On Sun, Nov 29, 2015 at 3:39 PM, Andrea Zanni <zanni.andrea84@gmail.com> wrote:
(...)

Unfortunately, for the time being, we hare scattered communities with no software support whatsoever: we don't have a Proofread page for Right-to-left languages, just imagine how much time it would take to redesign a MediaWiki for being a  multilanguage, unique digital library, in which anyone can contribute.
(...)

What about two multilanguage Wikisources? One for RTL languages, another for LTR languages. Or at least one multilanguage Wikisource for languages based on roman script

It's very disturbing to work on Wikisource if you is interested in a subject and is able to read in more than one language. On the very specif subject of "slavery on Brazil" you will find works in Portuguese, Spanish, English, French, German and Latin, with at least 90% of them never translated to another language.

Also is very annoying to create thousands of Author pages to simply show snippets of biographical data for persons mentioned in a given work, since any regular research work (current or ancient) *will* refer to works in more than one language; most of those persons don't have any work translated to your language, PD-old or copyrighted, but you still will need to create an Author page if you is interested to help readers to proper understand a given text.

BTW this need pointed me to create a new local innovation: a template to point the reader to the Wikisource edition that will have works by a given person ({{autor-idioma}}). And, to prevent good faith users uploading copyrighted translations, there's also the need to put an old local innovation from es.Wikisource (currently adopted in many subdomains, en.Wikisource including). See the

https://pt.wikisource.org/wiki/Autor:Bonaventure_des_P%C3%A9riers

Not to mention that each subdomain have their own way to sort and store data on Author namespace, that certainly is the crossing barrier for those that searchs for a subject on libraries OPACs to never adopt Wikisource as a search point (eventually going to Wikisource pages thanks to Google search): they don't have time to be trained on 60 subdomains website designs to research to find information in a language that they are able to read (or to translate using a software).

In this current subdomain approach, Wikisource will be able to proper store only fiction works (since language is the main criteria for adopting a fiction work to read), but never non-fiction works (neither non-fiction works that describes fiction works: the world was culturally globalized centuries ago).