19 March 2001
My subject, the "internationalization" of OpenOffice.org, derives from an ongoing debate over issues whose importance to Open Source projects demands substantive treatment. I want here, then, to briefly sketch out the argument that has been engaging participants on the discuss list of OpenOffice.org regarding the creation of versions of OpenOffice.org mailing lists (and even of the site) in various "major" languages. If my terms are a little coy, thats because there has not been so far (as of Friday, 16 March), any certain definition of what such a process of internationalization would entail, should it even go forward. But for those who want to read more on the issue, I urge you to refer to the extensive thread in the discuss mailing list.
Some background is in order to understand some of the issues that were raised by the discussions. OpenOffice.org traces its origins to the German software developed by the German company Star Division, and most of its Sun developers are German and speak German. Star Division was bought by the U.S.'s Sun Microsystems two years ago. Today, work on OpenOffice.org software is conducted primarily, if not exclusively, in English. That work is not done entirely in the U.S. It is in fact global, with the support staff located both in the San Francisco Bay Area and Ireland, and developers in Germany and throughout the world. From this perspective, then, it would seem that OpenOffice.org is anything but provincial; that it is, in fact, very international indeed.
The very fact that OpenOffice.org is so international raises some questions: Why is English (and not German, say) the predominant language? Does the de facto predominance of English in the website and in the mailing lists help or hinder the growth of the community? And, would it be reasonable (and feasible) to set up special mailing lists and website sections that would allow non-English speaking developers, users, and the merely curious to exchange ideas (if not necessarily code) efficiently?
As Guy Capra of the French site OpenOffice.fr (an independent shadow of OpenOffice.org) has argued, the advantage of linguistically diversifying OpenOffice.org content is that more developers might wish to join the community and participate in developing the code. According to this argument, not only would they find it easier to communicate amongst themselves in their native language but they would also, perhaps, feel less of a resentment to participating in yet another English-only project. Developers would not only feel more inclined to contribute code without feeling that they had to enter an English-only community but would be more inclined to contribute to a project that was more "theirs." The project would thus grow as it linguistically diversified.
But the opposition to this argument is quite strong and has been most precisely expressed by Sun's Bill Roth and Dietrich Schulten. Any dynamic open-source project always veers towards the chaotic and needs an organizing principle in order to stay focused. Otherwise, any project risks not just a centrifugal disintegration, but a forking of content (and perhaps even code) that ultimately diminishes the value of the anticipated product. Thus, if OpenOffice.org were to create satellite or parallel sites in various languages, the project as a whole might suffer because developers may either cease crossing linguistic gaps to communicate with other developers (French developers may not bother to communicate with, say, German developers)which is a fairly unlikely scenarioor there might just be the specter of a terrible confusion, as redundant lists and redundant messages in various languages start proliferating. And, for those developers working in a language other than English, they could use one of the Web's translating services (such as Alta Vista's BabelFish) to translate messages into English and back into their chosen tongue.
I would like to stop here this week. As I mentioned above, the discussion is still ongoing; Bill Roth, in fact, enabled a site where opinions on the issue can be polled. Next week, I will resume this debate and examine how other open-source projects have dealt with this problem (if at all). If you wish to discuss this, please feel free (in fact, feel encouraged) to post your responses to the OpenOffice.org discuss list. To post, you will have to subscribe. And we have made that part very easy.
9 March 2001 Creating the New Open Source
1 March 2001 Interview: Sander Vesik
22 February 2001 Allchin's Demagoguery, by Bill Roth, guest contributor
15 February 2001 Interview with Wilfredo Sánchez
9 February 2001 Organizing Open Source
1 February 2001 Open Source and Its Culture
23 January 2001 Community Action
16 January 2001 Quo Vadis OpenOffice.org?
9 January 2001 The 613 build: problems and opportunities
3 January 2001 Sun's open door