Photomuse
There has been quite a bit of buzz over the public "beta" of the PhotoMuse site, a new online database expected to eventually offer the combined collections of the George Eastman House and the International Center of Photography, almost all of it glowing. Certainly the partnership holds a great deal of promise sitting on the foundation of these two institutions. I love the idea of exposing the collections publicly, but as long as the technical infrastructure is directed by ICP the project is likely to fall far short of its aims. I don't know for sure that ICP is guiding the technical work, but after spending a few minutes with the PhotoMuse site I detect the same fingerprints that have continually marred ICPs previously deeply flawed site. The GEH site is even worse. (While writing this entry I discovered that ICP has quietly launched a completely new Web site– vastly improved and solving almost every major flaw that had plagued the previous incarnation. Still, they fail to link to Gallery Hopper in their resources section, so there is room for improvement.) Granted the site is a prototype (and was crashed by the deluge of attention from the NY Times' coverage of the partnership on Wednesday) so hopefully there is time to get some simple but effective changes made:
1) Put the search function right on the homepage
There is no reason the search function should be buried three pages down at the end of a string of increasingly inscrutable and unnecessary links. This also would have alleviated the traffic problem by reducing the number of pages a user has to load to do a search (or would it? I suspect a fair number of visitors gave up looking for the collection before they stumbled upon it.)
2) Spend a day with other photo databases and steal their best features
Has anyone in this project used Getty or Flickr? Doesn't look like it. As it’s backed by non-profits, I suspect this project is hobbled by an overly academic approach that translates into snail’s-pace development and a design-by-committee ethos. (Also evident in monolithic corps like IBM.) Getty, as a for-profit, lives and dies by its visitor’s ability to find just the right image so its search and presentation capabilities are well refined. Flickr grew as a small grassroots site for public storage and display of photographs and has a wonderfully user-friendly interface - a perfect model for PhotoMuse.
3) Minimize the use of extraneous images and animation.
What is the need to spell out "PhotoMuse" with a series of large, individual graphics that reveal images from the collection on rollover? Nice idea, but the execution is botched and detracts from the point of the site – searching the database. One small logo would have done the trick. And that little accordion graphic that cryptically links to the search interface is the equivalent of a circa 1996 flaming logo. Lose it.
MAN has declared the St. Louis Arts Museum site to be the worst of major American museum Web sites. It might be ugly, but at least you can find what you’re looking for.

I totally agree. I was just reminding myself recently how poor the GEH website is and how it does such a disservice to itself. I think in this case at least there was been a bit more attention paid to aesthetics but I wouldn't even go too far in that respect.
As you mention, there are numerous and varied models for image galleries/archives on the web. Photomuse doesn't seem to have benefitted from the strengths or weaknesses of those examples.
This actually seems like a project which didn't even have the "benefit" of a committee vetting process. It seems rather insular and untested in terms of flow and also how it deals with displaying both information and navigation.
Someone like Second Story should contact GEH and make them a proper website.