Learning Python and Scheme was fun. Learning Haskell feels like being in an 80s training montage. One day you're a nobody, the next day you're an FP-wielding superhero.
http://www.haskell.org/haskellwiki/Applications is good for inspiration.
On a side note, I'm learning Haskell through the book Real World Haskell, and it is the best resource I've found so far (I've looked through "Write yourself a scheme", "learn you a haskell for great good", and "Yet another Haskell Tutorial"). Of course, it is the first one I've actually had in printed form, and that probably made a huge difference.
As you mentioned, #haskell has been a tremendous resource. I would highly recommend anyone interested in learning haskell to lurk on that channel.
Statistics with Incanter is another piece of this pie that seems interesting and valuable. Hopefully I'll have written something useful with these technologies by this time next year.
I feel like I need a fresh perspective before I can approach Clojure -- otw I'll just end up writing a bunch of Java with parens.
For me, it's about "stop spending all my time on that one boundary condition that 0.1% of the users will hit" :)
It will let players of that game analyze their current situation and environment better and let them communicate with allies more effectively.
I've been "working on" that project for the last 3+ years but in fact I've been mostly "screwing around" learning Common Lisp and learning/implementing some advanced algorithms like rete and implementing some advanced web framework that I'll release as open-source in a few years but not before I implement some of the crazy ideas I have that will enable a completely revolutionary way of developing web apps.
It's stackoverflow but for product reviews, critical mass is getting there but not quite. I have some other features I want to add like "awards" which show off the best and worst rated products of the year, month, etc. It would be fun to publicly shame bad products/companies.
It looks like you're trying to make things as simple as possible, but I feel you really need some way for people to rate things other than good/bad (up/down in your case). You need a way for someone like me to be able to see that product A is considered better than product B, and no, number of up votes doesn't do that. All it tells me is that more people purchased item A and thought is was not rubbish. This may mean the product has been on the market longer, or just has better marketing. Seeing that 50 people gave product A 3 stars, but 10 people gave 5 stars to product B carries more information
Being able to see the distribution of votes is also important. Some products are divisive, with some people loving them (5 stars) and others hating them (1 star), and other products are just uniformly considered average (3 stars). Showing this distribution also tells me more about the product than just showing me products C and D both average 3 stars
Reviews. No idea how you'll do this, but you absolutely need a means of encouraging people to write good informative reviews. Allowing users to rate reviews seems to work for some places, and productreview has some sort of rewards scheme based around reviews. Given that no product is perfect, it's a well written review that allows me to see what the faults are and figure out if they're likely to be an issue given the way I intend to use it. It's the reviews that flesh out why some users gave a product 2 stars and others gave it 5. If, for example, I wanted to buy a mobile phone and I could see that the people who rated it 1 star did so because the sole carrier in that country has lousy coverage I'd know that didn't affect me as I have (being in Aus) 5 different carriers from which I can purchase said phone.
Oh, I like the clean layout of the site. looks good.