Interesting that this has appeared here now, since it's been around for quite a while (I first saw colleagues using this nearly 7 years ago). I've only heard good thing about it, though I've never used it myself (not being much of a mac person).<p>Other similar services include Mendeley [0] (now owned by Elsevier, I believe), gPapers [1] (less functionality, looks defunct), and good ol' bibtex files.<p>[0] - <a href="http://www.mendeley.com/" rel="nofollow">http://www.mendeley.com/</a><p>[1] - <a href="http://gpapers.org/" rel="nofollow">http://gpapers.org/</a>
I use BibDesk [1], which is open source and comes bundled with MacTeX. It is useful and very fast by itself, and it has a full AppleScript API which makes it extendible without having to dig into the code.<p>I wrote a couple of 'plugins' for it, one for looking up papers through Alfred, and another for quickly importing papers from the arXiv and from INSPIRE (the high-energy physics database).<p>[1] <a href="http://bibdesk.sourceforge.net/" rel="nofollow">http://bibdesk.sourceforge.net/</a>
In grad school, I wrote my own academic research "organizer" using Django (the framework makes it really easy). It takes a few weekends of work, but then you have something really cool and suited to your needs.<p>And it allows you to do stuff that commercial software couldn't do, like use a headless browser to fetch metadata from the ACM website using your credentials (hey, I pay for my membership, and I have very reasonable rate limits built in my system). Sadly that's why it's the one piece of software I'll likely never open source :'(<p>Colleagues were always surprised to hear about my system, but honestly I believe that any tool you'll use consistently throughout your life is worth building yourself.<p>I ended up dropping out of my PhD for startups :) but I'll probably go back to research in the long term, and am looking forward to adding more features to it.
I remember that OSX colleagues used to be quite enthusiastic about Papers. As a Linux and sometimes Windows user, I used to use Mendeley quite intensively, until I rediscovered Zotero, which is open source, and for my use case much better than the competition.<p>Zotero is cross-platform; it enables me to sync my papers in whichever way I choose and its bibliographic data extraction and PDF downloading work much better that that of Mendeley, because it does this from your computer, not via their servers, meaning you have access to all the fulltexts that your institution has access to.<p>I've used Zotero to write a bunch of papers, book chapters and part of a book. It's very solid.
Papers had two downfalls for me, which were that it had no functionality similar to Mendeley's sync to .bib option, and it made weird citekeys when you manually exported lists.<p>Using Pandoc and its referencing tools requires .bib files, and Mendeley's automatic .bib export means I can just type citekeys and expect them to work. And if you want your text to make sense as a markdown file, you want your citekey clean and informative. It's been a while, but I think Papers used citekeys that didn't conform to any sort of spec (something like :_blah instead of Author:DATE), and pandoc simply couldn't read them.<p>I use Mendeley now. Doesn't have the cool research tools figured out as well as Papers, but it does what it's supposed to do, which is to store my citations in a way that makes it easy to put them in a paper.
I tried to use this recently, even going as far as paying for the iOS version (whilst on the Mac trial). Unfortunately, it was too slow and clunky on my Mac, which is admittedly fairly old (mid 2009). The iOS app also crashed a lot, and wouldn't sync particularly well.<p>If they fixed these issues, it would be great. I ended up using Mendeley in the meantime.
Does it have usable bibtex support nowadays? The last time I tried using it (~4-5 years ago), that was the reason that I did not end up using it.<p>I build a small web application that full-text-indexed all of the papers I needed back then instead.
Anybody interested in a place to discuss papers and have a wiki-like area for things like links to other papers, links to algorithm implementations, summaries, links to blog posts, etc? If so, what would you hope to get it out of it?