I’m pleased to tell you that bkkeepr, my project to create a Last.fm-alike for reading (and more besides) now has an API.

An Application Programming Interface (API) is essentially a machine-readable version of an application, and more specifically, the data in contains. bkkeepr is first and foremost an application that does stuff with data, and bkkeepr.com is the human-readable version of that application. What an API does is allow third parties to build small applications, widgets and so on that utilise that data in new and different ways. (This is another post, but I pretty much believe that everything should have an API. And not just everything on the web. E.G.)

bkkeepr is itself built on Twitter‘s API, and uses data from a wide variety of other web services, including LibraryThing, Google, Amazon and others, via their APIs. So opening up bkkeepr’s data in turn is something I’ve intended to do from the beginning.

The bkkeepr API is very basic at the moment, allowing you to do two things: get all of a particular reader’s reading data, and get all the reading data about a particular book. There’s more information about it here.

The most important function of the API, to my mind, is that it gives users control over their own data. It’s not locked up in a database over which they have no control, and they can pull it out and store it elsewhere any time they like. I’ll say that again: it’s their data, not mine, and they should have control.

It also gives people the opportunity to build cool things. Here are some examples, some dull, some fun, that I’ve been thinking about:

• bkkeepr ‘bestsellers’ – hot books, favourite books, the most read. Charts, and suchlike.
• Find a friend – who’s reading similar things? I like this particularly as it would allow you to branch back to Twitter – or any other service – and find new people with similar interests there.
• Better widgets (because frankly, the current bkkeepr badge is pretty basic).
• Reading speed – silly, because I don’t think the speed at which you read means anything, as long as you read at all, but, like Dopplr’s personal velocity, it’s a neat data toy, and could be implemented fairly easily with book page numbers from Amazon’s Associates API.
• Pretty stuff with covers and calendars (in the spirit of the awesome LastGraph).
• I’d love to see booksites implement a ‘who’s reading this now?’ widget, but realistically I think bkkeepr needs a few more users to see that happen…

There are a tonne of possibilities, and I’d love to see people do interesting stuff. bkkeepr currently has just over 500 users – not many, but I’m hoping adding features like the API and its results will draw more in. To those who say that an API is just a way to outsource the development of an application to those with more time on their hands, I say: yes. Yes, it is. Have fun.

Flickr and many other services uses simple tagging to provide metadata around their content, but this system offers much more: additional content, outside the original system, curated by users, adding information back into the system.

I’d love to see a system like this for books. I search google and wikipedia all the time for additional information on things I discover between paper pages – imagine if this information could be aggregated and linked back to the original book, just like Adobe’s system. Googling dementia praecox from p. 31 of Eric Stanley Gardner’s The case of the rolling bones takes me to Wikipedia’s definition and further background reading on ataxia. Tagging these pages in del.icio.us or similar with not only information about them (ataxia, mentalillness) but why I arrrived at them (literaryreference, ericstanleygardner, perrymason, thecaseoftherollingbones) creates a network of metadata around the book which could be accessed by an ereader – or cross-referenced with other texts to create indexes of mental illness references in literature, medical references in crime novels, and so on.

The joy of this system is that it does not rely on the publisher and the reader agreeing on what’s important information in the book – publishers can still create indexes and concordances to their work, but readers can create and share their own indexes – so a mental health practitioners’ index to Perry Mason would contain differently weighted information to a policeman’s, for example. As with many of these ideas, non-fiction books would probably benefit from this much more than novels – can you imagine a cookbook where you got access to other readers’ researches as well as the authors and your own? – but I like to put fiction through these things too…

Sorry it’s been quiet around here. With London Lit Plus in full swing for the last couple of weeks, and a new job, it’s been a little hectic. However, we do have one important announcement.

Lit+ (litplus.com) is a new booktwo.org project: taking the London Lit Plus ethos – an open-access, distributed literary festival – and turning it into a template that anyone can use to set up their own festival. We’ll be using the same kind of tools – the power of the internet and free software – to create a resource for all.

We’ve already had plenty of outside interest in London Lit Plus and we want to use the momentum to build new and exciting literary cultures. We’ll need your help, so stay tuned.

