The key to handling this project successfully demanded using both the right resources and the right processes.
ULG instituted a specific process to address the diversity of file formats, to ensure that quality control standards were maintained throughout the translations. The files for this project were particularly complex in that they contained hidden elements - white text on white background and elements that were either invisible or not immediately visible.
For the source content embedded within the Flash files, ULG employed a proven Flash text extraction technology. The extracted text was independently verified by a localization engineer since Flash files can be created in many different ways, making automatic text extraction difficult and cumbersome.
A second text extraction tool was created to handle the specific task of extracting embedded content from ActionScript files.
ULG assigned localization engineers with extensive Flash experience to the initial text extraction. The engineers diligently documented every action taken in the project instructions in order to ensure that the post-translation phase would also be as smooth as possible.
At every step of this project, team members verified that all text was present, localized and corrected in all the learning objects, as verified against the English version of the training. ULG’s linguistic testers and quality control specialists were also instructed to do a full walkthrough of all of the training on a specifically set up test bed in the ULG in-house lab.
This allowed testers to view the source and target content side-by-side on localized operating systems. ULG’s online proprietary defect tracking system was used at every step of the project to efficiently handle any bugs that were found. Each bug could easily be tracked, assigned to the appropriate resource to fix, and regressed and closed by the person who originally logged it.