisource books that does not means nothing (digitized books in a given language? audiobooks in a given language? books about Wikisource in a given language?)
So, I'm proposing to adopt a standardized category schema. My first propose is as follows, any inputs are welcome. Lugusto (talk) 19:15, 31 January 2014 (UTC)
Works in multiples JPG/PNG/GIF/TIFF files should be placed in a category named according to the work title. The category title sould be placed in the same categories as in the multipage files
Multilanguage works should be placed on all relevant language categories and in an additional one, Multilanguage <etc> (multilanguage books, multilanguage journals)
<Language subdomain> Wikisource <anything> should not be nevermore used. Wikimedia Commons is a general file repository; InstantCommons makes all files available also to external wikis. The MediaWiki software currently offers appropriate ways to track global usage on Wikimedia wikis than in the past years. If really necessary, project ownership should be disclosed using templates like commons:Template:Original.
Categories for source of origin should be added only using templates, likes the commons:Template:IA does for archive.org.
Images extracted from book scans should be categorized as encyclopedial content (subject, year of production, creators). If a book contains many images (or if the book is widely known due to it's images), they can receive also a category named as <work title> extracted content