Christina writes on social networking services for researchers and why they fail. Particularly useful is the reference to relevant literature on diffusion of innovation - Cameron Neylon

why they fail? oh dear, i thought i was giving suggestions for why one would work. - Christina Pikas

well why one could work - I have only read the first bit in detail - bookmarked it to come back to later in detail... - Cameron Neylon

Haven't read the whole thing yet, but I submit that the challenge still is that there is no such thing as a "scientist". Physicists work differently from biologists, and even that might be generalizing a bit. Which leaves us with a problem, small populations, which makes building communities hard - Deepak Singh

Great post. Left a comment on your blog almost as long as your post, but probably not as well-written. I think value up-front is a key requirement for success, and discovery both through automated recommendation based on user activity(like last.fm or Amazon) and discovery based on expert recommendation(like Pandora or just sharing of links by experts here and Twitter) is precisely the kind of up-front value that such a network could provide. Would I sound like a broken record if I mention, again, Mendeley and their idea of being the last.fm for scientific papers? What they're doing here: http://www.mendeley.com/blog... is setting up for exactly that. A non-IF based article metric is certainly value up-front! - Mr. Gunn

i'll leave this on my blog, too, but value up front deals with perceived relative advantage - how much potential users think this thing beats competitors... @Deepak - absolutely - you can't just talk about "scientists" really, but i wonder if you can do a more generic thing that could work for interdisciplinary projects, and also maybe with sub-communities for different research areas - Christina Pikas

I mentioned it on my blog: http://ways.org/en... . - Daniel Mietchen