Last modified: 2010-10-26 20:51:20 UTC
...So that one could grab the parsed text as well as the raw wikitext. You might ask: "Why not use action=parse?" I'm working on a slideshow gadget, which should naturally display the description of the current file as well. If I would grab the wikitext, I could obviously get the content of the Description field of the {{Information}} template, but this would likely be some kind of wikitext. So to display it I would have to query the parse that for every image, which would be impractical as well as resource heavy (if the users goes through a 400 image category etc.)
Was already fixed in r75266. Note that I limited parsing to one article per request.