OmegaT 3.0.5 Review
November 1st, 2013, Roman Mironov
OmegaT release 3.0.5 has quite a few improvements that definitely deserve a closer look. This post discusses productivity gains from the new features.
I already touched upon this in detail, so I will refer you to that article and just summarize the main points here. It is now possible to have two types of glossaries in a project: a project-specific glossary and a general one. The Glossary pane displays terms from both. Although you could do the same in previous versions, you would not know which of the two glossaries a term is coming from. Translators could not easily choose the term that was right for their current project. Now, as terms from a project-specific glossary are highlighted in the Glossary pane, translators know they have priority for this project. And they still benefit from the big collection of terms in the general glossary. This allows reusing translated terms across projects more effectively.
The auto-completer can now display the source terms (Options => Auto-completion => Glossary => Display source terms), making it easier to distinguish between term suggestions. Two more views have been added as well:
- The first one is Auto-text (Options => Auto-completion => Auto-text). It allows creating short abbreviations, e.g. “asap,” that OmegaT will turn into full words upon selection, “as soon as possible” in this example. Right now, I do not see any benefits, because its functionality is the same as programs like Autohotkey provide. But whereas in OmegaT this functionality is limited to one program, Autohotkey makes is available system-wide, allowing me to create one “asap” entry and use it across all programs.
- The second view is Character table (Options => Auto-completion => Character table). It allows inserting into a translation various characters not found on a keyboard such as © or £. This one will definitely come in handy.
To move between the views after you have opened the Auto-completer (Ctrl+Space), press Ctrl+Page Down or Ctrl+Page Up.
MyMemory machine translator
OmegaT can now directly access human-created TMs made available by MyMemory (Options => Machine Translate => MyMemory) in addition to the machine translation output provided by the same service. In theory, this is very promising, since human translations matching your current original are of course valuable. In practice, the value of these translations is still unclear. I tested this service with English to Russian translations and found the results to be mostly unusable. For example:
Original: Be sure the System Monitor clock is correct.
MyMemory match: Убедитесь в том, чтобы заказать правильный запасных частей. (two grammar errors)
Looks like machine translation output, right? Of course, it would be ridiculous to expect perfect matches each time, but I do not expect such obvious errors in a human translation either.
Merging smaller TMX entries into big ones “on-the-fly” to match the current segment
For some time now, OmegaT has been able to analyze a TMX for big entries consisting of several sentences and attempt to break the big entry into smaller ones as you open a segment. As a result, it can match the opened segment with the one in the TMX, even though there is no one-to-one match in the TMX (there is just one big segment). With the release of 3.0.5, OmegaT can do it the other way around. When you disable Enable sentence-level segmenting in the Project Properties, OmegaT will attempt to merge smaller entries in the TMX into larger entries to match your current big segment opened in the Editor.
Current segment: She looked away. He sighed.
Two entries in an English to Russian translation memory:
1st entry: She looked away. Она отвернулась.
2nd entry: He sighed. Он вздохнул.
OmegaT assembles these two matches into one and presents you with this assembled match:
Original: She looked away. He sighed.
Translation: Она отвернулась. Он вздохнул.
This may sound complicated, but the good news is it happens behind the scenes, so you do not have to think about it at all.
The match statistics feature has been improved. Previously, you could have a lot of fuzzy matches displaying in the Fuzzy Matches pane, but much fewer matches in the match statistics. The match statistics have been made consistent with what displays in the Fuzzy Matches pane.
Version 3.0.5 is another excellent release, adding useful features to what is already a translator productivity powerhouse. For an open-source software, the overall progress over the last two years is stunning. For more information about OmegaT, check our posts about other functions offered by this translation memory program.