Software localization is the process by which the original version of a software application is translated and adapted to be used in a different market. A software localization project usually comprises some or all of the following tasks:

• Preparation of materials
• Draft translation
• Editing
• Generation and proofreading of user interface
• Client review of user interface
• Update of translation with client changes
• Generation of target help system
• Linguistic test
• Functional test
• Generation of user manual and remaining materials

Depending on the specific requirements of the client, our team can undertake all of the above tasks, or just a subset of them. In some instances, for example, the scope of our team’s contribution is limited to the linguistic part of the project, while the client undertakes the software engineering and project management. In these cases, the client provides a group of files (resource files, help files, text files, etc.) and, after we perform the translation and editing, the client incorporates them into the software package and tests the localized application. In other cases, our team plays a more active role in the localization project, by undertaking several or all of the above tasks. Ultimately, the decision about the scope of our participation is made by the client after evaluating the relative costs and availability of each of the parties involved.

Next we briefly describe each of the tasks mentioned above.

Preparation of materials

The localization specialist analyses project-specific client requirements and all materials comprising the software product. He collects any available product line glossaries, previous versions of the software product and its documentation, and similar software products already localized, and makes sure that all required software tools are available in the appropriate versions. He then designs a customized plan for the execution of the localization project in agreement with the client.

The different modules to be localized, namely, user interface, help system, installation wizard, getting started guide, user manual, readme files, content on packaging materials and labels, warranty card, and supplementary documentation, along with any tutorials, templates or sample files, are usually imported into a translation memory software in preparation for the translation. Text on splash screens and other non-editable text blocks embedded in bitmap figures are typed out and saved in a separate text file, which is also imported for processing with the translation memory engine along with the rest of the files. The translation memory engine, which keeps all previously translated materials in a database, automatically searches for segments of the imported files that match those in the database. This operation not only allows the reuse of previous translations and thus can result in significant savings, but it also provides a superior degree of translation and terminology consistency, helps preserve the original structure of the documents, and protects the software code against unintentional alterations. Available glossaries and previous versions of the software product already localized may also be aligned with their corresponding translations and imported in the translation memory.

Draft translation

The translator starts by browsing through the program and software documentation to be localized, as well as through other references provided, in order to get familiar with the software application. He then performs the translation using available glossaries and reference materials.

Usually, the user interface module is the first in the translation sequence, and it is later used as the main glossary for the translation of the rest of the components. This module includes all menus, dialog boxes, status and error messages, etc. Once the translation of the strings comprising the user interface module is finished, the translator checks the translation for completeness, grammar and spelling, and then passes it to the editor. The translator then continues with the translation of the remaining modules (help files, user manual, etc.), which are not going to be delivered to the editor until after receiving feedback from editor/client review on the translation of the user interface module.

Editing

The editor, another translator with equal or higher qualifications, reviews the translation comparing it to the source document. He checks for grammar, punctuation, spelling, and syntax. The editor also checks for inappropriate or inconsistent terminology, misunderstandings, mistranslations, omissions or additions, and inappropriate style or register, and makes any necessary changes to ensure that the translation is culturally acceptable and that the language is suitable for its purpose. Questions, comments and suggestions about the source document are brought to the client's attention during this stage. Once the editing of the strings comprising the user interface module is finished, the editor passes it to the localization specialist.

Generation of user interface

The localization specialist usually exports the translated text strings from the translation memory tool into its original resource files, and rebuilds the translated user interface module. As part of the process, he resizes menus and dialog boxes to accommodate expansion of translated text strings, checks for conflicts on hot key assignments, etc. Any necessary changes in the software code, such as measurement unit or currency conversions, are performed in close coordination with the client.

Proofreading of user interface

This "draft" user interface module is sent to a proofreader, who checks it for any remaining grammar, spelling, hyphenation, line breaks, and reformatting errors, to make sure that it meets professional translation standards and that it is ready to be delivered to the client for review.

Client review

The client performs an in-house review of the user interface and generates a report requesting changes in terminology and other changes. This process may require a second client revision after retrofeeding client changes in the new build of the user interface. At the end of this step, the client provides the approval of the user interface so that, from this point on, the user interface can be used as the approved reference for terminology in the translation of the help system, user manual and remaining materials.

Update of translation with client changes

The translation/editing of the remaining materials (help system, installation wizard, user guides, readme files, content on packaging materials and labels, warranty card, supplementary documentation, tutorials, templates, sample files) usually runs in parallel with the preparation of the "draft" user interface for client review. Therefore, changes, feedback and other comments from the client review are expected to be available for retrofeeding after completion of the translation and editing of those materials.

Generation of the help system

The localization specialist compiles the help system previously updated with client changes and links it to the user interface.

Linguistic test

This step includes the final proofreading of the text on each one of the user interface screens and help system, checking for any remaining grammar, spelling, hyphenation, line breaks, and reformatting errors, with particular attention to inconsistencies between user interface and help system, to make sure that it meets professional translation standards and that it is ready to be delivered to the client.

Functional test

The functionality of every option of the software is checked, to make sure that the translated software operates exactly as expected. This step is based on a testing procedure which is either provided by the client or designed along with the client.

Generation of user manual and remaining documentation

The user manual and remaining materials are produced as described in the general translation procedure which is outlined in the Translation section.