Syndicate content
  • English
  • Deutsch

Eclipse translations, Babel, FOLT and openTMS

Babel is the Eclipse translation project. But there is lot more going on in the translation area, in Japan at the Eclipse Japan WG and at FOLT here in Germany. The interesting thing about this is that it's all open-source, and all EPL - a huge potential for mutual benefits.

Let me first report on FOLT.

Yesterday I was the regular FOLT meeting in Sindelfingen. FOLT is the Forum Open Language Tools, an informal industry group of companies from the translation domain. They got together to create an open-source toolset for professional translators, translation agencies and organizations that translate on a large scale. Their plan in includes a translation memory server, import and export filters and web/desktop clients. All is based on standards, XLIFF, TMX etc.

FOLT has now made a first version of their translation server available. The first cut is still a bit rough. For convenience, the is a Ubuntu virtual machine available for download where everything is pre-configured. However, that's not recommended for productive use.
The download is at http://opentms.de/. Don't get confused by the names, FOLT is the group of companies and openTMS is the name of the toolset.
The code is, as mentioned above, available under the EPL. It is hosted at SourceForge, http://sourceforge.net/projects/open-tms/ The code base in Subversion seem a bit unstructured to me. I could not figure out how to get it to work in Eclipse. It needs more documentation and instructions how to set up a developer workspace.

What's in it? The first version of the translation server let's you translate an XLIFF file. It has an XMLRCP interface. At the meeting yesterday, a demo was shown how the server can be integrated into a CMS using the XMLRCP interface. There is not much else there at the moment, the import/export filters and the user interface are on the roadmap though. After all, it's open-source and contributions are welcome.

The news on their website are a bit outdated but the FOLT newsletter archive will give you an overview of recent activities and events.

The second interesting group working on translation tools is the Eclipse Japan Working Group. At EclipseCon 2009 they presented, together with Babel, what they are working on at the moment. They plan to contribute their work to the Babel project this summer. The slides of their presentation can be found at the EclipseCon 2009 website, http://www.eclipsecon.org/2009/sessions?id=504 The photo below is from that session.

Basically, they are doing the same as FOLT. Yes! Japan is far away and it is hard to meet. Japan is a very export-oriented economy, very much like Germany. No wonder, that translators are longing for good tools. For the Japanese efforts, the roadmap looks pretty similar to FOLT's: a translation memory server, import/export filters, user interface. Their goals also include the translation workflow/process which FOLT is currently not addressing.

The difference seems to be that they have an Eclipse background from localizing Eclipse itself to Japanese. This also shows also in their approach to the user interface which is, of course, Eclipse-based.

Here you see a photo from EclipseCon where Motoki MORI presents the UI of their translation tools. It shows an XLIFF editor in the middle, a tree of projects and files on the left, matches from the translation memory on the upper right, glossary below that, and validation messages at the bottom.

EclipseCon 2009EclipseCon 2009

This tool is not available yet but it is planned to be included in the Babel project in summer 2009.

What's the point? There are two industry groups working on open-source translation tools, in Japan and in Germany. Both are using the EPL. FOLT has a strong support by local companies but still not enough resources to build all by themselves. Or, it just takes more time.

Who would be interested in two different EPL translation memories if one would be enough? What about joining effort so that the available resources can be focused on the remaining large tasks like filters and a state-of-the art user interface?