The final day of Open Translation Tools 2007 found us tired yet eager for more. We followed a similar pattern to day 2 with parallel sessions dominating, punctuated by another SpeakGeeking session. By this day, I had a much clearer idea of what I needed (at least in regard to the challenges ahead for the eIFL-FOSS ILS project). Nevertheless it was still a hard choice to decide which sessions to attend.

The first session I joined was on business models in the open content sphere. I thought I might have something to contribute to this discussion since I have a reasonable grasp of business models in the open source world. In fact I'm not sure I was able to contribute all that much. Most of the revenue streams in the open content projects represented were from grants from foundations or trust or government. That, of course, is a worthy model. But it has challenges when it comes to sustainability. The other model we discussed was primarily support/training. I think everyone present sensed that there is a potential publishing business model around open content tantalizingly close. But I didn't feel we were able to flesh it out. We were, however, fortunate to have a couple of entrepreneurs in our group. They kept us tied to the real world. This was useful because in the end the success of a business model is tied to your ability to meet your payroll month in month out.

In the second session I joined up with the software development folks to talk about localization of software. I confess I felt more at home in this group. Every one (which the possible exception of me) was an experienced FOSS developer, comfortable with the tool chain used for localization, and ready to discuss the challenges they were facing. Fortunately I had learned enough on Day 2 of OTT07 to at least follow the conversation. The key challenge for most localizers of software is upstream software developers who have not planned for internationalization from the start. Clearly that ought to be a prerequisite for any FOSS development project (I now understand). But even if the project has designed for future localization efforts, that still won't avoid the inevitable challenge of language itself. What is the best Serbian word for “Inbox”? How do you make an interface that is built around the memes of office culture (e.g. files and folders) accessible to a potential audience where office experience is minimal? Does that matter? Should you incorporate a jargon term from an English interface if there is no simple substitute in the target language? Or does that undermine the natural evolution of your language? Hard questions to be sure.

The SpeedGeeking session on Day 3 was just as exciting as that on Day 2. Again, I will hold off describing all of the relevant projects here. But do expect to find posts on many of them in the coming weeks.

The final parallel session of the event saw us dividing into the original two groups: tool and open content. Contrary to what I had done on Day 1, this time I went with the tools folks. The focus for each group was on practical next steps. What emerged in the tools session was that a number of the developers had discovered possibilities for future development through collaboration with other projects present. The first tentative steps towards these were already taking place. Only the constraints of time, resources, and other commitments can hold back exciting new developments around the corner.

Open Translations Tools 2007 was an excellent event. Everyone participating came with knowledge and experience to share, and found others with knowledge and experiences from which they could benefit. The constant focus on participation (much spoken of elsewhere but rarely achieved) made the days exhausting but highly rewarding. To close out the day, Gunner asked us all to look around the room and really see each participant there. Definitely a group I was honoured to be a part of.