Images courtesy of Yaniv Golan and Robert Brook, via Creative Commons.

Firstly, it became clear to the steering committee that they had to relinquish all control of the project to the community in order for the community to flourish. There can be no half-measures in crowd-sourcing: you either let people do exactly what they want, or they won’t do it. This was demanded by the users, and the original editorial board had no choice to go along with it. In the end, they realised that this did energise the project.

However, freeing up the community also meant that the final book was not acceptable on delivery: “To yield an acceptable business book, it would be necessary to hire an accomplished professional author who would also handle the fact checking process.” This is not that unexpected, but it is a problem – particularly if that editor has to negotiate edits with X hundred authors…

There are some other insights as well – not least that the originally intended participants, tenured professors at Wharton and MIT, refused to participate, and that this didn’t make any difference in the long run – that make the Gilbane report worth reading in full. But it’s interesting to compare too with the Million Penguins project, as the final reports on that make pretty much the same observation: the journey was more interesting than the destination. It doesn’t matter whether you’re writing a business book or a novel; for the participants, the act of creation is more important than the end result.

Is this, then, the central quality of the wikibook? Crowdwriting – or many-to-many publishing, as Gilbane puts it – is inherently selfish. It’s not in the individual author’s best interest to make their shard integrate well with others – in fact, the opposite may be true. Altruism exists, but it’s balanced with self-promotion, of one kind or another. But perhaps we should pick apart that phrase, ‘many-to-many publishing’ – something there resists the urge to put an end to the endeavour. Like Wikipedia, is a wikibook – by its very nature – permanently unfinished? One of the core perceived attributes of book 1.0 is that it represents the “one true version” – yet many of our most culturally important books – think of the Bible, or the works of Shakespeare – exist in multiple versions.

Perhaps, if we are to understand the wikibook, we need to place it in the context of mythical texts, like the Bible and other religious and historical works. They share the same core attributes: multiple authors, disputed authorship, multiple versions, endless potential versions, authors and versions distributed across time and space and filterable by the reader/editor’s prejudices. The only thing we can do is add metadata to aid historians, tracking changes and creating concordances.

With the need for a “one true version” removed, we promote the reader to editor, and the relationship graph becomes truly many-to-many, instead of passing through the editorial bottleneck. Combine this with innovative licensing which allows for-profit publication of remixed text (CC3.0 now available), and you have the seeds of a new literary culture…

Ficlets is a new site for authoring CC-licensed text snippets which others can play with. It’s pretty cool, and what’s more amazing is it’s come out of AOL. It’s not dissimilar to Yarn, which I mentioned earlier:

ficlets are shorter than short stories. Well, no, actually, they are short stories, but they’re really short stories. Really short, as in there’s not a maximum word count … there’s actually a maximum character count (1,024). There is also a minimum character count, and the number of that beast is 64.

If you wish, we’ll provide you with inspiration (photos, themes, suggested beginnings and endings, even other ficlets), but you’re completely free to blaze your own trail. Now, here’s where the real fun comes in: Each and every ficlet is modular in that, though you may have written a stand-alone story with a beginning, middle, and ending, your fellow ficleteers may choose to write a prequel or sequel to your story. In this respect, you can think of ficlets as literary Legos.

All ficlets are covered under Creative Commons, which means that if you wrote it, you own it. Period.

To give you an idea of what you can do with 1,024 characters, that is the exact length of this “About Ficlets” description

They had to leave a period off the end there, but you get the idea. Ficlets is beautifully implemented and easy to use. There’s also a strange, cool imagination at work here – the ‘Inspiration‘ link pulls random photo sets out of Flickr to spark your creativity, as well as an ‘on this day’ snippet from The History Channel, and some random lines. It’s a beautifully simple mash-up that works very well.

I can’t help thinking of a Yahoo Pipes type GUI to stitch them together into persistent, save-able stories – the current implementation is more of a choose-your-ow-adventure deal, and I don’t have an AOL screen name to see exactly what AIMShare does (OpenID AOL?). But I’ll still be wasting plenty of time there…

[UPDATE: Thanks to Jason Garber in the comments for pointing out that Ficlets does take OpenID. Nice one, guys.]

