It's pretty sad that we (researchers, academics) need a startup and $2M for this. You could do it for free by adding some basic voting and commenting features to arxiv.

Could somebody explain what this is without the "last.fm" analogy, since I don't have the foggiest idea what "last.fm" is.

When you use Last.fm and install their software thingy, it tracks what music you are playing on your computer. (I think its called "scrobbling" or something).

I think the analogy is that their software will show what papers you have on your computer, which would indicate that you have read them and might reference them.

Thanks. Color me paranoid, but I would never install a program that looks at my HD and reports aspects of its contents to someone else.

Not only that: An important feature of Last.fm is its recommendation system, and Mendeley seems to have one as well.

I just tried it out and the interface is a bit clunky, but I like the idea. Indexing was far from perfect, and haven't tried out recommendations yet.

However, in general, the idea of doing passive (meaning, implicitly gathered) collaborative filtering / recommendations on research is something that I find interesting, being a recommender systems mensch and all. This is one of the problems I'd like to see solved as a way of pulling in research that's in my field that I'm not yet familiar with, so I hope it works out for them.

They haven't implemented the killer feature yet... the ability to highlight and annotate documents, and share those highlights. For instance, you read a paper and find some portion interesting, I want see that portion and then if I need more info I can read the whole thing.

Also, I want them to integrate with the university library system like google does so that I can get authenticated access to the documents.

This looks pretty cool; but the name first made me think of Mendeleyev and Mengele. The first one's probably not too bad, but the second one might create the wrong mental picture when you're talking about research.

Of course, I'm not great with naming domains myself, so maybe I'm the only one who would think that.

I think it's supposed to be Gregor Mendel, the Belgian monk who did experiments on peas to identify heritable traits. Pair a tall bean with another tall bean, you get a long offspring bean. This is where the phrase mendelian genetics comes from, and that's what I thought they named the company after.

I first thought of George's imaginary company on Seinfeld: that's Vandalay Industries.

doesn't work for either AMS journals or front.arxiv You may as well just use evernote's webclipping service, 'cause honestly if you're doing research and you don't know what your target community is, nor do you know who to ask for good papers to read, you probably shouldn't be doing research in that area

Independent observation: somebody really should provide a free database of academic papers. I'm graduating soon and not looking forward to paying ridiculous $29.95 a la carte fees.

In my experience public libraries subscribe to many of the academic databases. Often you are able to use them from home by first going to the library's web site and clicking the link to whatever you want to use, rather than going to the publisher/database website directly. Maybe you will have to enter a library card number. Some university libraries also provide services to alumni, either free or for a fee (in the case of my university's library, I believe it's an annual fee rather than a la carte).

In some areas, it is common for the researchers to put their papers up on their web pages or on the ArXiv. Alternatively, just write to the author. They will send you a copy.

...or write to anyone who is at a university. Skype / IRC are my standard mechanisms for such. I can usually get most papers inside of 10 minutes by pinging a couple people.