Hi Mikko,
Sounds awesome!
This request might be a stretch, but a preprocessor that parses out words (namely, %&$%* explatives
) for particular text fields out of an array (like the Badwords component) would be very helpful.
Here's another that's on my personal wish list, and it might take down a server or two to run on a larger site, but it'd open the door to some neat 'Web 2.0' features if a module were paired with this one. On updating a certain field, a new table with a counter tracks instances of words included within (including new ones), and accumulates them for output in modules.
Practical application: For a social web site, you have a field titled "Bands I hate the most." User 1 enters "ABBA." User 2 enters "ABBA, and anything Swedish." User 3 enters "Bee Gees, Beach Boys, The Swedish Chef, and ABBA" and so on. The cream (or the crap in this case) would float to the top, and I'd imagine anything like this would skip one or two-letter occurrences. From there, the module options are endless: bar charts, pie charts, tag matrices that look like Flickr, Delicious and others:
www.flickr.com/photos/tags/
I don't know if that's within the scope (I get out of hand sometimes), but the base from which these things can happen I'd imagine would start with a preprocessor triggered on defined events.