07 September 2009

In Praise of the Book Sprint

One of the things that I find fascinating about open source is the way it generates epiphenomena - things that don't really happen with conventional computing. Here's another one: the book sprint.

The event is another in the growing body of FLOSS Manuals Book Sprints, kicked off by our first meeting to write a manual for Inkscape. The aim of these sprints is to write a book in 5 days. Actually, we have done it it in shorter time – in February of this year we wrote a 260 page manual introducing newbies to the Command Line in 2 days. Though created quickly, these books are extremely well written texts: comprehensive, readable, and complete.

Needless to say, as well as being about free software, these creations are imbued with its spirit:

A 220 page manual in 5 days - not bad. And it's all free, libre and gratis. Some of the material is also now being translated by the FLOSS Manuals Finnish community, and we hope more translations will follow.

Present at the sprint was myself (Adam Hyde, founder of FLOSS Manuals), Jan Gerber (ffmpeg2theora developer), Jörn Seger (Ogg Tools developer), Holmes Wilson (FSF Campaigns manager) and Theora geeks Susanne Lang and David Kühling. A few popped in remotely to help out, for which we are always grateful – notably Silvia Pfeiffer and Ogg K.

In the end we have free documentation that you can read online, download as a PDF, or log in and improve. It's also available in dead tree format for those who'd like it on their shelf.

Follow me @glynmoody on Twitter or identi.ca.


guy said...

Oops, pet subject this one.

One of the aspects of FOSS that I have most respect for is the folks who do the "unglamorous" bits like documentation and packaging --- the bits that turn a pile of source code into a product: Just because it's FOSS doesn't mean it shouldn't be treated as product and those bits are the *first* things that a user will encounter. Good 'whole product' elements to a project (sorry, I've been reading the marketing books again) will make a user much more sympathetic to a project when they encounter bugs or problems. They make or break the 'FOSS is only free if your time is worthless' argument.

It's particularly impressive that volunteers are prepared to do this --- most of the commercial programmers I know loathe writing documentation, put it off for as long as they can, and write,... well let's just say they could do a bit better.

That said, blasting through documentation like this might be a way of getting something where previously you had nothing, but good documentation can take as much effort as good code and needs the same level of maintenance and testing (a cost which project managers seldom want to admit). Indeed, sometimes you only really think properly about a program when you come to document it, so better to start early.

Glyn Moody said...

@guy: excellent points, I quite agree. Maybe such sprints could form the basis of later refinements, in the standard "release early, release often" way...