For the last 6 months or so I've been working on a 'wikipedia for images' type site. I'm trying to fix all the UI deficiencies that bug me about other image sit

For the last 6 months or so I've been working on a 'wikipedia for images' type site. I'm trying to fix all the UI deficiencies that bug me about other image sites. As a personal project with no deadline I've had a lot of fun premature-optimising the crap out of it! For example the thumbnails are in strips of 50 for a huge reduction in http requests. Here's the test set I've been using- http://imagecorpus.com/marylin-monroe http://imagecorpus.com/marylin-monroe My plan is to initially scrape a lot of hard-to-browse public domain image sets, for example the Library of Congress daguerreotype collection http://lcweb2.loc.gov/ammem/daghtml/daghome.html http://lcweb2.loc.gov/ammem/daghtml/daghome.html. And then I'll solicit contributions from the public to bulk it out. I was going to wait till I was ready to launch before posting here, but that could be a while... Any criticism/suggestions or offers to help are very welcome!