I've done a bit of work looking into foss tools for other projects, and as far as document translation is concerned, OmegaT+ is a pretty decent starting point. However it's a desktop application, the translation base is managed locally, the parser is limited, and the fuzzy matching is sketchy. It also requires OOo, or a document base standardized on .odt files. However, with open document format, we are talking about xml, and that's easily transformed. On the web side, I've only seen .po editing systems and multilingual wikis. Can't go wrong with mediawiki for user contributed content, but for translations it's not a complete solution. OmegaT+ has a decent user interface for a document workflow, but again, version control isn't integrated. You'd kind of hope to see a versioned source document structure, without presuming a single language source document base, with branches for translations, and then an auditing tool that's as nice as some of the application localization tools for completeness metrics. I wouldn't try to get away from svn for document revisions, there always needs to be a versioned, canonical, source language document or there's no way to come up with metrics for the translations. I'm going to guess that aside from software localization, a lot of the stuff in the wild is ad hoc, as with your wikipedias, et al. My $0.02 ~David On Thu, Jun 4, 2009 at 5:35 PM, Matt Madia <mattmadia@xxxxxxxxx> wrote: > > Does anyone feel like researching how other online communities handle > website translations? > Surely, there's a fairly elegant solution out there. > > --mmadia > >