Achim Herrmann wrote:
You are right. In the last years more and more software file formats are based on XML or text files like JSON that can be translated with virtually any translation tool. The software UIs are build more dynamically and no static layout information is stored in the resource files. This will prevent that the software UI can be visualized, checked for truncations and adapted during the translation process with tools like SDL Passolo.
From the translator's point of view, the shift from C/C++ to C# was simply catastrophic.
Nevertheless I see an increasing demand from software manufacturers to add additional context information to the localization process. They are suffering from low translation quality when the files are translated using standard document translation tools without context.
Finally they start to realize it.
A lot of old tested GUIs were screwed up because of 100% matches with no context information and no extensive testing.
E.g. the S&R Window in MS Word in Polish has a blatant error i.e. Format is translated as verb (imperative) while a noun is expected.
The error was introduced in Word 2007 (using the .resx model from C#, no preview...),
BTW, the help text inheritated from Word 2003 contains the correct translation
Compared to specialized visual localization tools there is no ability to associate helpful comments for translators, to add pictures, screen shots or web links to single software strings.
Nowadays, probably the most effective tool like that I use is a special fancy Excel sheet provided by one of my customers...
This kind of "progress" is completely crazy.