It depends on what parts you're using. If you're using Mendeley mainly as a personal reference manager, Zotero is as good as Mendeley or better imo. But it doesn't have a web-based social layer like Mendeley does, for people who like that part.

I think Mendeley had more advantages over Zotero than the social aspect (which I never used). Mendeley could automatically incorporate your folder of orphan PDFs into your library, automatically fetching the bibliographic info. It had a nice dedicated desktop client for searching your library, correctly bib info, and adding notes and whatnot to papers, which just isn't as nice on a browser.

    > I think Mendeley had more advantages over Zotero than the social aspect
    > (which I never used). Mendeley could automatically incorporate your
    > folder of orphan PDFs into your library

So can Zotero.

http://www.zotero.org/support/retrieve_pdf_metadata

    > automatically fetching the bibliographic info.

https://github.com/zotero/translators

https://github.com/zotero/translation-server

(Those two repositories are the greatest things about Zotero, in my opinion. Holy crap.)

    > It had a nice dedicated desktop client for searching your library

Zotero standalone seems to be listed here very prominently:

http://www.zotero.org/download/

Thanks for the correction. I never downloaded it because it was described everywhere on the web as a browser extension. The existence of the standalone client is not mentioned on the homepage, where in fact it says "It lives right where you do your work—in the web browser itself."

Is the standalone client decent? Is the meta-data retrieval as good as Mendeley?

> Is the standalone client decent? Is the meta-data retrieval as good as Mendeley?

I think the standalone client is actually based on Gecko SDK and possibly xulrunner, so calling it standalone might be a little unfair of me. However, it certainly doesn't require Firefox itself...

Zotero retrieves metadata directly from a publisher's site with 100s of javascript plugins. Mendeley told me once that they just do OCR with some super-proprietary system they licensed. But I suspect that Mendeley now uses their servers to centralize millions of pdfs and their metadata, and probably uses that to identify and fingerprint the files in your collection... maybe.

Here's to hoping that Elsevier wont revoke your pdfs!

Why the use of the past tense? Mendeley still has all those features and more. :) Also, the open API we provide allows for building third-party applications like Android apps and such. But I digress. :)

(Disclosure: I'm part of the Mendeley team)

The past tense is presumably because the GP doesn't want to use a tool produced by Elsevier, so as far as they are concerned the tool no longer exists.

Yup. I was already very hesitant to invest a bunch of time in a tool that could be pulled out from underneath me by a for-profit company, and in fact I never integrated it into my work. With Elsevier owning it, I wouldn't consider risking it.

You might think a lot of things, but your description of Zotero is completely inaccurate. Zotero isn't a browser-based application, and it can do everything you describe.

No need to be a dick, anyone reading Zotero's homepage comes away with the message that this is a browser-based application only.

There is no mention anywhere of any other way of using it except on the download page.

The first time I went there I didn't click on the download button because I didn't want a browser-based app.

Okay, we've tried to remove the implication that Zotero is browser-based from the website. Let us know if you have a suggestion to make things clearer. (Also, sorry stakats is a dick :)

That's great news, I'm sure Zotero is getting a traffic bump these days.

You're new here so pro-tip: More concrete details, less snark, or you're gonna downvoted to oblivion.

Instead of just telling someone they're wrong, explain why. How is Zotero not a browser-based application, and how does it do everything OP described?

Sure! Zotero has been available as a standalone application for several years (and even when it only ran within Firefox in the early days, it was still a local application): http://www.zotero.org/download/

Zotero retrieves PDF metadata using a Google Scholar lookup, a method that Mendeley later adopted: http://www.zotero.org/support/retrieve_pdf_metadata

This process can also be applied to a watch folder using the third-party add-on ZotFile (Zotero supports extensions): http://www.columbia.edu/~jpl2136/zotfile.html

Zotero has always allowed all kinds of searching of metadata and content: http://www.zotero.org/support/searching

Zotero also includes notes (and these notes can be pulled from PDF annotations using ZotFile): http://www.zotero.org/support/notes

I liked BibDesk better myself, and some of my non-Mac colleagues had success with JabRef.