"I feel bad. The nice folks at Mendeley just sent me a T-shirt and I'm going to complain about their product. But I want to like Mendeley, really I do. I want it to take over from CiteULike as my online reference manager because, well I'm not sure really, I just feel it should. But I'm still struggling, every time I go back, I just don't get it. Sure it's pretty, but it's a lots less functional for me than CiteULike, which though rough round the edges, "just works". So I thought I might be able to make some progress by facing my demons. To do that, I made a list of all the things (I could currently think of), that I feel are holding me back so that people can tell me what an idiot I am. Yes, really - don't hold back." - AJCann from Bookmarklet

AJCann, what cool features should they add, to make 'your' communities split? You indicate being able to find FOAFs based on shared literature references... anything else? Mendeley aggregating comments from 3rd party sources (PostGenomic, Cb, ..., whatever general bookmarking service), like PLoS does with article metrics? - Egon Willighagen

Mendeley has had a lot of publicity and that in itself will attract people who I might want to track. The other possibility is that CiteULIke might go away, but that would be simpler, because then Mendeley would be the obvious network. The stats they provide right now are getting close to PLoS article level metrics, but again, detailed stats are a premium feature (or likely to be in future). Other than that? Mendeley is shiny, built around hoarding PDFs, has an EndNote-killer MSWord plugin... etc. None of those features appeal to me much, but they are persuasive. - AJCann

A follow up: http://chem-bla-ics.blogspot.com/2010... - Egon Willighagen

You pick it right at the top. You don't like PDFs. Mendeley is most effective for the usecase which includes keeping a pdf on your computer somewhere and wanting to read it elsewhere. I just don't think it fits your use case that well... - Cameron Neylon

Martin Fenner says he loves PDFs - but he still prefers Papers to Mendeley. - AJCann

True. But I can't help feel your objection would be a lot less if the interface put the URL links much more front and centre. I mean, you don't have to use the desktop client at all for what you want as far as I can see? - Cameron Neylon

I am not using the Mendeley client at all... just the web... - Egon Willighagen

I have been advised that most of the power currently lies in the desktop client. The web component is where the social (such as it is) lives. - AJCann

Will there ever be one reference manager that fits all needs, is fun to use, and free or affordable? Not very likely. So I think it's good that we have several alternatives, and that also includes Endnote. - Martin Fenner

+1 Martin - Graham Steel

The discussion desktop client vs. Web-based also very much depends on how you write your papers. I very much hope we will at some point have a collaborative web-based solution for paper writing (and Google Docs is definitely not the answer). This will change the reference management market completely. - Martin Fenner

The power lies in the client if you want to manage and/or annotate pdfs or format references in a document. It would be nice to have that annotation possibility for web based versions of papers I have to admit. That could get very interesting if there was an overlay style thing going on. - Cameron Neylon

+1 Martin - Jennifer Melinn

I like and use webapps for most things I do, but for heavy-duty tasks, it's just not there yet. We're going to have to wait for further developments around HTML5 before a web-based service can give similar performance to a desktop app, and we'd should also consider how we can serve everyone as opposed to just those who live where broadband access is easy-to-get. I had hopes for Wave (and I still have hopes for the Wave protocol). - Mr. Gunn

Seems to me that there are three spaces a reference manager can operate in: 1) desktop (incl. mobile), 2) web, 3) API. Mendeley is trying to cover all three, and it feels like their priority is 1 > 2 > 3 which is understandable. However, if they reordered things 3 > 2 > 1, then things would get a lot more interesting. If the web site and the desktop client were built on the same (open) API, others could expand/enhance the functionality (think of the twitter client ecosystem). I think Mendeley has gained traction through proving a free desktop client that has the functionality many need. The rest of the equation (web + API) isn't nearly as mature. - Roderic Page

I hear ya, Roderic. We've hired new people and are working as hard as we can to fill out the gaps in those areas. Hopefully the API as it matures will allow people to build stuff that we just don't have time to get to (also like Twitter). - Mr. Gunn

Google Docs is good for getting the text written, but for the markup and citations you have to use something else. - Mr. Gunn

Marius, I think online writing tools shouldn't just be copies of desktop word processors with collaborative features thrown in. So it's not just about adding reference management and a few other small things. A scientific paper is a very structured document and the writing tool should reflect that instead of offering font sizes, line heights and similar formatting options that have no meaning for a scientific manuscript. - Martin Fenner

Roderic, I would love to see reference management being built around APIs. But everybody would want to become the Twitter or Facebook of this API ecosystem, including Mendeley, Scopus and Web of Science. - Martin Fenner

I don't think there is a market for an affordable desktop application that understands the structure of a scientific document, runs on Windows, Mac and Linux, does reference management, and has collaborative features. Remember Publicon from Wolfram Research? So I think that this scientific writing tool has to be web-based, and HTML5 should be good and fast enough for it. Lemon8-XML is a good start http://pkp.sfu.ca/lemon8 - Martin Fenner

Martin, what a very, very interesting idea. In my undergrad classes, I use "templates" that guide students through writing scientific papers, and students generally like them, but what they'd really like is if I could have "work boxes" next to each bit where they could enter text, hit a button, and poof! Some kind of draft is created. I tried to do that using forms, but do not have the programming skills to create such forms (or, get them to go the right party -- that takes scripting knowledge I just don't have). - Mickey Schafer

Martin, the trick is who will do it well. Web of Science & Scopus will be too bound to their commercial interests to do it well, and have no community. Maybe I'm an outlier but I don't use either of these tools. Google is my friend. Mendeley is well positioned to do something in this area, but they'd need to deal with the open data issue, which so far they've not really addressed. - Roderic Page

Mickey, take a look at the Microsoft Word Article Authoring Add-in http://blogs.nature.com/mfenner... - Martin Fenner

Roderic, I completely agree. But will be interesting to watch how the new Elsevier SciVerse platform opens up to third-party applications. And then there is Google (which currently doesn't look that interested in scholarly literature) and Microsoft Academic Search. - Martin Fenner

Thanks, Martin. I was not aware of the add-in, though I'm not sure that the undergrads need that kind of power -- but using the template function in Word might work...hmm. This may be too much for my brain on a Friday afternoon! - Mickey Schafer

Just to play devil's advocate a little: I don't think it's a good idea to have a tool impose structure on a ms. My ideal would be much closer to a fast, lightweight text editor that gave good, intuitive control over formatting (then it's easy for me to structure my document for myself) and could interface with a reference manager to handle citations. The idea of using forms or templates to write a paper gives me hives! - Bill Hooker

Martin, I'm not sure how many 3rd party apps will wind up on Sciverse - my impression is that Elsevier considers it a closed system. I would also argue Google is very intererested in scholarly literature (Google Book Search) but not necessarily in schol. research journals. There are journals in GBS but no extra search fields of identifers for articles. I agree with everything else though. - Elizabeth Brown

We will see how 3rd party apps around SciVerse, Mendeley, CiteULike, Endnote and Zotero develop (they all have APIs). Twitter claims that 250.000 apps are using the Twitter API. - Martin Fenner