Some photos have unreadable EXIF info because it's not in Unicode. This is not really a bug in MediaWiki, but in the photos. But it would be nice to have a way to specify the encoding of the photo somewhere and have it show correctly, or convert the image file to Unicode using MediaWiki interface.
Version : 1.21.x Severity : enhancement
Example:
https://commons.wikimedia.org/wiki/File:Yemenites_at_Rosh_Haayin.jpg
Note: The exif standard specifically states that most fields should be in ASCII (although some standards recommend assuming utf-8 instead), with exception of some fields which do specify their encoding.
However, the photo in question's metadata is not exif formatted, but uses iptc-iim metadata. That format for metadata has a specific mechanism for denoting what encoding the image data is in (which we have support for). [See page 20 of http://www.iptc.org/std/IIM/4.1/specification/IIMV4.1.pdf ] (Of course, that mechanism is somewhat newish and possibly comes after the image was created. Additionally lots of such images don't have the encoding data specified).
I don't feel we should put something in the mediawiki layer if the file format itself supports the feature. Hence I suggest wontfix.
It doesn't seem we have had a lot of complaints about this. Also "some files"and 1 example isn't really a lot to go on 9 years later..
Considering Brian's suggestion of WONTFIX, I'm closing this. If people have more specific examples that defy the WONTFIX that Brian suggested, then we can always file new tickets.