Version:            (using KDE 4.1.3)
OS:                MS Windows
Installed from:    MS Windows

Currently there's an option to add illustrations to words. But user still has to find and download such illustrations. My idea is to make use of free-licensed images gathered by Wikimedia Commons (http://commons.wikimedia.org/). Because words have different meanings, it's not easy to automatically find the matching photo, so here's an another idea: utilise the same mechanism AmaroK uses to get CD covers: user clicks "find image", gets a dialog window with image previev and buttons: "save", "previous", "next". In this way user can choose between different illustrations of the same word (sometimes word can have many different meanings).

To avoid making a copyright violation, there should be some way of showing copyright information for each image downloaded from Wikimedia Commons. I suggest adding an URL poiting to image description page (which looks like http://commons.wikimedia.org/wiki/Image:Emmentaler.jpg) to dictionary XML. When showing the image, there should always be a button "show image info", starting a web browser with remembered URL.

One suggestion about where to get images from: don't use Wikipedia or Wikimedia Commmons. You should use Wiktionary, because it has images for words in many languages. All images used on Wiktionary come from Wikimedia Commons. I suggest checking different language versions of Wiktionary in the following order: de.wiktionary, pl.wiktionary, en.wiktionary, fr.wiktionary. That's because different language versions use image more or less eagerly. In my suggested AmaroK-way of checking images you don't have to check all language versions at once: you can visit first version, fetch all images, when user click "next" and reaches the end of all fetched - go and visit another language version, fetch more images etc.

Examples:
http://de.wiktionary.org/wiki/Turm
http://pl.wiktionary.org/wiki/kolumna