CAT OmegaT
Most of the things you want to know can be found at http://www.omegat.org/. Here you can download the program and get first instructions. If you are already familiar with the program you may be interested in reading My Problems with OmegaT.
Requirements
OmegaT works on all platforms (Linux, Mac, Windows), but you need
- a working Java environment (included in the package for Windows)
- (recommended) OpenOffice (download and info at http://www.openoffice.org/)
OmegaT understands files in *.txt and *.html (and some other weird extensions, but the best results are achieved in *.odt (the format of OpenOffice).
First steps
There is not much to learn before using OmegaT, but only by time you will see, what the program can and what cannot be achieved with it. You should read through some introductory notes, watch a short video or take a look at screencasts I made with BB Flashback.
In March 2011 I switched to Ubuntu installing version 10.10. Things are slightly different here. You can activate OmegaT using the Software Centre. That would install OmegaT 1.8 (the last stable version), but not with a complete German interface. The alternative is to download the compressed package and unpack it to a folder such as /user/.omegat. In here are two files that can launch OmegaT. One is simply called Omega and the other one is OmegaT.jar. If you make these files executable you can launch the program (currently version 2.2.3_4). To make a short cut on your desktop you can use Nautilus (the file manager), hold CTRL+SHIFT and drag the file with the mouse to the desktop. This, however, only works with OmegaT and leaves you with the option to first make a choice in a pop up window whether you want to execute the program.
Start with
Creating a project in OmegaT
OmegaT organizes your translation work in so-called translation projects. For each project OmegaT will create a set of folders. These are used to store the source documents to be translated, the glossaries you may wish to use, and any translation memories you may want to use. OmegaT also creates a “Target” folder which will contain the final translated documents.
It is recommendable that you create a folder for the projects you create with OmegaT. This may be a folder rooted in one of the main drives (c: etc.) or even on a usb-stick (e:\, f:\ etc.). You can name it anything you like such as "project", "omega", "translations" etc. Likewise you can give your project any name you want. You should save the project(s) in the folder you created and OmegaT will create the folders for the source files, the target files, the translation memory and glossaries you wish to use.
Before you save the "empty" project make sure that you selected the source and target language correctly.
Files to be translated with OmegaT
You can watch the creation of a source file in OpenOffice in the first video on Creating a project in OmegaT. In the video I have copied text into an empty text document, changed the language of the text and saved it in *.odt to a location of my projects.
Tip: If you save the files to be created into the folder source (e.g. ...omega/project/source) you do not have to import it.
If you need more written information take a look at Samuel Murray's introduction.
Translating with OmegaT
The left window of the screen is the place for your translation(s). The segments (sentences) of the source file will already be there (highlighted in green, if you marked that option in "Views"). You may now choose whether you want the original text to be included into the field for translation or whether it should be empty. This is done under
Options > Editing Behaviour...
Since any formatting of text (bold, italic etc.) is marked with so called tags it is advisable to include the original text into the field for translation, because the source and the target have to contain identical tags (otherwise OmegaT might not save your translation).
Sections for translation (and translated text) are highlighted in yellow (if you marked that option in "Views"). Hitting "Enter" will take you to the next segment. Hitting CTRL+Enter will take you one segment back. The same result is achieved with CTRL+N (next) and CTRL+P (previous). Each segment you translated will go into the translation memory. You can, but you do not have to save your translation memories (TM) separately (easiest way to do this is CTRL+S = save). At an interval of 10 minutes the TM will be saved automatically.
So called "fuzzy matches" of text will appear in the upper part of the right window indicating the percentage of identity (> 30%). Again under
Options > Editing Behaviour...
you can specify the percentage of identity for a match to be directly included as translated text (with the prefix you have chosen, it is [Fuzzy] at the beginning). You may want to have a low percentage at the beginning, but can increase that by the time your translation memory gets bigger.
Problem with segments
Details can now be found under My Problems with OmegaT
Advanced Options
There are some advanced options that you may want to use. There are certainly many more options, but these are the ones that I tried out.
Building a glossary
It is not really necessary to build a dictionary, because there is (currently for people without some programming skills) no way to directly include words or phrases out of this glossary. They show up in the lower right part of the window and you would have to copy and paste them into the document. The files that you create as glossaries have to be build with simple Tabs (first tab = source language, second tab = target language, third tab = explanations).
There is an additional difficulty to build glossaries. They have to be saved in a format that I had not seen anywhere else. Glossaries need the extension *.utf8 (this is not meant as the code in which to save the file). You can achieve this (window users) by using Notepad (Editor in German) and when you save the file opting for "all files" as the type of file and then type the full name with the extension (e.g. mydictionary.utf8).
You can also use OpenOffice and save the file with quotation marks (e.g. "mydictionary.utf8"). In both cases you should save the file encoded in UTF8 (otherwise special characters such as umlauts may not appear correctly).
Using spelling-check
Once you have installed OpenOffice you should have several dictionaries for checking the spelling of documents. More can be installed through a routine of the program itself. However, there is no existing dictionary for Turkish. In OmegaT do the following: Click "options"-"spelling". A separate window opens and if you cannot see the available languages you have to browse to your program OpenOffice.
For OpenOffice 3 the path in my case was
C:\Programme\Openoffice.org 3\share\uno-packages\cache\uno-packages\AD.tmp_\dict-en.oxt
There may be other folders such as AB.tmp and AF.tmp. You can see which one suits your needs best. If you enable "automatically check the spelling" then the misspelled words will automatically be highlighted.
Things are quite different under Linux. In my case using Ubuntu 10.10 (as of March 2011) the paths that were given as examples as results in Google did not work for me. By mere coincidence I found the dictionaries at
/usr/share/dictionaries-common
and could install the ones that I need (BTW: there is a dictionary for Kurdish, based on Turkish or the Latin alphabet, but no Turkish dictionary).
Tips from the Yahoo Group
Among the Yahoo Groups there is a group on OmegaT. As of May 2009 it had more than 1000 members and over 14000 messages. You can search the messages without being a member, but if you want to scroll through the messages you have to become a member (otherwise you can only see the 10 oldest and newest entries for each month). You might also want to have a look at the forum of translatorcafe on Omegat
Tip 1: If you start a new project and want to use the translation memory of as previous project you can look at this message or copy all the subfolders of your old project into the new one (overwrite if necessary).
Tip 2: For conversion of file (types) and creating reports apsic may have some solutions.
Videos for Beginners
There are two sets of videos. Five clips were created with a demo version of Instant Demo and, therefore, show a link on the right bottom. No voice was recorded (only the clicks can be heard). The films pause when explaining text appears, but can be stopped at any stage. That could not be achieved when displaying them in MediaWiki. Therefore, I've put them in the ordinary section of my home page.
The second set of videos was recorded with voice using BB Flashback. I have made four clips
- Creating a project in OmegaT length 5:19
- Translating with OmegaT length 5:25
- Build your own glossary in OmegaT length 2:32
- Creating your own segmentation rules length 5:15
The following clips were done with Instant Demo:
- Make new project and load correct file in OmegaT; 6:47
- Translating with memory in OmegaT, some files have already been translated; 5:41
- Create a glossary for OmegaT; 5:25
- See a (not so) talented translator at work, fuzzy matches, use of the glossary and spell check; 8:20
- Final corrections and creation of translated file(s); 6:48
I have also made a page for the standard introduction Samuel Murray prepared as creative commons licensed video on First steps in OmegaT