‘Booksprint’ is a working model for collaborative book authorship, and has since inspired a FLOSS tool called Booki. Its genius is that it takes the proven model of the ‘codesprint’ that open source software communities have used so successfully – to develop huge amounts of code in single intense bursts of focussed collective labour over one week of living, thinking, and working together – and applies it directly to book production. Wireless and F/OSS geek and grassroots technology generalist Thomas Krag introduced Booksprint to the Open Source Publishing Tools workshop as an inverse story about the matter-ing of publishing: “..an outsider’s view of this whole book thing.”
Moving On to Bookness
The Booksprint idea came about while Kraag was working with wire.less.dk, a non-profit he co-founded dedicated to establishing internet infrastructure using open wireless technologies in developing countries. Their company model – “two Danish geeks travelling the world” – was not at all scalable to the wireless networking they wanted to see developing. Limited attention was being given to existing manuals and didactic wikis (“it never occurred to us to ask why anyone else would use our modules when we never read anyone else’s modules”) while the labour involved in ongoing “teach the teacher” sessions was unrealistic. The net at that time in some parts of Africa was also so slow as to not handle simple file downloads of didactic materials. What seemed necessary was a singular authoritative book. It was not possible to put one together from existing quality published material because the book had to be on a free creative commons license, legally open to any translator who wished to translate it, and most importantly, it needed to be able to be legally re-sold locally, so that translators could benefit from their investment in translating it. Kraag realized he was not prepared for the task…
I didn’t want to write a book for 18 months, because at heart I’m pretty lazy, so I called a bunch of friends working on wireless networks that were already coming to a conference in London: “Can we stick around for a week afterwards and write this book?” Some of them said, “Are you crazy?” The reason I thought we could do it is through the existing production model of the codesprint. We sit together and it increases our efficiency. So I found 5000K which was enough for tickets, and to pay someone to go on holidays to Morocco and leave his house to us for a week.
The Conversion of a Genre of Text Labouring into Software Tools
The text still needed to be edited for 6 months, but ‘Wireless Networking in the Developing World’, has had 2 million downloads since and is in its 2nd edition. The process was tedious, using emailed and cut and paste files, proprietary software, and open source outputs. Better tools have developed since 2009 when Adam Hyde began using booksprint for FLOSS manuals and has since fully developed the booksprint model in to Bookie software, a robust and customised collaborative authoring tool. ‘How to Bypass Internet Censorship’ was written with Bookie directly in to the browser, with the crew pressing the “publish” button on the 7th day.
It doesn’t really matter if its not completely polished… it still feels great… The first day you write the table of contents and the index… on friday night you upload it to lulu and its done and it feels so good. a week of all nighters is so much more doable than a year of working on something.
Because Booki is open source, you can download it and adapt it as you want. FLOSS’ design has beautifully simple READ and WRITE interfaces and PUBLISH buttons down one side, and a chat section down the right, where you can talk with and share material between other authors writing alongside you remotely. It can handle versioning, tracking authorship (for attribution for different licences), Javi, wiki style sheets that generate html, and a whole bunch of other things. (Check out the van that the Booki crew built for Booki that can drive around and print books!) All the licensing is handled by the site itself and built on Hyde’s own licensing expertise.
Krag has not received government or local government funding for his work and instead relies on philanthropic investment since 2002. In the closing discussion, the huge issue of translation software was raised. Simon Worthington of Mute has done research into this and states there really isn’t much, especially not open source. A rare strong example is Pootle. Krag noted FLOSS manuals exists in 5 languages already and can do split views, but this is an area that needs a lot of work. One of the main problems is that professional translators have very established workflow methods – the bottom line is that have to send to translations in Microsoft Word as standard. Femke from OSP mentioned that the EU’s translation department has incredible tools – but only in Microsoft!! One of the exceptions is the Spanish local governments, which do some very good work with open translation, including machine translation – they have some of the same remits that the EU have which means they have to translate large amounts of government text. This may be somewhere to look for modelling solutions.
Floss Manuals Booki available here.