OmegaT 2.0.2 Beta / 1.8.1.03

OmegaT is a free translation memory application written in Java. OmegaT is a tool intended for professional translators.
OmegaT does not translate for you! (Software that does this is called "machine translation", and you will have to look elsewhere for it.)

Main features:
- Fuzzy matching.
- Match propagation.
- Simultaneous processing of multiple-file projects.
- Simultaneous use of multiple translation memories.
- External glossaries.
- Document file formats: OpenOffice.org/StarOffice (the latter has excellent conversion filters for Excel, MS Word, RTF), plain text, HTML.
- Unicode (UTF-8) support: can be used with non-Latin alphabets.
- Compatible with other translation memory applications (TMX Level 1).

System requirements:
-

Enhancements
Implemented requests:
- Support for alpha-3 language codes of ISO 639-2 and 639-3.
- ResX file filter.
- Display all the instances of a segment in searches.
- Polish segmentation rules.
- Dutch segmentation rules.
Other enhancements:
Source and target segments are now, by default, left justified for LTR languages and right justified for RTL languages depending on the languages of the project. It is still possible to switch with Shift+Ctrl+O, which has now three states:
- Justification according to languages (default).
- Justification on the left.
- Justification on the right.
- Ctrl+Page up and Ctrl+Page down go to the beginning/end of a document
- (without entering the segment).
- XLIFF filter: the segmentation of pre-segmented documents (tagged by
- is used.
- Update of some user interface texts for clarification purposes.
- Spanish localisation updated to 2.0.1 (UI, readme, instant start).
Bug Fixes:
- Glossary: unclear display if combined entries contain a ",".
- Bidi wrong justification.
- OmegaT hangs at random.
- Random hangs, exceptions in command line, nothing in log.
- Confused display when RTL segment starts with LTR characters.
- Spellchecker slow down.
- Over matching in segmentation rules.
- Pattern after in Text files segmentation rule contained "s+" instead
- of " +".
- The SkipMeta option in the HTML and XHTML filters was not initialised to
- the default value.