Don’t you love it when you think of something really cool, but you don’t have the skills to make it happen – and then you find out someone already has?

Back in October of last year, I suggested a couple of the projects that I’d like to see Booktwo build. One of these was Exquisite Corpus, an updated take on the old parlour game, Exquisite Corpse, where players took it in turn to add to a drawing or story created by the previous player. Sadly, we never managed to implement this.

However, we were very pleased to stumble upon Yarn, a collaborative story-writing tool created by the folks at The Daily Jolt, a US College network (and is somewhat proprietary for that, but hey). Yarn allows people to start their own stories, or add to those created by others, and they can also splice off at any point from existing stories to create choose-your-own-adventure-type tales. Go have a play.

Yarn isn’t presented as a cutting-edge literary tool, but it has the potential to be one. The other project we suggested late last year was Infintie Entries, a distributed wiki-novel. This was fulfilled up to a point by Penguins’ Million Penguins project, but the latter failed – in as much as it became unmanageable and was eventually euthanised – because it tried to synthesise a single voice – one version of the truth – from a wealth of disparate voices. If the wiki- or massively-authored novel is to succeed, it needs to embrace dissonance and non-linearity, it needs to take the path mapped out by avant-garde writers like J. G. Ballard and William Burroughs. Yarn, cutesy though it is, provides one possible framework to make this happen.

Nevertheless, it’s impressive that Penguin have stuck to it, and not thrown their hands in the air when the going got sticky – then again, they haven’t been showered with goatse images either. Instead, they’ve instigated a number of techniques, such as locking the wiki for a few hours each day to catch their breath, without interrupting the flow of the project. Preliminary results suggest that the open wiki as it stands is not the best vehicle for such an endeavour, but there’s no reason it wouldn’t work for smaller-scale projects – as indeed, projects like We>Me seems to suggest. The collaborative novel is off to a rocky start, but it’s not over yet.

This morning, Penguin announced the launch of A Million Penguins, a wikinovel project in association with De Montfort University.

Students from De Montfort’s MA in creative writing form the basis of the projected community of writers, which will edit and expand upon the short first chapter provided over a period of six weeks. (I think six weeks – the timescale is a little unclear. Rather sweetly, they’ve left lots of setup notes on their blog, such as the inspiration gained from this Lost fan wiki.) The students will also form the core moderators of the project, which should help prevent edit wars.

The project is another brainchild of Penguin’s Digital Publisher Jeremy Ettinghausen, who’s also behind Penguin’s extensive Second Life presence and other forward-thinking projects. Viking editor Jon (no surname given) will be guiding the project, acting as a regular editor, giving ideas on direction and revision. Knowing what editing one author’s work is like, I don’t envy the job of doing the same for a potentially massive authorbase (we need some new mass nouns here). He’s certainly open-minded about the end product, as long as it doesn’t turn out to be a “robotic – zombie – assassins – against – African – ninjas – in – space – narrated – by – a – Papal – Tiara type of thing”. Shame.

The Guardian obviously obviously got the wrong end of the stick as it reports that “Ettinghausen is emphatic that the experiment has not been set up by Penguin as an online literary Pop Idol,” and Editor John also stresses that “the wikinovel experiment is not a place to prove to Penguin we should publish your book.” This is clearly more in the spirit of the networked book (or, dare we say it, Booktwo’s own, terminally alpha projects), and it seems unlikely that Penguin will get a novel out of it that they would consider publishing in the mass market – “To be honest, we don’t know exactly what is going to happen or how this will turn out”- all the more kudos to them for trying it, then.

I’ll be keeping a regular eye on the project, and trying to make sense of it as it evolves. I’m still looking forward to the network novel that won’t play merely with authorship but with structure too; that explores the potential of technology to change not just how novels are written, but how they are read. In the end, the product of A Million Penguins is unlikely to look very different to a regular novel – but it’s a great place to start.

We-think is about the power of mass creativity, charting the rise of mass, participative approaches to innovation from science and open source software, to computer games and political campaigning. The website interface does not make dipping into the book very easy, sticking to a more linear style – there’s also a Wiki version – but it does make commenting on it very easy. We particularly like the fact that commentators seem to be line editing the book as well as commenting on its content. Oh, if people would that for my books.