Recently, I have been playing around with the various tools and services available to people who want to store and share their bibliographies and research materials online. Such services are currently blossoming, and for a quick overview of what’s out there, I’d suggest starting with this review written by Eugene Barsky at the University of British Columbia.
My aim in all of this was to explore the options for creating a hub where people interested in emerging technologies and their implications for society could share and discuss scholarly materials – books, book chapters, academic papers, lecture videos etc. I think that what I had in mind was a bit like PLoS Hubs. As PLoS write in their blog post introducing the PLoS hubs biodiversity project:
“The vision behind the creation of PLoS Hubs is to show how open-access literature can be reused and reorganized, filtered, and assessed to enable the exchange of research, opinion, and data between community members.”
So, were any of the existing tools and services powerful enough to mirror the functionality offered by PLoS Hubs? I looked at Mendeley, Zotero, Wizfolio, Scribd and Connotea. The answer I’ve come to so far is… not quite.
That’s not surprising. Note that PLoS Hubs works exclusively with open access literature, which means they can import whole journal articles to their platform without violating copyright. By contrast, a lot of the material I want to highlight on my hub is protected by copyright. So although I can easily collate links to materials, visitors to my hub are going to have to go back out to the web to read some of the material I’m highlighting, and that might leave them less inclined to return back to start a conversation about it. That’s without even going into the tricky subject of whether they benefit from access to the same institutional subscriptions I do: I could be pointing to materials that, for them, are locked behind a subscription paywall they have no credentials to vault.
Of course, I could always break copyright law: most of the platforms I tested (with the exception of Connotea) have facilities to upload and share files direct from the platform, with some element of free storage. The platforms are wise to this, and employ various strategies to dissuade their customers from breaking copyright terms, from showing prim pop-up windows (that I presume they hope have the effect of transferring the liability for the infringement from them to you) when you upload a file (Zotero, Scribd) or opt to share it (Wizfolio) to only allowing you to upload files through a convoluted process involving a desktop client (Mendeley).
Setting the problem of copyright aside for now, I discounted two of the services right off the starter blocks: Scribd because I really dislike their reading pane and because it was fully public, with no option to share references among a private group; and Connotea because it had no option to upload files even if they were files you were permitted to upload. I then set about testing the remaining three products – Wizfolio, Mendeley and Zotero. One of the ways I used to test them was in collating a bibliography from Gabriella Coleman’s fantastic contribution to the Atlantic syllabus series “The Anthropology of Hackers“.
Wizfolio
Why did I end up starting with Wizfolio? I think it’s because it has the fastest and cleanest interface. Perhaps I was exposed to too many FileMaker Pro-based database tools at an impressionable age, but the forms Wizfolio uses just felt right. I might not have thought that had I not had AdBlock running on Chrome – the service is ad supported, and if you look at it without AdBlock on, it’s a totally different experience.
Wizfolio’s web button “WizAdd” also performed best out of the three products, picking up that I wanted to save a journal article and automatically fetching all the right details when I visited a web page like this which listed it, where the other two products’ web buttons thought I wanted to save a web page.
Wizfolio’s auto-complete for bibliographic detail was also the most efficient, even though it kept trying to locate the titles of the papers I was saving in PubMed. Adding books to the bibliography was a pleasure, it took less than a second after I’d inputted the ISBN number before Wizfolio had found it through Google Book search and could auto-complete the rest of the details for me, as well as supply a jacket picture and link to a preview on Google Books. Sweet.
Unlike Mendeley, Wizfolio is totally web based. This is good in one sense: the time Mendeley spends syncing your desktop library with your web library is probably best spent on other things. But it’s bad in another sense: any collection you build up of actual files could be gone the moment Wizfolio decides to withdraw its service from you, so you ideally need to be maintaining a desktop mirror in parallel – easy if it’s just you, less easy if you’re building up a library collaboratively.
Finally, there is very little that’s social about Wizfolio. Yes, you can share your bibliographies with colleagues, and you can even make them public. But I’m not sure what public means in this context: Wizfolio didn’t generate a url for the collection, so I’m forced to provide only an image of the final result:
Notifications about the activities of your colleagues online were delivered in Windows Messenger-style floating pop ups, and no record of social activity was maintained. Colleagues were unable to leave notes on papers you had shared with them to start a conversation about the text. So although Wizfolio scored highly on user experience making and maintaining bibliographies, it was all a bit 1.0. It was clear I was going to have to look elsewhere to scratch the social itch. Luckily, Wizfolio let me export the references I’d inputted as a .RIS file, which I could then input to the other two services I was testing.
