I'm on commons mainly for support of Armenian projects. Particularly I have been uploading to commons scanned Armenian out-of-copyright texts, to be used on Armenian Wikisource with the ProofreadPage extension.
I run the bot TeakyBot, which is used for the sole purpose of automatically uploading the scanned Armenian Texts.
Since I have been asked several times about the file and category names used by me and my bot, I'd thought to explain it here. The names of the files are exactly the names of the works, followed by the number of the page. The respective image files are then categorized by the original Armenian name of the work, and the categories for the works are subsequently put into the Author categories (English names), and the mother category Scanned Armenian texts.
The practice of naming the files and categories for scanned texts by the respective names of the works in the original languages is widespread. These image files will be hardly ever used outside of the corresponding language projects, thus the ASCII naming of the files "so that everybody can use them" seems a bit superfluous here. Besides, the ASCII transliteration of non-latin based alphabets are rarely universal, and such transliteration may create unnecessary confusion in the local projects. This will force such files out of commons and into local projects, jeopardizing the role of commons as a universal media repository for all languages and projects.