Mendeley
I’ve alluded already to a lot of what I didn’t like about Mendeley – the baggy desktop client, the poor web import button. Although I’ve said I didn’t rate the auto-complete functionality for bibliographic details, that’s more about usability than it is about performance – the interface is just unpleasant. I simply don’t think I could get used to the separation of powers between the desktop client and the web instance of my library. But I’m not going to give up. I’m attending this talk from the Mendeley people on Friday, so maybe I’ll get a better feel for how to use the product there.
What I initially loved about Mendeley, though, was how social it felt. I spent a good bit of time filling in my profile details – a sure sign I thought I was going to make some friends. The interface is very Facebook. And I admit it, the Ajax felt good after an hour-or-so’s form-filling at Wizfolio.
After I’d imported the .RIS file from Wizfolio to the Mendeley desktop client, I added the items in it to a new public group library called “The Anthropology of Hackers”. I then synced the desktop client with the web client, went off and made myself a cup of tea, and came back to see how things were going. The result is this (click the picture to go through to the library and have a look around):
You can see that the references haven’t come through all that cleanly, but that’s probably my fault as much as Mendeley’s – I wasn’t all too thorough when I inputted them into Wizfolio. What’s really annoying, however, is the weakness of the discussion facility. Each paper has its own page with a url, but there’s no comment facility so you can’t start talking about what’s in there. You can comment on the status update displayed on the Overview page (pictured), where it shows you I’ve uploaded some references, but for me that’s not the best place to be starting a conversation, especially as the upload messages are presented in batches. What do I want? I’m not sure. A forum? Comments under the papers? Paper fan pages? I don’t know what it is but I know I want something more from a site that looks this social.
Zotero
I looked at Zotero last because it’s Firefox-only, and Firefox is no longer my browser of choice thanks to how baggy it’s got over the past few years. Like Mendeley (and, by the way, unlike Connotea) it also had no trouble reading the .RIS file from Wizfolio. It also had a number of other features I liked. For a start, unlike Mendeley, it let me create sub-folders in my shared collections, and didn’t restrict the number of people with whom I could share a private collection (with Mendeley, the limit is 10).
Zotero also allowed me to export my bibliographies in a wide variety of formats. Whereas Mendeley would only offer me a Word plugin or a BibTeX export, and Wizfolio restricted itself to exporting a .RIS file, Zotero would let me get at my data in .RDF, MODS, BibTeX, BibIX, .RIS – even as a Wikipedia citation template. This is exactly the kind of good practice you would expect from a FLOSS project, which is what Zotero is. Zotero will also produce straight text bibliographies that conform to a range of bibliographic conventions, to cut and paste wherever you fancy.
As far as social features are concerned, Zotero does have some good ones, although they were hard to find initially. As well as providing a permanent url for your various collections which is publicly accessible, you can create groups with whom you can share your collections, and you can even start discussion threads on the group pages (horray!). Here’s a publicly accessible (but closed membership) group I created as a test, with the Anthropology of Hackers bibliography listed as a collection (again, click the picture to go through to the library and have a look around):
Although it’s not as pretty as Mendeley, then, Zotero does appear to have more smarts. And of the three tools I looked at I think it comes out (so far) as the favourite contender for creating my hub.
Social bibliographies and collaborative reading: I’m doing it wrong
If you’ve played a part in writing any of the tools I’ve talked about in this post, or if you’re a protective power-user, please don’t feel aggrieved if I’ve got it wrong, and please feel free to leave guidance in the comments.
Although I’ve now spent quite a while in the company of these three tools, I feel like I’ve barely scratched the surface, and I certainly am not ready to call a winner. For one thing, it feels like most of these tools are still in development. I can imagine Mendeley will be working up some more engaging social features to live up to its compelling feel in the near future. I suspect Wizfolio will continue to integrate with professional journals to take its auto-complete functionality further towards the seamless end of the spectrum.
Until I’m clear what I want to achieve with this hub (who knows, I might turn out to be looking for a solution to a problem that doesn’t exist) I’m going to keep experimenting with all three of these tools and whatever else comes my way. And I’ll try and keep this blog updated with what I find.
The 27th Chaos Computer Congress, which wound up yesterday in Berlin, was predictably fantastic. It’s the second time I’ve elected to flee the bosom of my family the day after Christmas to spend a few days in a basement with the world’s nicest hackers, and I have no regrets. This year’s congress not only had more snow (contributing to a not unpleasant feeling that I might be stranded in Chaos forever) but also, thanks to Wikileaks, better journalists AND better spies. Here are my highlights:
The Barefoot Technologist 